Co(II) Amide, Pyrrolate, and Aminopyridinate Complexes: Assessment of their Manifold Structural Chemistry and Thermal Properties

A series of cobalt(II) (silyl)amides, pyrrolates and aminopyridinates were synthesized. Inspired by the dimeric bis(trimethylsilylamido)cobalt(II) complex ([Co(TMSA)2]2), facile salt metathesis employing the ligand 2,2,5,5-tetramethyl-1,2,5-azadisilolidinyl (TMADS) yielded its congener [Co(TMADS)2]2. Novel, heteroleptic Lewis adducts of the former resulted in unusual three- to four-fold coordination geometry around the metal center. Similarily, the salt [Co(TMADS)3Li(DAD)2] was isolated which demonstrates an ion separated Co(II) anion with silylamide ligation and Li+ counter cation. Transpyrrolylation using [Co(TMSA)2]2 was established for the synthesis of bis[N,N’-2-(dimethylaminomethyl)pyrrolyl]cobalt(II), and bis(N-2-(tert-butyliminomethyl)pyrrolyl)cobalt(II). Treatment of CoCl2 with two equivalents of lithiated N,N-dimethyl(N’-tert-butyl)ethane-1-amino-2-amide and N,N-dimethyl(N’-trimethylsilyl)ethane-1-amino-2-amide resulted in the respective Co(II) amido-amines. Reaction of CoCl2 with lithium 4-methyl-N-(trimethylsilyl)pyridine-2-amide yielded the first binuclear, homoleptic Co(II) aminopyridinate complex with a distorted trigonal bipyramidal coordination environment (τ5 = 0.533) for one central Co(II) ion and a weakly distorted tetrahedral coordination geometry (τ4 = 0.845) for the other. All of the new compounds were thoroughly characterized in terms of composition and structure. Finally, the key thermal characteristics of volatility and thermal stability were assessed using a combination of thermogravimetric analysis and complementary bulk sublimation experiments.


Introduction
Ever since Alfred Werner coined the term coordination complex in 1893, [1] and was able to exemplify his far-reaching theory using the example of hexol in 1911, [2] the coordination chemistry of the later transition metal cobalt (Co) expanded vastly and has been subject of extensive research. Among the many compound classes known to date, there are Co(II) (silyl)amides, pyrrolates and aminopyridinates. With our long-term research focusing on the synthesis of inorganic complexes and their evaluation as precursors for the application in chemical gas phase deposition processes such as metal organic chemical vapor deposition (MOCVD) and atomic layer deposition (ALD), [3,4] and our recent interest in the chemistry of the later transition metal Co, [5] we identified prior mentioned classes as promising model systems for our studies. Even though operational modes and conditions of MOCVD and ALD processes differ, their complementary and precise applicability for the deposition of a broad range of thin film materials have made them industrially indispensable in fields like microelectronics, optoelectronics and photovoltaics. [6] Choosing the right precursor, namely a compound that is volatilized during the process to serve as chemical source for the target material, for a specific application is of crucial importance hereby: The precursor needs to meet demanding requirements such as high thermal stability, sufficient volatility, high but controllable reactivity and lastly non-etching behavior. [3,4,7] Reviewing the list of commonly used Co precursors, it becomes apparent that all-nitrogen or predominantly nitrogen coordinated compounds are widely unexplored in terms of their general suitability as precursors and in terms of actual CVD/ALD process development. This is surprising as the two allnitrogen coordinated precursors Co(II) bis-(di-tert-butyl-acetamidinate), [8] and Co(II) bis-(ditert-butyl-diazadienyl) (Co(DAD)2) [9] have led to significant advancements in ALD and material science. While the likely best known Co(II) silylamide [Co(N(SiMe3)2)2]2 in the following referred to as [Co(TMSA)2]2 has not found application in this field yet, the recent usage of its Lewis base coordinated congener [Co(TMSA)2(THF)] in ALD [10] raises the question if all-nitrogen Lewis adducts of the type [Co(TMSA)2L] could be viable precursor candidates. In this regard, it is noteworthy, that a number of Co(II) silylamides and Lewis adducts adapted from them have already been studied in recent years in terms of their structural, electronic and magnetic properties. [11][12][13][14][15][16][17] Beginning with [Co(TMSA)2]2, Power and coworkers hereby focused on the synthesis and bonding characterization of low-coordinate Co(II) silylamides of the type [Co(TMSA)2L], [11] and [Co(N(SiMe3)Dipp)L], [15] with Dipp being 2,6diisopropylphenyl and L being a series of Lewis bases ranging from THF and pyridine to several phosphines derivates. Complementing these efforts, Rabu and coworkers, [16] as well as Layfield and coworkers, [12]  Thus it is consequential that an assessment of the thermal properties of these types of compounds was not a main objective until now. This is likewise applicable to Co(II) pyrrolate and aminopyridinate complexes of which a few examples have been reported but never been compared in terms of structure and thermal properties. [18][19][20][21][22] Aiming to broaden the knowledge on selected Co(II) (silyl)amides and some of their Lewis adducts as well as Co(II) pyrrolates and aminopyridinates, this study provides a detailed survey over and comparison of the manifold structural chemistry of in total eleven complexes. First, utilizing the silylamide ligand 2,2,5,5-tetramethyl-1,2,5-azadisilolidinyl, a close structural relative to [Co(TMSA)2]2 is presented. Novel Lewis adducts of the latter with N,N'-di-tertbutylethane-1,2-diimine, N,N-dimethylpyridine-2-amine and N-tert-butyl-1-(pyridine-2yl)methaneimine are reported that broaden the number of all-nitrogen low-coordinate Co(II) silylamides. Furthermore, a straightforward and simplified protocol for the synthesis of two related Co(II) pyrrolates by transpyrrolylation of [Co(TMSA)2]2 is described. Hereby, the title compounds, chelated by the N,N'-2-(dimethylaminomethyl)pyrrolyl, [19] and N-2-(tertbutyliminomethyl)pyrrolyl [21] ligand, have been reported prior but not yet been compared in terms of structure and thermal properties. Sharing a chelating nature, the two amido-amino ligands N,N-dimethyl(N'-tert-butyl)ethane-1-amino-2-amide and N,N-dimethyl(N'trimethylsilyl)ethane-1-amino-2-amide are moreover employed to synthesize novel 4-fold coordinated Co(II) amide complexes with distorted tetrahedral geometry. For one of them, the complete absence of silicon in the ligand sphere renders a rare example and has not been reported for amido-amino ligands of comparable size to the best of our knowledge. Lastly, a rare example of a homoleptic, binuclear Co(II) aminopyridinate complex with a four-fold and a five-fold coordinated Co(II) atom surrounded by four 4-methyl-N-(trimethylsilyl)pyridine-2amido ligands is described. The synthesis and solid state structures, obtained from single crystal X-ray diffractometry (SC-XRD), of in total eight of the title compounds are reported for the first time. Owing to their respective ligand systems and binding modes, the compounds are divided into four structural categories which allows the identification of multiple similarities and differences that are found to directly affect their thermal stability and volatility in thermogravimetric analysis (TGA) and bulk sublimation experiments. Thus, the present study offers a first structure -thermal property correlation that may be built upon for future, increased exploration of Co(II) (silyl)amides and related compounds as precursors for chemical gas phase deposition.

Synthesis:
To begin our investigation, we first prepared the well-known silylamide [Co(TMSA)2]2 (1), following the procedure described by Power and co-workers (Scheme 1). [11] Additionally, following a similar protocol the 2,2,5,5-tetramethyl-1,2,5-azadisilolidinyl ligated congener [Co(TMADS)2]2 (2), was obtained by facile salt metathesis reaction of CoCl2 and two equivalents of the respective lithiated silylamide in Et2O. In both cases, removal of solvent from the dark green solutions was followed by extraction with pentane, and subsequent purification by vacuum distillation. The previously reported silylamide 1 and the hitherto unreported 2 formed dark yellow to green vapors that solidified as dark red to dark brown chunks. We were also able to obtain X-ray quality crystals of 2 which confirmed its solid-state structure (further discussed below). In both cases volatilization occurred concomitant to fractional decomposition of the crude products. The resolidified distillates were then recrystallized from pentane to yield the two title compounds in yields of ca. 75% (1) and ca. 35% (2); the lower yield is likely due to increased thermal decomposition during distillation. The effective magnetic moment (μeff in Bohr magneton μB) of the dimeric cobalt compounds 1 and 2 was estimated using the Evans' method ( Table 1). [23] This resulted in values of 5.56 μB and 5.32 μB, respectively. Following the preparation of the dimeric compounds, we speculated that we could break the μ-amido bridging bonding motif via addition of ligands that would coordinatively saturate the cobalt center to yield monomeric complexes. Reaction of dimeric compound 1 at room temperature with a series of nitrogen substituted Lewis bases, namely N,N'-di-tert-butylethane-1,2-diimine (DAD), N,N-dimethylpyridine-2amine (DMAPY), 2,2-bipyridine (BPY), and N-tert-butyl-1-(pyridine-2-yl)methaneimine (IMPY) yielded the adducts (3 -6) that were crystallized from hydrocarbon solutions as highly colored solids (Scheme 2). It should be noted that the synthesis and crystal structure of the BPY adduct 5 has been previously reported, [17] yet information on its thermal properties were not provided. Prior to single-crystal X-ray diffraction (SC-XRD) analysis, we hypothesized that  [11,14,16] Exemplarily, the in situ procedure was tested for the reaction of 2 and DAD.   [13] containing the [Co(TMSA)3]anion and crown ether ligated Li + cation exhibited an even higher µeff value of 5.25 μB at room temperature for solid samples.
These high values can generally be ascribed to the emergence of magnetic anisotropy within Co(II) complexes and salts.
Pyrrolate type ligands such as N,N-dimethyl-1-(pyrrol-2-yl)methaneamine (AMPR) and N-tertbutyl-1-(pyrrol-2-yl)methaneimine (IMPR) have found use for in variety of transition metal complexes. [26] For the two Co(II) pyrrolate complexes Co(AMPR)2 (8) and Co(IMPR)2 (9) Drevs [19] and Holm [22] had originally described elaborate and time-consuming multistep synthesis procedures that we found to be avoidable by a direct transpyrrolylation approach employing 1 and the respective protonated pyrrole derivate in Et2O under mild refluxing conditions for several hours. This preparation method allowed to isolate the crude products nearly quantitatively with subsequent crystallization (8) or sublimation (9) giving the purified title compounds in yields of ~ 75 % -80 % as brown-red to red solids (Scheme 4). In a next step, we aimed to synthesize Co(II) complexes bearing chelating amido-amine ligands and decided to introduce the two closely related ligands N,N-dimethyl(N'-tert-butyl)ethane-1,2-diamine (H-TBUAEDMA) [27] and N,N-dimethyl(N'-trimethylsilyl)ethane-1,2-diamine (H-TMSAEDMA). [28] Employing a standard salt metathesis route, Co(TBUAEDMA)2 (10) and Co(TMSAEDMA)2 (11) were obtained as black and purple solids from 1:2 reaction of CoCl2 with the respective lithiated amido-amine in refluxing Et2O. Subsequent removal of solvent from the intensely colorized solution was followed by extraction of the crude product from pentane, filtration and again solvent removal (Scheme 4). Noteworthily, the high solubility of both title compounds in ethers and hydrocarbons was found to impede crude product purification, so vacuum sublimation was attempted instead. Here, despite the strong congruity of the ligand systems, a vast difference between 10 and 11 not in terms of volatility (for both 80 °C at 1 x 10 -1 mbar was applied) but regarding thermal stability became apparent for the first time. While the trimethylsilyl side chain substituted amide 11 was repeatedly obtained in good yields of ~ 80 % -85 %, the yields amounted only to ~ 30 % -35 % for 9 possessing the tertbutyl side chain substituted amides. This can be seen as a further illustration of the superior stabilization capability of β-silylamides in direct comparison to β-carboamides. [29] Evans' method forwarded solution magnetic moments of 4.85 μB (10) and 4.64 μB (11). With the intention to synthesize a monomeric, homoleptic Co(II) aminopyridinate complex, lithiated 4methyl-N-(trimethylsilyl)pyridine-2-amide (Li-TMSMAPY) was prepared according to a literature reported procedure, [30] and reacted with CoCl2 (2:1 ratio) in boiling Et2O for 12 h (Scheme 5).
After solvent removal, re-dissolution in pentane and filtration, a clear dark green solution was obtained from which SC-XRD quality crystals grew upon storage at -30 °C overnight. A second batch of crystalline material was obtained from the concentrated mother liquor upon cooling (overall yield 75 %). X-ray analysis revealed the dimeric nature of the title compound and displayed rare and deviating coordination environments for the two Co(II) ions. A more detailed description is provided in the subsequent section. Considering the dimeric nature of 12 [Co(TMSMAPY)2]2 in the solid state and hypothesizing 12 to be a dimer in solution as well, the effective magnetic moment was estimated to be 5.21 μB. It is thus similar to the magnetic moments determined for the dimeric 1 and 2 despite the fact that the coordination environment for the Co(II) ions differs notably.

Structural Analysis:
All complexes for which solid state structures are presented were found to either crystallize in monoclinic or orthorhombic space groups as it has exemplarily been observed for several of Lewis adducts of the general type [Co(TMSA)2L]. [11,14] Further information on the crystallographic data and data acquisition parameters are displayed in Table 2. For selected compounds, namely 4 and 10 -12, additional comments on the disorders found in the measured crystals can be found in the SI (section 4). Counterfeiting the prior reported [Co(TMSA)2]2 (1), [Co(TMADS)2]2 (2), whose solid state structure is illustrated in Figure 1, is arranged as a dimer. It exhibits two three-coordinate Co (II) ions that are bridged through two amido ligands while additionally being bonded to a terminal amido ligand each. With 2.6133(5) Å, the Co···Co interatomic distance between both ions is slightly elongated in comparison to 1 where this distance was found to be 2.5864 (5) Å. Yet, the bridging and terminal Co−N binding lengths were slightly shorter with 1.9957(8) Å and 1.8907(7) Å respectively ( Table 3). Consideration of the Co2N2 core structures revealed a pronounced difference between 1 and 2: Contrasting the perfectly plain arrangement of the rhomb spanned between Co1, N2, Co1' and N2' in 1, the nitrogen atoms N2 and N2' in closely related 2 vaulted upwards of the plane that can be defined between Co1 and Co1'. Accordingly, a notable dihedral angle . [31] In opposition to this, the Lewis base was found to bind with only one of the imine nitrogen atoms in a monodentate fashion, while the other imine moiety was averted from the central ion (Figure 2). Table 3: Selected Interatomic Distances and Angels of Complexes 1, [11] and 2.

[Co(TMSA)2]2 2 [Co(TMADS)2]2
Bond Lengths / Interatomic Distances (Å) Similarly, the dimethylamide moiety of the DMAPY ligand in complex 4,  Table 3). The monodentate imine nitrogen of the Lewis base in 3 exhibited a distance of 2.0892(7) Å to the Co(II) ion, which is only slightly longer than the bridging Co−N bonds found in 1 that complete the trigonal square planar coordination. Contrasting this the  [Co(TMSA)2(IMPY)] (6) drawn with thermal ellipsoids at 50 % probability level. Hydrogens are omitted for clarity. The structure of 5 is reproduced from the data provided by Müller and co-workers. [17] It is a consequence of a preferential tilt of the DMAPY ligand towards one of the TMSA ligands.
Somewhat different angles of this kind for [Co(TMSA)2L] type Lewis adducts have been observed prior. [11] Just as its prior reported congener [Co(TMSA)2(BPY)] (5), [Co(TMSA)2(IMPY)] (6) was found to feature a four-fold, distorted tetrahedral coordination environment around the central ion. For both compounds, the increased spatial demand of the Lewis base was found to elongate the Co−N bond lengths between the Co(II) ion and the TMSA ligands. to the Co(II) -pyridine nitrogen distance in 6 (Co1−N4; 2.1248(21) Å). The distance between the ion and the imine nitrogen amounts to 2.1425(21) thus being only slightly longer. In comparison to 3 however, the bond is notably longer. The distorted tetrahedral coordination environment in 5 and 6 is reflected by τ4 [32] and τ4' [33] values of 0.802 and 0.801 for the BPIY adduct and 0.783 as well as 0.762 for the IMPY adduct, respectively. The latter is thus slightly more distorted which may be attributed to the influence of the spatially demanding tert-butyl group. When one plane is defined between the Co(II) ion and the TMSA nitrogen atoms on the one hand and another plane created between the central atom and the Lewis base nitrogen atoms, an angle between both planes can be measured that reflects the twist of the Lewis base towards the Co(TMSA)2 fragment. For 3 -6, this angle was found to be the least for 4 (42.95 °) and the largest for 6 (77.01 °), thus directly correlating to the spatial demand of the respective Lewis base ( Table 4).
Another aspect in which the four Lewis adducts were compared are their solid state packing and short contacts, whereby latter are defined as distances between individual atoms of different molecules smaller than the sum of their van der Waals radii. Thus, they may indicate weak   Table 6).

[Co(TMADS)3Li(DAD)2]
Bond Lengths / Interatomic Distances (Å) Σ 0 for Co1 (°) τ4 [32] / τ4' [33] for Li1 359.97 (7)  it can be seen from Table 6, not only the bonding distances except from the above discussed ones but also the bond angles in 8 and 9 were found to be very similar which masks to some extend the pronounced structural difference between the two complexes: Lacking a conjugated This disorder may be referred to as fractional superposition, was more present in 10 than in 11 and a structural solution could be found by description of disordered ligands with two parts. and Co(TMSAEDMA)2 (11) drawn with thermal ellipsoids at 50 % probability level. Hydrogens are omitted for clarity. The structure of 9 is reproduced from the data provided by Wei. [21] For the structure of 10, the superposition of partially disordered ligands is omitted for clarity.
In the subsequent discussion and in Table 6 bond lengths and angles for only one part are considered for clarity. The Co1−N1 amide bond lengths were found to be 1.9165(113) Å and 1.9421(33) Å, respectively (Figure 4). Interestingly, these bond lengths are shorter than those seen for TMS-amide bond lengths in the heteroleptic, tetrahedral complexes 5 and 6 and rather comparable to TMS-amide bond lengths found in complexes, 1 -4, which exhibit tricoordinated Co(II) ions. The Co1−N2 bonds describing the distance between the Co(II) ions and the chelating dimethyl-amines amounted to 2.1767(114) Å (10) and 2.1478(35) Å (11) which is longer than the respective distance found in 7. τ4 [32] / τ4' [33] 0.708 / 0. , in which the TMSMAPY ligands adopted the expected chelating binding mode. [20] In a prior study Glatz and Kempe had introduced the related N 2 ,N 6bis(trimethylsilyl)pyridine-2,6-diamide ligand (BTMSAPY) for the complexation of Co(II). [34] They obtained a binuclear, symmetrical complex, Co2(BTMSAPY)3, in which both Co(II) ions were surrounded by a slightly distorted tetrahedral coordination sphere. Hereby, one of three BTMSAPY ligands adopted a μ2-coordination utilizing both side chain amide-nitrogen and the pyridine-nitrogen atoms, while the other two were μ2-coordinating with only one side chain amide-nitrogen and pyridine-nitrogen atom. Reminiscence of these structural features is essential when considering our homoleptic and binuclear complex [Co(TMSAMAPY)2]2 (12).
The solid state structure of complex 12 is depicted in Figure 5 whereby the pyridine ring carbon atoms thermal ellipsoids have been omitted for clarity and only their connectivity is shown.  bond belonging to the μ2-coordinating TMSMAPY ligand was found to be remarkably short and in the same range as amido-nitrogen Co−N bonds.  Table 7, the τ5 value for central ion Co1 was calculated to be 0.533 while the τ4 and τ4' values for ion Co2 were determined to be 0.845 and 0.842. Interestingly, an only minor distorted tetrahedral coordination environment was established around the latter Co(II) ion, potentially on the expense of the coordination environment of the former that found itself right between a square pyramidal and a trigonal bipyramidal coordination geometry. τ5 [35] for Co1 τ4 [32] / τ4' [33] (or Co2 Even though the exact mechanism of evaporation for neither 1 nor 2 is known, it appeared manifest that the TMADS ligand in 2 impeded volatilization until a notable fraction of the analyte decomposed. Contrasting this, 1 apparently undergoes a conversion into a volatile and thermally stable product, presumably the monomer, upon heating. and co-workers [10] still evidences the complexation approach to be viable for the synthesis of enhanced Co(II) ALD precursors. The most pronounced contrast in terms of thermal stability and volatility was observed for the two pyrrolato complexes 8 and 9 as it can be seen in Figure   6 c. While the latter demonstrated clean one-step evaporation with a residual mass of only 5.5 % and an onset temperature of around 203 °C, compound 8 with its non-π-conjugated pyrrole dimethylaminoethyl side chain barely volatilized intactly. At this stage, the mass loss was fully ascribed to precursor decomposition and release of organic, volatile decomposition products.
Bulk sublimation experiments, described in the subsequent section should confirm this initial assessment. Lastly, as illustrated by Figure 6 d, amido-amine compound 10 with its tert-butyl amido ligand moiety outperformed the trimethylsilyl derivate 11 in terms of volatilization onset (140 °C vs. 169 °C) but showed inferior thermal stability evidenced by a notably higher residual mass (21.6 % vs. 4.7 %). Dimeric compound 12 was subjected to TGA as well and the experimental data can be found in the SI. Unsurprisingly, the mass loss was found to be dominated by a multitude of overlapping decomposition events with no evidence for partial intact precursor evaporation which rendered 12 to be the least suitable CVD/ALD precursor candidate.
While TGA is a most valuable tool to initially assess the performance of a precursor candidate, it can be deceptive as the evaporation conditions typically differ from evaporation parameters applied to precursors in CVD and ALD processes. More precisely, TGA only probes a few milligrams of an analyte at atmospheric pressure and not multi-gram quantities under roughing pump vacuum conditions. Additionally, the adjustment of carrier gas flows can have a strong impact on the observed volatilization behavior. Consequently, we probed compounds 1 -12 under more realistic process conditions: 300 mg of each compound were subjected to sublimation at around 1 x 10 -1 mbar at the lowest temperature at which sublimation was observed. The recovered substances were first weighted and then subjected to elemental analysis (EA) to assess their purity compared to non-sublimed reference samples. Table 8 summarizes the respective sublimation temperatures Ts and the bulk sublimation recovery yields alongside the residual masses from TGA and melting points TM obtained from melting point measurements. Expectedly, compounds 1, 9 and 11 which had already demonstrated the best performance in TGA were recovered with high yields ranging from 78 % (9) up to over 90 % (1, 11), whereby the recovered substances EA did not differ from the reference samples (difference in C, H and N content in % smaller than 0.3 %). Quantities of 300 mg of the compounds were filled into Schlenk flasks with attached sublimation fingers and sublimed at given TS in vacuum (1 x 10 -1 mbar). b Determination not possible prior to decomposition of the sample.
Likewise, recovered material from compounds 3 and 4 was found to be pure by EA. However, the recovery rates were already significantly lower with around 50 % and 60 % and decomposed material was observed at the bottom of the sublimation set-up. Compounds 2, 6 and 10 were sublimed with low recovery yields and only for 10 the conducted EA was found to not significantly differ from the reference. Samples of 5, 8 and 12 could not be recovered by sublimation as decomposition occurred upon careful and stepwise heat-up to a maximum sublimation temperature of 120 °C. In summary, the suitability of candidates 1, 9 and 11 for CVD as well as ALD applications in terms of their fundamental thermal properties was validated while the sublimation experiments allowed to identify 3 and 4 more clearly as potential CVD precursors. A significant advantage of candidate 9 over all other competitors of the investigated compound series is the absence of silicon in the ligand sphere which render it more appealing for process implementation as e.g. the contamination of metallic cobalt thin films by silicon is unwanted in semiconductor industry. [36] Conclusions: Conclusively, a series of Co(II) (silyl)amides and Lewis adducts thereof as well as Co(II) pyrrolates and aminopyridinates was synthesized.  Utilizing the reagents above, the ligands listed in the following have been synthesized according to literature reported procedures. Prior to use, their purity was assessed by 1 H-NMR spectroscopy: N,N'-di-tert-butylethane-1,2-diimine (DAD), [37] N-tert-butyl-1-(pyridine-2yl)methaneimine (IMPY), [38] 4-methyl-N-(trimethylsilyl)pyridine-2-amine (H-TMSAPY), [30] N,N-dimethyl-1-(1H-pyrrol-2-yl)methaneamine (H-AMPR), [39] N-tert-butyl-1-(1H-pyrrol-2yl)methaneimine (H-IMPR), [40] N,N-dimethyl(N'-tert-butyl)ethane-1,2-diamine (H-TBUAEDMA), [27] N,N-dimethyl(N'-trimethylsilyl)ethane-1,2-diamine (H-TMSAEDMA) [28] . 1

Additional Crystallographic Details and Images
Specific details on individual disorders and structure refinements: Compound 4: For this compound, the DMAPY ligand was found to possess two strongly intercontorted positions. This double-position could not be described by a space group with higher symmetry, however and was thus identified as significant distortion. Besides, both TMSA ligands were found to be systematically disordered whereby the Si(CH3)3 groups exhibited rotatory distortions. and the nitrogen atoms positional distortions. Satisfying structure solution was only achieved by assigning two positions to the central Co(II) ion as well and by describing the entire molecule by two different parts differentiating the distortions. Compound 11: The main disorder found in this compound is strongly related to the one found for compound 10. It needs to be highlighted again, that several samples were subjected to single crystal XRD analysis and that for all of them the same type of disorder was found. In analogy to the TBUAEDMA ligand, the TMSAEDMA ligand can fold itself around the Co(II) central ion in two different ways that need to be described by two parts to obtain a satisfying structure solution. The disorder was found for eight of thirteen molecules in the unit cell.