A diethylhydroxylaminate based mixed lithium/beryllium aggregate

1 Introduction

This work resulted from our continued efforts to investigate the coordination chemistry of the Be²⁺ cation [1–5] and that of hydroxylamine ligands toward different metal ions (e.g., Na, K [6], Al [7, 8], Ga [8, 9], Zn [10, 11], Cd [12], Si [13–15], Ge [16], Sn [17] and the lanthanides [18–21]). Here we have put our attention to the coordination of Be²⁺ to a hydroxylamine ligand. Although the knowledge on the coordination chemistry of many elements of the periodic table with hydroxylamine species appears to be well developed, only one hydroxylaminato beryllium compound is known [5] as yet. We report the finding of a second representative of this compound class, which, however, unexpectedly contains both lithium and beryllium, a motif that only rarely has been described in the literature (10 entries were found in the CSD in the November 2014 version [22], and eight out of these are lithium fluoroberyllates [23]).

2 Results and discussion

Anhydrous beryllium dichloride is a commercially available and a convenient starting material for the preparation of beryllium compounds. It reacts exothermically with diethyl ether to form a soluble complex (forming a two-phase mixture of a solution of the complex in ether and a solution of ether in the complex), which readily undergoes metathesis reactions with lithium compounds. We intended to utilize this scheme when attempting the following two-step, one-pot reaction:

(1)2 Et2NOH →1.nBuLi2.BeCl2 Be(ONEt2)2 + 2 LiCl↓ + 2 BuH↑ (1)

Samples of the reaction mixture showed a complex signal pattern in the ⁹Be-NMR spectrum, indicating the presence of several different beryllium containing species in ethereal solution. The formation of such a complex reaction mixture might be attributed to a fine energetical balance between several species of very similar thermodynamic stability under the chosen reaction conditions. Nevertheless, it was possible to isolate a small amount of a highly air and moisture sensitve colorless homogeneous crystalline material. A crystal specimen suitable for X-ray diffraction could be isolated, and a solution and refinement of the crystal structure was possible. Figure 1 shows a plot of the solid-state structure of this compound.

Fig. 1:

Molecular structure of 1 in the crystal (Ortep style plot from Platon [24, 25]) with atomic numbering scheme. Displacement ellipsoids are drawn at 10 % probability level. Methyl fragments of the ethyl groups and hydrogen atoms are omitted for clarity. One of the n-butyl groups is disordered, the alternative carbon atom positions are not shown here.

Because the crystals were only weakly diffracting, a result of both the absence of heavy atoms and also the thin plate morphology of the crystals, the structure refinement resulted in rather high R values (R = 13.27 % and wR(F²) = 41.90 %). Hence, we discuss the structure on a semi-quantitative basis in the following. The molecular composition of the compound is Be(ONEt₂)₂ [(LiONEt₂)₂ⁿBuBeONEt₂]₂ (1); it crystallizes in the triclinic space group P1̅, where each unit cell comprises two formula units (Z = 2). Between these formula units, no other than van der Waals interactions could be detected. Each molecule of 1 shows an approximate C₂ axis of symmetry passing through the central beryllium atom Be1. Be1 shows a strongly distorted tetrahedral coordination environment of four surrounding hydroxylaminato oxygen atoms with all four Be–O bond lengths above 1.60 Å, and O–Be–O angles ranging from 102° to 124° for O11–Be1–O41 and O31–Be1–O41, respectively. Thus, all these Be–O distances are significantly longer than the Be–O distances of 1.493(5) to 1.600(5) Å in the dimeric beryllium-bis(isopropyl)hyxdroxylamine (2) [5]). The distortion of the BeO₄ tetrahedron can be described as a partial planarization of the tetrahedron under approximate S4 symmetry. From this central partially flattened BeO₄ tetrahedron two opposing ladder type structures emerge that are based on “steps” of metal–oxygen subunits. Each of the two ladders is composed of four such metal–oxygen steps: Be–O, O–Li, Li–O and O–Be. The first contains the central beryllium atom connecting the two ladders, and the terminal ones contain the Be atoms bound to the n-butyl groups. The alkyl substitution of the terminal Be atoms Be2 and Be3 leads to a formal residual valency of 1. In this way the three beryllium atoms in this hetero bimetallic aggregate can realize a ladder motif, which is well known from lithium chemistry [26]. Each hydroxylaminato ligand has the nitrogen lone pair as a second potential donor site. This is utilized to coordinate lithium atoms, except the two outermost hydroxylaminato ligands. The nitrogen atoms of the two inner ligands, N12 and N22, are coordinating the lithium atoms Li61 and Li81, thereby forming five-membered ring systems, a structural motif that also appears in 2 [5]. However, for the five-membered rings in 1, a significant deviation from planarity is observed (the sum of internal angles is 527°/526° vs. 540° for an ideally planar pentagon as found in 2). The nitrogen atoms of the remaining four hydroxylaminato ligands are coordinating in the more common η² mode, forming four three-membered rings with the metal atoms, which are also coordinated by oxygen atoms of the same ligand. Aparently the distortion from a regular ladder-type structure is mainly influenced by the additional coordination of the nitrogen atoms. Note that all metal cations are coordinated by oxygen atoms, but only lithium cations are coordinated additionally by nitrogen. This can be interpreted as a preference of beryllium for the harder Lewis bases as compared to the relatively softer lithium cation.

The n-butyl groups bound to Be2 and Be3 are positioned equatorially relative to the aggregate and show a straight chain conformation. One of them (attached to Be2) shows a disorder over two carbon atom positions, which was resolved in the structure model.

The assignment of beryllium and lithium atoms was checked by various permutations in the crystal structure model, which all led to significantly worse fits or even unstable refinements. The assignment given here is further confirmed by a DFT [RI-DFT(BP86)/SV(P)] structure optimization [27–30] yielding a minimum structure in very close agreement with the molecluar structure determined by X-ray diffraction in the solid state. This indicates that intermolecular forces are only weakly disturbing the molecular structure in the solid state. A possible explanation of this could be the number of alkyl groups in the periphery, partially shielding the molecule from intermolecular interactions.

3 Experimental