Synthesis and Structural Characterization of Dinitrogen Chromium Complexes with Triamidoamine Ligands Possessing Bulky Substituents, and Nitrogen Fixation by These Complexes

1. Introduction

Nitrogen fixation occurs primarily in nature via nitrogenase enzymes and industrially through the Haber-Bosch process [1,2]. However, as the Haber-Bosch process requires high-temperature, high-pressure reaction conditions and hydrogen derived from fossil fuels, active research is underway into nitrogen fixation under milder conditions without hydrogen as the proton source [3]. Among various studies on nitrogen fixation, research on dinitrogen activation using transition metal complexes has been vigorously pursued since the Schrock’s group first succeeded in catalytic nitrogen fixation using complexes centered on molybdenum, the metal present in nitrogenase enzymes [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29].

Group VI metals such as molybdenum and tungsten are frequently employed as representative transition-metal complexes for dinitrogen activation, and highly interesting researches have been reported to date [12,13,16,18,20,29,30,31,32,33,34,35,36,37,38,39,40,41]. However, research utilizing chromium, which also belongs to Group VI, has been scarcely conducted; consequently, studies on the activation of dinitrogen molecules using chromium are highly significant [29,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70]. Among these, although few in number, the dinitrogen activation reactions using chromium complexes have been reported, with some reporting the production of ammonia, hydrazine, hydrazine derivatives, or tris(trimethylsilyl)amine (N(SiMe₃)₃) [44,46,47,48,49,52,55,56,58,59,60,61,62,63,64,65,66,67,68,69,70]. The oxidation states of the chromium ions in these chromium-dinitrogen complexes are primarily I, II or III. However, recently, dinitrogen complexes in which the formal oxidation state of the chromium ion is –I have been reported [63]. We also synthesized several dinitrogen complexes containing Cr(IV) ions, which had not previously been used to activate dinitrogen, and reported that these complexes react with a proton source to produce ammonia and hydrazine [71]. However, as numerous unknowns remain in research concerning dinitrogen activation using chromium ions, further studies on dinitrogen activation employing diverse chromium ions are required.

Previously, we synthesized and structurally characterized dinitrogen chromium(IV) complexes bearing triamidoamine ligands and investigated their nitrogen fixation reactions yielding ammonia and hydrazine. However, the terminal substituent of the triamidoamine ligand was limited to a benzyl group. Therefore, in this study, the terminal substituents of the ligands were changed to the bulky tren derivatives, tris(2-(3-pentylamino)ethyl)amine (H₃L^Pen) and tris(3-dicyclohexylmethylaminoethyl)amine (H₃L^Cy), and the effect of substituents on chemical properties and reactivities with dinitrogen were investigated.

2. Results and Discussion

2.1. Preparation of Bulky Tren Ligands and Their Chromium Complexes, 1, 2, 3, 4, 5(THF), and 6

Two tren derivatives with bulky substituents, H₃L^Pen and H₃L^Cy, were synthesized by the previously reported method, and several chromium complexes were prepared using these ligands [72]. These tren derivatives were lithiated with 3 equivalents of n-BuLi, yielding Li₃[L^Pen] and Li₃[L^Cy], respectively.

Li₃[L^Pen] reacted with CrCl₃ in Et₂O under N₂ to yield the dinitrogen dichromium complex, [{Cr(L^Pen)}₂(μ-N₂)] (1) (Scheme 1). Recrystallization of complex 1 from Et₂O solution gave dark red crystals. Reacting complex 1 with 10 equivalents of K metal in Et₂O under N₂ atmosphere yielded the dichromium complex containing two dinitrogen ligands and two K⁺ ions, [{CrK(L^Pen)(μ-N₂)(Et₂O)}₂] (2) (Scheme 1). Recrystallization of complex 2 from Et₂O solution gave dark green crystals. The Cr(IV) complex bearing Cl^– ions, [CrCl(L^Pen)] (3), was also synthesized as the Cr(IV) complex without dinitrogen ligands by the previously reported method [71].

On the other hand, reacting Li₃[L^Cy] with CrCl₃ in THF under N₂ afforded the mononuclear chromium complex [Cr(L^Cy)] (4), which lacks a dinitrogen ligand. Recrystallizing complex 4 from n-hexane solution yielded dark red crystals. Reacting complex 4 with K metal and 18–crown–6 ether in THF under N₂ atmosphere gave the dinitrogen chromium complex containing a K⁺ ion, [CrK(L^Cy)(μ-N₂)(18–crown–6)(THF)] (5(THF)) (Scheme 2). The Cr(IV) complex bearing Cl^– ion, [CrCl(L^Cy)] (6), was also synthesized as the Cr(IV) complex without dinitrogen ligands by the same method as 3 (Scheme 3). Recrystallization of complex 6 from cyclopentylmethyl ether solution yielded dark violet crystals.

2.2. Crystal Structures of Complexes 1, 2, 4, and 5(THF)

Crystal structures of 1 and 2 are shown in Figure 1 together. The crystal parameters and the selected bond lengths and angles of 1 and 2 are listed in Tables S1 and 1, respectively. Complex 1 is a dichromium complex possessing a dinitrogen ligand as the bridging ligand. Each Cr atom forms a trigonal bipyramidal geometry with three amido nitrogen atoms of [L^Pen]^3- located on a trigonal plane and another amino nitrogen of [L^Pen]^3- and dinitrogen molecule at the axial positions. The N–N bond length is 1.185(3) Å, which is significantly elongated compared to free nitrogen molecule (1.10 Å) [20], and is almost identical to those (1.188(4) and 1,185(7) Å) of the previously reported complex bearing a benzyl group as the substituent on the triamidoamine ligand ([L^Bn]^3-), [{Cr(L^Bn)}₂(μ-N₂)] (Figure 2) [71]. The Cr–N_N2 bond length of 1 is 1.8468(15) Å, which is longer than those (1.805(2) and 1.804(4) Å) of the previously reported dichromium complex with benzyl groups. This is thought to be due to the increased steric repulsion resulting from the pentyl group having greater spatial extension than the benzyl group. Therefore, the Cr–N_N2 bond of 1 is considered to possess the properties of a weak multiple bond. Furthermore, the N–N distance is shorter than that (1.226(3) Å) of the dinitrogen-divanadium complex with the same triamidoamine ligand [72]. This is thought to be due to the difference in the back-donating ability of the metal ions to the dinitrogen ligand. That is: When the π-back donation is large, electrons enter the antibonding orbitals of dinitrogen from the metal, weakening the N–N bond and then reducing the ν(N–N) stretching vibration. Comparing dinitrogen-dichromium complex 1 and dinitrogen-divanadium complex [72], the ν(N–N) stretching vibration in the dichromium complex is 1715 cm⁻¹, while in the divanadium complex it is 1412 cm⁻¹, indicating the divanadium complex is more weakened. This indicates that the π-back-donation is larger in the dinitrogen-divanadium complex than in the dinitrogen-dichromium complex. Consequently, the M–N_N₂ distance in the divanadium complex should be shorter than in the dichromium complex. Indeed, the V–N_N₂ bond distance (1.7935(14)) Å is shorter than the Cr–N_N₂ bond (1.8468(15) Å), clearly reflecting the difference in the π-back donating ability of the metal ions.

Complex 2 is a dinuclear Cr complex containing two potassium ions and two dinitrogen molecules. Each Cr atom forms a trigonal bipyramidal structure coordinated with the ligand [L^Pen]^3- and dinitrogen molecule, similar to complex 1. The dinitrogen ligand formed a bridge between the chromium and the two potassium ions, coordinating with one potassium ion in an end-on mode and with the other potassium ion in a side-on mode. The N–N bond length in 2 is 1.172(9) Å, showing a tendency to be shorter than that of 1. This N–N bond length is longer than that (1.1624(19) Å) observed in the previously reported complex [{CrNa(L^Bn)(µ-N₂)(Et₂O)}₂] (Figure 2) [71], which possesses a similar structure with terminal substituents of a benzyl group and a sodium ion, but with a potassium ion substituted instead. This is thought to be due to the large ionic radius of the potassium ion (1.37–1.64 Å) in comparison with sodium ion (0.99–1.39 Å) [73]. The potassium ion sandwiched between the two amido nitrogen atoms of the ligand migrated towards the dinitrogen ligand, resulting in one dinitrogen ligand forming coordination bonds with two potassium ions. The Cr–N_N2 bond length is 1.762(8) Å, indicating that 2 has stronger multiple bond character than 1.

Table 1. Selected Bond Lengths (Å) and Angles (deg) for 1 and 2.

1
Cr1–N2	1.8468(15)			Cr1–N3	1.8965(16)		Cr1–N4		1.8906(18)
Cr1–N5	1.8876(17)			Cr1–N6	2.1110(17)		N2–N2i		1.185(3)
Cr•••Cri	4.8774(7)
N2–Cr1–N3	98.11(7)			N2–Cr1–N4	99.76(7)		N2–Cr1–N5		97.25(7)
N2–Cr1–N6	177.95(7)			N3–Cr1–N4	117.87(7)		N3–Cr1–N5		118.37(7)
N3–Cr1–N6	80.89(7)			N4–Cr1–N5	117.51(7)		N4–Cr1–N6		82.29(7)
N5–Cr1–N6	81.71(7)			N2i–N2–Cr1	177.1(2)
2
Cr1–N3		1.762(8)	Cr1–N5			2.096(7)		Cr1–N6		1.931(7)
Cr1–N7		1.901(8)	Cr1–N8			1.857(8)		N3–N4		1.174(9)
N3–K2i		2.870(8)	N4–K2i			3.101(8)		N6–K2i		2.983(8)
N7–K2i		3.108(7)	N4–K2			2.658(8)		K2–O9		2.722(7)
Cr1•••K2i		3.252(3)	Cr•••K2			5.551(3)
N3–Cr1–N5		179.0(3)	N3–Cr1–N6			96.7(3)		N3–Cr1–N7		96.9(3)
N3–Cr1–N8		97.2(3)	N5–Cr1–N6			82.5(3)		N5–Cr1–N7		83.0(3)
N5–Cr1–N8		83.8(3)	N6–Cr1–N7			115.9(3)		N6–Cr1–N8		121.4(3)
N7–Cr1–N8		118.4(3)	N3–K2i–N4			22.25(18)		N3–K2i–N4i		100.7(2)
N3–K2i–N6		56.3(2)	N3–K2i–N7			54.4(2)		N3–K2i–O9i		167.0(2)
N4–K2i–N4i		78.5(2)	N4–K2i–N6			74.1(2)		N4–K2i–N7		132.5(2)
N4–K2i–O9i		166.4(2)	N4i–K2i–N6			139.9(2)		N4i–K2i–N7		132.5(2)
N4i–K2i–O9i		90.7(2)	N6–K2i–N7			64.4(2)		N6–K2i–O9i		110.7(2)
N7–K2i–O9i		121.2(2)	N4–N3–Cr1			175.6(7)		N3–N4–K2		168.5(7)

Table 1. Selected Bond Lengths (Å) and Angles (deg) for 1 and 2.

1
Cr1–N2	1.8468(15)			Cr1–N3	1.8965(16)		Cr1–N4		1.8906(18)
Cr1–N5	1.8876(17)			Cr1–N6	2.1110(17)		N2–N2i		1.185(3)
Cr•••Cri	4.8774(7)
N2–Cr1–N3	98.11(7)			N2–Cr1–N4	99.76(7)		N2–Cr1–N5		97.25(7)
N2–Cr1–N6	177.95(7)			N3–Cr1–N4	117.87(7)		N3–Cr1–N5		118.37(7)
N3–Cr1–N6	80.89(7)			N4–Cr1–N5	117.51(7)		N4–Cr1–N6		82.29(7)
N5–Cr1–N6	81.71(7)			N2i–N2–Cr1	177.1(2)
2
Cr1–N3		1.762(8)	Cr1–N5			2.096(7)		Cr1–N6		1.931(7)
Cr1–N7		1.901(8)	Cr1–N8			1.857(8)		N3–N4		1.174(9)
N3–K2i		2.870(8)	N4–K2i			3.101(8)		N6–K2i		2.983(8)
N7–K2i		3.108(7)	N4–K2			2.658(8)		K2–O9		2.722(7)
Cr1•••K2i		3.252(3)	Cr•••K2			5.551(3)
N3–Cr1–N5		179.0(3)	N3–Cr1–N6			96.7(3)		N3–Cr1–N7		96.9(3)
N3–Cr1–N8		97.2(3)	N5–Cr1–N6			82.5(3)		N5–Cr1–N7		83.0(3)
N5–Cr1–N8		83.8(3)	N6–Cr1–N7			115.9(3)		N6–Cr1–N8		121.4(3)
N7–Cr1–N8		118.4(3)	N3–K2i–N4			22.25(18)		N3–K2i–N4i		100.7(2)
N3–K2i–N6		56.3(2)	N3–K2i–N7			54.4(2)		N3–K2i–O9i		167.0(2)
N4–K2i–N4i		78.5(2)	N4–K2i–N6			74.1(2)		N4–K2i–N7		132.5(2)
N4–K2i–O9i		166.4(2)	N4i–K2i–N6			139.9(2)		N4i–K2i–N7		132.5(2)
N4i–K2i–O9i		90.7(2)	N6–K2i–N7			64.4(2)		N6–K2i–O9i		110.7(2)
N7–K2i–O9i		121.2(2)	N4–N3–Cr1			175.6(7)		N3–N4–K2		168.5(7)

For complex 4 synthesized using Li₃[L^Cy], accurate results could not be obtained from X-ray crystal structure analysis. However, because the approximate structure was already known, 25% probability ellipsoids are displayed for the carbon, nitrogen, and chromium atoms in the ORTEP display of 4 (Figure 3). The crystal parameters and the selected bond lengths and angles of 4 are listed in Tables S1 and S2, respectively. Complex 4 revealed a trigonal monopyramidal coordination geometry, with the axial position being vacant. The chromium complex with a similar structure has previously been reported by Filippou’s group [74]. Also, the previously reported vanadium complex with the same ligand, [V(L^Cy)] [72], also possessed an axial vacancy, but one cyclohexane ring was positioned axially, thereby preventing the coordination of other molecules. In contrast, the cyclohexane ring in complex 4 is not located axially, resulting in a structure with many gaps, as shown in Figure 4. Consequently, complex 4 is very air sensitive.

Next, reacting complex 4 with metallic potassium and 18-crown-6 ether under N₂ yielded the dinitrogen chromium complex 5(THF). Fortunately, a single crystal of 5(THF) was obtained from THF-n-hexane solution and analyzed by X-ray diffraction. The crystal structure is shown in Figure 5, and the crystal parameters and the selected bond lengths and angles are listed in Tables S1 and 2, respectively. The Cr complex adopts a trigonal bipyramidal structure, forming a trigonal plane with three amido nitrogen atoms of [L^Cy]^3- ligand, with the axial positions occupied by another amine nitrogen of [L^Cy]^3- ligand and a dinitrogen ligand. The opposite side of the dinitrogen ligand is bonded to a potassium ion coordinated by an 18-crown-6 ether ligand from the planar side and a THF molecule from the apical position, resulting in the N₂ ligand forming a bridge between the chromium and potassium atoms, with an N–N bond length of 1.162(3) Å. The N–N bond length is slightly shorter than that (1.167(2) Å) of the previously reported chromium–dinitrogen complex containing a benzyl group, [CrK(L^Bn)(μ-N₂)(18-crown-6)] (Figure 2) [71]. The Cr–N_N2 and K–N_N2 bond lengths of 5(THF) were 1.761(2) and 2.647(2) Å, respectively. These bond lengths were also shorter compared with the Cr–N_N2 (1.7678(15) Å) and K–N_N2 bond lengths (2.7822(16) Å) of the previously reported chromium complex [CrK(L^Bn)(μ-N₂)(18-crown-6)] [71].

Table 2. Selected Bond Lengths (Å) and Angles (deg) for 5(THF).

5(THF)
Cr1–N3	1.761(2)	Cr1–N5	1.912(2)	Cr1–N6	1.911(2)
Cr1–N7	1.907(2)	Cr1–N8	2.125(2)	K2–N4	2.647(2)
K2–O9	2.867(2)	K2–O10	2.775(2)	K2–O11	2.866(2)
K2–O12	2.767(2)	K2–O13	2.852(2)	K2–O14	2.774(2)
K2–O15	2.707(2)	N3–N4	1.162(3)	Cr1•••K2	5.5547(8)
N3–Cr1–N5	97.44(10)	N3–Cr1–N6	97.40(9)	N3–Cr1–N7	96.46(10)
N3–Cr1–N8	179.24(9)	N5–Cr1–N6	118.77(9)	N5–Cr1–N7	119.16(9)
N5–Cr1–N8	82.52(9)	N6–Cr1–N7	117.56(10)	N6–Cr1–N8	83.27(9)
N7–Cr1–N8	82.91(9)	N4–N3–Cr1	179.6(3)	N4–K2–O9	100.44(7)
N4–K2–O10	95.50(7)	N4–K2–O11	101.65(7)	N4–K2–O12	87.16(7)
N4–K2–O13	96.66(7)	N4–K2–O14	91.30(7)	N4–K2–O15	172.33(9)
N3–N4–K2	171.8(2)

Table 2. Selected Bond Lengths (Å) and Angles (deg) for 5(THF).

5(THF)
Cr1–N3	1.761(2)	Cr1–N5	1.912(2)	Cr1–N6	1.911(2)
Cr1–N7	1.907(2)	Cr1–N8	2.125(2)	K2–N4	2.647(2)
K2–O9	2.867(2)	K2–O10	2.775(2)	K2–O11	2.866(2)
K2–O12	2.767(2)	K2–O13	2.852(2)	K2–O14	2.774(2)
K2–O15	2.707(2)	N3–N4	1.162(3)	Cr1•••K2	5.5547(8)
N3–Cr1–N5	97.44(10)	N3–Cr1–N6	97.40(9)	N3–Cr1–N7	96.46(10)
N3–Cr1–N8	179.24(9)	N5–Cr1–N6	118.77(9)	N5–Cr1–N7	119.16(9)
N5–Cr1–N8	82.52(9)	N6–Cr1–N7	117.56(10)	N6–Cr1–N8	83.27(9)
N7–Cr1–N8	82.91(9)	N4–N3–Cr1	179.6(3)	N4–K2–O9	100.44(7)
N4–K2–O10	95.50(7)	N4–K2–O11	101.65(7)	N4–K2–O12	87.16(7)
N4–K2–O13	96.66(7)	N4–K2–O14	91.30(7)	N4–K2–O15	172.33(9)
N3–N4–K2	171.8(2)

2.3. Crystal Structures of Complexes 3 and 6

The crystal structure of Cr^IVCl complex with [L^Pen]^3– or [L^Cy]^3–, 3 or 6, is shown in Figure 6 together, and the crystal parameters and the selected bond lengths and angles are listed in Tables S1 and 3, respectively. Crystal structural data for 3 has been reported previously [71] and the ORTEP views are also shown in Figure 6 of this paper. The coordination structures around the Cr ion in 3 and 6 have both trigonal bipyramidal geometry. The three amine nitrogen atoms of the bulky ligand are arranged in a triangular plane around Cr atom, while another amido nitrogen atom of the ligand and the chlorine atom are positioned axially. The Cr–Cl bond lengths in 3 and 6 are almost the same (2.3145(6) Å for 3, 2.3147(7) Å for 6). However, comparing the Cr–N_amine and Cr–N_amido bond lengths in these complexes, the Cr–N_amine (2.0446(16) Å) and Cr–N_amido (av) (1.866 Å) of 3 are shorter than those (Cr–N_amine 2.088(2) Å, Cr–N_amido (av) 1.8828 Å) in 6. This is thought to be due to the steric repulsion between the terminal substituents of the ligand in 6 being greater than that in 3. Previously, Filippou’s group has reported a similar chromium complex with terminal trimethylsilyl groups on the triamidoamine ligand, however, no structural trend could be found among these three Cr(IV)–Cl complexes [74].

Table 3. Selected Bond Lengths (Å) and Angles (deg) for 3 and 6.

3[a]
Cr1–Cl2	2.3145(6)	Cr1–N3	2.0446(16)	Cr1–N4	1.8661(16)
Cr1–N5	1.8680(17)	Cr1–N6	1.8650(17)
Cl2–Cr1–N3	179.51(5)	Cl2–Cr1–N4	96.55(5)	Cl2–Cr1–N5	97.03(5)
Cl2–Cr1–N6	97.09(6)	N3–Cr1–N4	82.98(7)	N3–Cr1–N5	83.08(7)
N3–Cr1–N6	83.28(7)	N4–Cr1–N5	119.03(8)	N4–Cr1–N6	119.39(7)
N5–Cr1–N6	117.32(8)
6
Cr1–Cl2	2.3147(7)	Cr1–N3	2.088(2)	Cr1–N4	1.883(2)
Cr1–N5	1.8874(19)	Cr1–N6	1.878(2)
Cl2–Cr1–N3	179.50(6)	Cl2–Cr1–N4	96.80(6)	Cl2–Cr1–N5	96.44(6)
Cl2–Cr1–N6	97.20(7)	N3–Cr1–N4	83.05(8)	N3–Cr1–N5	83.23(8)
N3–Cr1–N6	83.29(9)	N4–Cr1–N5	119.49(9)	N4–Cr1–N6	117.96(9)
N5–Cr1–N6	118.39(9)

^[a] The bond lengths and angles of 3 are described with modifications to the original atom numbering scheme reported in the previously paper [71].

Table 3. Selected Bond Lengths (Å) and Angles (deg) for 3 and 6.

3[a]
Cr1–Cl2	2.3145(6)	Cr1–N3	2.0446(16)	Cr1–N4	1.8661(16)
Cr1–N5	1.8680(17)	Cr1–N6	1.8650(17)
Cl2–Cr1–N3	179.51(5)	Cl2–Cr1–N4	96.55(5)	Cl2–Cr1–N5	97.03(5)
Cl2–Cr1–N6	97.09(6)	N3–Cr1–N4	82.98(7)	N3–Cr1–N5	83.08(7)
N3–Cr1–N6	83.28(7)	N4–Cr1–N5	119.03(8)	N4–Cr1–N6	119.39(7)
N5–Cr1–N6	117.32(8)
6
Cr1–Cl2	2.3147(7)	Cr1–N3	2.088(2)	Cr1–N4	1.883(2)
Cr1–N5	1.8874(19)	Cr1–N6	1.878(2)
Cl2–Cr1–N3	179.50(6)	Cl2–Cr1–N4	96.80(6)	Cl2–Cr1–N5	96.44(6)
Cl2–Cr1–N6	97.20(7)	N3–Cr1–N4	83.05(8)	N3–Cr1–N5	83.23(8)
N3–Cr1–N6	83.29(9)	N4–Cr1–N5	119.49(9)	N4–Cr1–N6	117.96(9)
N5–Cr1–N6	118.39(9)

^[a] The bond lengths and angles of 3 are described with modifications to the original atom numbering scheme reported in the previously paper [71].

2.4. Raman and IR Spectra of 1, 2, and 5(THF)

To investigate the reactivity of the dinitrogen molecule bound to Cr atom in complexes 1, 2, and 5(THF), vibrational spectroscopic measurements were carried out on them. Complex 1 was studied by Raman spectroscopy in cyclohexane solution using laser light excited at 532 nm. The ν(¹⁴N–¹⁴N) stretching vibration of complex 1 was detected at 1715 cm⁻¹. When ¹⁵N₂ was used in the reaction instead of ¹⁴N₂, the ν(¹⁵N–¹⁵N) stretching vibration was obtained at 1651 cm⁻¹, exhibiting an ideal isotope shift (Figure S1). This value is significantly smaller than free dinitrogen (2331 cm^-1) [20], clearly indicating that the dinitrogen ligand has been activated, but it is larger than the ν(¹⁴N–¹⁴N) stretching vibration (1412 cm^-1) of the dinitrogen divanadium complex possessing the same triamidoamine ligand [72]. As mentioned in Section 2.1, this is also refelected in the N–N bond length and is thought to be due to the fact that the back-donation from the metal to the dinitrogen ligand is smaller for Cr than V.

Complexes 2 and 5(THF), which possess the dinitrogen ligands bridging between the chromium ion and potassium ion, were investigated by IR spectroscopy (ATR method). The peaks assignable to ν(¹⁴N–¹⁴N) stretching vibration were observed as two bands at 1787 and 1743 cm^-1 for 2 (Figure S2), and as one band at 1824 cm^-1 for 5(THF) (Figure S3). When ¹⁵N-labeled dinitrogen chromium complexes of 2 and 5, [{CrK(L^Pen)(μ-¹⁵N₂)(Et₂O)}₂] (2’) and [CrK(L^Cy)(μ-¹⁵N₂)(18-crown-6)(THF)] (5’(THF)), were employed, these peaks shifted to 1728 and 1687 cm^-1 for 2’ and to 1757 cm^-1 for 5’(THF), respectively. These findings indicate that the stretching vibration frequencies of dinitrogen molecules have shifted to the lower energy side due to the isotope effect, and the presence of two peaks in 2 suggests that dinitrogen exists in two different bridging modes between Cr and K. In the previously reported complex [{CrNa(L^Bn)(μ-N₂)(Et₂O)}₂], which possesses a sodium ion instead of a potassium ion and features benzyl groups at the terminals of the triamidoamine ligand, the ν(¹⁴N–¹⁴N) stretching vibration was observed at 1813 cm^-1 [71]. Therefore, based on the crystal structure of 2 described in Section 2.2 (Figure 1), the two peaks detected at 1787 and 1743 cm^-1 for 2 are considered to correspond to the dinitrogen ligand bridged by a chromium ion and one potassium ion in an end-on mode, and the dinitrogen ligand bridged by a chromium ion and two potassium ions in an end-on and a side-on modes, respectively.

The ν(N–N) stretching vibration (1824 cm^-1) of 5(THF) was observed in a higher energy region than that (1807 cm^-1) of the previously reported complex [CrK(L^Bn)(μ-N₂)(18-crown-6)] [71]. This is consistent with the elongated N–N bond lengths of the respective complexes being 1.162(3) Å for 5(THF) and 1.167(2) Å for [CrK(L^Bn)(μ-N₂)(18-crown-6)], respectively, with complex 5(THF) exhibiting a slightly shorter bond length [71].

2.5. IR Spectra of 2 and 5(THF) in Solution

To further elucidate the structures of 2 and 5(THF) in solution, the IR spectra were measured in toluene and THF. Those of 2 are shown in Figures S4 and S5, respectively, and those of 5(THF) are shown in Figures S6 and S7, respectively. The ν(N–N) stretching vibration peaks of 2 appeared as a single peak at 1790 in toluene and at 1805 cm^-1 in THF, respectively. However, in the solid state, as mentioned above, two peaks (1787 and 1743 cm⁻¹) were observed in the lower energy region compared to the solution state. This finding suggests that the dinitrogen ligand in solution is bridging between the K⁺ ion and the Cr(IV) ion in end-on fashion, although the dinitrogen ligands in the crystal structure of 2 involved in both end-on and side-on fashions.

On the other hand, for complex 5(THF), the ν(N-N) stretching vibration peak was observed as a single peak at 1805 cm^-1 in toluene (Figure S6) and at 1809 cm^-1 in THF (Figure S7), respectively, both of which were observed in lower energy side compared to that (1824 cm^-1) in the solid state. As explained in Section 2.6 below, in toluene or THF solution, the THF ligand is thought to be removed from complex 5(THF), forming a THF-free complex 5.

15. N NMR Spectra of 2 and 5(THF)

To further examine the coordination mode of dinitrogen molecules with Cr atom in complexes 2 and 5(THF) in solution, ¹⁵N NMR spectra were measured in C₆D₆ solution using complexes 2’ and 5’(THF) with ¹⁵N-labeled N₂. The peaks of ¹⁵N-labeled dinitrogen ligands in both complexes were observed at –1.9 and –32 ppm for 2’ (Figure S8), and at 8.8 and –27.5 ppm for 5’(THF) (Figure S9). The peaks in low and high magnetic field regions for 2’ and 5’ (THF) are well consistent with those of the previously reported chromium complexes [{CrK(L^Bn)(μ-N₂)}₄(Et₂O)₂] (1.79 and –71.26 ppm) [71] and [CrK(L^Bn)(μ-N₂)(18-crown-6)] (4.79 and –25.22 ppm) [71], so they were assigned to the N_α atom bonded to Cr on the low magnetic field side of the coordinated N₂, and to the N_β atom on the opposite side on the high magnetic field side, respectively. The two peaks in complex 2’ exhibited both small shoulder peaks on the lower magnetic field side. To investigate the reason, spectra were aquired by varying the number of scans (10,000, 20,000, 30,000, 40,000, and 50,000 scans). It was confirmed that the shoulder peaks gradually changed from a small double peak to a shoulder peak (Figure S10). Considering the IR spectral behavior of 2 described above, these peaks are likely to originate from species in which the dinitrogen ligand coordinates to the potassium ion in a side-on mode. Therefore, these spectra suggest that species in which the dinitrogen ligand coordinates to the potassium ion in the side-on mode are thermally unstable in solution and gradually transform into the end-on mode species.

By the way, in the spectrum of 5’(THF) [CrK(L^Cy)(μ-¹⁵N₂)(18–crown–6)(THF)], small peaks were observed at 5, –24, and –72 ppm. The peak at –72 ppm is thought to be due to the amine N atoms of the ligand [L^Cy]^3-, because it exhibits the largest peak area among the three peaks. A single crystal was accidentally precipitated from the NMR sample tube used for the ¹⁵N NMR spectral measurement of 5’(THF), so its crystal structure analysis was carried out. As a result, the THF ligand coordinated to the potassium ion in 5’(THF) was removed, forming the complex [CrK(L^Cy)(µ-¹⁵N₂)(18–crown–6)] (5’). The N–N bond length of the dinitrogen ligand in 5’ was slightly elongated from 1.162 (3) Å to 1.170 (4) Å as compared with that in 5’(THF) (Figure S11, Table 2, and Table S3). Considering that most of THF ligands in 5(THF) are removed even in THF solution, the fact that stretching vibrations in the low-energy region were observed in solution, as mentioned in Section 2.5 above, is consistent with the elongation of the N–N bond distance in the crystal structure. Therefore, the strong peaks at 8.8 and –27.5 ppm in ¹⁵N NMR spectrum (Figure S9) of 5’(THF) are thought to originate from the N₂ ligand of complex 5’ without bound THF. Furthermore, the minor peaks at 5 ppm and –24 ppm in Figure S9 are considered to originate from the N₂ ligand of a small amount of 5’(THF) in which THF is coordinated to the potassium ion.

2.6.1. H NMR Spectra of 1 – 6 

The ¹H NMR spectrum of 1 in C₆D₆ exhibited paramagnetic shifts with broadened peaks at 12, 5, 0.8, and –38 ppm (Figure S14). The magnetic moment (μ_eff) of 1 in solution was measured by the Evans method and estimated to be μ_eff = 2.87 μ_B at room temperature. This result suggests that the spin state of 1 is S = 1, similar to that of [{Cr(L^Bn)}₂(μ-N₂)] reported previously [71]. On the other hand, the peaks in ¹H NMR spectra of 2 (Figure S15) and 5(THF) (Figure S18) in C₆D₆ were observed as simple spectra in the diamagnetic region, despite the complexity of their crystal structures. These findings clearly indicate that complexes 2 and 5(THF) exhibit high symmetry, specifically C₃ symmetry, in solution. As mentioned above, in C₆D₆, most of the THF ligands in 5(THF) were released, likely forming [CrK(L^Cy)(μ-¹⁴N₂)(18-crown-6)] (5). And these are also consistent with those for the previously reported complexes, [{CrNa(L^Bn)(μ-N₂)}₂] and [CrK(L^Bn)(μ-N₂)(18-crown-6)], suggesting a very strong antiferromagnetic interaction between the Cr ion and the N₂ ligand with S = 0 spin state [71]. Furthermore, the ¹H peaks of ligands H₃L^Pen and H₃L^Cy were also observed in the ¹H NMR spectra of 2 and 5(THF) (Figures S15 and S18). This indicates that these complexes are decomposed in solution at room temperature, even under anaerobic conditions.

The ¹H NMR spectra of 3 and 6 in C₆D₆ showed paramagnetic shifts with broadened peaks at 20, 12, 0.4, –36, and –69 ppm for 3 (Figure S16), and at 13, 3.6, 3.1, 1.7, 1.5, 1.4, 1.1 and –41 ppm for 6 (Figure S19). The magnetic moments (μ_eff) of these complexes in solution were estimated to be 2.84 and 2.96, respectively. Therefore, these complexes were shown to be high-spin d² metal complexes [71]. The ¹H NMR spectrum of Cr(III) complex 4 exhibited paramagnetism, with broadened peaks at 3 – 0, –18, –35, and –66 ppm (Figure S17). The magnetic moment of 4 calculated by the Evans method was μ_eff = 3.82 μ_B at room temperature, corresponding to a spin state of S = 3/2. This value is very close to that reported previously by Filippou’s group for a Cr(III) complex bearing a trimethylsilyl group as the terminal substituent of a triamidoamine ligand which has a structure similar to 4 (μ_eff = 3.8 μ_B) [74].

2.7. Comparison of Physico-chemical Properties (X-ray, Raman, IR, NMR) of 1, 2 and 5(THF) with Previously Reported Chromium Dinitrogen Complexes

The physicochemical properties (Cr–N_N2/Å, N–N/Å, ν(N–N)/cm^-1, ¹⁵N NMR/ppm) of dinitrogen ligands coordinated to Cr complexes obtained in this study were compared with those of previously reported Cr complexes [71] to examine the relationships among them. The results are summarized in Table 3. We expected to observe the trans influence between the Cr–N_N2 bond length and the N–N bond length, but no such an influence was observed. Rather, the longer the Cr–N_N2 bond length between two Cr atoms, the longer the N–N bond length tended to be as well. Conversely, in Cr complexes with nitrogen ligands sandwiched between Cr and K, the Cr–N_N2 bond length tended to be shorter, and the N–N bond length also tended to be shorter. In contrast, a reasonably consistent trend was observed between the N–N bond lengths and the ν(N–N) stretching vibrations. Especially, as the N–N bond length increased, the ν(N–N) stretching vibration decreased, and as the N–N bond length decreased, the ν(N–N) stretching vibration increased. This relationship has been reported previously [18] and is natural when considering the relationship between force and spring in vibrational phenomena.

Next, we considered that observing the NMR spectra of dinitrogen ligands bound to Cr would allow us to infer the electronic state of the bound N atom of N₂ ligands, and thus compared the ¹⁵N peaks of the dinitrogen ligands in the chromium dinitrogen complexes. While the ¹⁵N NMR spectrum of the N₂ molecule sandwiched between two Cr atoms was not measured, in the dinitrogen ligands sandwiched between Cr and K, a clear difference in chemical shifts was observed between N_α and N_β. This is because electrons flow from Cr to the N₂ molecule via a π-back bonding and the N_β atom, reduced by K, carries a negative charge. The nitrogen ligand forms an ionic bond with K⁺, attracting electrons to N_β and increasing its negative charge, leading to the detection of N_b in high magnetic fields. The N_α signal observed on the low magnetic field side generally appears between approximately 9 ~ −2 ppm, though with some variation. In contrast, the N_β signal is observed on the high magnetic field side around –30 ppm. However, in the previously reported complex [{CrK(L^Bn)(μ-N₂)}₄(Et₂O)₂], it was observed at −71 ppm, a significantly higher magnetic field side. This is likely due to a substantial accumulation of electron density on the N_β in this complex.

2.8. Reactions of Chromium Complexes with K[C₁₀H₈] and HOTf Under N₂

The reactions of complexes 1 – 6 with K[C₁₀H₈] and HOTf in THF were carried out under N₂ atmosphere, yielding hydrazine and a small amount of ammonia in all cases. The quantifications of hydrazine and ammonia were carried out using the ¹H NMR and p-dimethylaminobenzaldehyde methods, respectively. The quantification results and the details are shown in Table 4 and Table S4, respectively. Table 4 presents also the results from the previously reported nitrogen fixation experiments using Cr(IV) complexes with [L^Bn]^3- ligand for comparison. Although the yields of ammonia and hydrazine produced in this study were both low in all complexes, hydrazine was formed in greater quantities than ammonia. Here, two interesting findings can be noticed from Table 4. They are described below.

One is that there is a significant difference in the yield ratio of ammonia to hydrazine between the current Cr complexes containing [L^Pen]^3- or [L^Cy]^3- and the previous Cr complexes containing [L^Bn]^3-. That is, the yield of ammonia was significantly higher in the previous study than in this one [71]. The major difference between these two systems is that the carbon at the substituent-binding site of triamidoamine ligands [L^Pen]^3- and [L^Cy]^3- is a tertiary carbon, whereas the carbon in ligand [L^Bn]^3- is a secondary carbon. The former is densely packed, while the latter is not so much. Here, let’s trace the dinitrogen reduction reaction catalyzed by Cr complexes. First, the Cr(III) complexes bind with dinitrogen molecules to form chromium(IV) dinitrogen complexes. Next, under N₂ atmosphere, when the dinitrogen molecule bound to chromium atom is reduced by K[C₁₀H₈], an intermediate is formed in which the generated K⁺ ions interact with the dinitrogen ligand. Subsequently, protons are supplied to this intermediate, yielding ammonia and/or hydrazine. That is, if the reduction of the coordinated dinitrogen ligands is complete, ammonia is formed, however, if it is incomplete, hydrazine is produced. At this time, potassium, which binds to the coordinated dinitrogen in side-on mode, is considered more effective in cleaving the N-N bond. Consequently, ligands that promote the formation of more spatially sparse chromium complexes are considered advantageous for ammonia production. The Cr complexes coordinated with [L^Pen]^3- or [L^Cy]^3- result in more densely packed dinitrogen chromium complexes than those with [L^Bn]^3-. This likely prevented the supply of protons and potassium ions after dinitrogen reduction, thereby preventing ammonia formation.

Another is that complex 5(THF) exhibited the highest yield of hydrazine among complexes 1 – 6. Complex 4 also produced hydrazine in a reasonably large yield, though not as high as 5(THF). This is understandable considering that complex 4 is a precursor into 5(THF). The high yield of hydrazine in complexes 4 and 5(THF) is plausible base on the preceding explanation. Specifically, the Cr complex bearing the [L^Cy]^3- ligand, which is sterically bulkier than [L^Pen]^3-, adopts a highly hindered structure. This prevents the formation of a dinuclear Cr complex, resulting instead in the mononuclear complex 4. Due to this highly hindered structure, even when dinitrogen binds to chromium ions, it forms structure such as 5(THF) without forming dinuclear chromium complexes, where dinitrogen bridges between Cr and K⁺. Furthermore, the sterically hindered structure of complex 5 prevents K⁺ ions from interacting sufficiently with the dinitrogen from the side. As a result, the dinitrogen cannot be sufficiently activated, leading to hydrazine production. The previously reported Cr complex containing [L^Bn]^3- produced ammonia as the main product, despite having a structure similar to 5(THF). This is attributed to its spatially more open structure.

To clarify whether the above interpretations are correct, we tracked the reaction intermediates formed when 1, 3, 4, and 6 reacted with a reducing agent in THF and C₆D₆ solutions using IR and ¹H NMR spectroscopy (Figures S20–S27). Note that K[C₁₀H₈] is unsuitable for IR spectroscopy in this study, therefore K metal was used as the reducing agent. First, when complexes 1 and 4, bearing [L^Pen]^3– and [L^Cy]^3– ligands respectively, were reacted with K metal under N₂ atmosphere, new IR active peaks appeared at 1805, 1746, and 1582 cm⁻¹ for complex 1 (Figure S20) and at 1801 and 1576 cm⁻¹ for complex 4 (Figure S22), respectively. The peak at 1805 cm⁻¹ detected in the reaction with 1 corresponded to the ν(¹⁴N–1⁴N) stretching vibration peak (1805 cm⁻¹) of complex 2 measured in THF (spectra B and C in Figure S20). The IR peak at 1801 cm⁻¹ detected in the reaction with complex 4 (ν(¹⁵N-¹⁵N): 1743 cm⁻¹) was close to the peak of 5(THF) measured in THF (1809 cm⁻¹) (spectra B and D in Figure S22). The minor peak at 1746 cm⁻¹ (spectrum B in Figure S20) detected in the reaction with complex 1 closely resembled the IR peak (1743 cm⁻¹) (spectrum A in Figure S2) of complex 2 measured in the solid state. These findings suggest that complexes 1 and 4 react with K metal under N₂ atmosphere to adopt structures similar to those seen in the crystal structures of complexes 2 (Figure 1) and 5(THF) (Figure 5). This means that the dinitrogen ligand coordinated to the chromium ion interacts with the potassium ion via a side-on mode and/or an end-on mode, similar to their crystal structures.

Regarding the broad peaks appearing around 1582 and 1576 cm⁻¹ after the reduction reactions of complexes 1 (Figure S20) and 4 (Figure S22), respectively, the small peaks observed at approximately the same positions in the solid-state IR spectra of complexes 2 and 5(THF), at 1576 and 1593 cm⁻¹ respectively, did not shift even when ¹⁵N-labeled N₂ was used (Figure S2 and S3), suggesting that these peaks are unrelated to the dinitrogen ligand.

Furthermore, when reacting 1 or 4 with metal K, the ¹H NMR peaks for complex 1, except for the Et₂O peak, matched those of 2, the ¹H NMR peaks for 4 could not be assigned due to its paramagnetism (Figures S21 and S23). These results indicate that complex 4, possessing sterically hindered groups, generates only compounds that the dinitrogen binds to the chromium ion and potassium ion in a bridging fashion during the reaction of the complex 4 with K metal under N₂ atmosphere. On the other hand, complex 1, which possesses a low-stereochemically barrier ligand [L^Pen]^3–, appears to generate not only such compounds but also a small amount of compounds where the dinitrogen ligand bridges a chromium ion and two potassium ions in end-on and side-on modes.

In addition, IR spectra of the solutions obtained by the reactions of Cr(IV) complexes 3 and 6 with K metal in THF under N₂ atmosphere revealed new peaks at 1806 and 1801 cm⁻¹, respectively, which are considered to originate from the ν(N–N) stretching vibrations (Figures S24 and S26). The former peak at 1806 cm⁻¹ obtained in the reaction of 3 closely matched that of complex 2 (1805 cm⁻¹) measured in THF (Figure S5 and spectrum D in Figure 24) and also that (1805 cm⁻¹) obtained after reacting complex 1 with K metal in THF (Figure S20 and spectrum C in Figure 24). The latter peak at 1801 cm⁻¹ obtained in the reaction of 6 matched the peak of the species obtained after reacting complex 4 with K metal in THF (spectra B and C in Figure S26).

Furthermore, the ¹H NMR spectra of complex solutions of 3 and 6 reacting with metallic potassium under N₂ atmosphere were investigated. The ¹H NMR spectrum obtained after the reaction of 3 with K metal (Figure S25) was consistent with the spectrum of 2 in THF solution and also agreed well with the spectrum obtained after the reaction of 1 with K metal (Figure S21). The latter spectrum (Figure S27) after the reaction of 6 with K metal resembled that of 5(THF) in C₆D₆ (Figure S18), although it could not be identified. Therefore, it is considered that complexes 3 and 6 also formed species similar to the structures of 2 and 5(THF), respectively, when reacted with K metal under N₂ atmosphere. The findings regarding the formation behavior of intermediates in the reduction reaction of Cr complexes in solution well reflect the two aforementioned explanations concerning nitrogen fixation.

3. Materials and Methods

3.1. General Procedures

All manipulations were carried out under an inert atmosphere (N₂ or Ar) using either a vacuum/gas manifold or an MBraun MB 150-G glovebox. Reagents and solvents employed were commercially available. All anhydrous solvents were purchased from Kanto Chemical Co., Inc. The ligands, tris(2-(3-pentylamino)ethyl)amine (H₃L^Pen) and tris(2-dicyclohexylmethylaminoethyl)amine (H₃L^Cy2), were synthesized according to the literature methods [72]. Complex 3, [CrCl(L^Pen)], was synthesized according to the literature method [71].

3.2. Physical Measurements

¹H and ¹³C NMR spectra were recorded on a JEOL JNM-ECA500 FT NMR spectrometer (500 MHz (¹H), 125.77 MHz (¹³C)), and ¹⁵N NMR spectra were recorded on a JNM-ECA600 FT NMR spectrometer (60.815 MHz (¹⁵N)). All NMR spectra were acquired in C₆D₆ or DMSO-d₆ at 298 K. ¹H and ¹³C chemical shifts were referenced using residual protonated solvent resonances (C₆D₆: 7.16 ppm (¹H) and 128.06 ppm (¹³C), DMSO-d₆: 2.50 ppm (¹H)). ¹⁵N chemical shifts were externally referenced to HCONH₂ (–266.712 ppm (¹⁵N)). FT IR spectra were recorded on an Agilent Cary 630 FT IR spectrophotometer. The IR spectra of the crystals were recorded using the ATR method, while those for the solution were recorded by the transmission method using a CaF₂ cell. Elemental analyses were recorded on a Perkin Elmer CHN–900 elemental analyzer. The effective magnetic moments (μ_eff) of solution samples of 1, 4 and 6 were calculated from solution magnetic susceptibility data, which were determined based on the chemical shift of the methyl proton in toluene (Evans method) [75,76]. Resonance Raman spectroscopy was performed using a JASCO NRS–5500 spectrometer with 532 nm laser as excitation source, and the Raman shifts were calibrated using a Si plate (520 cm^-1).

3.3. X-ray Crystallography Procedures

The X-ray diffraction data for complexes 1, 2, 4, 5(THF), 5’ and 6 were collected on a Rigaku R-AXIS RAPID II diffractometer using multi-layer mirror-monochromatized CuKα (λ = 1.54178 Å) radiation. Crystallographic data and experimental details are summarized in Table S1. The calculations were performed with the Olex2 software package [77]. All structures were solved by intrinsic phasing using the ShelXT program [78], and the remaining non-hydrogen atoms were located from subsequent difference Fourier maps. The structures were refined on F² using full-matrix least-squares minimization with ShelXL [79]. All non-hydrogen atoms were refined anisotropically, unless otherwise stated. Hydrogen atoms were placed at idealized positions and refined using a riding model, unless otherwise stated. Complex 5’ was obtained from the sample vial during the ¹⁵N-NMR measurement of the sample obtained under ¹⁵N₂ atmosphere. Therefore, complex 5’ contains ¹⁵N₂, but it should be noted that the structural refinement in its crystal structure analysis was performed assuming these atoms as ¹⁴N atoms (Tables S1, S3, Figure S7). CCDC-2502967 (1), 2502968 (2), 2218004 (3), 2502969 (4), 2502970 (5(THF)), 2504580 (5’), 2502971 (6) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.acm.ac.jk/data_request/cif, or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

3.4. Synthesis of [{Cr(L^pen)}₂(μ-¹⁴N₂)] (1)

A 20 mL Schlenk flask was charged with H₃L^Pen (1.0 g, 2.8 mmol) and Et₂O (4 mL) and cooled to –60 ˚C under ¹⁴N₂ atmosphere, to which n-butyllithium (3.2 mL, 8.4 mmol, 2.6 M in hexane) was added via syringe. After 15 min, the reaction mixture was slowly warmed to 25 ˚C and was stirred for 1 h at room temperature. The reaction mixture was then cooled to –60 ˚C, and the resulting suspension was transferred via cannula to another Schlenk flask containing CrCl₃ (0.44 g, 2.8 mmol). The reaction mixture was again slowly warmed to 25 ˚C and was stirred overnight. The solvent was removed in vacuo, and the residue was extracted with Et₂O (6 mL). The extract was filtered through Celite. The Et₂O extract was transferred to a 20 mL Schlenk flask and stored in a freezer at –35 ˚C. [{Cr(L^pen)}₂(μ-¹⁴N₂)] was obtained as dark red crystals, which were washed with acetone and dried in vacuo (yield 0.56 g, 48%). ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) 12.14, 5.00, 0.79, –38.3. FT IR (ATR, cm^–1): 2956, 2924, 2898, 2870, 2851, 2822, 2797, 1446, 1370, 1349, 1340, 1329, 1277, 1260, 1237, 1208, 1163, 1148, 1128, 1111, 1085, 1047, 1025, 1016, 997, 954, 904, 895, 861, 835, 814, 800, 755, 744, 652, 613, 598, 561, 533, 524, 477. Anal. Calcd. for C₄₂H₉₀Cr₂N₁₀•1.25H₂O: C, 58.53; H, 10.82; N, 16.26. Found: C, 58.51; H, 10.81; N, 16.17. μ_eff = 2.87 μ_B (Evans method).

3.5. Synthesis of [{Cr(L^pen)}₂(μ-¹⁵N₂)] (1’)

¹⁵N-labeled dinitrogen complex 1 (1’) was prepared following the same procedure described for complex 1, using ¹⁵N₂ instead of ¹⁴N₂. Complex 1’ was obtained as dark red crystals (yield 0.37 g, 31%). FT IR (ATR, cm^–1): 641 (ν(Cr–¹⁵N)).

3.6. Synthesis of [{CrK(L^pen)(μ-¹⁴N₂)(Et₂O)}₂] (2)

A 10 mL vial was charged with complex 1 (500 mg, 0.60 mmol) and Et₂O (2.0 mL) under ¹⁴N₂. Then, potassium metal (235 mg, 6.0 mmol, 10 equiv.) was added, and the mixture was stirred vigorously at room temperature overnight. The reaction mixture was filtered through Celite, and the filtrate was stored in the freezer at –30 ˚C to afford dark green crystals. The crystals were collected, washed with hexane, and dried in vacuo (yield 473 mg, 72%). This compound decomposes quickly upon exposure to air, making it impossible to measure elemental analysis. ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) = 1.12 (t, 12H, CH₃–_Et2O), 1.26 (t, 36H, CHCH₂–CH₃), 1.65 (t, 12H, –CH₂–_tren), 1.87 (m, CHCH₂–CH₃), 1.93 (m, 12H, CHCH₂–CH₃), 3.27 (t, 8H, –CH₂–_Et2O), 3.34 (t, 12H, –CH₂–_tren), 4.31 (m, 6H, –CHCH₂CH₃). ¹³C{¹H} NMR (125.77 MHz, C₆D₆, 298 K): δ(ppm) = 68.18, 51.93, 51.30, 29.36, 13.39. FT-IR (ATR, cm^–1): 2956, 2922, 2848, 2797, 2654, 1787, 1743, 1576, 1457, 1448, 1442, 1372, 1340, 1331, 1314, 1299, 1280, 1260, 1237, 1206, 1148, 1128, 1100, 1081, 1044, 1025, 995, 947, 908, 894, 863, 829, 796, 755, 595, 559, 537, 522, 479, 444.

3.7. Synthesis of [{CrK(L^pen)(μ-¹⁵N₂)(Et₂O)}₂] (2’)

¹⁵N-labeled dinitrogen complex 2 (2’) was prepared following the same procedure described for complex 2, using ¹⁵N₂ instead of ¹⁴N₂. Complex 2’ was obtained as dark green crystals (yield 427 mg, 65%). ¹⁵N NMR (60.815 MHz, C₆D₆, 298 K): δ(ppm) = –1.9 (N_α), –32 (N_β). FT IR (ATR, cm^–1): 1728, 1687 (ν(¹⁵N–¹⁵N)).

3.8. Synthesis of [Cr(L^Cy)] (4)

Complex 4 was prepared following the same procedure described for complex 1, using H₃L^Cy instead of H₃L^pen. Additionally, THF was used as the reaction solvent, and hexane was used for recrystallization, in place of Et₂O (yield 301 mg, 28%). ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) = 2.8, 2.6, 2.3, 2.1, 1.9, 1.8, 1.7, 1.5, 1.3, 0.9, 0.3, –18, –35, –66. FT IR (ATR, cm^–1): 2954, 2919, 2874, 2848, 2796, 1603, 1528, 1446, 1377, 1347, 1325, 1305, 1280, 1258, 1234, 1208, 1144, 1131, 1096, 1072, 1031, 999, 967, 952, 891, 850, 833, 788, 781, 764, 725, 708, 691, 621, 602, 567, 485, 449. Anal. Calcd. for C₄₅H₈₁CrN₄•0.5(H₂O): C, 73.12; H, 11.18; N, 7.58. Found: C, 73.33; H, 11.47; N, 7.51. μ_eff = 3.82 μ_B (Evans method).

3.9. Synthesis of [CrK(L^Cy)(μ-¹⁴N₂)(18-crown-6)(THF)] (5(THF))

A 15 mL vial was charged with complex 4 (0.20 g, 0.27 mmol) and THF (4 mL) under ¹⁴N₂ atmosphere. Potassium metal (0.11 g, 2.7 mmol, 10 equiv.) was added, and the resulting mixture was stirred at R.T. for overnight. The mixture solution was filtered through Celite. A solution of 18-crown-6-ether (79.3 mg, 0.30 mmol) in THF (1 mL) was added to the filtrate, and the mixture was stirred at R.T. for 1 min. Hexane (6 mL) was slowly added to the reaction mixture. The solution was allowed to stand at –35 ˚C to give complex 5 as green crystals (172 mg, 51%). This compound decomposes quickly on exposure to air, making it impossible to measure elemental analysis. ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) = 4.58 (br, 3H, N–CH–Cy), 3.62 (br, 6H, –CH₂–_Cy), 3.58 (t, 4H, O–CH₂–_THF), 3.53 (br, 12H, –CH₂–_{18-crown-6-ether}), 3.20 (br, 12H, –CH₂–_{18-crown-6-ether}), 2.38 (t, 12H, –CH₂–_tren), 2.10 – 1.10 (m, 66H, –CH₂–_Cy, –CH₂–_n-hexane),1.42 (m, 4H, –CH₂–_THF). FT IR (ATR, cm^–1): 2911, 2844, 2745, 1824, 1465, 1444, 1364, 1349, 1334, 1327, 1284, 1256, 1249, 1236, 1219, 1193, 1178, 1141, 1131, 1113, 1105, 1068, 1040, 1016, 997, 962, 909, 891, 874, 839, 827, 800, 786, 775, 753, 727, 704, 695, 626, 615, 596, 578, 550, 533, 488, 447, 440, 432, 440, 432, 432.

3.10. Synthesis of [CrK(L^Cy)(μ-¹⁵N₂)(18-crown-6)(THF)] (5’(THF))

¹⁵N-labeled dinitrogen complex 5(THF) (5’(THF)) was prepared following the same procedure described for complex 5, using ¹⁵N₂ instead of ¹⁴N₂. Complex 5’(THF) was obtained as green crystals (yield 74.2 mg, 22%). ¹⁵N NMR (60.815 MHz, C₆D₆, 298 K): δ(ppm) = 8.8 (N_α), –27.5 (N_β). FT IR (ATR, cm^–1): 1757 (ν(¹⁵N–¹⁵N)).

3.11. Synthesis of [CrK(L^Cy)(μ-¹⁴N₂)(18-crown-6)] (5)

Complex 5 was obtained as a green, THF-free crystal by leaving a C₆D₆ solution of 5(THF) in a NMR sample tube at room temperature for several days (yield 10%). ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) = 4.58 (br, 3H, Cy–CH–Cy), 3.62 (br, 6H, –CH₂–_tren), 3.53 (br, 12H, –CH₂–_Cy), 3.21 (br, 24H, –CH₂–_{18-crown-6-ether}), 2.39 (br, 18H, –CH₂–_Cy), 2.10 – 1.10 (m, 66H, –CH₂–_Cy, –CH₂–_tren, –CH₂–_n-hexane). FT IR (ATR, cm^–1): 1813 (ν(¹⁴N–¹⁴N)).

3.12. Synthesis of [CrK(L^Cy)(μ-¹⁵N₂)(18-crown-6)] (5’)

¹⁵N-labeled dinitrogen complex 5 (5’) was obtained as a green crystal by leaving the C₆D₆ solution of 5’ in a NMR sample tube at room temperature for several days (yield 9.5%). ¹⁵N NMR (60.815 MHz, C₆D₆, 298 K): δ (ppm) = 8.8 (N_α), –27.5 (N_β). FT IR (ATR, cm^–1): 1753 (ν(¹⁵N–¹⁵N)).

3.13. Synthesis of [CrCl(L^Cy)] (6)

Complex 6 was synthesized by the literature method using [L^Cy]^3- instead of [(Me₃SiNCH₂CH₂)₃N]^3- and recrystallized from cyclopentylmethylether solution of 6 in cyclopentylmethylether (yield 39%) [74]. ¹H NMR (500 MHz, C₆D₆, 298 K): δ(ppm) = 13, 3.6, 3.1, 1.7, 1.5, 1.4, 1.1, -41.0. FT IR (ATR, cm^–1): 2915, 2846, 2818, 1608, 1446, 1351, 1325, 1293, 1280, 1258, 1234, 1221, 1210, 1193, 1183, 1142, 1129, 1088, 1073, 1027, 1016, 967, 952, 936, 921, 893, 865, 852, 833, 800, 786, 755, 704, 626, 600, 580, 565, 546, 518, 483. Anal. Calcd. for C₄₅H₈₁ClCrN₄•1.25(C₆H₁₂O): C, 70.78; H, 10.86; N, 6.29. Found: C, 70.81; H, 11.13; N, 6.37. μ_eff = 2.96 μ_B (Evans method).

3.14. Reactions of 1, 2, 3, 4, 5(THF), and 6 with K[C₁₀H₈] and HOTf

A 5 mL THF solution of the reducing agent (K[C₁₀H₈]) freshly prepared from metallic potassium (5.6 mg, 0.14 mmol, 12 eq.) and naphthalene (18 mg, 0.14 mmol, 12 equiv.) in a 20 mL Schlenk flask was added to a 5 mL THF solution of each chromium complex (5.95 × 10^-2 mmol) at –78 ˚C, and the mixture was stirred for 1 h under N₂. HOTf (trifluoromethanesulfonic acid, 42 mg, 0.28 mmol) was added to reaction mixture of chromium complex under vigorous stirring, and the resultant solution was slowly warmed to 25 ˚C. After the solution was stirring for 1 h at room temperature, the solvents were removed under reduced pressure to give a white solid containing ammonium and hydrazinium salts. The residue in the Schlenk tube was washed with diethyl ether and then extracted with H₂O (5 mL). The aqueous extract was evaporated, and the residue was analyzed using ¹H NMR spectroscopy (for NH₃) or p-dimethylaminobenzaldehyde method (for N₂H₄).

3.15. NH₃ Quantification Procedure

Ammonium salts were quantified using ¹H NMR spectroscopy. The quantification of NH₄⁺ was carried out using the method reported by Ashley and coworkers [80]. The 14NH₄⁺ resonance was integrated using the vinyl proton of 2,5-dimethylfuran in DMSO-d₆ sealed in the capillary as the reference (d: 5.83, s, 2H). This quantification was calibrated using a standard 2.6×10^-2 M solution of NH₄⁺ in DMSO-d₆ (Figure S26 and Table S4).

3.16. N₂H₄ Quantification Procedure

The quantification of N₂H₆²⁺ was carried out using the method reported by Ashley and coworkers [80]. The residues were dissolved in H₂O (10 mL), and the resulting solution was diluted tenfold with H₂O. Aliquots of the diluted solution were analyzed for N₂H₄ via UV-vis spectroscopy using a standard spectrophotometric method. This involved reacting the sample with an acidic p-dimethylaminobenzaldehyde solution and generating a yellow azine dye with a characteristic absorption maximum (λ_max) at 458 nm. The N₂H₄ content in the aliquot was quantified by comparison to the calibration curve (Figure S27).

4. Conclusions

In this study, six chromium complexes with the triamidoamine derivative ligand bearing bulky substituents at their terminal positions, [L^Pen]^3- or [L^Cy]^3-, were synthesized and structurally characterized. Using these chromium complexes, the effect of the ligand substituents on the nitrogen fixation reactivity was investigated. When employing chromium complex bearing slightly less bulky ligand [L^Pen]^3–, the dinitrogen dichromium complex, Cr(IV)-N₂-Cr(IV) (1), was prepared under N₂ atmosphere. And further reduction of this complex with reductant K metal under N₂ atmosphere yielded the dinitrogen tetranuclear Cr₂K₂ complex, Cr(K)-(μ-N₂)₂-KCr (2). On the other hand, when the chromium complexes bearing more bulky ligand [L^Cy]^3– was employed, the steric hindrance prevented the formation of dichromium dinitrogen complexes, unlike that obtained with chromium complexes bearing the [L^Pen]^3– ligand. Instead, an axially vacant Cr(III) complex 4 was obtained. However, interestingly, when complex 4 was reduced with K metal in the presence of 18-crown-6 ether under N₂ atmosphere, a dinuclear Cr(IV)-N₂-K(18-crown-6 ether)(THF) complex (5(THF)) was obtained. This formation is attributable to the sterically smaller ligand of the 18-crown 6-ether. The formations of dinitrogen chromium complexes 1, 2 and 5(THF) were also confirmed by Raman and IR spectroscopy. The Cr(IV)Cl complexes 3 and 6, each possessing [L^Pen]^3– and [L^Cy]^3–, respectively, and lacking N₂, were also synthesized and structurally characterized to study the nitrogen fixation reactivity.

For chromium complexes 2 and 5(THF) with ¹⁵N-labeled N₂ (2’ and 5’(THF)), ¹⁵N NMR and solution IR spectroscopic measurements were performed, revealing that the potassium ions, which coordinated in both side-on and end-on modes seen in the crystal of 2, gradually transitioned to coordination exclusively in the end-on mode in solution. Furthermore, the results obtained from the ¹⁵N NMR spectrum of 5’(THF) revealed that the THF ligand is removed from complex 5(THF) in solution, leading to the formation of complex 5. Using single crystals accidentally obtained from the NMR sample tube, X-ray structural analysis was performed to yield complex 5 without THF, confirming the structure as predicted in solution.

When complexes 1 and 3 bearing the [L^Pen]³⁻ ligand reacted with potassium in solution under N₂ atmosphere, substances exhibiting IR and ¹H NMR spectra similar to those of complex 2 were obtained. Furthermore, the spectra obtained when complexes 4 and 6, bearing the [L^Cy]^3– ligand, reacted with metallic K under N₂ atmosphere were also similar to those of 5(THF).

Finally, nitrogen fixation reactivity was studied using complexes 1, 2, 3, 4, 5(THF), and 6 under N₂, employing K[C₁₀H₈] as a reductant and HOTf as a proton source. The results showed that hydrazine and small amounts of ammonia were obtained in all cases. However, unlike the nitrogen fixation reaction previously reported using Cr complexes with [L^Bn]³⁻ ligand [71], the ammonia yield decreased in all cases compared to the previous study [71]. And the Cr complexes 4 and 5, which possess the bulky ligand [L^Cy]³⁻, generated significantly more hydrazine than the others. These results are likely due to the fact that while the [L^Pen]³⁻ and [L^Cy]³⁻ ligands used in this study possess tertiary carbons, [L^Bn]³⁻ has a secondary carbon. This is thought to arise from the steric hinderance of the [L^Pen]³⁻ and [L^Cy]³⁻ ligands inhibiting the supply of K⁺ and protons. The findings obtained in this study indicate that when Cr complexes of triamidoamine ligands bearing bulky substituents react with N₂ in the presence of K, the resulting dinitrogen chromium complexes are influenced by K bound via ligand-dependent interaction modes. The results obtained here will provide important insights into dinitrogen activation and nitrogen fixation using chromium complexes, which have been scarcely reported to date.