Multinuclear copper(I) guanidinate complexes

Article abstract of DOI:10.1021/ic201602m

A series of multinuclear Copper(I) guanidinate complexes have been synthesized in a succession of reactions between CuCl and the lithium guanidinate systems Li{L} (L = Me₂NC(ⁱPrN)₂ (1a), Me₂NC(CyN)₂ (1b), Me₂NC( ^tBuN)₂(1c), and Me₂NC(DipN)₂ (2d) (ⁱPr = iso-propyl, Cy = cyclohexyl, ^tBu = tert-butyl, and Dip = 2,6-disopropylphenyl) made in situ, and structurally characterized. The di-copper guanidinates systems with the general formula [Cu₂{L} ₂] (L = {Me₂NC(ⁱPrN)₂} (2a), {Me₂NC(CyN)₂} (2b), and {Me₂NC(DipN) ₂} (2d) differed significantly from related amidinate complexes because of a large torsion of the dimer ring, which in turn is a result of transannular repulsion between adjacent guanidinate substituents. Attempts to synthesis the tert-butyl derivative [Cu₂{Me₂NC( ^tBuN)₂}₂] result in the separate formation and isolation of the tri-copper complexes [Cu₃{Me₂NC( ^tBuN)₂}₂(μ-NMe₂)] (3c) and [Cu₃{Me₂NC(^tBuN)₂} ₂(μ-Cl)] (4c), both of which have been unambiguously characterized by single crystal X-ray diffraction. Closer inspection of the solution state behavior of the lithium salt 1c reveals a previously unobserved equilibrium between 1c and its starting materials, LiNMe₂ and N,N′-di-tert-butyl-carbodiimide, for which activation enthalpy and entropy values of δH^? = 48.2 ± 18 kJ mol^-1 and δS^? = 70.6 ± 6 J/K mol have been calculated using 1D-EXSY NMR spectroscopy to establish temperature dependent rates of exchange between the species in solution. The molecular structures of the lithium complexes 1c and 1d have also been determined and shown to form tetrameric and dimeric complexes respectively held together by Li-N and agostic Li...H-C interactions. The thermal chemistry of the copper complexes have also been assessed by thermogravimetric analysis.

Full text of DOI:10.1021/ic201602m

Article
pubs.acs.org/IC
Multinuclear Copper(I) Guanidinate Complexes
Alexander M. Willcocks,† Thomas P. Robinson,† Christopher Roche,† Thomas Pugh,†
Stephen P. Richards,†,‡ Andrew J. Kingsley,‡ John P. Lowe,† and Andrew L. Johnson*,†
†Department of Chemistry, University of Bath, Bath, BA2 7AY, United Kingdom
‡SAFC-Hitech, Power Road, Bromborough, Wirral, CH62 3QF, United Kingdom
S
* Supporting Information
ABSTRACT: A series of multinuclear Copper(I) guanidinate
complexes have been synthesized in a succession of reactions
between CuCl and the lithium guanidinate systems Li{L} (L =
Me2NC(iPrN)2 (1a), Me2NC(CyN)2 (1b), Me2NC(tBuN)2 (1c),
and Me2NC(DipN)2 (2d) (iPr = iso-propyl, Cy =
t
cyclohexyl, Bu = tert-butyl, and Dip = 2,6-disopropylphenyl)
made in situ, and structurally characterized. The di-copper guani-
dinates systems with the general formula [Cu2{L}2] (L = {Me2NC-
(iPrN)2} (2a), {Me2NC(CyN)2} (2b), and {Me2NC(DipN)2} (2d) differed significantly from related amidinate complexes because of a
large torsion of the dimer ring, which in turn is a result of transannular repulsion between adjacent guanidinate substituents. Attempts
to synthesis the tert-butyl derivative [Cu2{Me2NC(tBuN)2}2] result in the separate formation and isolation of the tri-copper complexes
[Cu3{Me2NC(tBuN)2}2(μ-NMe2)] (3c) and [Cu3{Me2NC(tBuN)2}2(μ-Cl)] (4c), both of which have been unambiguously
characterized by single crystal X-ray diffraction. Closer inspection of the solution state behavior of the lithium salt 1c reveals a
previously unobserved equilibrium between 1c and its starting materials, LiNMe2 and N,N′-di-tert-butyl-carbodiimide, for which
activation enthalpy and entropy values of ΔH⧧ = 48.2 18 kJ mol−1 and ΔS⧧ = 70.6 6 J/K mol have been calculated using 1D-
EXSY NMR spectroscopy to establish temperature dependent rates of exchange between the species in solution. The molecular
structures of the lithium complexes 1c and 1d have also been determined and shown to form tetrameric and dimeric complexes
respectively held together by Li−N and agostic Li···H−C interactions. The thermal chemistry of the copper complexes have also been
assessed by thermogravimetric analysis.
nitrogen atoms, and are capable of imparting a significant
degree of electronic and steric stabilization.1
INTRODUCTION
■
Anionic bridging ligands such as guanidinates, amidinates, and
triazenides (Chart 1) continue to evolve as versatile N,N′-donor
ligands in coordination chemistry largely because of the range
of derivatives available through substitution at the terminal
Amidinate, guanidinate, and triazenide ligands have been
widely used with Group 11 metals to form a range of multi-
nuclear complexes,2 with varying coordination arrangements,
which typically support Cu(I)−Cu(I), Ag(I)−Ag(I), and
Au(I)−Au(I) interactions between formally closed shell metal
centers (d10-d10).3 Such complexes have been the subject of
considerable attention for some time, specifically with respect
to the ability to control such metallophilic bonding.2h,j,4 More
recently these systems have attracted attention as potential
metal deposition precursors,2d−g,l,5 catalysts6 and as model
systems for the study of spin delocalization superexchange
pathways.2b,c
Chart 1. General Formulae of the Amidinate, Guanidinate
and Triazenide Anions, and the Resonance Forms of the
1,2,2,3-Tetra-alkyl Guanidinate Ligand
a
The guanidinate ligand systems have available to them a
zwitterionic iminium/diamide type resonance structure, arising
from the delocalization of the lone pair on the uncoordinated
nitrogen into the {CN2} component of the ligand (Chart 1).
This delocalization results in an imposed coplanarity of the
{NR2} and {CN2} components of the ligand, which can result
in significant steric repulsion between the nitrogen substituents,
and can have an effect on the projection of the two N-donor
a
(A) 1,3-diazaallyl resonance form, (B) the iminium/diamide
Received: July 26, 2011
Published: December 14, 2011
resonance form, and (C and D) the effect of π-delocalization on the
steric interactions and N-orbital projections in the ligand.
© 2011 American Chemical Society
246
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
synthesis of [{Me2NC(iPrN)2}2Cu2], 2a, (Scheme 1):2g This
involved the dropwise addition of a stoichiometric amount of
the appropriate carbodiimide to a solution of LiNMe2 in tetra-
hydrofuran (THF) to form the lithiated guanidinates (1a−d) in
situ. This solution was then added to a stoichiometric amount
of CuCl, in the absence of light. After extraction into hexane
and recrystallization at −28 °C, the desired Cu(I) guanidinate
complex was isolated and fully characterized.
Chart 2. General Formulae of the Three Distinct Classes of
Guanidinate Anion (m/n = 0 or 1)
In the case of the isopropyl substituted guanidinate system
1
atom orbitals. Rotation about the C−NR2 bond, such that the
{NR2} substituent is no longer coplanar with the {CN2} com-
ponent alleviates both steric repulsion and constriction of the
ligands bite-angle, but with the loss of delocalization over the
{CN3} backbone.
This central {CN3} core unit, derived from the biologically
important guanidine molecule, [HN=C(NH2)2],7 is present in
three distinct classes of guanidinate ligand; specifically bicyclic
guanidinate systems,8 1,1,3,3-tetra-alkyl guanidinate systems,2i,9
and 1,2,2,3-tetra-alkyl guanidinate systems,2g,5 all three of which
have been exploited in the coordination chemistry of Group 11
metals. (Chart 2).
Copper(I) amidinate dimers have been reported as potential
chemical vapor deposition (CVD) and atomic layer deposition
(ALD) precursors,2d,e and more recently Barry and co-workers
have described the synthesis, molecular structure, and gas phase
thermolysis of the copper(I) guanidinate complexes [Me2NC-
(iPrN)2Cu]2 and [iPrN(H)C(iPrN)2Cu]2.2g,5
2a, both H and 13C NMR data are consistent with that pre-
viously reported by Barry et al.2g As part of their study, a single
crystal data set was collected at 120 K. For structural compar-
ison with subsequent complexes described in this paper, the
single crystal X-ray diffraction analysis of 2a was also under-
taken. As all structures in this study where measured at a tem-
perature of 150 K we have included this data here as a direct
comparison to subsequent structures within this study.
The 1H NMR spectrum of the cyclohexyl substituted guanidinate
complex, 2b, in C6D6, shows the presence of four multiplet reso-
nances associated with the four hydrogen environments in the
{C6H11} moiety between δ = 1.33−2.60 ppm and a sharp singlet
at δ = 3.04 ppm for the {NMe2} protons. The 13C NMR spectra
also contains four resonances between δ = 25.2 and 54.5 ppm for
the four unique carbon environments in the {C6H11} moiety and
resonances at δ = 58.7 and 161.7 ppm corresponding to the
methyl amino carbon and the NCN central carbon, respectively.
Single crystal X-ray diffraction analysis reveals the molecular
structure of 2b to be a di-copper complex, which crystallizes
with a full molecule in the asymmetric unit, in the triclinic P1
space group, and is structurally similar to the isopropyl substituted
guanidinate complex previously reported. Selected bond lengths,
bond angles, and torsion angles for the di-copper complexes 2a
and 2b are shown in Table 1, and the molecular structures of the
two complexes are shown in Figure 1.
In this report, we expand this family of complexes and describe
the synthesis and characterization of a series of multinuclear cop-
per complexes bearing 1,2,2,3-tetra-alkyl/aryl guanidinate ligands
̅
i
t
of the general formula {(NR)2CNMe2} where R = Pr, Cy, Bu
and 2,6-diisopropylphenyl. These compounds were characterized
by single-crystal X-ray diffraction. In all cases complexes 1H and
13C NMR studies and microanalysis were consistent with the
formulation derived from single crystal data.
These complexes share many of the gross structural features
of previously structurally characterized di-copper amidinate,
guanidinate, and triazenide complexes.2j In each case, two metal
RESULTS AND DISCUSSION
■
centers are bridged by two guanidinate ligands in a μ,η1,η1-fashion
with geometries about the two-coordinate metal centers that
are approaching linearity [for 2a: N(1)−Cu(1)−N(4) = 174.67
(9)°; for 2b: N(1)−Cu(1)−N(4) = 170.59(11)° and N(2)−
Cu(1)−N(5) = 170.58(11)°]. The deviation from linearity is
presumably caused by variation in the internuclear separation
of the Cu(I) centers (Figure 1, A-B). In both complexes, the
metal−metal distances (see Table 1) which are comparable
to related di-copper amindinate, guanidinate, and triazenide
complexes,2j are significantly shorter than the sum of the
Synthesis and Characterization. Inspired by the work of
Gordon et al.,10−15 our initial interest was to synthezise a range
of guanidinate complexes and assess the effect of varying the
substituents on the nitrogen atoms of the {CN2} unit, on the
precursors suitability toward copper ALD and CVD.
The di-copper(I) guanidinates [{Me2NC(iPrN)2}2Cu2],
[{Me2NC(CyN)2}2Cu2], [{Me2NC(tBuN)2}2Cu2], and [{Me2-
NC(DippN)2}2Cu2] (Dip = 2,6-disopropylphenyl) were identified
as the initial target complexes, and their synthesis was attempted
following a modification of the literature procedure for the
Scheme 1
247
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
compared to the isopropyl analogue, 2a, as highlighted by the
torsion angles reported in table 1.
Table 1. Selected Geometric Data for Complexes 2a and 2b
2a
2b
Similar, but significantly smaller twisting of guanidinate
ligands has previously been attributed to delocalized π-bonding of
the exocyclic {NR2} moiety in delocalized π-bonding with the
{CN2} section of the guanidinate ligand. As noted earlier this should
result in an imposed (and approximate) coplanarity of the {NR2}
and {CN2} units in the ligand (Chart 1, structure B) as well as partial
pyramidalization of the amide nitrogen atoms.20 A close examination
of the structural parameters for complexes 2a and 2b reveals that a
change in the alkyl substituent (iPr to Cy) on the guanidinide ligand
has little overall effect on the geometry about the {N−C(NMe2)−
N} core of the ligand, with the C−N bond between the exocyclic
{NR2} group and the {CN2} unit showing no evidence of increased
multiple bond character [for 2a: C(1)−N(3) 1.406(3) Å; C(21)−
N(6) 1.384(5) Å, and for 2b: C(1)−N(3) 1.394(4) Å; C(21)−
N(6) 1.387(4) Å] or an increase in the pyramidal nature of the
amide nitrogen atoms (see Table 2) as the alkyl substituents change.
Selected Bond Lengths (Å)
1.8745(19) Cu(1)−N(1)
1.8743(19) Cu(1)−N(4)
Cu(2)−N(2)
Cu(1)−N(1)
Cu(1)−N(4)
1.878(2)
1.881(3)
1.872(2)
1.876(2)
2.4399(5)
1.343(3)
1.466(4)
1.353(4)
1.472(4)
1.394(4)
1.347(4)
1.467(4)
1.343(4)
1.469(4)
1.387(4)
Cu(2)−N(5)
Cu(1)−Cu(1A)
N(1)−C(1)
N(1)−C(2)
2.4254(6)
1.335(3)
1.484(3)
Cu(1)−Cu(2)
N(1)−C(1)
N(1)−C(2)
N(2)−C(1)
N(2)−C(8)
N(3)−C(1)
N(4)−C(21)
N(4)−C(22)
N(5)−C(21)
N(5)−C(28)
N(6)−C(21)
N(3)−C(1)
N(4)−C(21)
N(4)−C(22)
1.406(3)
1.344(3)
1.476(3)
N(6)−C(21)
1.384(5)
Selected Bond Angles (deg)
Table 2. Additional Geometric Data for Complexes 2a and 2b
N(1)−Cu(1)−N(4)
174.67(9)
N(1)−Cu(1)−N(4)
N(2)−Cu(2)−N(5)
N(1)−C(1)−N(2)
N(4)−C(21)−N(5)
170.59(11)
170.58(11)
120.2(3)
2a
2b
N(1)−C(1)−N(1A)
N(4)−C(21)−N(4A)
Cu(1)−N(1)−C(2)
120.3(3)
120.2(3)
Selected Torsion Angles (deg)
119.9(3)
N(1)−C(1)−N(3)−C(5)
N(1A)−C(1)−N(3)−C(5)
43.9
N(1)−C(1)−N(3)−C(14) 43.0
117.02(16) Cu(1)−N(1)−C(2)
Cu(2)−N(2)−C(8)
120.07(8)
120.38(8)
120.84(19)
120.3(2)
136.1 N(2)−C(1)−N(3)−C(14) 137.7
N(1)−C(1)−N(3)−C(15) 35.7
Cu(1)−N(4)−C(22)
Cu(1)−N(4)−
N(2)−C(1)−N(3)−C(15) 143.6
C(22)
N(4)−C(21)−N(6)−C(25) 51.9
N(4)−C(21)−N(6)−C(25A) 128.1 N(5)−C(21)−N(6)−C(34) 140.2
N(4)−C(21)−N(6)−C(35) 37.0
N(4)−C(21)−N(6)−C(34) 40.1
Cu(2)−N(5)−
120.3(2)
C(28)
Selected Torsion Angles (deg)
N(1)−Cu(1)−
Cu(1A)−N(1A)
18.2
15.4
16.5
N(1)−Cu(1)−
29.0
29.7
27.0
N(5)−C(21)−N(6)−C(35) 142.6
Cu(2)−N(2)
Sum of Angles (about the Atom) (deg)
N(4)−Cu(1)−
N(4)−Cu(1)−
Cu(2)−N(5)
N(1)
N(4)
357.43 N(1)
N(2)
358.56
356.52
356.78
357.51
Cu(1A)−N(4A)
N(1)−N(4)−N(4A)−
N(1)−N(4)−
358.16 N(4)
N(5)
N(1A)
N(5)−N(2)
van der Waals radii16 of Cu (1.40 Å) suggesting the possible
presence of d10-d10 metallophilic bonding. While considerably
weaker than related aurophilic interactions, closed shell
cuprophilic interactions are now well established.2j,17 While a
full discussion of the nature of these interactions is beyond the
scope of this paper,18 these interaction are considered to play
an important role in both solid state self-assembly and photo-
luminescence of polynuclear Cu(I) complexes.17b,d,19 The
Cu−N bond distances in both 2a and 2b are within the expected
ranges for terminal Cu(I)−N bonds (see Table 1). Similarly the
N−C bond lengths within the guanidinate ligands are also com-
We attribute the greater twisting of the {Cu2N4C2} plane in the
cyclohexyl complex, 2b, to the greater steric repulsion between
adjacent amide nitrogen substituents.
Using an identical procedure to that used in the formation of
2a and 2b, the reaction of N,N′-di-tert-butyl-carbodiimide with
LiNMe2 was used to firm the in situ complex lithium guanidi-
nate complex, Li[(tBuN)2CNMe2], 1c. Reaction of this reaction
mixture with CuCl in THF, followed by extraction into hexane
and recrystallization resulted in the formation of colorless
1
crystals. H NMR spectroscopy of the product showed the
parable to those observed in related systems.5 7
,
presence of four singlet resonances at δ = 1.41, 1.48, 2.41, 3.09
ppm, in an 9:9:6:3 ratio, respectively, suggesting the presence of
two different environments for the tert-butyl groups and two
different {NMe2} environments which is consistent with the for-
mation of a system of the general formula [Cu3{Me2NC(tBuN)2}2-
(NMe2)], 3c, rather than the desired di-copper system, [(Me2NC-
(tBuN)2)Cu]2.
Analysis of 3c by single crystal X-ray diffraction confirms the
complex to be a tri-copper cluster in which a central {Cu3}
triangle is bridged by two {(tBuN)2CNMe2} ligands and by one
{NMe2} unit, as shown in Figure 2. Selected bond lengths
and angles are shown in Table 3. The complex 3c crystallizes
in the monoclinic space group P21/n, with a full molecule in
the asymmetric unit cell. Successive attempts to synthesize the
desired di-copper complex using the method outlined above
resulted in the continual isolation of the tri-copper amide
Unlike structurally related copper(I) amidinate complexes,
which in the solid state have essentially planar bicyclic cores,2c,d
the central eight-membered {Cu2N4C2} ring formed by the two
{NCN}-ligands and the {Cu2} unit in 2a and 2b deviate sig-
nificantly from planarity. The root mean square (rms) deviation
from the least-squares plane {Cu2N4C2} is 0.197(1) Å for 2a,
with the atoms N(1) and N(4) (also N(1)A and N(4)A)
moving out of the plane by 0.286 Å and 0.270 Å, respectively.
The rms deviation of the {Cu2N4C2} plane for 2b is even larger
(0.335(2) Å) with the atoms N(1), N(2), N(4), and N(5)
moving out of the plane by 0.468, 0.474, 0.477, and 0.477 Å
respectively. This deformation of the eight-membered ring can
clearly be seen by viewing the complexes down the Cu−Cu axis
(Figure 1C-D). Also noticeable in Figure 1 is the increased
deformation of the {Cu2N4C2} plane in the cyclohexyl complex, 2b,
248
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
Figure 1. Diagram showing the molecular structures of the complexes 2a (A) and 2b (B): Figure 1A shows the molecular structure of one molecule
of 2a (50% probability ellipsoids). Hydrogen atoms have been omitted for clarity. Symmetry transformations used to generate equivalent atoms:
−x, y, 1/2 − z.; Figure 1B shows the molecular structure of 2b (50% probability ellipsoids). Hydrogen atoms and solvent of crystallization have been
omitted for clarity. Figures 1C and 1D show the molecular structures of the complexes viewed down the Cu−Cu axis emphasizing the deformation
of planarity in the eight membered {Cu2N4C2} core.
complex, presumably a result of the continued presence of free
{NMe2} in solution. In an attempt to eliminate the possibility
of an error in the stoichiometry of the reaction, or the presence
of free Li[NMe2], the lithium guanidinate salt, 1c, was synthe-
sized from the addition of a stoichiometric amount of N,N′-di-
tert-butyl-carbodiimide to a suspension of lithium amide, LiNMe2,
in hexane and isolated by crystallization at −5 °C (Scheme 2).
As part of our study, the molecular structure of the lithium
salt, 1c, was determined by single crystal X-ray diffraction. The
solid state structure of 1c is shown in Figure 3 and selected
bond lengths and angles in Table 4. In the solid state (and in
the absence of donor solvent) 1c exists as a tetrameric complex,
[Li4{(tBuN)2CNMe2}4], crystallizing in monoclinic space
group C2/c with half of the tetrameric unit in the asymmetric
unit cell, the second half being generated by a 2-fold rotational
axis. Within the tetramer there are two different lithium
atom coordination environments: Li(1), which is coordinated
to by two nitrogen atoms of different guanidinate ligands
[Li(1)−N(2): 1.960(2) Å, Li(1)−N(5): 1.969(2) Å, N(2)−Li(1)−
N(5): 157.83(12)°] and Li(2) which is similarly coordinated to
two nitrogen atoms of different guanidinate ligands [Li(2)−
N(1): 1.919(2) Å, Li(2)−N(4A): 1.907(2) Å, N(1)−Li(1)−
N(4A): 170.80(15) °] but with slightly closer Li−N contacts.
Each lithium atom is further supported by long agostic Li···H−C
interactions21 between the lithium centers and neighboring
intramolecular methyl groups. Li(1) has four such interactions
from about 2.19 to 2.28 Å, whereas Li(2) has four longer agostic
interactions from about 2.45 to 2.72 Å, which is presumably a
result of closer Li−N contacts.
The 1H NMR spectrum of pure and isolated 1c, in C6D6, shows
the presence of several resonances, at δ = 1.14, 1.26, 1.36, and 1.41
ppm in the “tert-butyl” region and δ = 2.57, 2.65, and 2.82 ppm in
the “NMe2” region of the spectrum, suggesting the presence of
more than one species in solution. Additionally, the relative
intensities of these resonances appear to change with relative
concentration of 1c in solution, suggestive of possible equilibrium
249
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
Scheme 3
22
1
(as expected) in the H NMR spectrum, in both C6D6 and
THF, suggesting any equilibrium between starting materials and
products lies far toward the right-hand side of the equation.
In an attempt to elucidate any equilibrium processes present,
variable temperature NMR studies of a sample of 1c in d8-THF
1
were investigated. In contrast the room temperature H NMR
spectrum of 1c in C6D6, the spectrum of 1c in d8-THF shows
only four resonances: δ = 1.09 and 2.58 ppm which correspond
to the {t-butyl} and {NMe2} groups of 1c, and resonances at
δ = 1.23 and 2.75 ppm correspond to free N,N′-di-tert-butyl-
Figure 2. Diagram showing the molecular structures of the tetrameric
lithium guanidinate complex 1c (50% probability ellipsoids). Hydro-
gen atoms on methyl groups not involved in supporting Li···H−C
interactions and methyl carbons attached to C(6), C(26), C(6A), and
C(26A) have been omitted for clarity. Symmetry transformations used
to generate equivalent atoms: −x, y, 1/2 − z.
6
carbodiimide and LiNMe2 respectively. The Li NMR spectra
shows only one resonance at δ = 1.21 ppm indicative of a rapid
exchange of lithium ions between 1c and LiNMe2. Between 320
1
and 270 K the H spectra of 1c in d8-THF shows a significant
change in the relative intensities of resonances: at 320 K the relative
intensity ratios of 1c to free N,N′-di-tert-butyl-carbodiimide is
approximately 1:0.66. A reduction in temperature to 270 K
results in an attendant change in the relative intensity ratio to
Table 3. Selected Geometric Data for 1c
Selected Bond Lengths (Å) for 1c
1
Li(1)−N(2)
Li(1)−N(5)
Li(2)−N(1)
Li(2)−N(4A)
1.960(2)
1.969(2)
1.919(2)
1.907(2)
Li(1)−C(3)
2.581(5)
2.270
1:0.33. Below 270 K the H NMR spectrum of 1c in d8-THF
Li(1)−H(3A)
Li(1)−H(3C)
Li(1)−C(23)
Li(1)−H(23A)
Li(1)−H(23C)
Li(2)−C(4)
Li(2)−H(4A)
Li(2)−H(4C)
Li(2A)-C(25)
Li(2A)−H(25A)
Li(2A)−H(25C)
shows a splitting of the resonances associated with the complex
1c into several new resonances, presumably because of a
freezing out of E and Z isomerization processes. Kinetic studies
of the exchange between N,N′-di-tert-butyl-carbodiimide and 1c
were studied using selective 1D-EXSY NMR spectroscopy follow-
ing the method of Nikonov et al.23 Spectra, run at 320, 326, 332,
and 338 K respectively, were acquired at several mixing times,
and a graph of signal intensity of the exchange peak, (normalized
to the signal intensity of the irradiated peak) versus mixing time
was found to be linear at short mixing times. Values for the rate
constants (k) were taken from the slope of the graph. An Eyring
plot of ln(k/T) versus 1/T provided an activation enthalpy
2.192
2.580(5)
2.278
C(1)−N(1)
C(1)−N(2)
C(1)−N(3)
C(21)−N(4)
C(21)−N(5)
C(21)−N(6)
1.3388(13)
1.4811(13
2.152
2.747(5)
2.531
1.3417(13)
1.3275(13)
2.449
2.835(5)
2.720
2.509
Selected Bond Angles (deg) for 1c
N(1)−C(1)−N(2)
N(4)−C(21)−N(5)
120.59(8)
124.82(9)
N(2)−Li(1)−N(5)
N(1)−Li(2)−N(4A)
170.80(15)
157.83(12)
of ΔH⧧ = 48.2
18 kJ mol−1, and an activation entropy of
6 J/K mol for the reaction (see Supporting
ΔS⧧ = 70.6
Information for more detail).
Although the insertion of carbodiimides into lithium amide
bonds has been widely studied and used extensively in syn-
thesis,1a this study is to the best of our knowledge the first
example of a reversible insertion of carbodiimides into a Li−N
bond. Computational studies by Woo and Richardson have
suggested that carbodiimide insertion into lithium amide com-
plexes proceeds with relatively modest activation energies (less
than 23 kcal mol−1), with the ability of the migrating amide
group to participate in delocalized π-bonding with the {NCN}
backbone being a significant factor in the stability of inter-
mediate species within the amide transfer.24 In related work
Chang and co-workers have shown experimentally N,N′-di-tert-
butyl-carbodiimide insertion into Al−NR2 bonds is unfavorable
and reversible, an observation that has been rationalized in
computational studies by Barry and co-workers on the basis of
steric destabilization of intermediate transition states.25
Reaction of the isolated, and pure, lithium salt 1c with a an excess
of CuCl (1:6) in THF, followed by extraction into hexane
and recrystallization at −28 °C yielded a crop of pale yellow crystals.
Scheme 2
between 1c and the starting materials (Scheme 3). Correspondingly,
the 6Li NMR spectrum shows the presence of two distinct broad
resonances at 1.07 and 1.28 ppm respectively reinforcing this
hypothesis. These observations are in direct contrast to di-
isopropyl derivative 1a, which displays only three resonances
250
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
Figure 3. Diagram showing the molecular structures of the complexes 3c (A) and 4c (B) (50% probability ellipsoids). Hydrogen atoms have been
omitted for clarity.
Table 4. Selected Geometric Data for 3c and 4c
Scheme 4
3c
4c
Selected Bond Lengths (Å)
Cu(1)−Cu(2)
Cu(2)−Cu(3)
Cu(1)−Cu(3)
Cu(1)−N(1)
Cu(1)−N(4)
Cu(2)−N(2)
Cu(3)−N(5)
Cu(2)−N(7)
Cu(3)−N(7)
C(1)−N(3)
2.7257(3)
2.7290(3)
2.6790(3)
Cu(1)−Cu(2)
Cu(2)−Cu(3)
Cu(1)−Cu(3)
2.7448(6)
2.7221(5)
2.6927(6)
1.8850(16)
1.8884(16)
1.8901(17)
1.8902(16)
2.1431(7)
2.1510(7)
1.351(2)
1.8827(15) Cu(1)−N(1)
1.8776(15) Cu(1)−N(4)
1.8912(14) Cu(2)−N(2)
1.8913(14) Cu(3)−N(5)
1.8956(15) Cu(2)−Cl(1)
1.8958(15) Cu(3)−Cl(1)
ligands and one chlorine atom to form [Cu3{Me2NC(tBuN)2}2-
(μ-Cl)] (4c) (Scheme 4) which crystallizes in the triclinic
1.392(2)
C(1)−N(3)
space group P1 and contains a full molecule in the asym-
C(21)−N(6)
1.391(2)
C(21)−N(6)
1.387(2)
̅
metric unit cell. The molecular structure of 4c is also shown
Figure 3, and selected bond lengths and angles are shown in
Table 4.
The triangular {Cu3} cores (Figure 4) of both tri-copper
complexes, 3c and 4c, have approximately isoscelean geometry
(see Table 4 for Cu−Cu bond lengths). Surprisingly in both
cases the shortest Cu−Cu distance is bridged by a guanidinate
ligand rather than a {NMe2} or {Cl} moiety, as is the case in
the structurally related trimethylsilyl-substituted guanidinate
derivative [Cu3{Me2NC(Me3SiN)2}2(μ-NMe2)].26
In both complexes the Cu−Cu edges are significantly longer
than that observed in either 2a or 2b. As with 2a and 2b, the
Cu−N distances are in the range expected for comparable
copper-amidinate and -triazenide complexes,2j (see Table 4)
The bridging guanidinate ligands of both complexes form two
twisted five-membered {Cu2N2C} rings. As an example, Figure 3
shows the skeleton framework of the complex 3c viewed down the
Cu(2)−Cu(3) and N(7)−Cu(1) vectors, illustrating the twisting
that is experienced by the {L3} ligand in both 3c and 4c. A com-
parison of the torsion angles observed in [Cu3(L3)2(μ-NMe2)];
N(1)−Cu(1)−Cu(2)−N(2) (42.63(7)°), N(4)−Cu(1)−Cu(3)−
N(5) (36.61(7)°) and [Cu3(L3)2(μ-Cl)]; N(1)−Cu(1)−Cu(2)−
N(2) (42.47(7)°), N(4)−Cu(1)−Cu(3)−N(5) (37.96(7)°),
indicate significantly more distortion within the tri-copper com-
plexes than in the dicopper systems (cf. 18.2° and 15.4° for 2a,
29.0° and 29.7° for 2b]). It is also notable that the smaller torsion
angles are associated with the short Cu(1)−Cu(3) edge in both
Selected Bond Angles (deg)
Cu(1)−Cu(2)−
58.832(8)
60.645(8)
60.532(8)
Cu(1)−Cu(2)−
59.016(14)
60.912(15)
60.072(8)
Cu(3)
Cu(3)
Cu(2)−Cu(3)−
Cu(2)−Cu(3)−
Cu(1)
Cu(1)
Cu(3)−Cu(1)−
Cu(3)−Cu(1)−
Cu(2)
Cu(2)
Cu(2)−N(7)−Cu(3)
N(1)−Cu(1)−N(4)
N(2)−Cu(1)−N(7)
N(5)−Cu(1)−N(7)
N(1)−C(1)−N(2)
N(4)−C(21)−N(5)
92.07(6)
Cu(2)−Cl(1)−Cu(3) 78.680(19)
160.55(7)
175.79(7)
171.59(6)
N(1)−Cu(1)−N(4)
N(2)−Cu(1)−Cl(1)
N(5)−Cu(1)−Cl(1)
164.36(7)
168.82(5)
166.73(5)
117.71(17)
117.95(16)
117.64(16) N(1)−C(1)−N(2)
118.54(15) N(4)−C(21)−N(5)
Selected Torsion Angles (deg)
N(1)−Cu(1)−
Cu(2)−N(2)
42.63(7)
N(1)−Cu(1)−
42.47(7)
37.96(8)
Cu(2)−N(2)
N(4)−Cu(1)−
36.61(7)
N(4)−Cu(1)−
Cu(3)−N(5)
Cu(3)−N(5)
1
The H NMR spectrum of the product revealed the presence of
three sharp singlet resonances at δ = 1.35, 1.40, and 2.38 ppm, res-
pectively in a 9:9:6 intensity ratio indicating two different magnetic
environments for the tert-butyl groups. The absence of an additional
resonance corresponding to addition {NMe2} was consistent with
the possible formation of the desired [Cu2(L3)2] complex.
However, determination of the product’s molecular structure
revealed the complex to be a new tri-copper complex, with a com-
parable central {Cu3}-triangle bridged by two {(tBuN)2CNMe2}
251
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
Figure 4. Diagram showing the core of 3c: a view along the approximate Cu(2)−Cu(3) vector (A), and down the N(7)−Cu(1) vector (B)
pictographically illustrating twisting of the guanidinate ligands. Selected angles: N(1)−Cu(1)−Cu(2)−N(3): 42.63(7)°, N(4)−Cu(1)−Cu(3)−
N(6): 36.61(7)°.
clusters. The increased twisting observed in the tert-butyl derivatives
is again attributed to the increased steric repulsion experienced
within the ligand sphere of the complexes compared to the
cyclohexyl and iso-propyl derivatives respectively.
In both 3c and 4c one unique edge of the {Cu3} cluster is
bridged by either a {NMe2} moiety [Cu(2)−N(7); 1.9856(15) Å,
Cu(3)−N(7); 1.8958(15) Å] or {Cl} atom [Cu(2)−Cl(1);
2.1431(7) Å, Cu(3)−Cl(1); 2.1510(7) Å] with bond lengths
that are comparable to related Cu(I) systems with edge bridging
amide groups27 or chlorine atoms.28 The longer bond lengths
between the Cu(2)−Cu(3) edge and the bridging units ({NMe2}
or {Cl}) are also reflected in the Cu(2)−E−Cu(3) angle [E =
N(7); 92.07(6), E = Cl(1); 78.680(19)].
reported despite the commercial availability of N,N′-di-tert-
butyl-carbodiimide.
Initial attempts to synthesize the 2,6-diisopropylphenyl
substituted guanidinate derivative 2d focused on the in situ
formation of the lithium guanidinate salt 1d in reactions that
are analogous to those used in the synthesis of 2a and 2b.
However, this reaction resulted in the formation of a highly air
and oxygen-sensitive reaction mixture which would rapidly
react with trace amounts of oxygen to form an intensely green
colored solution. In an attempt to reduce the overall reaction
time with Cu(I) salts to a minimum, the lithium precursor 1d
was synthesized and isolated prior to reaction with copper
chloride.
In the case of 3c the {NMe2} moiety is orientated approximately
perpendicular [89.25(15)°] to the plane of the tri-copper unit,
reflecting the three-center two-electron (3c-2e) bonding between
Cu(1), Cu(3) and the {NMe2} unit. This is a common feature in
tri- and tetra-copper based clusters containing edge bridging amide
groups,27,29 The range for Cu−Cu distances bridged by 3c-2e
bonds has previously been shown to be 2.702(2)−2.742(2) Å
in planar [Cu4(NR2)4] systems27 which are comparable to the
Cu(1)−Cu(3) bridged edges in both 3c and 4c.
Successive attempts to produce the di-copper complex [Cu2-
{Me2NC(tBuN)2}2] by reacting 1c with varying excess amounts
(2:1 and 3:2) of CuCl in different solvents (diethylether or
tetrahydrofuran), resulted in further isolation of the tri-copper
system 4c.
Addition of a stoichiometric amount of N,N′-bis-2,6-di-
isopropylphenyl-carbodiimide to a suspension of lithium amide,
LiNMe2, in hexane, resulted in immediate reaction and forma-
tion of a colorless solution. Filtration of the reaction mixture,
to remove insoluble material, followed by cooling to −5 °C
resulted in the formation of a crop of colorless crystals
(Scheme 5).
Scheme 5
It is our belief that the steric demands caused by the bulky
tert-butyl substituents is such that the putative dimeric complex
[{Me2NC(tBuN)2}2Cu2] is unstable, leading to an opening of
the metallo-ligand framework and the incorporation of an
additional {CuNMe2} or {CuCl} unit to form the a tri-copper
systems with reduced steric strain and twist. It is not clear why
these products should be formed over a possible tri-copper
complex [Cu3[{Me2NC(tBuN)2}3] as is the case in related
silver amidinate, guanidinate, and triazenide systems.2a,j,k
Analysis of the lithium salt by single crystal X-ray diffraction
reveals the system to be a dimeric complex that can be represented
formulaically as [Li2{Me2NC(DipN)2}2] (Dip = 2,6-diisopropyl-
phenyl), and which crystallizes in the monoclinic space group
P21/n, containing two independent whole molecules of the di-
lithium salt in the asymmetric unit cell. The two dimer molecules
differ only in the relative orientations of the 2,6-diisopropylphenyl
substituents and are equivalent, within experimental error, to each
other. Figure 5 shows the molecular structure of one of the two
dimer molecules of 1d in the asymmetric unit cell, and is taken as
representative of both molecules that are present. Selected bond
lengths and angles for 1d are reported in Table 5.
1
Variable temperature H NMR studies of both tri-copper
complexes reported here show no evidence of equilibrium
between dimeric and trimeric species or evidence of ligand
scrambling at elevated temperatures.
In this context, it is interesting to note that while cyclohexyl
and iso-propyl copper(I) amidinates complex have been studied
by Gordon and co-workers as CVD and ALD precursors,2d,e the
related tert-butyl substituted copper(I) amidinate analogues,
t
that is, {(tBu-N)2CR} (R = Me or Bu), have never been
252
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
bent coordination environment [Li(2)−N(2): 1.928(4) Å and
Li(2)−N(5): 1.923(4) Å; N(2)−Li(2)−N(5): 122.31(18)°].
Although both lithium environments are different, the Li−N
contacts lie within the range expected for lithium amidinate and
guanidinate systems.21b,30 The coordination sphere of the more
exposed Li(2) metals center is completed by three agostic inter-
actions (see Figure 5) [between 2.194 and 2.298 Å] to ligand
hydrogen atoms.
Reaction of the 1d with stoichiometric amounts of CuCl in
THF, and in the absence of light, yielded a pale yellow
solution which was highly susceptible to oxidation by trace
amounts of oxygen. Removal of the solvent under reduced
pressure followed by extraction into hot hexane, and filtration
yielded a pale yellow solution which on standing at −28 °C
produced yellow crystals. 1H NMR spectroscopy of the
product showed the presence of four doublet resonances at
δ = 0.71, 1.19, 1.29, and 1.95 ppm, one singlet resonance at
1.99 ppm, two septet resonances at 3.53 and 3.81 ppm, and a
range of multiplets between 6.96 and 7.14 ppm in the relative
intensity ratio of 3:3:3:3:3:1:1:3, suggesting a coordination
environment for the guanidinate ligand system in which each
methyl and methine group attached to a 2,6-diisopropyl
phenyl substituent is in a different magnetic environment,
consistent with the formation of the di-copper complex 2d.
A similar resonance pattern has been observed for the
related 2,6-diisopropyl phenyl substituted copper triazenide
complex.2j
Crystals of 2d suitable for single crystal X-ray diffraction have
been grown from two different solvents, hexane and toluene. In
both cases poor quality crystals have produced low quality data.
This, combined with inherent disorder with the isopropyl
groups on the ligand, has meant that to-date it has not proved
possible to find a structural model that allows us to place any
confidence on the bond lengths and angles for the complex.
However, after collecting and partially solving four data sets, we
are confident that the geometry of the heavy atoms is that
shown in Figure 6.
Figure 5. Diagram showing the molecular structures of the dimeric
lithium guanidinate complex 1d (50% probability ellipsoids). Hydro-
gen atoms not involved in supporting Li···H−C interactions and
isopropyl groups {CHMe2} attached to C(3), C(7), C(19), C(37),
C(45), and C(49) have been omitted for clarity.
Table 5. Selected Geometric Data for 1d
Selected Bond Lengths (Å) for 1d
Li(1)−N(1)
Li(1)−N(2)
Li(1)−N(4)
Li(1)−N(5)
Li(2)−N(2)
Li(2)−N(5)
Li(1)···Li(2)
1.977(3)
2.293(3)
1.979(3)
2.253(3)
1.928(4)
1.923(4)
2.452(5)
N(1)−C(1)
N(2)−C(1)
N(3)−C(1)
N(4)−C(31)
N(5)−C(31)
N(6)−C(31)
1.324(2)
1.424(2)
1.381(2)
1.325(2)
1.368(2)
1.375(2)
While little confidence can be placed on the accuracy of bond
lengths and angles in 2d, it is clear that the complex is struc-
turally comparable to both 2a and 2b, adopting a dimeric
structure in the solid state. Unlike 2a and 2b, there is no clear
evidence of twisting of the guanidinate ligand, as can be seen
from Figure 7 which shows views down the N(3)−C(1)−C(31)−
N(6) vector and the Cu(1)−Cu(2) vector respectively. It can also
be clearly seen from Figure 7 that relief from the steric repulsion
induced by the close proximity of the 2,6-diisopropylphenyl
groups occurs as a result of pyramidalization of the nitrogen atom
to which these groups are attached.
Li(2)−C(20)
Li(2)−H(20)
Li(2)−C(38)
Li(2)−H(38)
2.882(7)
2.257
Li(2)−C(39)
Li(2)−C(39B)
2.865(7)
2.298
2.809(7)
2.194
Selected Bond Angles (deg) for 1d
N(1)−Li(1)−N(4)
N(2)−Li(1)−N(5)
N(2)−Li(2)−N(5)
174.82(19)
95.77(13)
122.31(18)
N(1)−Li(1)−N(2)
64.43(10)
64.96(10)
70.42(13)
71.42(13)
N(4)−Li(2)−N(5)
Li(1)−N(2)−Li(2)
Li(1)−N(5)−Li(2)
N(1)−C(1)−N(2)
N(4)−C(31)−N(5)
116.66(14)
116.02(14)
Thermolysis. The copper guanidinate compounds 2a, 2b,
2d, 3c, and 4c were studied to determine their thermal
chemistry.
Within the molecular dimer there are two different lithium
coordination environments. Li(1) is coordinated by two chelat-
ing guanidinate ligands. While the ligand based around C(1)
possesses much more clearly localized bonding, the ligand based
around C(31) is more delocalized, as can be seen from the rela-
tive C−N bond lengths (Table 5).
The relative orientation of the system is such that there are two
short Li−N contacts [Li(1)−N(1): 1.977(3) Å and Li(1)−N(4):
1.967(3) Å] in an environment that is approaching linearity
[N(1)−Li(1)−N(4): 174.82(19)°] and two longer Li−N con-
tacts [Li(1)−N(2): 2.293(3) Å and Li(1)−N(5): 2.253(3) Å]
with a more bent geometry [N(2)−Li(1)−N(5): 95.77(13)°].
The second lithium atom, Li(2), is coordinated by the two
more distant nitrogen atoms, N(2) and N(5), in a somewhat
Thermogravimetric analysis (TGA) data, under argon, were
collected for each compound to determine their volatility
and appraise their decomposition profiles. While all five
compounds showed decomposition profiles with residues
less than 25%, and indicated on the basis of mass loss the
formation of copper metal, the decomposition TGA profiles
of the tri-copper complexes 3c and 4c, and the aromatic
containing complex 2d, suggest the compounds decompose
in a series of multistep processes over a temperature range
of 150−350 °C.
The TGAs of 2a and 2b are shown in Chart 3. Barry et al.
have previously reported the onset of volatility for 2a as 105 °C,
253
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
Chart 3. Thermogravimetric Analyses of 2a (Light Line) and
2b (Dark Line)
Figure 6. Molecular structure of 2d, showing the gross structural
features of the complex within the disordered structure. Only the
heavy atoms are shown, drawn as arbitrary sized spheres.
CONCLUSIONS
■
Reported herein are the structures of several novel Cu(I) com-
plexes containing metal−metal interactions supported by
progressively more sterically demanding guanidinate ligands.
Reactions of CuCl with lithium guanidinate complexes Li[{Me2-
NC(iPrN)2}], Li[[{Me2NC(CyN)2}], and Li[[{Me2NC-
(DipN)2}] provide the homoleptic complexes 2a, 2b, and 2d,
respectively, which have been structurally characterized. For
the complexes 2a and 2b, increasing the relative size of the
guanidinate substituent has a significant effect on the molecular
structure of the resultant dimers, specifically with respect to the
twisting of the guanidinate substituents and loss of planarity
across the dimer. In the case of 2d, steric relief is achieved by
pyramidalization of the aryl substituted nitrogen atoms of the
guanidinate ligand, a feature which has previously been ob-
served in the related triazenide system.2j
a value that is consistent with our results.2g However, in our
hands TGA mass residue levels for 2a are approximately 25%,
which is slightly lower than the value that corresponds to the
decomposition of the species to provide pure copper (27%),
and significantly higher than the residue levels of 1.8% reported
by Barry et al.
In contrast, 2b shows a considerably higher onset of mass
loss at approximately 176 °C. After the onset of volatilization,
both systems show relatively sharp and clean decomposition
profiles to give stable residues. For 2b the stable residue
reached is higher than a value that would correspond to pure
copper, 26% (cf. 20%), indicating some contamination in the
decomposition product.
Figure 7. Views of the complex 2d down the N(3)−N(6) (A) and Cu(1)−Cu(2) (B) vectors respectively, highlighting the relative planarity of the
central {Cu2N4C2} core and the lack of twisting of the guanidinate about the Cu−Cu bond.
254
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
CN-tBu), 167.6 (NCN of 1c); 6Li NMR (C6D6, 155.5 MHz)
In contrast, the reaction of the tert-butyl substituted guanidi-
nate system 1c (formed in situ) with CuCl results in the seren-
dipitous formation of a novel Cu(I) tri-metallic system, 3c.
Closer inspection of the solution state properties of 1c reveal
a previously unreported equilibrium between the lithium guani-
dinate complex 1c and free N,N′-di-tert-butyl-carbodiimide and
LiNMe2, for which activation enthalpy and activation entropy
values have been calculated using 1D-EXSY NMR spectros-
copy. The observation that 1c participates in this equilibrium
process explains why {NMe2} is available in solution for incor-
poration into the complex 3c.
δ: 1.21 (s, 1Li).
Synthesis of (1d). Complex 1d was synthesized in an analogous
fashion to 1c using N,N′-bis-2,6-diisopropylphenyl-carbodiimide
1
(0.72 g, 2.0 mmol). Yield = 0.81 g, 98%. H NMR (C6D6, 300.22
MHz) δ: 1.04 (d, 12H, J = 7 Hz, (CH3)CH(CH3)), 1.22 (d, 12H, J =
7 Hz, (CH3)CH(CH3)), 2.23 (s, 6H, N(CH3)2), 3.43 (m, 4H, (CH3)-
CH(CH3)), 6.98−7.06 (m, 6H, CHAr); 13C NMR (C6D6, 75.49 MHz)
δ: 23.4 (CH(CH3)2), 25.6 (CH(CH3)2), 28.5 (CH(CH3)2), 40.2
(N(CH3)2), 122.5 (CH), 124.2 (CH), 124.2 (CH), 141.5 (C-CH-
(Me)2), 146.8 (C-CH(Me)2), 144.3 (N-C), 167.5 (NCN). 6Li NMR
(C6D6, 155.5 MHz) δ: 2.83 (s, 1Li).
Synthesis of (2b). In a dry Schlenk, tetrahydrofuran (40 mL) was
added to lithium dimethylamide (0.77 g, 15.0 mmol) to give a pale
yellow slurry. N,N′-dicyclohexyl-carbodiimide (3.09 g, 15.0 mmol) was
then added dropwise via syringe, and the reaction mixture left to stir
for 2 h. The resulting pale yellow solution was transferred by cannula
to a Schlenk containing copper(I) chloride (1.49 g, 15.0 mmol), and
the reaction mixture allowed to stir for 18 h, after which time the
volatiles were removed under reduced pressure. Hexane (20 mL) was
added to the resultant residue and left to stir for 15 min, and was
subsequently removed under reduced pressure. This process was
repeated twice to remove any residual tetrahydrofuran. The product
was then extracted into hexane and filtered through Celite, and the
resultant filtrate was concentrated. Storage of the solution at −28 °C
facilitated the growth of colorless crystals suitable for single crystal X-
ray diffraction (2.90 g, 62%). Mp 152 °C. Elemental analysis: calcd. for
C30H58Cu2N6 (%):C, 57.39, H: 8.99, N, 13.38, found: C, 56.92, H,
8.91, N, 13.01%. 1H NMR (C6D6, 300.22 MHz) δ: 1.33 (m, 8H, CH2),
1.68 (m, 16H, (CH2)2), 2.06 (m, 16H, (CH2)2), 2.60 (s, 12H,
N(CH3)2), 3.04 (m, 4H, NCH); 13C NMR (CDCl3, 75.49 MHz)
δ: 25.2 (CH2), 25.6 ((CH2)2), 40.0 ((CH2)2), 54.5 (NCH), 58.7
(N(CH3)2), 161.7 (NCN).
The precise reasons for this difference in reactivity between
the guanidinate systems 1a−1d (1a, 1b, and 1d show no
evidence of equilibrium processes) are not fully understood, but
it is possible to speculate that there is a fine balance between
steric and electronic considerations. Complexes containing the
tert-butyl guanidinate unit {Me2NC(tBuN)2}, are expected to
be highly sterically hindered, which goes some way to explaining
the observed equilibria of the lithium complex 1c. This steric
hindrance is relieved somewhat by twisting of the guanidinate
ligand as observed in the tri-copper systems 3c and 4c.
Despite the fact that our initial intent was to synthesis the
dimeric system [Cu2{Me2NC(tBuN)2}2], we now believe that
the steric repulsion between adjacent tert-butyl groups across
the central {Cu2} unit is too great for this system to be stable.
Twisting of the guanidinate backbone, along with scavenging of
free “CuNMe2” or “CuCl” allows the steric strain to be relieved
resulting in the formation and isolation of the tri-metallic
systems [Cu3{Me2NC(tBuN)2}2(μ-NMe2)] (3c) and [Cu3{Me2-
NC(tBuN)2}2(μ-Cl)] (4c).
Synthesis of (2d). To a dry Schlenk, tetrahydrofuran (40 mL) was
added to a mixture of lithium dimethylamide (0.77 g, 15.0 mmol) and
N,N′-bis-2,6-diisopropylphenyl-carbodiimide (5.44 g, 15.0 mmol) to
give a pale yellow solution. The reaction mixture was left to stir for 2 h
after which the solution was transferred by cannula to a Schlenk
containing copper(I) chloride (1.49 g, 15.0 mmol). The reaction
mixture was allowed to stir for 18 h, after which time the volatiles were
removed under reduced pressure. Hexane (20 mL) was added to the
resultant residue and left to stir for 15 min, after which time the hexane
was removed under reduced pressure. The process was repeated twice to
remove any residual tetrahydrofuran. The product was then extracted
into hexane and filtered through Celite. The resultant filtrate was
concentrated, and storage of the material at −28 °C facilitated the growth
of yellow crystals suitable for single crystal X-ray diffraction (5.10 g,
72.4%). Mp 111 °C (dec). Elemental analysis: calcd. for C54H80Cu2N6
(%): C, 68.97, H: 8.58, N, 8.94, found: C, 69.08, H, 8.34, N, 8.84%. 1H
NMR (C6D6, 300.22 MHz) δ: 0.71 (d, 12H, J = 7 Hz, (CH(CH3)2),
1.19 (d, 12H, J = 7 Hz, CH(CH3)2), 1.28 (d, 12H, J = 7 Hz,
CH(CH3)2),1.59 (d, 12H, J = 7 Hz, CH(CH3)2), 1.99 (s, 12H,
N(CH3)2), 3.53 (sept, 4H, J = 7 Hz, CH(CH3)2), 3.81 (sept, 4H, J =
7 Hz, CH(CH3)2), 6.96−7.14 (m, 12H, CHAr); 13C NMR (CDCl3,
75.49 MHz) δ: 22.9 (CH(CH3)2), 23.1, (CH(CH3)2), 24.2 (CH-
(CH3)2), 25.7 (CH(CH3)2), 27.9 (CH(CH3)2), 28.1 (CH(CH3)2),
40.3 (N(CH3)2), 123.6 (CH), 124.1 (CH), 124.2 (CH), 142.7 (C-CH-
(Me)2), 142.8 (C-CH(Me)2), 144.3 (N-C), 168.4 (NCN).
Synthesis of (3c). In a dry Schlenk, tetrahydrofuran (20 mL) was
added to lithium dimethylamide (0.36 g, 7.0 mmol). To this, N,N′-di-
tert-butyl-carbodiimide (1.35 mL, 7.0 mmol) was added dropwise, and
the reaction mixture left to stir for 2 h. The clear solution was then
transferred via cannula to another Schlenk containing copper(I)
chloride (0.69 g, 7.0 mmol) and allowed to react for 2 h, after which
time the volatiles were removed under reduced pressure, and hexane
(20 mL) was added to the resultant residue. This was left to stir for
15 min and then the volatiles were removed under reduced pressure.
The process was repeated twice more to remove any residual
tetrahydrofuran. Further hexane (20 mL) was added, and the slurry
was filtered through Celite to remove any insoluble materials followed
While all five Cu(I) complexes described here have been
analyzed by thermogravimetric analysis, only 2a and 2b show
potential as CVD or ALD precursors because of their clean
decomposition pathways and relative onsets of volatilization.
EXPERIMENTAL SECTION
■
General Procedures. Elemental analyses were performed by
Elemental Microanalysis Ltd., Okehampton, Devon, U.K. H and 13C
1
NMR spectra were recorded on a Bruker Avance 300 MHz FT−NMR
spectrometer, as saturated solutions in d2-CD2Cl2; chemical shifts are
quoted in units of parts per million (ppm), relative to Me4Si (1H, 13C);
coupling constants are in hertz (Hz).
All reactions were carried out under an inert atmosphere, and in the
absence of light, using standard Schlenk techniques. Solvents were dried
over activated alumina columns using an Innovative Technology solvent
purification system (SPS) and degassed under an argon atmosphere. All
other reagents were purchased from commercial sources.
Thermogravimetric analysis (TGA) of the complexes was per-
formed at SAFC Hitech, Bromborough, U.K., using a Shimadzu TGA-51
Thermogravimetric Analyzer. Data points were collected every second at
a ramp rate of 20 °C min−1 in a flowing (50 mL min−1) N2 stream.
Synthesis of (1c). In a dry Schlenk, hexane (20 mL) was added to
lithium dimethylamide (0.10 g, 2.0 mmol) to give a cream/pale
yellow suspension. To this, N,N′-di-tert-butyl-carbodiimide (0.39 mL,
2.0 mmol) was added dropwise, and the reaction mixture left to stir for
2 h. The resulting clear solution was heated to reflux and filtered, via
cannula, into a clean Schlenk. Storage of the reaction mixture at −5 °C
resulted in the formation of a crop of colorless crystals which were
isolated by filtration. (0.36 g, 87%). 1H NMR (d8-THF, 300.22 MHz)
t
δ: 1.09 (s, 9H, C(CH3)3 of 1c), 1.23 (s, 9H, C(CH3)3 of free Bu-
NCN-tBu), 2.58 (s, 6H, N(CH3)2 of 1c), 2.75 (s, 6H, N(CH3)2
of free LiNMe2); 13C NMR (d8-THF, 75.49 MHz) δ: 31.4 (C(CH3)3
t
of free Bu-NCN-tBu), 33.5 (C(CH3)3 of 1c), 42.3 (C(CH3)3 of
1c), 47.9 (C(CH3)3 of free tBu-NCN-tBu) 51.5 (N(CH3)2 of 1c),
t
54.7 (N(CH3)2 of free LiNMe2), 139.2 (NCN of free Bu-N
255
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Table 6. Crystallographic Data for the Complexes 1c, 1d, 2a, 2b, 2d , 3c and 4c
Article
a
a
1c
1d
2a
2b
2d
3c
4c
empirical formula
formula weight
T/K
C44H96Li4N12
821.09
C54H80Li2N6
827.12
C18H40Cu2N6
467.64
C30H56Cu2N6
627.89
C54H80Cu2N6
940.32
C24H54Cu3N7
631.39
C22H48Cl1Cu3N6
622.73
150(2)
150(2)
150(2)
150(2)
150(2)
150(2)
150(2)
crystal system
space group
a (Å)
monoclinic
C2/c
monoclinic
P21/n
orthorhombic
C2221
triclinic
monoclinic
C2/c
monoclinic
P21/n
triclinic
P1
̅
P1
̅
16.0800(2)
15.8510(2)
22.4240(3)
13.1600(2)
20.7540(2)
38.1730(3)
11.3540(2)
21.2909(6)
9.9530(2)
10.4150(3)
11.3900(3)
16.0510(4)
97.087(2)
107.597(2)
112.555(2)
1613.01(7)
2
43.1820(4)
13.0740(2)
31.9510(3)
13.6340(1)
15.6710(2)
14.8520(2)
10.215(2)
10.644(2)
14.732(3)
103.876(2)
107.091(2)
98.077(2)
1447.4(5)
2
b (Å)
c (Å)
α (deg)
β (deg)
109.363(2)
92.730(2)
117.080(2)
94.945(1)
γ (deg)
V
5392.23(12)
8
10414.06(15)
2406.01(9)
4
16060.8(2)
12
3161.44(6)
4
Z
4
ρcalc mg/m−3
μ(Mo-Kα), mm−1
F(000)
1.011
0.060
1824
1.055
0.061
3616
1.291
1.780
992
1.293
1.167
0.833
6048
1.326
2.023
1336
1.429
1.346
2.296
672
652
crystal size, mm
0.25 × 0.20 × 0.15 0.32 × 0.20 ×
0.20 × 0.17 ×
0.15
0.22 × 0.10 ×
0.08
0.20 × 0.15 ×
0.10
0.40 × 0.25 ×
0.10
0.25 × 0.25 × 0.10
0.20
θ range (deg)
8.53 to 28.27
44134
4.44 to 27.47
143043
3.59 to 27.49
14843
3.10 to 28.25
18809
8.52 to 27.10
3.25 to 28.29
66451
8.17 to 27.44
19713
reflections collected
independent reflections
[R(int)]
6472 [0.1049]
23687 [0.0804] 2758 [0.0554]
7863 [0.0833]
7827 [0.0580]
6405 [0.0558]
reflections observed [I >
2σ(I)]
5297
14089
2548
4894
6516
5707
max., min transmission
goodness-of-fit
0.9910, 0.9851
1.050
0.9879, 0.9808
1.018
0.7761, 0.7172
1.050
0.8999, 0.7561
0.985
0.8233, 0.4984
1.039
0.8029, 0.5975
1.100
final R1 (wR2) [I > 2σ(I)] 0.0451 (0.1181)
final R1 (wR2) (all data) 0.0571 (0.1268)
0.0575 (0.1214) 0.0309 (0.0746) 0.0479 (0.0945)
0.1151 (0.1445) 0.0376 (0.0776) 0.1035 (0.1148)
0.0291 (0.0659) 0.0308 (0.0758)
0.0403 (0.0710) 0.0364 (0.0788)
Largest diff. peak and hole, 0.254 and −0.145 0.248 and
0.630 and
0.642 and
0.473 and
−0.447
0.546 and −0.609
e Å−3
−0.264
−0.473
−0.756
a
Crystallographic data for the complex 2d is reported here but has not been deposited with the CCDC.
by removal of the volatiles in vacuo. The resulting yellow crude prod-
uct was purified by recrystallization from hexane −28 °C, to afford the
final product as colorless crystals (1.10 g, 71% based on copper). Mp
97 °C (dec). Elemental analysis: calcd. for C24H56Cu3N7; C, 45.66, H,
all structures, data were collected on a Nonius Kappa CCD dif-
fractometer at 150(2) K using Mo-Kα radiation (λ = 0.71073 Å).
Structure solution and refinements were performed using SHELX8631
and SHELX9732 software, respectively. Corrections for absorption were
made in all cases. For all complexes, hydrogen atoms were included at
calculated positions.
1
8.62, N, 15.53, found: C, 44.65, H, 8.24, N, 15.53%. H NMR (C6D6,
300.22 MHz) δ: 1.41 (s, 18H, NC(CH3)3), 1.48 (s, 18H, NC(CH3)3,
2.41 (s, 12H, CN(CH3)2, 3.09 (s, 6H, CuN(CH3)2). 13C NMR (C6D6,
75.49 MHz) δ: 33.4 (NC(CH3)3), 34.7 (NC(CH3)3), 42.6 (CN-
(CH3)2), 49.3 (CuN(CH3)2), 54.4 (NC(CH3)3), 55.9 (NC(CH3)3),
168.3 (NCN).
ASSOCIATED CONTENT
* Supporting Information
The X-ray crystallographic files in CIF format. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
Synthesis of (4c). To a dry Schlenk, tetrahydrofuran (20 mL) was
added to 1c (0.82 g, 1.0 mmol) to give a colorless solution. The
solution was then was then transferred via cannula to another Schlenk
containing copper(I) chloride (0.59 g, 6.0 mmol) and allowed to react
for 2 h, after which time the volatiles were removed under reduced
pressure and dry hexane (20 mL) was added to the resultant residue.
This was left to stir for 15 min and then the volatiles were removed
under reduced pressure. The process was repeated twice more to
remove any residual tetrahydrofuran. Further hexane (20 mL) was
added, and the slurry was filtered through Celite to remove any
insoluble materials followed by removal of the volatiles in vacuo. The
resulting yellow crude product was purified by recrystallization from
hexane, at −28 °C, to afford the product as colorless crystals (0.90 g,
72% based on copper). Mp 102 °C (dec). Elemental analysis: calcd.
for C22H48Cu3N6Cl; C, 42.43, H, 7.77, N, 13.50, found. C, 42.74, H,
AUTHOR INFORMATION
Corresponding Author
*E-mail: A.L.Johnson@bath.ac.uk. Phone: 44 (0)1225 384467.
Fax: 44 (0)1225 386231.
■
ACKNOWLEDGMENTS
■
We acknowledge the financial support of the University of Bath
and the EPSRC for the award of a CASE for New Academics
(A.M.W.) and studentships to T.P.R. and T.P.
REFERENCES
■
1
7.79, N, 13.26%. H NMR (C6D6, 300.22 MHz) δ: 1.35 (s, 18H,
(1) (a) Edelmann, F. T. Adv. Organomet. Chem. 2008, 57, 183−352.
(b) Mohamed, A. A. Coord. Chem. Rev. 2010, 254, 1918−1947.
(c) Schmidt, J. A. R.; Arnold, J. J. Chem. Soc., Dalton Trans. 2002,
3454−3461. (d) Bailey, P. J.; Pace, S. Coord. Chem. Rev. 2001, 214,
91−141. (e) Bailey, P. J.; Grant, K. J.; Mitchell, L. A.; Pace, S.; Parkin,
A.; Parsons, S. J. Chem. Soc., Dalton Trans. 2000, 1887−1891.
NC(CH3)3), 1.40 (s, 18H, NC(CH3)3, 2.38 (s, 12H, CN(CH3)2.
13C NMR (C6D6, 75.49 MHz) δ: 33.0 (NC(CH3)3), 34.4 (NC(CH3)3),
42.4 (CN(CH3)2), 54.4 (NC(CH3)3), 55.7 (NC(CH3)3), 168.4 (NCN).
Crystallography. Experimental details relating to the single-
crystal X-ray crystallographic studies are summarized in Table 6. For
256
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257

Inorganic Chemistry
Article
(2) (a) Gordon, R. G.; Lim, B. S.; Rahtu, A.; Park, J. S. Inorg. Chem.
2003, 42, 7951−7958. (b) Jiang, X.; Bollinger, J. C.; Lee, D. J. Am.
Chem. Soc. 2005, 127, 15678−15679. (c) Jiang, X.; Bollinger John, C.;
Baik, M.-H.; Lee, D. Chem. Commun. 2005, 1043−1045. (d) Li, Z.;
Barry, S. T.; Gordon, R. G. Inorg. Chem. 2005, 44, 1728−1735. (e) Li,
Z.; Rahtu, A.; Gordon, R. G. J. Electrochem. Soc. 2006, 153, C787−
C794. (f) Chen, T.; Xu, C.; Baum, T. H.; Hendrix, B. C.; Cameron, T.;
Roeder, J. F.; Stender, M.; Advanced Technology Materials, Inc.,
U.S.A. U.S. Patent Appl. Publ. WO-2007142700, 2006-US62709, 2007.
(g) Coyle, J. P.; Monillas, W. H.; Yap, G. P. A.; Barry, S. T. Inorg.
Chem. 2008, 47, 683−689. (h) Mohamed, A. A.; Abdou, H. E.; Mayer,
A.; Fackler, J. P. Jr. J. Cluster Sci. 2008, 19, 551−559. (i) Bunge, S. D.;
Steele, J. L. Inorg. Chem. 2009, 48, 2701−2706. (j) Johnson, A. L.;
Willcocks, A. M.; Richards, S. P. Inorg. Chem. 2009, 48, 8613−8622.
(k) Barry, S. T.; Whitehorne, T. J. J.; Coyle, J. P.; Mahmood, A.;
Monillas, W. H.; Yap, G. P. A. Eur. J. Inorg. Chem. 2011, 3240−3247.
(l) Krisyuk, V.; Aloui, L.; Prud’homme, N.; Sysoev, S.; Senocq,
F.; Samelor, D.; Vahlas, C. Electrochem. Solid-State Lett. 2011, 14,
D26−D29.
(3) (a) Pyykko, P. Chem. Rev. 1997, 97, 597−636. (b) Pyykko, P.;
Mendizabal, F. Chem.Eur. J. 1997, 3, 1458−1465. (c) Pyykko, P.;
Runeberg, N.; Mendizabal, F. Chem.Eur. J. 1997, 3, 1451−1457.
(4) (a) Irwin Michael, D.; Abdou Hanan, E.; Mohamed Ahmed, A.;
Fackler John, P. Jr. Chem. Commun. 2003, 2882−2883. (b) Maier, S.;
Hiller, W.; Straehle, J.; Ergezinger, C.; Dehnicke, K. Z. Naturforsch., B:
Chem. Sci. 1988, 43, 1628−1632.
(5) Coyle, J. P.; Johnson, P. A.; DiLabio, G. A.; Barry, S. T.; Mueller,
J. Inorg. Chem. 2010, 49, 2844−2850.
(6) Shibayama, K.; Seidel, S. W.; Novak, B. M. Macromolecules 1997,
30, 3159−3163.
(7) Guanidines: Historical, Biological, Biochemical and Clinical Aspects
of the Naturally Occurring Guanidino Compounds; Mori, A., Cohen,
B. D., Lowenthal, A., Eds.; Plenum Press: New york, 1985.
(8) (a) Cotton, F. A.; Feng, X. J.; Timmons, D. J. Inorg. Chem. 1998,
37, 4066−4069. (b) Oakley, S. H.; Coles, M. P.; Hitchcock, P. B. Inorg.
Chem. 2004, 43, 5168−5170. (c) Oakley, S. H.; Soria, D. B.; Coles, M.
P.; Hitchcock, P. B. Dalton Trans. 2004, 537−546. (d) Oakley, S. H.;
Coles, M. P.; Hitchcock, P. B. Inorg. Chem. 2003, 42, 3154−3156.
(e) Coles, M. P.; Hitchcock, P. B. Polyhedron 2001, 20, 3027−3032.
(9) (a) Bunge, S. D.; Bertke, J. A.; Cleland, T. L. Inorg. Chem. 2009,
48, 8037−8043. (b) Bunge, S. D.; Ocana, J. A.; Cleland, T. L.; Steele,
J. L. Inorg. Chem. 2009, 48, 4619−4621.
(20) Kenney, A. P.; Yap, G. P. A.; Richeson, D. S.; Barry, S. T. Inorg.
Chem. 2005, 44, 2926−2933.
(21) (a) Braga, D.; Grepioni, F.; Biradha, K.; Desiraju, G. R. J. Chem.
Soc., Dalton Trans. 1996, 3925−3930. (b) Jin, G. X.; Jones, C.; Junk,
P. C.; Lippert, K. A.; Rose, R. P.; Stasch, A. New J. Chem. 2009, 33,
64−75.
(22) Cameron, T. M.; Xu, C. Y.; Dipasquale, A. G.; Rheingold, A. L.
Organometallics 2008, 27, 1596−1604.
(23) Nikonov, G. I.; Shirobokov, O. G.; Gorelsky, S. I.; Simionescu,
R.; Kuzmina, L. G. Chem. Commun. 2010, 46, 7831−7833.
(24) Woo, T. K.; Rowley, C. N.; Ong, T. G.; Priem, J.; Richeson,
D. S. Inorg. Chem. 2008, 47, 9660−9668.
(25) Rowley, C. N.; DiLabio, G. A.; Barry, S. T. Inorg. Chem. 2005,
44, 1983−1991.
(26) Guo, D. L.; Qiao, X. L.; Tong, H. B.; Zhou, M. S. Acta
Crystallogr., Sect. E 2009, 65, m405−u572.
(27) Gambarotta, S.; Bracci, M.; Floriani, C.; Chiesivilla, A.; Guastini,
C. J. Chem. Soc., Dalton Trans. 1987, 1883−1888.
(28) Kohn, R. D.; Pan, Z.; Mahon, M. F.; Kociok-Kohn, G. Chem.
Commun. 2003, 1272−1273.
(29) (a) Niemeyer, M. Acta Crystallogr., Sect. E 2001, 57, m491−
m493. (b) Chen, H.; Olmstead, M. M.; Shoner, S. C.; Power, P. P.
J. Chem. Soc., Dalton Trans. 1992, 451−457.
(30) (a) Chivers, T.; Parvez, M.; Schatte, G. J. Organomet. Chem.
1998, 550, 213−220. (b) Coles, M. P.; Hitchcock, P. B. Chem.
Commun. 2005, 3165−3167. (c) Ong, T. G.; O’Brien, J. S.; Korobkov,
I.; Richeson, D. S. Organometallics 2006, 25, 4728−4730. (d) Rudolf,
D.; Kaifer, E.; Himmel, H. J. Eur. J. Inorg. Chem. 2010, 4952−4961.
(e) Schmidt, J. A. R.; Arnold, J. Chem. Commun. 1999, 2149−2150.
(31) Sheldrick, G. M. SHELX-86, Computer Program for Crystal
Structure Determination; University Of Gottingen: Gottingen,
̈
̈
Germany, 1986.
(32) (a) Sheldrick, G. M. SHELX-97, Computer Program for Crystal
Structure Refinement; Univerity of Gottingen: Gottingen, Germany,
̈
̈
1997; (b) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112−122.
(10) Lim, B. S.; Rahtu, A.; Gordon, R. G. Nat. Mater. 2003, 2, 749−754.
(11) Li, Z. W.; Gordon, R. G.; Farmer, D. B.; Lin, Y. B.; Vlassak, J.
Electrochem. Solid-State Lett. 2005, 8, G182−G185.
(12) Li, Z. W.; Gordon, R. G. Chem. Vap. Deposition 2006, 12,
435−441.
(13) Li, Z. W.; Rahtu, A.; Gordon, R. G. J. Electrochem. Soc. 2006,
153, C787−C794.
(14) Dai, M.; Kwon, J.; Halls, M. D.; Gordon, R. G.; Chabal, Y. J.
Langmuir 2010, 26, 3911−3917.
(15) Ma, Q.; Guo, H. S.; Gordon, R. G.; Zaera, F. Chem. Mater. 2010,
22, 352−359.
(16) Bondi, A. J. Phys. Chem. 1964, 68, 441−451.
(17) (a) Fu, W. F.; Gan, X.; Che, C. M.; Cao, Q. Y.; Zhou, Z. Y.;
Zhu, N. N. Y. Chem.Eur. J. 2004, 10, 2228−2236. (b) Jakle, F.;
Sundararaman, A.; Zakharov, L. N.; Rheingold, A. L. Chem. Commun.
2005, 1708−1710. (c) Yin, Y. G.; Zhang, J. X.; He, J.; Hu, M. H.; Li,
D.; Huang, X. C. Inorg. Chem. 2008, 47, 3471−3473. (d) Merz, K. M.;
Hoffmann, R. Inorg. Chem. 1988, 27, 2120−2127.
(18) (a) Tsipis, C. A.; Karagiannis, E. E. Organometallics 2010, 29,
847−859. (b) Tsipis, A. C. Organometallics 2010, 29, 354−363.
(c) Schwerdtfeger, P.; Assadollahzadeh, B. Chem. Phys. Lett. 2008, 462,
222−228. (d) Schwerdtfeger, P.; Hermann, H. L.; Boche, G. Chem.
Eur. J. 2001, 7, 5333−5342.
(19) (a) Siemeling, U.; Vorfeld, U.; Neumann, B.; Stammler, H. G.
Chem. Commun. 1997, 1723−1724. (b) Mendizabal, F.; Reyes, D.;
Olea-Azar, C. Int. J. Quantum Chem. 2006, 106, 906−912.
257
dx.doi.org/10.1021/ic201602m | Inorg. Chem. 2012, 51, 246−257