^MeL₂Zn₂(μ-1,6-Ph₂-N₆) – a building block for new hexazene complexes

Abstract

The zinc hexazene complex ^MeL₂Zn₂(μ-1,6-Ph₂-N₆) 1 (^MeL = HC[C(Me)N(2,4,6-Me₃C₆H₂)]₂) is a suitable hexazene transfer reagent in reactions with main group metal and transition metal complexes containing M–Me units. The reactions proceed with elimination of ^MeLZnMe and the resulting complexes were characterized by NMR and IR spectroscopy and single crystal X-ray diffraction (5, 8). Quantum chemical calculations were performed to investigate the electronic structure of 5′ and 8′ in more detail and to identify the absorption bands of the hexazene unit.

---

Introduction

Metal complexes containing a metal-bonded hexazene dianion, [RNNNNNNR]²⁻, which have recently received growing interest, are still rare and only eight complexes have been reported to date (Scheme 1). They are typically synthesized by one-electron reductive coupling reactions of two organic azide molecules RN₃, in particular the sterically demanding 1-azidoadamantane (AdN₃), with iron(I)^1,2 and magnesium(I)^3,4 complexes. These strongly reducing metal complexes are typically kinetically stabilized by monoanionic, N,N′-chelating β-diketiminato or guanidinato groups. In addition, an iron(II) hexazene complex was recently prepared by reaction of AdN₃ with a less reducing bis(alkoxide) iron(II) complex.⁵

Scheme 1 Structurally characterized metal hexazene complexes.

According to density functional theory studies, the reaction proceeds via initial formation of a dimeric iron azide reaction intermediate, in which the azide group is monoreduced to a bridging azide radical and the iron centers are oxidized to the formal +3 oxidation state. The iron azide then reacts to the hexazene complex.⁵ In contrast, the reaction of the Fe(I) complex (^MeL′Fe)₂N₂ (^MeL′ = HC[C(Me)N(2,6-i-Pr₂C₆H₃)]₂) with AdN₃ proceeds through a different reaction mechanism as can be already derived from the observation, that this reaction is strongly solvent-dependant. The hexazene-bridged diiron(II) complex [(^MeL′Fe)₂(μ–η²:η²-PhN₆Ad)] was obtained in high yields in non-coordinating solvents,¹ while the tetrazene complex ^MeLFe(AdNNNNAd) containing a monoanionic radical tetrazene ligand (AdNNNNAd˙⁻) and a Fe(II) metal center were obtained in the presence of coordinating solvents such as t-Bu-pyridine.⁶ QM/MM computations showed that the formation of [(^MeL′Fe)₂(μ–η²:η²-PhN₆Ad)] could not result from simple radical dimerization of two molecules of L^MeFe(N₃Ad˙) since there is not a significant thermodynamic driving force for this dimerization.⁷ Instead, the diiron organoazide intermediate (^MeLFe)₂(μ-N₃Ad) is initially formed, which gives the hexazene complex upon addition of a second AdN₃ molecule. In contrast, displacement of N₂ in L^MeFeNNFeL^Me by a coordinating solvent yields a monometallic complex, which reacts with AdN₃ with formation of non-isolable L^MeFe^II(N₃Ad˙⁻) containing a radical monoanion, which eliminates N₂ with subsequent formation of the Fe(III) imido complex L^MeFe=NAd. The imido complex then reacts in a dipolar 1,3-cycloaddition reaction to the tetrazene complex as was previously observed in reactions of other low-valent metal complexes with organic azides.^8–11 The reactions most likely involve the initial formation of a transient metal imide (M=NR) intermediate, which subsequently undergoes a dipolar 1,3-cycloaddition reaction with a second equivalent of the azide RN₃ as was shown for isolated metal-imido complexes.^12–15

We recently reported on reactions of the one-electron reducing Zn(I) agent (^MeL)₂Zn₂ (^MeL = HC[C(Me)N(2,4,6-Me₃C₆H₂)]₂) with organic azides including the sterically less demanding PhN₃, yielding the first Zn(II) hexazene complex [(^MeLZn)₂(μ–η²:η²-PhN₆Ph)] 1.¹⁶ The solid state structure of 1 is almost identical to those of other hexazene complexes and the bonding situation in the dianionic hexazene unit is best described with a central N–N single bond and two allyl-like, monoanionic N₃ moieties with delocalized π-electrons. In remarkable contrast to neutral hexaazadienes,^17–21 metal hexazene complexes are thermally very stable due to the sterically demanding organic substituents. To the best of our knowledge, the chemical reactivity of hexazene complexes has not been investigated in detail, to date. Herein we show that 1 with the sterically less demanding Ph-substituted hexazene moiety is a powerful hexazene transfer reagent in reactions with transition metal and main group metal complexes. Moreover, quantum chemical calculations were performed to identify the N₆ vibrations in the resulting complexes.

Results and discussion

Reactions of 1 with equimolar amounts of Me₂Zn, Me₃Al and MeLi (Scheme 2) in C₆D₆ occurred with elimination of ^MeLZnMe and subsequent formation of the heteroleptic hexazene complex 2 and the heterobimetallic hexazene complexes 3 and 4. Organozinc complexes are long known transfer reagents and their capability to serve as starting reagents for the synthesis of heterometallic complexes has recently been shown.²²

Scheme 2 Synthesis of the heteroleptic hexazene complex 2 and heterobimetallic hexazene complexes 3 and 4 by metal exchange reactions of [(^MeLZn)₂(μ–η²:η²-N₆Ph₂)] 1 with equimolar amounts of different metal alkyl complexes (MeLi, Me₂Zn, Me₃Al).

The formation of the homobimetallic hexazene complex [(^MeLZn)(μ–η²:η²-N₆Ph₂)(ZnMe)] 2 and the heterobimetallic hexazene complexes [(^MeLZn)(μ–η²:η²-N₆Ph₂)(AlMe₂)] 3 and [(^MeLZn)(μ–η²:η²-N₆Ph₂)(Li)] 4 was confirmed by in situ¹H NMR spectroscopy. Since 2 and 3 show a comparable solubility as the by-product ^MeLZnMe, numerous attempts to separate 2 and 3 from ^MeLZnMe by fractional crystallisation from different organic solvents failed. Therefore, pure samples of the heteroleptic complexes 2 and 3 could not be isolated. In contrast, pure 4 was isolated by filtration since it is almost insoluble in organic solvents.

The in situ¹H NMR spectra of 2 and 3 (Fig. 1) showed the expected resonances of the by-product ^MeLZnMe as well as a set of resonances of the ^MeL substituent, the Ph groups of the hexazene moiety and the metal-Me substituent in a relative intensity of 1 : 2 : 1 (2) and 1 : 2 : 2 (3), respectively, clearly proving the substitution of only one ^MeLZn unit of 1 by a ZnMe (2) and AlMe₂ (3) unit. Since 4 is completely insoluble in organic solvents, an in situ¹H NMR spectrum of the reaction of 1 and one equivalent of MeLi only showed the resonances of resulting ^MeLZnMe. Unfortunately, crystals of 2–4 suitable for single crystal X-ray diffraction studies were not obtained.

Fig. 1 ¹H NMR spectra of 1 (top), ^MeLZnMe (middle) and the reaction of 1 and Me₂Zn (bottom), demonstrating the substitution of only one ^MeLZn moiety by a MeZn unit.

Detailed time-dependant in situ¹H NMR spectroscopic studies showed, that 2 slowly rearranged in solution into the corresponding homoleptic complexes [(^MeLZn)₂(μ–η²:η²-PhN₆Ph)] 1 and [MeZn)₂(μ–η²:η²-PhN₆Ph)] 5 and after five days, the formation of 1 and 5, which crystallized in the NMR tube as deep red crystals due to its limited solubility, was complete. The resulting ¹H NMR spectrum only showed resonances of 1 and ^MeLZnMe. As a consequence, we investigated the substitution of both ^MeLZn moieties in 1 by reaction with two equivalents of Me₂Zn and Me₃Al (Scheme 3), respectively.

Scheme 3 Synthesis of homobimetallic hexazene complexes 5 and 6 by metal exchange reactions of 1 with two equivalents of Me-substituted metal complexes and the donor-stabilized complex 8.

The formation of two equivalents of ^MeLZnMe in the resulting reaction mixtures was clearly confirmed by in situ¹H NMR spectroscopy. In addition, one set of resonances due to the AlMe₂ (6) groups and the Ph groups of the hexazene unit in a relative intensity of 1 : 1 are visible, pointing to the formation of the homobimetallic hexazene complex [Me₂Al)₂(μ–η²:η²-PhN₆Ph)] 6. Since 5 is insoluble in non-coordinating deuterated solvents (benzene, toluene, fluorobenzene) even at elevated temperatures, it immediately formed a red crystalline powder in the NMR tube and as a consequence, NMR data could not be obtained in these solvents. In contrast, 5 dissolved easily in coordinating solvents such as thf-d₈ with a color change to yellow. The ¹H NMR spectrum of the resulting, most likely thf-coordinated, complex show the expected resonances of the Zn–Me and Ph groups in a relative intensity of 1 : 1. In the ¹³C NMR spectrum the Zn–Me₃ signal was not visible as is often observed for Zn–Me compounds.

The transmetalation reactions occurred stepwise as was proven by in situ¹H NMR spectroscopy for the reaction of 1 with Me₂Zn (Fig. 2). Addition of one equivalent of Me₂Zn to 1 yielded 2, which then reacted with a second equivalent of Me₂Zn with formation of 5 and ^MeLZnMe, respectively. 5 was also obtained from the reaction of 2 with one equivalent of Me₂Zn.

Fig. 2 ¹H NMR spectra of 1 +one equivalent of Me₂Zn (top), +two equivalents of Me₂Zn (middle), 2 with one equivalent of Me₂Zn (bottom), clearly demonstrating the stepwise substitution reaction of the ^MeLZn moieties.

In remarkable contrast to the reactions of 1 with two equivalents of ZnMe₂ and AlMe₃, its reaction with two equivalents of MeLi was completely different. 1 reacts with the first equivalent MeLi with formation of ^MeLZnMe and 4, which then reacts with a second equivalent of MeLi to ^MeLLi and 7 (Scheme 3). No further reaction was observed with a third equivalent of MeLi.

Single crystals of 5 were obtained from a solution of 1 in C₆D₆, which was carefully stratified with a solution of ZnMe₂, upon standing (without stirring) at ambient temperature for 6 hours. An X-ray diffraction study revealed the formation of the homobimetallic hexazene complex [(MeZn)₂(μ–η²:η²-PhN₆Ph)] 5 (Fig. 3), which crystallizes in the monoclinic space group C2/c with the molecule placed on a centre of inversion.

Fig. 3 Solid state structure of 5 (thermal ellipsoids are shown at 50% probability levels); H atoms are omitted for clarity.

The N₄Zn metallacycle is almost perfectly planar (r.m.s. deviation from the least-squares plane 0.056(6) Å). The central N₆Ph₂ hexazene unit adopts a bridging position and is connected to two three-coordinated Zn atoms. The Zn–C bond lengths (Zn–C 1.953(3) Å) are in the range of known zinc methyl complexes.^23–25 The different N–N bond lengths within the hexazene unit give insight into the electron localization within the complex. The central N3–N3′ bond (1.378(5) Å) is significantly elongated compared to the almost identical N1–N2 (1.305(3) Å) and N2–N3 (1.299(3) Å) bonds, hence the hexazene unit is best described as a dianionic ligand with a central N3–N3′ single bond and two allyl-like, monoanionic N₃ moieties containing delocalized π-electrons between N1, N2 and N3. The N1–N2–N3 bond angle of 118.4(2)° and the N–N bond lengths within the hexazene unit are comparable to those previously observed in hexazene complexes (Table 1).^1–5,16 In contrast to as-described hexazene complexes, the π-electrons with the five-membered ring of dianionic tetrazene complexes MN₄ (Ni, Fe, Al, Cr) are not delocalized. According to singly crystal X-ray diffraction studies and computational studies, the tetrazene complexes have a central N–N double bond and two N–N single bonds.^6,26–28

Table 1 Bonding parameters of structurally characterized hexazene complexes

	N1–N2	N2–N3	N3–N3′	N1–N2–N3
a This work; L = CH3CN; Pipiso = [(DipN)2C(cis-NC5H8Me2-2,6)]^–.
(^MeL′Fe)2N6Ada2 ^1	1.304	1.305	1.411	117.107
(^t-BuL′Fe)2N6Ad2 ^1	1.307	1.318	1.408	116.485
[(Pipiso)Fe]2N6Ada2 ^2	1.300	1.309	1.415	117.487
[(RO)2Fe]2N6Ada2 ^5	1.295	1.308	1.399	117.592
(^MeL′Mg)2N6Ada2 ^3	1.301	1.314	1.410	117.827
(^t-BuL′Mg)2N6Ada2 ^4	1.304	1.312	1.419	117.984
(^MeLMg)2N6Ada2 ^4	1.303	1.310	1.424	117.853
(^MeLZn)2N6Ph21 ^16	1.304	1.302	1.403	117.350
(MeZn)2N6Ph25 ^a	1.305	1.299	1.377	118.374
[Me(L)Zn]2N6Ph2 8^a	1.310	1.297	1.402	117.352

---

Intermolecular Zn⋯N interactions (2.275(2) Å) between adjacent hexazene complexes in 5, which result from the electron deficiency of the threefold-coordinated Zn atom, result in the formation of ladder-like one-dimensional polymeric structure (Fig. 4).

Fig. 4 Intermolecular interactions in 5 leading to ladder-type 1-dimensional chains in the solid state (thermal ellipsoids are shown at 50% probability levels); H atoms are omitted for clarity.

The intermolecular Zn⋯N interactions can be disrupted in reactions with stronger Lewis bases as was exemplarily shown in the reaction of 5 with two equivalents of CH₃CN, resulting in the formation of the Lewis base-coordinated hexazene complex (MeZn)₂(μ–η²:η²-PhN₆Ph)(CH₃CN)₂8. Crystals of 5 were suspended in C₆D₆, yielding a deep red suspension, which immediately turned to a yellow solution after addition of CH₃CN. Since the solubility of the resulting CH₃CN-stabilized complex [(MeZn)₂(μ–η²:η²-N₆Ph₂)(CH₃CN)] 8 is significantly better than that of base-free 5, ¹H and ¹³C NMR spectra could be recorded. 8 shows the expected resonances due to the metal–Me substituent, the Ph groups of the hexazene moiety and the coordinating acetonitrile in a relative intensity of 1 : 1 : 1. The NMR signal of the acetonitrile molecule (2.12 ppm) is significantly shifted to lower field compared to free CH₃CN (0.58 ppm).

Single crystals of 8 were obtained from a concentrated solution in C₆D₆ after storage at +4 °C for 12 hours (Fig. 5). 8 crystallizes in the monoclinic space group P2₁/n with two molecules in the unit cell. The zinc atoms in 8 are slightly out of the plane of the N₄Zn rings (r.m.s. deviation from the least-squares plane 0.142(3) Å). The N–N bond lengths (N1–N2 1.310(16) Å, N2–N3 1.2968(16) Å and N3–N3′ 1.402(2) Å) and the N1–N2–N3 bond angle of 117.35(11)° within the hexazene unit are comparable to those observed in 1 and 5.

Fig. 5 Solid state structure of 8 (thermal ellipsoids are shown at 50% probability levels); H atoms are omitted for clarity.

The individual molecules of 8 form chains via CH⋯π interactions. It is a trifurcated interaction involving all three hydrogen atoms of the acetonitrile. The individual bond angles are comparably small (Table 2), however the whole group is nearly perpendicular to the plane of the phenyl ring (angle between best plane of the ring and C8–C9 vector: 83.4°). The neighboring molecules within the chain are related by inversion (Fig. 6). The chains are oriented along [−110] and in turn interconnected by CH⋯π interactions via screw-axis and glide-plane symmetry. Again a hydrogen of the acetonitrile is involved. Its interaction with the phenyl ring's π-systems shows the typical orientation of a strong interaction. The second non-classical hydrogen bond is formed between the phenyl ring (H5) and the π-electrons of the hexazene unit (represented by the N1–N2–N3–N3′ centroid). The values of its geometric parameters suggest a similar strength. The remaining voids in the packing are filled with benzene molecules.

Fig. 6 CH⋯π interactions in 8.

Table 2 CH⋯π interactions

DH⋯A^a	Symmetry	H⋯A/Å	DH⋯A/°	χ ^b/°
a az: centroid of the hexazene unit as defined below. Ph: centroid of the phenyl ring. b χ is the angle between the best plane of the ring and the H/centroid vector.
C9–H9a⋯Ph	−x + 1, −y, −z + 1	2.9976	101.27	72.7
C9–H9b⋯Ph	−x + 1, −y, −z + 1	3.4053	77.34	63.4
C9–H9c⋯Ph	−x + 1, −y, −z + 1	3.0481	98.11	76.1
C9–H9b⋯Ph	−x + 3/2, y − 1/2, −z + 3/2	2.7031	152.87	88.2
C5–H5⋯az	x − 1/2, −y + 1/2, z − 1/2	3.1145	133.98	68.9

---

Quantum chemical calculations

In order to identify the spectroscopic signatures of the hexazene unit, we have carried out density functional theory calculations with the BP86 functional²⁹ including a third generation dispersion correction (DFT + D3)^30,31 using a quadruple-zeta basis set³² for the free monomers of 5′ and 8′. In case of 5 this represents a drastic approximation since the bridging function of the zinc atoms connecting different monomers within the crystal thus is ignored. Geometry optimization leads to a C_2h symmetrical structure for 5′ (Zn–N1 2.035, Zn–N3 2.001, N1–N2 1.306, N2–N3 1.308, N3–N3′ 1.394 Å, N1–N2–N3 116.6°) and to a structure symmetry C_i for 8′ (Zn–N1 2.062, Zn–N3 2.031, N1–N2 1.311, N2–N3 1.301, N3–N3′ 1.400 Å, N1–N2–N3 116.4°). The most important difference between the quantum chemical gas phase structure of 8′ and the corresponding crystal structure is a bending of the acetonitrile molecules towards the phenyl rings to form intramolecular CH–π bonds. According to a natural population analysis³³ the total charge on the N₆ unit is −1.80e for 5′ (−0.50e for N1, −0.38e for N3) and −1.73e for 8′ (−0.50e for N1, −0.34e for N3) while the charge on Zn amounts to 1.37e for 5′ and 1.42e for 8′, respectively.

The most intense line above 1000 cm⁻¹ in the experimental IR spectrum of 8 at 1262 cm⁻¹ according to the DFT + D3 calculations has to be ascribed to an antisymmetric linear combination of N1–C (phenyl) and N1′–C′(phenyl′) stretching vibrations (with contributions from phenyl ring deformations and a slight N1–N2 stretch contribution), computed at 1269 cm⁻¹. The second most intense peak at 1203 cm⁻¹ is ascribed to the antisymmetric N1–N2/N1′–N2′ stretch vibration (with a contribution from C–C–H_ortho bend of the phenyl rings), computed at 1201 cm⁻¹. We assign the experimental line at 1341 cm⁻¹ to the antisymmetric N2–N3/N2′–N3′ stretch vibration (with phenyl ring deformation contributions), which in the theoretical spectrum is located at 1352 cm⁻¹. Note that the line at 1301 cm⁻¹ corresponds to an in plane deformation of the phenyl rings, calculated at 1299 cm⁻¹. Finally the first intense line below 1000 cm⁻¹ at 943 cm⁻¹ is assigned to an antisymmetric linear combination of the N1–N2–N3 and N1′–N2′–N3′ bending modes (again with contributions of in plane phenyl ring deformations), computed at the same frequency. These findings essentially agree with those for the Zn–hexazene complex 1.¹⁶

The assignment of the experimental IR spectrum of 5 is less unambiguous due to the fairly different structures of 5 in the crystal and 5′ in the gas phase: the intense line at 1258 cm⁻¹ is probably the antisymmetric C–N1/C′–N1′ stretch vibration, computed at 1264 cm⁻¹. However in the computed spectrum this line has a much lower intensity than that of the N2–N3/N2′–N3′ stretch vibration computed at 1294 cm⁻¹, which seems to be absent in the experimental spectrum, possibly due to the involvement of the N2 atom in a long bond to the zinc atom of a neighbouring molecule. The two small peaks at 1198 cm⁻¹ and 1343 cm⁻¹ on the other hand are likely to be ascribed to the antisymmetric N1–N2/N1′–N2′ and N2–N3/N2′–N3′ stretch vibrations, computed at 1204 and 1347 cm⁻¹, respectively. Again we find the antisymmetric N1–N2–N3/N1′–N2′–N3′ bending mode in the computed spectrum at 947 cm⁻¹ as the most likely explanation for the small experimental peak at 954 cm⁻¹. One may speculate that the intense absorption bands at 1079 and 1011 cm⁻¹ are due to Zn–N stretch vibrations along the corresponding weak Zn–N bonds between two neighbouring molecules in the crystal of 5.

The most intense line at 1065 cm⁻¹ in the experimental Raman spectrum of 5, which is also found in the spectrum of 8, according to our calculations corresponds to the symmetrical linear combination of the N1–N2–N3 and N1′–N2′–N3′ bending vibrations with some admixture of N3–N3′ stretch, computed at 1031 and 1028 cm⁻¹, respectively. The neighbouring line at 1084 cm⁻¹, the most prominent feature in the experimental spectrum of 8, however is missing in the spectrum of 5. Thus it is likely due to an acetonitrile ligand vibration, the best candidate of which is CH₃-wagging mode. The latter is computed at 1021 cm⁻¹ but might be strongly shifted due to the different environment of the methyl group in the crystal of 8. The intense Raman lines in the experimental spectra at 1295 (5) and 1289 cm⁻¹ (8) were identified in our computations as the symmetric C–N1/C′–N1′ stretch vibrations, calculated at 1279 (5′) and 1290 cm⁻¹ (8′), respectively. The symmetrical linear combination of the N2–N3 and N2′–N3′ stretch vibrations is computed to have frequencies of 1361 (5′) and 1370 cm⁻¹ (8′), respectively, i.e., in a frequency range where in the experimental spectrum it is difficult to distinguish lines from noise. However, based on the computed frequencies of 1194 (5′) and 1192 cm⁻¹ (8′), we assign the experimental Raman bands at 1211 (5) and 1209 cm⁻¹ (8) to the symmetrical N1–N2/N1′–N2′ stretch vibration. The band at 938 cm⁻¹ can be assigned with the help of the computed line position of 922 cm⁻¹ to the symmetrical linear combination of acetonitrile C–C stretch vibrations, and the experimental bands at 895 (5) and 897 cm⁻¹ (8) correspond to the N3–N3′ stretch vibration, found at 882 (5′) and 860 cm⁻¹ (8′) in our calculations.

Conclusions

We demonstrated the promising potential of [(^MeLZn)₂(μ–η²:η²-PhN₆Ph)] 1 to serve as building block for the synthesis of both hetero- and homobimetallic main group metal and transition metal hexazene complexes. Our particular reaction pathway offers a general access to metal hexazene complexes and doesn't requires any reduction potential of the metal fragment that should be incorporated into the hexazene complex, which is in remarkable contrast to the previously reported route using low-valent reducing reagents such as Fe(I), Mg(I) and Zn(I) complexes. Quantum chemical calculations at the DFT + D3 level yielded structures (5′, 8′), which were in good agreement with experimental ones for the central N₆ unit and which allowed an assignment of the most prominent features related to the N₆ unit in the IR and Raman spectra.

Experimental

Synthetic procedures

All manipulations were performed in a glovebox under an Ar atmosphere or using standard Schlenk techniques. Solvents were carefully dried over Na/K and degassed prior to use. 1 was prepared according to literature method.¹⁶¹H (300 MHz) and ¹³C{¹H} (75.5 MHz) NMR spectra (δ in ppm) were recorded using a Bruker Avance DPX-300 spectrometer and are referenced to the trace of the respective protonated solvent impurities present in the deuterobenzene (C₆D₅H, ¹H: δ = 7.154; ¹³C: δ = 128.0) and deuterotetrahydrofuran (C₄D₈O, ¹H: δ = 3.580; ¹³C: δ = 67.4). IR spectra were measured in an ALPHA-T FT-IR spectrometer equipped with a single reflection ATR sampling module. The spectrometer was placed in a glovebox so as to perform the FT-IR measurements in inert gas atmosphere. Raman spectra were measured 100 s with a Bruker Senterra Microscope at 785 nm and 20 mW. The microanalyses were performed at the elemental analysis laboratory of University of Duisburg-Essen.

(^MeLZn)N₆Ph₂(Li) 4. 1 (90 mg, 90 μmol) was dissolved in toluene and MeLi (1.6 M in Et₂O, 55 μL, 90 μmol) was dropwise added at −40 °C. After warming to ambient temperature, the reaction mixture was stirred for 12 h, yielding an orange suspension. 4 was isolated by filtration as an orange solid, which was dried in vacuo. 4 is insoluble in any organic solvents, hence NMR spectra could not be recorded. However, in situ¹H NMR spectroscopy studies showed the formation of the by-product ^MeLZnMe.

Yield: 43 mg (74.3%). – M.P.: 273 °C. – C₃₅H₃₉N₈ZnLi (642.94): calcd C 65.32, H 6.07, N 17.42; found C 65.1, H 6.11, N 17.49. – IR: ν = 2962, 2913, 1605, 1549, 1516, 1452, 1395, 1351, 1323, 1258, 1199, 1085, 1012, 973, 858, 793, 758, 723, 693, 633, 491, 452, 390 cm⁻¹. – ¹H-NMR (300 MHz, C₆D₆, 25 °C): ^MeLZnMe: δ = −0.61 (s, 3H, Zn–CH₃), 1.63 (s, 6H, CCH₃), 2.12 (s, 12H, o-CCH₃), 2.14 (s, 6H, p-CCH₃), 4.97 (s, 1H, γ-CH), 6.81 (s, 4H, m-H).

(MeZn)₂N₆Ph₂5. 1 (60 mg, 60 μmol) was dissolved in C₆D₆ and ZnMe₂ (1.2 M in toluene, 100 μL, 120 μmol) was added. The reaction vessel is mixed over night, yielding a red suspension. 5 was isolated by crystallization at room temperature and filtration as red crystals. The crystals were suitable for a single crystal X-ray diffraction study.

Yield: 18 mg (86%). – M.P.: 217 °C (decomp.). – C₁₄H₁₆N₆Zn₂ (398): calcd C 42.21, H 4.02, N 21.10; found C 42.3, H 4.03, N 21.14. – IR: ν = 2962, 2906, 2819, 1595, 1485, 1398, 1343, 1257, 1197, 1079, 1010, 954, 861, 790, 724, 688, 661, 496, 398 cm⁻¹. – Raman: ν = 359, 391, 946, 1150, 1286, 1578, 1746, 1810, 2130, 2349 cm⁻¹. – ¹H-NMR (300 MHz, C₆D₆, 25 °C): ^MeLZnMe: δ = −0.62 (s, 6H, Zn–CH₃), 6.93 (m, 1H, Ph-p-H), 7.11 (m, 2H, Ph-m-H), 7.41 (m, 2H, Ph-o-H). The solubility of 5 in C₆D₆ and toluene-d₈ was too low to obtain suitable NMR spectra. However, 5 is soluble in coordinating solvents such as thf-d₈. ¹H-NMR (300 MHz, thf-d₈, 25 °C): δ = −0.55 (s, 3H, Zn–CH₃), 7.01 (m, 2H, Ph-p-H), 7.29 (m, 4H, Ph-m-H), 7.51 (m, 4H, Ph-o-H). – ¹³C-NMR (75 MHz, thf-d₈, 25 °C): δ = 119.38 (p-Ph), 124.60 (m-Ph), 129.89 (o-Ph), 150.31 (i-Ph).

(Me₂Al)₂N₆Ph₂6. 1 (60 mg, 60 μmol) was dissolved in C₆D₆ at −40 °C and AlMe₃ (5.74 μL, 120 μmol) was slowly added. The reaction vessel stirred over night, yielding an orange suspension. 6 was isolated by filtration as an orange solid, which was dried in vacuo.

Yield: 16 mg (84%). – M.P.: 224 °C. – C₁₆H₂₂N₆Al₂ (351.96): calcd C 54.55, H 6.25, N 23.87; found C 55.1, H 6.27, N 24.01. – IR: ν = 2962, 2914, 1594, 1554, 1485, 1453, 1397, 1344, 1257, 1197, 1087, 1012, 946, 859, 793, 755, 719, 690, 634, 495, 455, 392 cm⁻¹. – ¹H-NMR (300 MHz, C₆D₆, 25 °C): δ = −0.62 (s, 12H, Al-CH₃), 6.88 (m, 4H, Ph-o-H), 7.26 (m, 4H, Ph-m-H), 7.46 (m, 2H, Ph-p-H). – ¹³C-NMR (75 MHz, C₆D₆, 25 °C): δ = −18.05 (Al-CH₃), 124.46 (p-Ph), 132.19 (m-Ph), 133.31 (o-Ph), 152.85 (i-Ph).

(MeZn)N₆Ph₂(Li) 7. Method A : 1 (90 mg, 90 μmol) was dissolved in toluene at −40 °C and MeLi (1.6 M in Et₂O 112 μL, 180 μmol) was slowly added. The reaction mixture was stirred for 12 h, yielding an orange suspension. 7 was isolated by filtration as a yellow solid, which was dried in vacuo. Method B: 4 (29 mg, 47 μmol) was suspended in toluene at −40 °C and MeLi (1.6 M in Et₂O 29 μL, 47 μmol) was slowly added. The reaction mixture was stirred for 12 h, yielding an orange suspension. 7 was isolated by filtration as a yellow solid, which was dried in vacuo. 7 is insoluble in any organic solvents, hence NMR spectra couldn't be recorded. However, ¹H-NMR spectroscopy studies showed the formation of the by-products ^MeLZnMe and ^MeLLi, respectively.

Yield: 15 mg (53%). – M.P.: 326 °C. – C₁₃H₁₃N₆ZnLi (324.94): calcd C 48.01, H 4.00, N 25.85; found C 48.1, H 4.03, N 25.91. – IR: ν = 2960, 1625, 1481, 1453, 1259, 1200, 1084, 1013, 981, 857, 792, 695, 582, 440 cm⁻¹. – ¹H-NMR (300 MHz, C₆D₆, 25 °C): ^MeLLi: δ = 1.69 (s, 6H, C–CH₃), 2.04 (s, 12H, o-CCH₃), 2.30 (s, 6H, p-CCH₃), 4.79 (s, 1H, C–H), 6.93 (s, 4 H, Mes-H). ^MeLZnMe: δ = −0.61 (s, 3H, Zn–CH₃), 1.63 (s, 6H, CCH₃), 2.12 (s, 12H, o-CCH₃), 2.14 (s, 6H, p-CCH₃), 4.97 (s, 1H, γ-CH), 6.81 (s, 4H, m-H).

(MeZn)₂N₆Ph₂(CH₃CN)₂8. 5 (30 mg, 30 μmol) was suspended in 1 mL C₆D₆ and 0.1 mL of CH₃CN was added. The reaction mixture was stirred for 12 h, yielding an orange solution. Crystals of 8 suitable for a single crystal X-ray diffraction study were obtained after 48 h from a concentrated solution (0.5 mL) upon storage at 4 °C.

Yield: 13.3 mg (92.4%). – M.P.: 237 °C (decomp.). – C₁₈H₂₂N₈Zn₂ (481.17): calcd C 44.93, H 4.61, N 23.29; found C 45.02, H 4.60, N 23.42. – IR: ν = 3026, 2906, 2276, 1586, 1453, 1340, 1301, 1262, 1203, 1072, 999, 943, 896, 810, 755, 721, 667, 629, 547, 510, 493, 471, 442 cm⁻¹. – Raman: ν = 222, 244, 356, 389, 946, 1032, 1152, 1286, 1325, 1570, 1734, 1807, 2134, 2349 cm⁻¹. – ¹H-NMR (300 MHz, C₆D₆, 25 °C): δ = −0.36 (s, 6H, Zn–CH₃), 2.12 (s, 6 H, NCCH₃), 7.00 (m, 4H, Ph-o-H), 7.28 (m, 4H, Ph-m-H), 7.76 (m, 2H, Ph-p-H). – ¹³C-NMR (75 MHz, C₆D₆, 25 °C): δ = −15.04 (Zn–CH₃), 1.03 (NCCH₃), 119.60 (NCCH₃), 124.58 (p-Ph), 129.23 (m-Ph), 130.08 (o-Ph), 150.07 (i-Ph).

Crystallography

Crystallographic data of 5 and 8 were collected on a Bruker AXS SMART diffractometer (MoK_α radiation, λ = 0.71073 Å) at 100(1) K. The structures were solved by Direct Methods (SHELXS-97)³⁴ and refined anisotropically by full-matrix least-squares on F² (SHELXL-2014).^35,365 was refined as non-merohedral twin based on HKL5 data. Absorption corrections were performed semi-empirically from equivalent reflections on basis of multi-scans (Bruker AXS APEX2, TWINABS). Hydrogen atoms were refined using a riding model or rigid methyl groups.

Acknowledgements

S.S. and G.J. gratefully acknowledge financial support by the University of Duisburg-Essen.

Notes and references

1. R. E. Cowley, J. Elhaïk, N. A. Eckert, W. W. Brennessel, E. Bill and P. L. Holland, J. Am. Chem. Soc., 2008, 130, 6074.
2. L. Fohlmeister and C. Jones, Aust. J. Chem., 2014, 67, 1011.
3. S. J. Bonyhady, S. P. Green, C. Jones, S. Nembenna and A. Stasch, Angew. Chem., 2009, 121, 3017 ( Angew. Chem. Int. Ed. , 2009 , 48 , 2973 ).
4. S. J. Bonyhady, C. Jones, S. Nembenna, A. Stasch, A. J. Edwards and G. J. McIntyre, Chem. – Eur. J., 2010, 16, 938.
5. J. A. Bellow, P. D. Martin, R. L. Lord and S. Groysman, Inorg. Chem., 2013, 52, 12335.
6. R. E. Cowley, E. Bill, F. Neese, W. W. Brennessel and P. L. Holland, Inorg. Chem., 2009, 48, 4828.
7. R. E. Cowley, N. J. DeYonker, N. A. Eckert, T. R. Cundari, S. DeBeer, E. Bill, X. Ottenwaelder, C. Flaschenriem and P. L. Holland, Inorg. Chem., 2010, 49, 6172.
8. C. Cui, H. W. Roesky, H.-G. Schmidt and M. Noltemeyer, Angew. Chem., 2000, 112, 4705 ( Angew. Chem., Int. Ed. , 2000 , 39 , 4531 ).
9. N. J. Hardman and P. P. Power, Chem. Commun., 2001, 1184.
10. H. Zhu, Z. Yang, J. Magull, H. W. Roesky, H.-G. Schmidt and M. Noltemeyer, Organometallics, 2005, 24, 6420.
11. M. T. Mock, C. V. Popescu, G. P. A. Yap, W. G. Dougherty and C. G. Riordan, Inorg. Chem., 2008, 47, 1889.
12. A. H. Obenhuber, T. L. Ganietti, X. Berrebi, R. G. Bergman and J. Arnold, J. Am. Chem. Soc., 2014, 136, 2994.
13. K. E. Meyer, P. J. Walsh and R. G. Bergman, J. Am. Chem. Soc., 1995, 117, 974.
14. A. M. Geer, C. Tejel, J. A. López and M. A. Ciriano, Angew. Chem., 2014, 126, 5720 ( Angew. Chem., Int. Ed. , 2014 , 53 , 5614 ).
15. R. I. Michelman, R. G. Bergman and R. A. Andersen, Organometallics, 1993, 12, 2741.
16. S. Gondzik, S. Schulz, D. Bläser, C. Wölper, R. Haack and G. Jansen, Chem. Commun., 2014, 50, 927.
17. F. R. Berson, The High Nitrogen Compounds, Wiley, New York, 1984.
18. O. Nuyken, C. Scherer, A. Baindl, A. R. Brenner, U. Dahn, R. Gärtner, S. Kaiser-Röhrich, R. Kollefrath, P. Matusche and B. Voit, Prog. Polym. Sci., 1997, 22, 93.
19. K. A. Hofmann and H. Hock, Chem. Ber., 1911, 44, 2946.
20. D. Mackay, D. D. McIntyre and N. J. Taylor, J. Org. Chem., 1982, 47, 532.
21. C. M. Fitchett, C. Richardson and P. Steel, J. Org. Biomol. Chem., 2005, 3, 498.
22. T. Kaczorowski, I. Justyniak, D. Prochowicz, K. Zelga, A. Kornowicz and J. Lewiński, Chem. – Eur. J., 2012, 18, 13460; W. Clegg, E. Crosbie, S. H. Dale-Black, E. Hevia, G. W. Honeyman, A. R. Kennedy, R. E. Mulvey, D. L. Ramsay and S. D. Robertson, Organometallics, 2015, 34, 2580.
23. S. Marks, T. K. Panda and P. W. Roesky, Dalton Trans., 2010, 39, 7230.
24. M. S. Hill and P. B. Hitchcock, Dalton Trans., 2002, 4694.
25. A. Kasani, R. McDonald and R. G. Cavell, Organometallics, 1999, 18, 3775.
26. S. W. Lee, G. A. Miller, C. F. Campana and W. C. Trogler, Inorg. Chem., 1988, 27, 1215.
27. H. P. Zhu, Y. Zhi, J. Magull, H. W. Roesky, H.-G. Schmidt and M. Noltemeyer, Organometallics, 2005, 24, 6420.
28. W. H. Monillas, G. P. A. Yap and K. H. Theopold, Inorg. Chim. Acta, 2011, 369, 103.
29. (a) A. D. Becke, Phys. Rev. A, 1988, 38, 3098; (b) J. P. Perdew, Phys. Rev. B: Condens. Matter, 1986, 33, 8822.
30. S. Grimme, WIREs Comput. Mol. Sci., 2011, 1, 211.
31. F. Furche, R. Ahlrichs, C. Hättig, W. Klopper, M. Sierka and F. Weigend, WIREs Comput. Mol. Sci., 2014, 4, 91.
32. (a) F. Weigend, F. Furche and R. Ahlrichs, J. Chem. Phys., 2003, 119, 12753; (b) F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297.
33. (a) A. E. Reed, R. B. Weinstock and F. Weinhold, J. Chem. Phys., 1985, 83, 735; (b) A. E. Reed, L. A. Curtis and F. Weinhold, Chem. Rev., 1988, 88, 899.
34. G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystallogr., 1990, 46, 467.
35. G. M. Sheldrick, SHELXL-2014, Program for the Refinement of Crystal Structures, University of Göttingen, Göttingen (Germany), 2014. (see also; G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystallogr., 2008, 64, 112 ).
36. shelXle, A Qt GUI for SHELXL, C. B. Hübschle, G. M. Sheldrick and B. Dittrich, J. Appl. Crystallogr., 2011, 44, 1281.

---

Footnotes

† Dedicated to Prof R. Boese on the occasion of his 70th birthday.

‡ Electronic supplementary information (ESI) available: Crystallographic data of 5 and 8; ¹H (2–8), ¹³C (5, 6, 8), IR (2–8) and Raman spectra (5, 8); quantum chemical calculations of 5′ and 8′. CCDC 1408085 (5) and 1408086 (8). For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5dt02423d

---