Homoleptic nickel(II) and cobalt(II) complexes of N,N′-dialkyl-β-dialdiminato ligands as potential chemical vapor deposition precursors

Abstract

A series of volatile cobalt(II) and nickel(II) N,N′-dialkyl-bis(β-dialdiminato) complexes of stoichiometry M[RN(CH)₃NR]₂ (where R = methyl, ethyl, iso-propyl, or tert-butyl) have been prepared and characterized. NMR, IR, and UV-vis spectra are reported, along with their magnetic properties and crystal structures. IR and crystallographic data support the conclusion that all these complexes contain divalent metal ions and closed-shell monoanionic ligands. NMR studies indicate that some spin density is delocalized directly into the π-orbitals of the β-dialdiminate ligand, as seen in other N,N′-diaryl and β-diketiminato compounds of cobalt(II) and nickel(II), and magnetic measurements afford magnetic moments that are larger than the spin-only values owing to orbital contributions to the moments. In the UV-Vis spectra, the higher energy spin-allowed d–d transition can be observed in the visible (or near visible) region. Overall, these new cobalt(II) and nickel(II) compounds are similar to their N,N′-diaryl and β-diketiminato analogues, but have lower molecular masses and thus are more volatile, making them potentially useful precursors for the chemical vapor deposition (CVD) of cobalt- or nickel-containing films.

---

Introduction

β-Diketiminate (nacnac) anions have been widely investigated as ancillary ligands in transition metal chemistry owing to their modularity and ease of synthesis, as well as their ability to stabilize a wide variety of metal oxidation states.^1–3 In particular, β-diketiminate complexes of first row transition metals such as cobalt(II) and nickel(II) often show interesting catalytic activity.^4–7 Closely related to β-diketiminate groups are their β-dialdiminate analogues in which the imine carbon bears hydrogen atoms rather than other (usually organic) substituents (Fig. 1, R² = R⁴ = H). Although structurally the two ligand classes are similar, β-dialdiminate complexes are usually more volatile than their β-diketiminate analogues owing to their lower molecular masses.

Fig. 1 A generic β-diimine.

Most of the known metal complexes of β-dialdiminate ligands bear aryl substituents on the two nitrogen atoms; in contrast, metal complexes of N,N′-dialkyl-β-dialdiminates have been little studied. There are a few N,N′-dialkyl-β-dialdiminate compounds of divalent Co, Ni, Cu, Zn and Pd in which the central methine (3-position) is substituted with a nitro group.^8,9 In addition, some transition metal complexes of tetradentate ligands incorporate N,N′-dialkyl-β-dialdiminate units.^10–13 None of these complexes, however, contains unsubstituted N,N′-dialkyl-β-dialdiminate ligands with small aliphatic alkyl groups on nitrogen. Recently we reported the synthesis and characterization of a series of magnesium(II) and zinc(II) N,N′-dialkyl-β-dialdiminato compounds that are examples of coordination complexes in this class.¹⁴

Here, we focus on the synthesis and characterization of monomeric cobalt(II) and nickel(II) N,N′-dialkyl-β-dialdiminate complexes; NMR, IR, and UV-vis spectra are reported, along with their magnetic properties and crystal structures. Owing to their relatively high volatilities, these compounds may prove useful as precursors for CVD or atomic layer deposition (ALD) of cobalt- and nickel-containing films.^15–18

Results and discussion

We have previously described the synthesis of N,N′-dialkyl-1,3-propanedialdiminium chlorides, N,N′-dialkyl-1,3-propanedialdimines, and lithium N,N′-dialkyl-1,3-propanedialdimin-ates, in which the N-alkyl groups are methyl, ethyl, iso-propyl, or tert-butyl.¹⁹ In the present paper, we make use of these reagents to prepare bis(β-dialdiminato) complexes of cobalt(II), 1a–1d, and nickel(II), 2a–2d (Scheme 1). The cobalt complexes are orange to red-orange, whereas the nickel compounds are all dark brown.

Scheme 1 Synthetic routes to the new cobalt(II) and nickel(II) compounds.

All eight compounds were prepared in moderate-to-good yields, depending on the metal and the substituent on nitrogen, by one or both of two routes: (i) salt metathesis from a metal(II) halide (or metal(II) acetylacetonate) and isolated samples of lithium N,N′-dialkyl-1,3-propanedialdiminate (1a, 52%; 1b, 70%; 1c, 89%; 1d, 45%; 2d, 46%), or (ii) salt metathesis from a metal(II) halide and sodium N,N′-dialkyl-1,3-propanedialdiminate prepared in situ from the N,N′-dialkyl-1,3-propanedialdiminium chloride and sodium ethoxide (1a, 78%; 1b, 82%; 2a, 75%; 2b, 80%; 2c, 69%). All compounds can be isolated by sublimation, crystallization, or vacuum distillation. The ethyl and iso-propyl compounds are liquids or low-melting solids at room temperature, but the methyl and tert-butyl compounds are solids. All eight compounds can be handled for at least an hour in air without decomposition, but over longer periods (several hours to several days) the solids discolor and agglomerate, presumably due to hydration or hydrolysis.

The IR spectra exhibit two strong peaks due to N=C and C=C stretching modes, one between 1580 and 1595 cm⁻¹ and the other between 1506 and 1516 cm⁻¹. These frequencies are very similar to those of 1563–1605 cm⁻¹ and 1501–1530 cm⁻¹ reported for N,N′-diaryl-β-dialdiminato complexes of cobalt(II) and nickel(II).^20–24 These complex lack higher-frequency N=C bands in the 1650 to 1770 cm⁻¹ region that are seen when the ligands are reduced to their radical dianionic state.²⁵

Crystallographic studies

All compounds have been crystallographically characterized except the N-iso-propyl-substituted nickel compound 2c, which is low melting and does not crystallize well. Comparisons of their structural features with those of similar compounds are given in the SI. In the solid state, all compounds are monomeric and adopt pseudo-tetrahedral geometries in which the ligands form six-membered rings with the metal center (Fig. 2 and 3). The N-methyl compounds 1a and 2a are isomorphous (space group Pbcn), and are also isomorphous with the zinc analogue.¹⁴ The N-tert-butyl compounds 1d and 2d are isomorphous with one another (space group I4̄2m) and also with the zinc analogue.¹⁴

Fig. 2 Molecular structures of the cobalt compounds 1a–1d. Thermal ellipsoids drawn at the 50% probability density level except hydrogen atoms, which are represented by arbitrarily-sized spheres. Hydrogen atoms on the alkyl arms have been omitted for clarity.

Fig. 3 Molecular structures of the nickel compounds 2a–2b, and 2d. Thermal ellipsoids drawn at the 50% probability density level except hydrogen atoms, which are represented by arbitrarily-sized spheres. Hydrogen atoms on the alkyl arms have been omitted for clarity.

In all the compounds, the π-bonds in the planar NCCCN ligand backbone are delocalized so that the two N=C distances are equal, as are the two C=C distances. The N=C distances of ∼1.32 Å are consistent with the conclusion from the IR spectra that the ligands have a closed-shell monoanionic electronic configuration; for comparison, the N=C distances lengthen to ∼1.35 Å when the ligands are reduced to their radical dianionic form.²⁵

For the cobalt compounds 1a–1d, the M–N bond distances are 1.963(1) for 1a vs. 1.987(1) Å for 1d, with intermediate values for 1b and 1c (Table 1). Although the M–N bond distances increase slightly with increasing steric bulk of the substituent on nitrogen, the differences are too small to establish this trend as chemically significant. The ligand bite angle (defined as the N–M–N angle formed by a single β-dialdiminate ligand) also increases as the N-substituent becomes larger: from 96.14(5)° for the N-methyl compound 1a to 99.62(7)° for the N-tert-butyl compound 1d, with the N-ethyl and N-iso-propyl compounds 1b and 1c again having intermediate values. Similar behaviour is also seen in the nickel N-methyl, N-ethyl, and N-tert-butyl compounds 2a, 2b, and 2d: the M–N bond distance increases very slightly from 1.942(1) Å in 2a to 1.979(1) Å in 2d, and the ligand bite angle becomes larger from 93.10(7)° for 2a to 96.38(8)° for 2d, with the values for 2b lying in between. For the same substituent on the nitrogen, the M–N bond distances are systematically about 0.01–0.03 Å longer in the cobalt compounds than in the nickel compounds, reflecting the slightly larger ionic radius of cobalt(II) compared to nickel(II) (0.72 Å and 0.69 Å, respectively).²⁶ The metal–nitrogen distances are consistent with the metal centers all being high-spin, a conclusion that has been confirmed by magnetic measurements (see below).

Table 1 Selected crystal structure parameters for compounds 1a–1d, 2a–2b, and 2d

Compound	τ4	Dihedral angle (°)	M–N (Å)	Bite angle (°)
1a	0.84	81.66(5)	1.963(1)	96.14(5)
1b	0.88, 0.89	88.22(5), 86.58(5)	1.964(5) to 1.971(1)	96.96(7) to 97.45(6)
1c	0.89	87.43(4)	1.970(1) to 1.981(1)	97.48(5), 97.54(5)
1d	0.93	90	1.987(1)	99.62(7)
2a	0.76	76.04(5)	1.942(1)	93.10(7)
2b	0.73, 0.84	72.8(2), 89.4(3)	1.937(9) to 1.965(6)	92.8(4) to 94.0(4)
2d	0.90	90	1.979(1)	96.38(8)

---

For high-spin ions in tetrahedral coordination environments, Jahn–Teller distortions are not expected for d⁷ cobalt(II) but should occur for d⁸ nickel(II). Despite this difference, the complexes described here have very similar geometries irrespective of the metal ion. All exhibit one or both of two kinds of small distortions away from perfect tetrahedral geometry: (1) the N–M–N angles are not all the same (although this distortion is simply due to the chelating nature of the β-dialdiminato ligands), and (2) except for the N-tert-butyl compounds 1d and 2d the two ligand planes are not perfectly orthogonal. The dihedral angle between two mean ligand planes – defined by the six-membered rings of the ligands – should be 90° for perfect tetrahedral geometry (and equals this value in the tert-butyl substituted complexes 1d and 2d), but when the N-alkyl group is methyl the dihedral angle is 81.66(5)° in the cobalt compound 1a, and 76.04(5)° in the nickel compound 2a. Similar deviations away from the orthogonal geometry are also seen for the nacnac analogues of 1a and 2a in which the nitrogen atoms bear small methyl groups: the interligand dihedral angle is 76.98(4) and 72.30(5)° for the cobalt(II)²⁷ and nickel(II)²⁸ analogues, respectively.

Similar conclusions are reached from comparisons of the τ₄ values (a parameter that ranges from 1 for perfect tetrahedral geometry to 0 for perfect square-planar geometry):²⁹ the distortion away from tetrahedral symmetry is largest for the N-methyl compounds 1a and 2a, which have τ₄ = 0.84 and 0.76, respectively. As the substituents on nitrogen become larger, the geometry more closely approaches tetrahedral symmetry, so that the N-tert-butyl compounds 1d and 2d have τ₄ values of 0.93 and 0.90, respectively (Table 1).

Because the cobalt(II) compounds should not distort owing to Jahn–teller effects, the <90° dihedral angles seen for 1a–1c in the solid state may be due to packing effects. There are, however, some differences in the UV-Vis spectra between 1a–1c and 1d (see below), which suggest that the interligand dihedral angles seen in the crystal structures may persist in solution.

NMR studies

The ¹H and ¹³C{¹H} NMR resonances of the paramagnetic cobalt(II) and nickel(II) compounds 1a–1d and 2a–2d (Fig. 4, 5 and Tables S1.1, S1.2) all have large line widths and are shifted significantly away from the shifts seen for the analogous diamagnetic compounds of zinc and magnesium.¹⁴ The assignments of the NMR resonances were deduced from the peak integrals, chemical shifts, and linewidths, and from comparisons with previously reported compounds.^23,30–32

Fig. 4 ¹H NMR resonances for the cobalt(II) compounds 1a–1d in C₆D₆ at room temperature. The sharp singlet at δ 7.16 is due to the solvent. The scale of the x- and y-axes varies from region to region.

Fig. 5 ¹H NMR resonances for the nickel(II) compounds 2a–2d in C₆D₆ at room temperature. The sharp singlet at δ 7.16 is due to the solvent. The scale of the x- and y-axes varies from region to region.

The broadest and most shifted ¹H NMR resonances seen for 1a–1d and 2a–2d are due to the aldimine N=CH protons, which are deshielded and appear near δ 490 for the Co compounds and near δ 370 for the Ni analogues; the full widths at half maximum (FWHM) for this resonance vary from 1200 to 6800 Hz. The central backbone β-CH protons are shielded instead of deshielded: they appear near δ −84 ppm for the cobalt compounds 1a–1d and near δ −140 for the nickel compounds 2a–2d. In the ¹³C NMR spectra, the broadest and most shifted resonance is not due to the imine N=CH carbon but rather due to the central backbone β-CH carbon: this resonance appears near δ 1250 for 1a–1d and between δ 626 and 741 for 2a–2d. The imine N=CH carbon appears between δ 560 and 690 for 1a–1d, and between δ −327 and −484 for 2a–2d. Thus, for both the cobalt and nickel compounds, the chemical shifts (ranging from most deshielded to most shielded) vary in the order NCH, α-alkyl, β-alkyl, central CH.

The ¹H NMR spectra of isostructural cobalt(II) and nickel(II) compounds have long been of interest because they can be analyzed to distinguish between the contact (i.e., through-bond) and pseudo-contact (also called dipolar; i.e., through-space) contributions to the observed paramagnetic shifts.^32,33 This analysis is possible whenever one of the two ions is magnetically anisotropic but the other is not: the pseudo-contact contribution is zero when the ion is magnetically isotropic. For perfectly tetrahedral complexes cobalt(II) and nickel(II), only nickel has a pseudo-contact contribution to the ¹H NMR shifts, but the situation is different if the symmetry is lowered.

The ground state of cobalt(II) in a tetrahedral (i.e., isotropic) ligand field is an orbitally non-degenerate ⁴A₂(F) state. But in a lower-symmetry axial ligand field, the higher energy triply-degenerate terms are split and then can interact with the orbitally non-degenerate ground state term. This mixing (via spin–orbit coupling) introduces magnetic anisotropy into the ground state, which increases as the axial distortion becomes larger.³³ As a result, in axial ligand fields the g values for cobalt(II) complexes can be anisotropic and often lie in the range of 2.2 to 2.4, usually with a large zero-field spitting of the spin-quartet.³⁴ For comparison, nickel(II) in a tetrahedral ligand field has an orbitally degenerate ³T₁ ground state and shows anisotropic g-values in the EPR spectra (usually in the range of 2.0–2.3); small distortions will lift this degeneracy, however, which can lead to g values that are isotropic (g_x = g_y = g_z) or nearly so.³⁴

Thus, in axially distorted tetrahedral cobalt(II) and nickel(II) compounds, the pseudo-contact shift is often large for cobalt but is small or zero for nickel.³⁵ This behavior has been seen for a series of N,N′-diaryl-β-dialdiminate compounds of cobalt(II) and nickel(II): both the contact and pseudo-contact terms affect the ¹H NMR chemical shifts of the cobalt compounds, whereas only the contact term contributes significantly to the observed ¹H NMR chemical shifts of the nickel compounds.^32,33 As we will show in the next paragraph, the ¹H NMR spectra of compounds 1a–1d and 2a–2d are consistent with this same conclusion: the observed paramagnetic ¹H NMR shifts of the Co β-dialdiminate complexes 1a–1d have both contact and pseudo-contact contributions, whereas the ¹H NMR shifts of the Ni β-dialdiminate complexes 2a–2d are mostly caused by the contact interaction alone.

For the contact shift, delocalization of spin density from the metal can occur through σ-orbitals or through π-orbitals (or both), and it is well understood that these two mechanisms have different effects on the NMR chemical shifts as a function of the number of bonds between the paramagnetic metal center and the ¹H nucleus.^35,36 For compounds 1a–1d and 2a–2d, the paramagnetic shifts of the protons in the backbone of the β-dialdiminate ligands are as follows: deshielded (N=CH), shielded (β-CH), and deshielded (N=CH). This alternation in the sign of the paramagnetic shift along the CH–CH–CH ligand backbone indicates that spin density is delocalized directly into the π-orbitals of the β-dialdiminate ligands, and that in these compounds the contact interaction is the dominant factor in determining the ¹H NMR chemical shifts in the ligand backbone.^35,36

The ¹H NMR chemical shifts of the cobalt(II) and nickel(II) complexes with the same N-substituents show the following trends: the N=CH resonances in the cobalt(II) compounds 1a–1d are deshielded by ca. 120 ppm relative to the nickel(II) compounds 2a–2d, and the β-CH resonances are deshielded by ca. 55 ppm. This deshielding effect is due in large part to the pseudo-contact interaction, which as stated above should be operative only in the cobalt complexes.

The ¹³C NMR shifts of the N=CH and β-CH resonances do not show the shielded/deshielded alternation seen in the ¹H NMR shifts of their attached protons. This observation suggests that the carbon atoms are more strongly affected by the contact interaction mediated by σ-orbitals, which is reasonable because the N=CH and β-CH carbon atoms in the ligand backbones are connected to the metal atoms by two or three bonds, respectively, vs. three or four for the hydrogen atoms.

UV-Vis studies

The cobalt compounds 1a–1d are all bright orange to red in the solid state, and make deep-orange to pale-yellow solutions in pentane depending on the concentration. The spectra of the cobalt compounds all feature intense (ε = 25 000–33 000 M⁻¹ cm⁻¹) absorption bands at ∼330–350 nm due to ligand-based π → π* (and other) transitions (Fig. 6 and Fig. S5.1), as has been seen for similar cobalt(II) bis(β-diketiminate) compounds.³⁷ Compounds 1a–1d also show three relatively weaker bands at 440–510 nm (ε = 750–1300 M⁻¹ cm⁻¹) in the visible region, with the three separate bands being most clearly resolved in the spectrum of the tert-butyl complex 1d.

Fig. 6 UV-Vis spectra of the cobalt compounds 1a–1d in pentane.

Three d–d transitions are expected for ideal tetrahedral high-spin cobalt(II) complexes: the lowest energy transition ⁴A₂ → ⁴T₂ is expected in the IR (near 3000–4000 cm⁻¹ or 2500–3300 nm), the transition ⁴A₂ → ⁴T₁(F) is expected in the near-IR (around 5000–8000 cm⁻¹ or 1250–2000 nm), and the highest-energy transition ⁴A₂ → ⁴T₁(P) is expected in the visible region (around 13 000–18 000 cm⁻¹ or 500–800 nm).^38,39 Previous studies of high-spin cobalt(II) compounds also show that the absorptivities of the d-d bands in the near-IR spectra are generally on the order of 10² M⁻¹ cm⁻¹ whereas the bands in the visible region are more intense, with absorptivities ranging up to 10³ M⁻¹ cm⁻¹. Based on these expectations, we can assign the bands at 440–510 nm for 1a–1d to the ⁴A₂ → ⁴T₁(P) transition, split into several components owing to spin–orbit coupling and the lowering of the molecular symmetry from T_d to D_2d.

As seen for 1a–1d, the ⁴A₂ → ⁴T₁(P) transition in other tetrahedral or near-tetrahedral cobalt(II) compounds is also split into multiple components by spin–orbit or Jahn–Teller effects. For example, the ⁴A₂ → ⁴T₁(P) transition appears as several bands between 726 and 776 nm in stibine and phosphine coordination complexes,⁴⁰ as several bands between 678 and 769 nm for cobalt(II) halide salts,⁴¹ and as several bands between 404 to 513 nm for a bis(benzoxazole-2-yl)methanide cobalt complex.⁴² For cobalt(II) compounds with N₄ ligand sets like those in 1a–1d, the highest-energy ⁴A₂ → ⁴T₁(P) absorption bands generally appear between about 17 500 to 25 000 cm⁻¹ or about 400 to 570 nm,^43–46 which fully encompasses the range seen for 1a–1d.

The UV-Vis spectra of similar cobalt(II) compounds (see SI) often are not sufficiently well resolved to exhibit the three distinct d–d absorptions seen for 1d and a few other compounds, as discussed above. Instead, what is often seen is one broad band occasionally accompanied by broad shoulders, as we see for 1a–1c. The poor resolution of the d–d bands in cobalt(II) complexes is sometimes due to the overlapping of these bands with charge transfer and other kinds of absorptions, and sometimes due to stereochemical flexibility in solution, so that the spectrum consists of overlapping contributions from a manifold of different conformers. We believe the latter phenomenon is most likely responsible for the better resolution of the UV-Vis peaks of 1d vs. those in 1a–1c: owing to the large steric size of the tert-butyl substituents on nitrogen, which disfavor twisting of the two ligands with respect to one another along the molecular 2-fold symmetry axis, 1d is more conformationally locked in solution than its analogues with smaller N-alkyl groups. Thus, the splitting of the three bands seen for 1d is probably due to spin–orbit coupling effects, rather than to Jahn–Teller distortions.

The nickel compounds 2a–2d are all dark brown in the solid state, but dissolve to afford yellow solutions in pentane. The intensities and absorption energies of tetrahedral(II) nickel complexes are generally similar to their those of their cobalt(II) analogs. For 2a–2d, intense ligand-based π → π* transitions (∼330–370 nm, ε = 13 000–22 000 M⁻¹ cm⁻¹) appear at about the same energy as seen for the cobalt compounds, and there also are bands in the mid-visible region between 430 and 490 nm (Fig. 7 and Fig. S5.2). The spectra of 2a–2d also contain weaker features at ∼600 nm.

Fig. 7 UV-Vis spectra of the nickel compounds 2a–2d.

Ideal tetrahedral high-spin nickel(II) complexes have three spin-allowed d-d transitions: the lowest energy transition ³T₁(F) → ³T₂(F) is expected in the IR (near 3000–4000 cm⁻¹ or 2500–3300 nm), the ³T₁(F) → ³A₂(F) transition is expected in the near-IR (near 5000–10 000 cm⁻¹ or 1000–2000 nm), and the highest-energy ³T₁(F) → ³T₁(P) transition is expected in the visible region (near 11 000–18 000 cm⁻¹ or 500–900 nm).³⁹ Based on these expectations, we can assign the bands at 430–490 nm for 2a–2d to the ³T₁(F) → ³T₁(P) transition, split into several components owing to the lowering of the molecular symmetry from T_d to D_2d. This ³T₁(F) → ³T₁(P) transition appears at about the same energy for the methyl, ethyl, iso-propyl substituted complexes 2a–2c, but is about 30 nm lower in energy (and somewhat more intense) for the tert-butyl compound 2d, probably because the two ligand planes in the latter are more exactly orthogonal. The weaker features at ∼600 nm may be due to transitions to spin singlet states (¹D), which are often observable in the regions between the spin allowed bands.³⁹

In certain nickel(II) salts the ³T₁(F) → ³T₁(P) transition appears at 11 000 to 17 000 cm⁻¹ (600 to 900 nm), and the ³T₁(F) → ³A₂(F) transition appears at 7000 to 9000 cm⁻¹ (1000 to 1400 nm).⁴⁷ Some other tetrahedral nickel(II) phosphine halide complexes also absorb in the 500–900 nm range,⁴⁸ exhibiting bands with fine-structure giving line shapes that are similar to those seen for 2a–2d.

Magnetic and EPR studies

The direct current (dc) magnetic susceptibilities of compounds 1a, 1b, 1d, 2a, and 2d were measured from 2 to 300 K (Fig. 8) in an applied field of 0.1 T. The room temperature (300 K) χT values for the cobalt(II) complexes 1a, 1b, and 1d are 2.26, 2.54, and 2.34 cm⁻³ mol⁻¹ K, respectively (corresponding to magnetic moments of 4.25, 4.51, and 4.33μ_B). These susceptibilities are larger than the spin-only χT value of 1.88 cm⁻³ mol⁻¹ K (= 3.88μ_B) for a single molecular spin center with S = 3/2 and g = 2.00, but closely resemble the magnetic moment of 4.30μ_B for an axially-distorted tetrahedral S = 3/2 spin center with a g value near 2.22. Large magnetic moments (even up to 5.00μ_B) have been seen for monomeric high-spin tetrahedral cobalt(II) complexes, owing to an orbital contribution to the magnetic moment. As discussed above in the context of the ¹H NMR spectra, the orbital contribution to the moment is caused by spin–orbit coupling, which splits the ⁴T₁ and ⁴T₂ states into four states each; two of the eight states have the same symmetry as the ⁴A₂ ground state which leads to mixing of the levels. This mixing introduces orbital angular momentum into the ground state, thereby increasing the magnetic moment^49,50 and shifting the g values to 2.2–2.4 assuming that the g factor is isotropic (g_x = g_y = g_z).³⁴

Fig. 8 χT versus T plot for bis(N,N′-dimethyl-1,3-propanedialdiminato)cobalt(II), (1a), bis(N,N′-diethyl-1,3-propanedialdiminato)cobalt(II) (1b), bis(N,N′-di(tert-butyl)-1,3-propanedialdiminato)cobalt(II) (1d), bis(N,N′-dimethyl-1,3-propanedialdiminato)nickel(II) (2a) and bis(N,N′-di(tert-butyl)-1,3-propanedialdiminato)nickel(II) (2d).

For 1a, the χT value remains relatively constant from 300 down to 100 K, then decreases gradually upon lowering the temperature further to 40 K and then more steeply to a value of 0.92 cm⁻³ mol⁻¹ K at 2 K. Similarly, χT for compound 1b decreases gradually from 300 K to about 30 K and then decreases sharply to a value of 1.51 cm⁻³ mol⁻¹ K at 2 K, whereas the χT value for compound 1d is relatively unchanged from 300 K to 20 K, with an abrupt drop to 1.81 cm⁻³ mol⁻¹ K at 2 K. This behavior is likely caused by the presence of zero-field splitting (ZFS), which strongly affects the magnetic properties of high-spin cobalt(II) centers,^51,52 and/or orbital angular momentum contributions to the observed magnetic moment.^{40,42,53–55} The temperatures at which the steep decrease in χT occurs are in the order 1d (∼20 K) < 1b (∼30 K) < 1a (∼40 K), indicating that the ZFS in 1d is smaller than for complexes 1a and 1b. The sloping of the high-temperature portion of the curve for 1b may be due to presence of a small amount of a ferromagnetic impurity; it is unlikely that the intrinsic magnetic properties of 1b should differ significantly from those of 1a and 1d. All of these data are consistent with 1a, 1b, and 1d having high-spin, axially distorted pseudo-tetrahedral cobalt(II) metal centers.⁵⁶

The nickel complexes 2a and 2d have room temperature (300 K) χT values of 1.27 and 1.29 cm⁻³ mol⁻¹ K, respectively (corresponding to magnetic moments of 3.19, and 3.21μ_B). These values are larger than the spin-only χT value of 1.00 cm⁻³ mol⁻¹ K (= 2.83μ_B) for a single molecular spin center with S = 1 and an isotropic g value of 2.00 but closely match those for an S = 1 center with g = 2.26. Similar values are seen for other pseudo-tetrahedral nickel(II) compounds.^28,42,57 Rigorously tetrahedral nickel(II) compounds have an orbitally degenerate ground state (³T₁) and can have magnetic moments up to 4.2μ_B,⁴⁸ but relatively modest distortions from T_d symmetry can remove the orbital degeneracy and lower the magnetic moment (the lower limit for the high-spin state being the spin-only value of 2.8μ_B).

All SQUID χT data reported here were fit to various models using the following Hamiltonian, which includes both Zeeman and ZFS terms:

where D is the zero-field splitting and g is the Landé g-factor. An additional temperature independent paramagnetism (TIP) term was required in some cases (see SI for full details).

The fits of the magnetic data for the cobalt(II) compounds give the following results: all three compounds 1a, 1b, and 1d are best considered to have an axial set of g-values in which the g-anisotropy is largest for the N-methyl compound 1a (g_x = g_y = 2.46; g_z = 1.59) and smallest for the N-tert-butyl compound 1d (g_x = g_y = 2.25; g_z = 2.19). Evidently, the anisotropy becomes smaller as the compound becomes more nearly perfectly tetrahedral. The zero-field splitting parameter D is largest and negative for the N-methyl compound 1a (−11 cm⁻¹) and smallest and positive for the N-tert-butyl compound 1d (+5 cm⁻¹).

The fits of the magnetic data for the nickel(II) compounds are best interpreted in terms of an isotropic or an axial but nearly-isotropic set of g-values, with g between 2.25 and 2.29 and D between +51 and +59 cm⁻¹. The N-tert-butyl compound 2d shows has the largest g-values and the largest zero-field splitting parameter.

Frozen toluene solutions of 1a gave very weak X-band EPR signals near g_eff = 2.05 and g_eff = 6.5; the nickel(II) compound 2a under similar conditions gave a weak signal near g_eff = 2.05. It is possible that some of these signals may have been due to impurities. For the cobalt compound, the absence of a well-defined EPR spectrum is most likely due to rapid spin relaxation caused by strong spin–orbit coupling; nickel(II) compounds are commonly EPR silent at X-band frequencies at 77 K.

Volatility studies of the New N,N′-dialkyl-β-dialdiminato nickel and cobalt compounds

Thermogravimetric analysis (TGA) studies of these new bis(N,N′-dialkyl-β-dialdiminato) complexes were carried out to determine their potential suitability as CVD precursors (Fig. 9 and 10). All eight compounds 1a–1d and 2a–2d exhibit smooth volatilization curves with no evidence of decomposition. The residual masses were <5% for the cobalt compounds 1a–1d and <1% for the nickel compounds 2a–2d.

Fig. 9 TGA ramp experiments for the cobalt(II) compounds 1a–1d under 1 atm of N₂; the heating rate was 10 °C per minute.

Fig. 10 TGA ramp experiments for the nickel(II) compounds 2a–2d under 1 atm of N₂; the heating rate was 10 °C per minute.

The enthalpies of vaporization, which were measured for all eight compounds from a series of isothermal experiments⁵⁸ (see Tables S3.2 and S3.3), indicate that the most volatile compounds are the N-methyl substituted compounds 1a and 2a, which have ΔH_vap = 66 and 69 kJ mol⁻¹, respectively. The N-ethyl and N-iso-propyl compounds 1b–1c and 2b–2c have similarly small enthalpies of vaporization of between 68 and 74 kJ mol⁻¹ whereas the N-tert-butyl substituted compounds 1d and 2d are significantly less volatile, with ΔH_vap = 103 and 105 kJ mol⁻¹, respectively. As expected, the temperature required to achieve 1 Torr of vapor pressure (a useful metric for CVD experiments) increases with increasing molecular mass: 100, 114, 145, and 184 °C for the cobalt methyl, ethyl, iso-propyl, and tert-butyl compounds, respectively, and 85, 120, 152, and 183 °C for the nickel methyl, ethyl, iso-propyl, and tert-butyl compounds, respectively. As expected, the N-methyl substituted compounds 1a and 2a are the most volatile of these eight new compounds. The volatilization onset temperatures under 1 atm of N₂ also follow this same trend (see Table S3.1).

Commercially available cobalt(II) and nickel(II) amidinate precursors sublime at 35–70 °C at 0.05–0.07 Torr.⁵⁹ No onset temperatures or enthalpies of vaporization were reported, but these sublimation conditions are similar to those employed during the purification of 1a–1c and 2a–2c. Nickel(II) and cobalt(II) complexes of N,N′-dialkyl β-ketoiminates have also been investigated as CVD precursors. The N-iso-propyl substituted nickel complex has an enthalpy of vaporization of 76.73 kJ mol⁻¹, and a temperature of 156 °C required to achieve 1 Torr of vapor pressure; the analogous cobalt complex is similarly volatile.^16,17 From these values, we conclude that 1a–1c and 2a–2c are more volatile than these β-ketoiminato complexes, whereas compounds 1d and 2d are somewhat less volatile. Overall, the volatility and thermal stability of the N,N′-di-alkyl-bis(β-dialdiminato) complexes 1a–1d and 2a–2d make them potentially useful CVD precursors, because their volatilities are similar to, or exceed, those of commercially available cobalt and nickel CVD precursors.

Conclusions

The electronic, magnetic, and thermal characteristics of the cobalt(II) and nickel(II) N,N′-dialkyl-β-dialdiminato complexes presented here are similar to those of their β-diketiminato analogues, except they are more volatile, a property that may prove useful for the deposition of nickel and cobalt containing films. Studies of the use of these new compounds as CVD precursors will be reported separately.

Experimental section

All operations were carried out in vacuum or under argon using standard Schlenk and glove box techniques unless otherwise stated. All glassware was oven-dried before use. Solvents were distilled under nitrogen from sodium (toluene), sodium/benzophenone (pentane, diethyl ether, and THF), calcium hydride (dichloromethane), or magnesium/iodide (methanol) before use. Metal halides (CoBr₂, CoCl₂, NiBr₂) and Ni(acac)₂ were purchased from Sigma-Aldrich, Strem, or Alfa Aesar and used as received. N,N′-dialkyl-1,3-propanedialdiminium chlorides, N,N′-dialkyl-1,3-propanedialdimines, and lithium N,N′-dialkyl-1,3-propanedialdiminates were prepared by literature routes.¹⁹ Deuterated benzene (Cambridge Isotope Laboratories) was distilled from calcium hydride and stored over 3 Å molecular sieves. All other compounds were purchased as reagent grade and used as received.

Elemental analyses were performed by the School of Chemical Sciences Microanalytical Laboratory at the University of Illinois at Urbana-Champaign. FTIR spectra were acquired on a Thermo Nicolet IR200 spectrometer as Nujol mulls between KBr plates and processed using the Omnic software package. NMR spectra were recorded on a Varian Unity Inova 500NB spectrometer at 11.75 T, a Varian VXR 500 spectrometer with Unity Inova 500 console at 11.75 T, or a B600 Bruker NEO spectrometer at 14.1 T equipped with a Bruker Avance Neo console and 600 MHz, 5 mm, BBO-BB Prodigy CryoProbe. Chemical shifts are reported in δ units, more positive shifts to higher frequency relative to TMS, referenced to residual solvent peaks. Due to the fast relaxation of the ¹H nuclei (particularly the N=CH protons, which were unobservable with a standard 90° pulse of 8 μs), ¹H NMR spectra were collected with a short pulse (2 or 5 μs). Melting points and decomposition temperatures were determined in closed capillaries under argon on a Thomas–Hoover Unimelt apparatus.

UV-visible spectra were recorded on a Varian Cary 50 Bio spectrophotometer or an Agilent Technologies 8454 spectrophotometer. Thermogravimetric analyses were performed on a TA Instruments Q50 TGA. Solid state magnetic data were collected on a Quantum Design third-generation Magnetic Property Measurement System (MPMS3), which uses a Superconducting Quantum Interference Device (SQUID), and were modeled using the MagProp program in Dave 2.5.⁶⁰

Bis(N,N′-dimethyl-1,3-propanedialdiminato)cobalt(II), 1a

Method A: To CoBr₂ (0.12 g, 0.54 mmol), sodium ethoxide (0.15 g, 2.2 mmol), and N,N′-dimethyl-1,3-propanedialdiminium chloride (0.17 g, 1.2 mmol) was added methanol (8 mL) and the mixture stirred for 12 h. The solution was evaporated to dryness in vacuum, and the residue was extracted with pentane (2 × 10 mL); the resulting orange extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was sublimed (40 °C, 10 mTorr) to yield red-orange crystals. Yield: 0.11 g (78%). Method B: To a suspension of CoBr₂ (0.088 g, 0.40 mmol) in diethyl ether (5 mL) at −78 °C was added a solution of lithium N,N′-dimethyl-1,3-propanedialdiminate (0.090 g, 0.86 mmol) in diethyl ether (15 mL), and the solution warmed slowly to room temperature with stirring over 12 h. The solution was evaporated to dryness in vacuum and the residue extracted with pentane (2 × 10 mL); the resulting orange extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was sublimed (40 °C, 10 mTorr) to yield red/orange crystals. Yield: 0.053 g (52%). M.p. 103 °C. Anal. calc. for C₁₀H₁₈N₄Co(253.21): C, 47.4; H, 7.16; N, 22.1; found: C, 47.4; H, 7.25; N, 21.9. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 486 (br, FWHM = 1400 Hz, 4H, N=CH), 216 (br, FWHM = 400 Hz, 12H, CH₃), −87 (br, FWHM = 220 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 1240 (br, FWHM = 460 Hz, β-CH), 691 (br, FWHM = 290 Hz, N=CH), 358 (br, FWHM = 250 Hz, CH₃). IR (KBr, cm⁻¹): 1590 (s), 1508 (s), 1394 (m), 1362 (m), 1326 (s), 1193 (w), 1125 (m), 1095 (m), 723 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 246 (7200), 291 (11 000), 329 (25 000), 342 (27 000), 380 (4700) sh, 456 (990) sh, 472 (1200), 502 (750) sh.

Bis(N,N′-diethyl-1,3-propanedialdiminato)cobalt(II), 1b

Method A: To CoBr₂ (0.16 g, 0.73 mmol), sodium ethoxide (0.20 g, 3.0 mmol), and N,N′-diethyl-1,3-propanedialdiminium chloride (0.24 g, 1.5 mmol) was added methanol (7 mL) and the mixture stirred for 12 h. The solution was evaporated to dryness in vacuum and extracted with pentane (2 × 10 mL); the resulting orange extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was sublimed (45 °C, 10 mTorr) to yield red-orange crystals. Yield: 0.19 g (82%). Method B: To a suspension of CoBr₂ (0.081 g, 0.37 mmol) in diethyl ether (5 mL) at −78 °C was added a solution of lithium N,N′-diethyl-1,3-propanedialdiminate (0.089 g, 0.67 mmol) in diethyl ether (15 mL), and the solution was warmed slowly to room temperature and then stirred for 4 days. The reaction solution was evaporated to dryness in vacuum, and the residue was extracted with pentane (2 × 10 mL); the resulting orange extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was sublimed (38 °C, 10 mTorr) to yield orange crystals. Yield: 0.080 g (70%). M.p. 46 °C. Anal. calc. for C₁₄H₂₆N₄Co (309.32): C, 54.36; H, 8.47; N, 18.1; found: C, 54.14; H, 8.44; N, 17.9. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 482 (br, FWHM = 1700 Hz, 4H, N=CH), 217 (br, FWHM = 520 Hz, 8H, CH₂), −23 (br, FWHM = 300 Hz, 12H, CH₃), −86 (br, FWHM = 270 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 1226 (br, FWHM = 610 Hz, β-CH), 610 (br, FWHM = 260 Hz, N=CH), 430 (br, FWHM = 160 Hz, CH₃), 325 (br, FWHM = 260 Hz, CH₂). IR (KBr, cm⁻¹): 1584 (s), 1508 (s), 1336 (m), 1262 (m), 1135 (m), 1041 (w), 725 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 249 (6300), 289 (9700), 329 (25 000), 344 (28 000), 382 (4900) sh, 447 (990) sh, 471 (1300), 501 (800) sh.

Bis(N,N′-di(iso-propyl)-1,3-propanedialdiminato)cobalt(II), 1c

To a suspension of CoBr₂ (0.55 g, 2.5 mmol) in diethyl ether (90 mL) at −78 °C was added a solution of lithium N,N′-di(iso-propyl)-1,3-propanedialdiminate (0.82 g, 5.1 mmol) in diethyl ether (40 mL), and the solution allowed to slowly warm to room temperature with stirring over 12 h. The reaction solution was evaporated to dryness in vacuum and the residue extracted with pentane (4 × 10 mL); the resulting extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was vacuum distilled (bath temperature of 260 °C, 10 mTorr) to yield an orange oil that solidified upon cooling. Yield: 0.81 g (89%). M.p. ca 20 °C. Anal. calc. for C₁₈H₃₄N₄Co (365.42): C, 59.16; H, 9.38; N, 15.33; found: C, 58.99; H, 9.45; N, 15.51. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 487 (br, FWHM = 2900 Hz, 4H, N=CH), 212 (br, FWHM = 940 Hz, 4H, N–CH), −6.5 (br, FWHM = 400 Hz, 24H, CH₃), −85 (br, FWHM = 420 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 1247 (br, FWHM = 1400 Hz, 2C, β-CH), 598 (br, FWHM = 300 Hz, 4C, N=CH), 452 (br, FWHM = 180 Hz, 8C, CH₃), 231 (br, FWHM = 360 Hz, 4C, N–CH). IR (KBr, cm⁻¹): 1580 (s), 1509 (s), 1336 (s), 1312 (s), 1288 (s), 1199 (m), 1141 (m), 1116 (m), 1017 (w), 958 (w), 730 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 247 (6000), 289 (9300), 331 (25 000), 347 (30 000), 380 (6500), 443 (880) sh, 473 (1300), 508 (810). Crystals of 1c suitable for X-ray crystallography were grown from saturated pentane solutions at –20 °C.

Bis(N,N′-di(tert-butyl)-1,3-propanedialdiminato)cobalt(II), 1d

To a suspension of CoCl₂ (0.071 g, 0.55 mmol) in diethyl ether (15 mL) at −78 °C was added a solution of lithium N,N′-di(tert-butyl)-1,3-propanedialdiminate (0.19 g, 1.0 mmol) in diethyl ether (20 mL), and the solution allowed to slowly warm to room temperature with stirring over 12 h. The reaction solution was evaporated to dryness in vacuum and the residue extracted with pentane (2 × 15 mL); the resulting extracts were filtered, combined, concentrated and stored at −20 °C to yield orange crystals. Yield: 0.081 g (39%). A second crop of crystals grown from the mother liquor yielded an additional 0.013 g, for a total yield of 0.094 g (45%). M.p. 252 °C. Anal. calc. for C₂₂H₄₄N₄Co (421.53): C, 62.7; H, 10.0; N, 13.3; found: C, 62.8; H, 9.97; N, 13.2. ¹H NMR (500 MHz, C₆D₆, 20 °C): δ 493 (br, FWHM = 4000 Hz, 4H, N=CH), 9.06 (br, FWHM = 730 Hz, 36H, CH₃), −80 (br, FWHM = 670 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 561 (br, FHMW = 550 Hz, 4C, N=CH), 481 (br, FWHM = 350 Hz, 12C, CH₃), 87 (br, FWHM = 550 Hz, 4C, N–C). ¹H NMR (600 MHz, CDCl₃, 20 °C): δ 492 (br, FWHM = 6800 Hz, 4H, N=CH), 8.95 (br, FWHM = 990 Hz, 36H, CH₃), −79 (br, FWHM = 830 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, CDCl₃, 20 °C): δ 1260 (br, FWHM = 1800 Hz, 2C, β-CH), 553 (br, FWHM = 530 Hz, 4C, N=CH), 474 (br, FWHM = 360 Hz, 12C, CH₃), 95 (br, FWHM = 630 Hz, 4C, N–C). IR (KBr, cm⁻¹): 1582 (m), 1513 (s), 1327 (s), 1241 (m), 1190 (m), 1171 (m), 928 (m), 736 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 290 (8900), 340 (33 000), 351 (31 000), 383 (6100), 445 (950), 471 (1300), 505 (1000).

Bis(N,N′-dimethyl-1,3-propanedialdiminato)nickel(II), 2a

To NiBr₂ (0.31 g, 1.4 mmol), sodium ethoxide (0.38 g, 5.5 mmol), and N,N′-dimethyl-1,3-propanedialdiminium chloride (0.38 g, 2.8 mmol) was added tetrahydrofuran (15 mL) and the mixture was stirred for 12 h. The solution was evaporated to dryness in vacuum, and the residue was extracted with pentane (2 × 10 mL); the resulting dark brown extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was sublimed (50 °C, 10 mTorr) to yield dark brown crystals. Yield: 0.26 g (75%). M.p. 92 °C. Anal. calc. for C₁₀H₁₈N₄Ni (252.969): C, 47.5; H, 7.17; N, 22.2; found: C, 47.5; H, 7.18; N, 21.9. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 374 (br, FWHM = 1200 Hz, 4H, N=CH), 258 (br, FWHM = 200 Hz, 12H, CH₃), −143 (br, FWHM = 100 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 741 (br, FWHM = 340 Hz, 2C, β-CH), 603 (br, FWHM = 310, 4C, CH₃), −327 (br, FWHM = 100 Hz, 4C, N=CH). IR (KBr, cm⁻¹): 1595 (s), 1506 (s), 1409 (m), 1394 (m), 1362 (m), 1322 (s), 1186 (w), 1123 (w), 1100 (w), 725 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 318 (16 000), 348 (17 000), 425 (2800), 471 (2900), ∼610 (230).

Bis(N,N′-diethyl-1,3-propanedialdiminato)nickel(II), 2b

To NiBr₂ (0.69 g, 3.1 mmol), sodium ethoxide (0.88 g, 13 mmol), and N,N′-diethyl-1,3-propanedialdiminium chloride (1.0 g, 6.4 mmol) was added tetrahydrofuran (55 mL) and the mixture stirred for 7 days. The solution was evaporated to dryness in vacuum, and the residue was extracted with pentane (4 × 10 mL); the resulting dark brown extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was vacuum distilled (bath temperature of 130 °C, 10 mTorr) to yield a brown oil. Yield: 0.77 g (80%). Anal. Calc. for C₁₄H₂₆N₄Ni (309.075): C, 54.4; H, 8.48; N, 18.1; found: C, 54.3; H, 8.24; N, 18.1. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 365 (br, FWHM = 2000 Hz, 4H, N=CH), 252 (br, FWHM = 300 Hz, 8H, CH₂), 5.7 (br, FWHM = 100 Hz, 12H, CH₃), −141 (br, FWHM = 100 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 721 (br, FWHM = 360 Hz, 2C, β-CH), 569 (br, FWHM = 260 Hz, 4C, CH₂), 400 (br, FWHM = 70 Hz, 4C, CH₃), −364 (br, FWHM = 120 Hz, 4C, N=CH). IR (KBr, cm⁻¹): 1591 (s), 1507 (s), 1484 (m), 1335 (m), 1261 (m), 1134 (m), 1038 (w), 726 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 317 (18 000), 339 (18 000), 354 (18 000), 428 (2600), 477 (2800), 600 (140). Crystals of 2b suitable for X-ray crystallography were grown from saturated pentane solutions at –20 °C.

Bis(N,N′-di(iso-propyl)-1,3-propanedialdiminato)nickel(II), 2c

To NiBr₂ (0.72 g, 3.3 mmol), sodium ethoxide (0.92 g, 13 mmol), and N,N′-di(iso-propyl)-1,3-propanedialdiminium chloride (1.2 g, 6.4 mmol) was added tetrahydrofuran (50 mL) and the mixture stirred for 7 days. The solution was evaporated to dryness in vacuum, and the residue was extracted with pentane (3 × 10 mL); the resulting dark brown extracts were filtered, combined, and evaporated to dryness in vacuum. The residue was vacuum distilled (bath temperature of 220 °C, 10 mTorr) to yield a brown oil that solidified upon cooling. Yield: 0.81 g (69%). M.p. 59 °C. Anal. calc. for C₁₈H₃₄N₄Ni (365.181): C, 59.2; H, 9.38; N, 15.3; found: C, 59.0; H, 9.27; N, 15.2. ¹H NMR (500 MHz, C₆D₆, 20 °C): δ 372 (br, FWHM = 3000 Hz, 4H, N=CH), 237 (br, FWHM = 500 Hz, 4H, N-CH), 12.2 (br, FWHM = 100 Hz, 24H, CH₃), −136 (br, FWHM = 100 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 692 (br, FWHM = 390 Hz, 2C, β-CH), 583 (br, FWHM = 230 Hz, 4C, N–CH), 429 (br, FWHM = 110 Hz, 8C, CH₃), −395 (br, FWHM = 140 Hz, 4C, N=CH). IR (KBr, cm⁻¹): 1588 (s), 1508 (s), 1479 (s), 1364 (m), 1344 (m), 1309 (m), 1284 (s), 1273 (s), 1195 (m), 1145 (m), 1110 (m), 1090 (w), 1043 (w), 1018 (w), 956 (w), 730 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 232 (14 000), 318 (15 000), 336 (16 000), 357 (16 000), 442 (3000), 487 (2400) sh, 586 (150).

Bis(N,N′-di(tert-butyl)-1,3-propanedialdiminato)nickel(II), 2d

To a suspension of Ni(acac)₂ (0.18 g, 0.69 mmol) in diethyl ether (10 mL) at −78 °C was added a solution of lithium N,N′-di(tert-butyl)-1,3-propanedialdiminate (0.27 g, 1.4 mmol) in diethyl ether (20 mL), and the solution allowed to slowly warm to room temperature with stirring over 12 h. The reaction solution was evaporated to dryness in vacuum and the residue extracted with pentane (3 × 10 mL); the resulting dark brown extracts were filtered, combined, concentrated and stored at −20 °C to yield dark brown crystals. Yield: 0.12 g (40%). A second crop of crystals grown from the mother liquor yielded an additional 0.017 g, for a total yield of 0.14 g (46%). M.p. 251 °C. Anal. calc. for C₂₂H₄₂N₄Ni (421.287): C, 62.7; H, 10.1; N, 13.3; found: C, 62.7; H, 10.0; N, 13.3. ¹H NMR (600 MHz, C₆D₆, 20 °C): δ 333 (br, FWHM = 3000 Hz, 4H, N=CH), 19.6 (br, FWHM = 200 Hz, 36H, CH₃), −135 (br, FWHM = 200 Hz, 2H, β-CH). ¹³C{¹H} NMR (600 MHz, C₆D₆, 20 °C): δ 634 (br, FWHM = 200 Hz, 4C, N–C), 626 (br, FWHM = 460 Hz, 2C, β-CH), 425 (br, FWHM = 130 Hz, 12C, CH₃), −484 (br, FWHM = 140 Hz, 4C, N=CH). IR (KBr, cm⁻¹): 1588 (m), 1516 (m), 1320 (m), 1186 (m), 1167 (m), 917 (w), 737 (m). UV-Vis (pentane, λ/nm, ε/M⁻¹ cm⁻¹): 233 (15 000), 257 (8300) sh, 318 (19 000) sh, 337 (22 000), 367 (13 000), 464 (3600) sh, 494 (3600), 585 (150).

Author contributions

All authors have given approval to the final version of the manuscript.

Conflicts of interest

The authors declare no competing financial interests.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: ¹H and ¹³C{¹H} NMR spectra, IR spectra, TGA data, SQUID data, EPR data, UV-Vis data, crystallographic studies, and comparisons of structural features of interest for compounds 1a–1d and 2a–2d. See DOI: https://doi.org/10.1039/d6dt01165a.

CCDC 2348010–2348015 and 2348314 (for 1a–1d, 2a, 2b, and 2d) contain the supplementary crystallographic data for this paper.^61a–g

Acknowledgements

We thank the National Science Foundation (CHE 24-00099) and the University of Illinois (fellowship for C. E. S.) for support of this research. We thank Dr Danielle Gray and Dr Toby Woods of the G. L. Clark X-ray Laboratory at the University of Illinois at Urbana-Champaign for collecting the X-ray diffraction data. We thank Professors Mirica and Olshansky for access to UV-Vis instrumentation. We thank the NMR Laboratory at the University of Illinois at Urbana-Champaign for technical support. The Bruker 500-Mz NMR spectrometer was obtained with the financial support of the Roy J. Carver Charitable Trust, Muscatine, Iowa, USA. The research was carried out in part in the Materials Research Laboratory Central Research Facilities, University of Illinois. The authors acknowledge the use of facilities and instrumentation at the Materials Research Laboratory Central Research Facilities, University of Illinois, partially supported by NSF through the University of Illinois Materials Research Science and Engineering Center DMR-1720633.

References

1. L. Bourget-Merle, M. F. Lappert and J. R. Severn, Chem. Rev., 2002, 102, 3031–3066.
2. R. L. Webster, Dalton Trans., 2017, 46, 4483–4498.
3. C. Camp and J. Arnold, Dalton Trans., 2016, 45, 14462–14498.
4. C. Chen, M. B. Hecht, A. Kavara, W. W. Brennessel, B. Q. Mercado, D. J. Weix and P. L. Holland, J. Am. Chem. Soc., 2015, 137, 13244–13247.
5. D. Kim, C. Chen, B. Q. Mercado, D. J. Weix and P. L. Holland, Organometallics, 2020, 39, 2415–2424.
6. N. J. Moghaddam, M. Gil-Sepulcre, J.-W. Wang, J. Benet-Buchholz, C. Gimbert-Suriñach and A. Llobet, Inorg. Chem., 2022, 61, 16639–16649.
7. S. Pahar, V. Sharma, B. Mahata, C. P. George, H. Sharma, K. Vanka and S. S. Sen, Inorg. Chem., 2022, 61, 17370–17377.
8. V. P. Kurbatov, S. V. Serbina and O. A. Osipov, Russ. J. Gen. Chem., 1973, 43, 702–705.
9. V. P. Kurbatov, S. V. Serbina and O. A. Osipov, Russ. J. Inorg. Chem., 1973, 18, 1414–1416.
10. A. F. Ghiron, R. K. Murmann and E. O. Schlemper, Inorg. Chem., 1985, 24, 3271–3278.
11. R. K. Murmann, J. Chipley and P. Droege, J. Chem. Crystallogr., 1998, 28, 465–473.
12. V. Sinha, B. Pribanic, B. de Bruin, M. Trincado and H. Grützmacher, Chem. – Eur. J., 2018, 24, 5513–5521.
13. F. Casas Cancela, PhD thesis, ETH Zürich, 2021.
14. C. E. Schmit and G. S. Girolami, Polyhedron, 2024, 256, 116985.
15. A. E. Kaloyeros, Y. Pan, J. Goff and B. Arkles, ECS J. Solid State Sci. Technol., 2019, 8, P119–P152.
16. D. Zywitzki, D. H. Taffa, L. Lamkowski, M. Winter, D. Rogalla, M. Wark and A. Devi, Inorg. Chem., 2020, 59, 10059–10070.
17. K. J. Puring, D. Zywitzki, D. H. Taffa, D. Rogalla, M. Winter, M. Wark and A. Devi, Inorg. Chem., 2018, 57, 5133–5144.
18. D. Zywitzki, R. Schaper, E. Ciftyürek, J.-L. Wree, D. H. Taffa, D. M. Baier, D. Rogalla, Y. Li, M. Meischein, A. Ludwig, Z. Li, K. Schierbaum, M. Wark and A. Devi, Adv. Mater. Interfaces, 2021, 8, 2100949.
19. C. E. Schmit and G. S. Girolami, Helv. Chim. Acta, 2023, 106, e202200152.
20. N. M. Tsybina, V. G. Vinokurov, T. V. Protopopova and A. P. Skoldinov, Zh. Obshch. Khim., 1966, 36, 1372–1374.
21. J. Takaichi, Y. Morimoto, K. Ohkubo, C. Shimokawa, T. Hojo, S. Mori, H. Asahara, H. Sugimoto, N. Fujieda, N. Nishiwaki, S. Fukuzumi and S. Itoh, Inorg. Chem., 2014, 53, 6159–6169.
22. V. P. Kurbatov, S. V. Serbina, T. G. Lesnyak and O. A. Osipov, Zh. Obshch. Khim., 1970, 40, 2360–2363.
23. R. Knorr, R. Zölch and K. Polborn, Heterocycles, 1995, 40, 559–576.
24. C. L. Honeybourne and G. A. Webb, Chem. Commun., 1968, 739–740.
25. C. E. Schmit, PhD thesis, University of Illinois Urbana-Champaign, 2024.
26. R. D. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751–767.
27. D. Piryazev, CCDC 1577787: Experimental Crystal Structure Determination, 2017,  DOI:10.5517/ccdc.csd.cc1pytct.
28. M. M. Khusniyarov, E. Bill, T. Weyhermüller, E. Bothe and K. Wieghardt, Angew. Chem., Int. Ed., 2011, 50, 1652–1655.
29. L. Yang, D. R. Powell and R. P. Houser, Dalton Trans., 2007, 955–964.
30. P. Healy, M. Bendall, D. Doddrell, B. Skelton and A. White, Aust. J. Chem., 1979, 32, 727–735.
31. R. Knorr, H. Hauer, A. Weiss, H. Polzer, F. Ruf, P. Löw, P. Dvortsák and P. Böhrer, Inorg. Chem., 2007, 46, 8379–8390.
32. C. L. Honeybourne and G. A. Webb, Chem. Phys. Lett., 1968, 2, 426–428.
33. C. L. Honeybourne and G. A. Webb, Mol. Phys., 1969, 17, 17–31.
34. R. L. Carlin, Magnetochemistry, Springer-Verlag, Germany, 1986, pp. 52–69.
35. W. D. Horrocks, R. C. Taylor and G. N. LaMar, J. Am. Chem. Soc., 1964, 86, 3031–3038.
36. When unpaired spin density in a metal orbital of σ-symmetry with respect to the ligands delocalizes into ligand σ-orbitals, the magnetic field experienced by the protons is enhanced by the electron spin magnetic field, and the proton resonances are all deshielded. This deshielding effect decreases rapidly as the number of bonds between a given proton and the metal increases. In contrast, when unpaired spin density in a metal orbital of π-symmetry with respect to the ligands is delocalized into ligand π-orbitals through dπ-pπ bonding, every other atom in the π-system develops a positive spin density, in which the electron spin is in the same direction as the applied magnetic field. Spin correlation causes the remaining atoms to have negative spin density. Positive spin density at a given atom in the π-system gives rise to shielding of a directly-attached hydrogen atom.
37. D. W. Randall, S. D. George, P. L. Holland, B. Hedman, K. O. Hodgson, W. B. Tolman and E. I. Solomon, J. Am. Chem. Soc., 2000, 122, 11632–11648.
38. C. L. Ballhausen, in Introduction to Ligand Field Theory, McGraw-Hill Book Company, Inc., 1962, pp. 226–293.
39. A. B. P. Lever, Inorganic Electronic Spectroscopy, Elsevier Science Publishers, Amsterdam, 2nd edn, 1986.
40. W. V. Taylor, B. K. Cashman, Z.-L. Xie, K. K. Ngo and M. J. Rose, Inorg. Chem., 2022, 61, 6733–6741.
41. F. A. Cotton, D. M. L. Goodgame and M. Goodgame, J. Am. Chem. Soc., 1961, 83, 4690–4699.
42. C. M. Legendre, D. Lüert, R. Herbst-Irmer and D. Stalke, Dalton Trans., 2021, 50, 16810–16818.
43. Y. Nishida, N. Oishi and S. Kida, Inorg. Chim. Acta, 1979, 32, 7–10.
44. S. G. McGeachin, Can. J. Chem., 1968, 46, 1903–1912.
45. G. A. Webb and C. P. Richards, J. Inorg. Nucl. Chem., 1976, 38, 165–167.
46. C. P. Richards and G. A. Webb, J. Inorg. Nucl. Chem., 1969, 31, 3459–3464.
47. D. M. L. Goodgame, M. Goodgame and F. A. Cotton, J. Am. Chem. Soc., 1961, 83, 4161–4167.
48. F. A. Cotton, O. D. Faut and D. M. L. Goodgame, J. Am. Chem. Soc., 1961, 83, 344–351.
49. R. H. Holm and F. A. Cotton, J. Chem. Phys., 1959, 31, 788–792.
50. R. H. Holm and F. A. Cotton, J. Chem. Phys., 1960, 32, 1168–1172.
51. R. Boča, Coord. Chem. Rev., 2004, 248, 757–815.
52. D. Kowalkowska-Zedler, N. Nedelko, K. Kazimierczuk, P. Aleshkevych, R. Łyszczek, A. Ślawska-Waniewska and A. Pladzyk, RSC Adv., 2020, 10, 29100–29108.
53. M. Böhme, S. Ziegenbalg, A. Aliabadi, A. Schnegg, H. Görls and W. Plass, Dalton Trans., 2018, 47, 10861–10873.
54. Y.-Q. Zhai, Y.-F. Deng and Y.-Z. Zheng, Dalton Trans., 2018, 47, 8874–8878.
55. M. R. Saber and K. R. Dunbar, Chem. Commun., 2014, 50, 12266–12269.
56. L. Devkota, D. J. SantaLucia, A. M. Wheaton, A. J. Pienkos, S. V. Lindeman, J. Krzystek, M. Ozerov, J. F. Berry, J. Telser and A. T. Fiedler, Inorg. Chem., 2023, 62, 5984–6002.
57. D. Lomjanský, C. Rajnák, J. Titiš, J. Moncoľ, L. Smolko and R. Boča, Inorg. Chim. Acta, 2018, 483, 352–358.
58. D. M. Price, Thermochim. Acta, 2001, 367–368, 253–262.
59. B. S. Lim, A. Rahtu, J.-S. Park and R. G. Gordon, Inorg. Chem., 2003, 42, 7951–7958.
60. R. T. Azuah, L. R. Kneller, Y. Qiu, P. L. Tregenna-Piggott, C. M. Brown, J. R. Copley and R. M. Dimeo, J. Res. Natl. Inst. Stand. Technol., 2009, 114, 341–358.
61. (a) CCDC 2348010: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt98y; (b) CCDC 2348011: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt99z; (c) CCDC 2348012: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt9b0; (d) CCDC 2348013: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt9c1; (e) CCDC 2348014: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt9d2; (f) CCDC 2348015: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jt9f3; (g) CCDC 2348314: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2jtm22.

---