To bend or not to bend: Experimental and computational studies of structural preference in Ln(TpiPr2)2 (Ln = Sm, Tm)

Article abstract of DOI:10.1021/ic501816v

The synthesis and characterization of Ln(Tp^iPr2)₂ (Ln = Sm, 3Sm; Tm, 3Tm) are reported. While the simple ¹H NMR spectra of the compounds indicate a symmetrical solution structure, with equivalent pyrazolyl groups, the solid-state structure revealed an unexpected, "bent sandwich-like" geometry. By contrast, the structure of the less sterically congested Tm(Tp^Me2,4Et)₂ (4) adopts the expected symmetrical structure with a linear B-Tm-B arrangement. Computational studies to investigate the origin of the unexpected bent structure of the former compounds indicate that steric repulsion between the isopropyl groups forces the Tp ligands apart and permits the development of unusual interligand C-H···N hydrogen-bonding interactions that help stabilize the structure. These results find support in the similar geometry of the Tm(III) analogue [Tm(Tp^iPr2)₂]I, 3Tm⁺, and confirm that the low symmetry is not the result of a metal-ligand interaction. The relevance of these results to the general question of the coordination geometry of MX₂ and M(C₅R₅)₂ (M = heavy alkaline earth and Ln(II), X = halide, and C₅R₅ = bulky persubstituted cyclopentadienyl) complexes and the importance of secondary H-bonding and nonbonding interactions on the structure are highlighted.

Full text of DOI:10.1021/ic501816v

Article
pubs.acs.org/IC
To Bend or Not To Bend: Experimental and Computational Studies of
Structural Preference in Ln(TpiPr ) (Ln = Sm, Tm)
2 2
Aurelien Momin,§,∥ Lee Carter,†,# Yi Yang,† Robert McDonald,† Stephanie Essafi (nee Labouille),§,△
́
́
́
Franco̧
is Nief,§ Iker Del Rosal,‡ Andrea Sella,*,⊥ Laurent Maron,*,‡ and Josef Takats*,†
†Department of Chemistry, University of Alberta, Edmonton, AB, Canada T6G 2G2
‡Laboratoire de Physique et Chimie des Nanoobjets, INSA, Universite
Cedex, France
́
Paul Sabatier, 135 Avenue de Rangueil, 31077 Toulouse
§Laboratoire de Chimie Molec
́
ulaires, CNRS, Ecole Polytechnique, 91128 Palaiseau, France
⊥Department of Chemistry, University College London, 20 Gordon Street, London, WC1H 0AJ, United Kingdom
S
* Supporting Information
ABSTRACT: The synthesis and characterization of Ln(TpiPr2)2 (Ln = Sm, 3Sm; Tm, 3Tm) are
1
reported. While the simple H NMR spectra of the compounds indicate a symmetrical solution
structure, with equivalent pyrazolyl groups, the solid-state structure revealed an unexpected, “bent
sandwich-like” geometry. By contrast, the structure of the less sterically congested Tm(TpMe2,4Et)2 (4)
adopts the expected symmetrical structure with a linear B−Tm−B arrangement. Computational
studies to investigate the origin of the unexpected bent structure of the former compounds indicate
that steric repulsion between the isopropyl groups forces the Tp ligands apart and permits the
development of unusual interligand C−H···N hydrogen-bonding interactions that help stabilize the
structure. These results find support in the similar geometry of the Tm(III) analogue [Tm(TpiPr2)2]I,
3Tm+, and confirm that the low symmetry is not the result of a metal−ligand interaction. The
relevance of these results to the general question of the coordination geometry of MX2 and M(C5R5)2
(M = heavy alkaline earth and Ln(II), X = halide, and C5R5 = bulky persubstituted cyclopentadienyl)
complexes and the importance of secondary H-bonding and nonbonding interactions on the structure are highlighted.
related alkaline earth metal compounds with the very bulky
C5 Pr5,
INTRODUCTION
■
17,18
C5Ph5,19 and CpBIG (CpBIG = C5(4-n-Bu-
i
The structure and bonding of the s-block and the lanthanides
are usually described as predominantly ionic. The absence, in
most cases, of specific directional interactions means that it is
typical to expect high-symmetry structures.1 Deviations from
these symmetries have often given rise to vigorous discussion.
For example, the discovery that, in the gas phase, the heavy
alkaline earth halides adopted a bent geometry at the metal2,3
led to a long debate over the relative contributions of
polarization and d-orbital participation.4
C6H4)5)20−22 all adopt parallel-ring sandwich arrangements.
Harder attributes the S10 symmetry in the latter compounds to
favorable, nonclassical C−H···π(phenyl) ring interactions.
Another noteworthy feature of these complexes is the
observation that the metal sits away from the center of the
molecule, allowing for dynamics; these alkaline earth and
lanthanide structures have been discussed and rationalized in
terms of a polarization model.22,23
Such structural diversity is even more pronounced for
complexes based on Trofimenko’s hydrotris(pyrazolyl)-
borates,24,25 which are also significantly influenced by the
substitution pattern of the ligand set. (The abbreviation we will
use for these ligands is that proposed by Trofimenko: Tp stands
for tris(pyrazolyl)borate; R and R′ are substituents at the 3- and
5-positions of the pyrazolyl group, respectively, and the 4-
substituent is denoted by the superscript 4R.)
Among the lanthanides, the classic example of this was
provided by Evans with his report of the iconic molecule
[Sm(η-C5Me5)2(THF)2]5 and its unsolvated analogue, [Sm(η-
C5Me5)2],5−7 which have very similar bent metallocene
structures. The isolation of the Eu and Yb analogues,8 followed
by those of Ca, Sr, and Ba,9,10 engendered several computa-
tional studies11−15 to try to understand the nature of the
distortion from the C5-symmetric linear sandwich structure.
These studies have led to the view that the distortion arises
predominantly from the attractive dispersion/van der Waals
interactions.11,16
Although these distortions are observed consistently, the
energy barrier separating the bent and linear sandwich
geometries is rather small since the geometrical preference
can be easily manipulated by changing the steric profile of the
cyclopentadienyl ligands. Thus, bis-cyclopentadienyl Ln(II) and
Some time ago two of us reported the synthesis,
spectroscopic properties, and solid-state structures of Ln-
(TpMe2
) (Ln = Sm, 1Sm; Yb, 1Yb) and some related
2
complexes.26−29 We anticipated that the greater cone angle of
TpMe2 (239°)24 compared to the C5Me5 ligand (142°)30 would
Received: July 31, 2014
Published: October 27, 2014
© 2014 American Chemical Society
12066
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
Figure 1. RT 1H NMR spectra of Ln(TpiPr2)2. (a) Ln = Sm (3Sm) in toluene-d8, #silicone grease; (b) Ln = Tm (3Tm) in benzene-d6, #pentane. In
both spectra, * is residual solvent peaks.
i
confer significantly greater steric control over the metal center.
This proved to be the case, and the more congested
Ln(TpMe2)2 complexes were isolated as THF-free materials;
the solid-sate structure proved to be highly symmetrical,
adopting a trigonal antiprismatic coordination geometry with
colinear B−Ln−B atoms and S6 molecular symmetry, a
“sandwich-like” structure. This high symmetry was, however,
accompanied by an unexpected lack of solubility, although
solubility could be increased substantially by introducing an
ethyl substituent in the 4-position of the pyrazolyl groups.
This coordination environment was found to give a good
balance of kinetic stabilization and reactivity, allowing the
isolation of a variety of seven-coordinate complexes, Sm-
(TpMe2)2X,28 and even eight-coordinate, Sm(TpMe2)2XY,
complexes26,31 from reaction with reducible substrates.
conformational preference for the Pr substituents to place the
C−H bond pointing toward the B and metal centers,33,34 it
presents an interior to the metal center analogous to that of the
TpMe2 ligand, while its exterior resembles that of TptBu,Me, and
with steric demand expected to fall somewhere between the
two. [Although cone angles are sometimes used in the
discussion of the steric demand of scorpionate ligands, care
must be exercised. The most consistent values seem to be based
on the structures of a series of monomeric Tl(I) scorpionate
complexes,24,25 but even these values should be used with
caution. Thus, the relevant values for the present purpose are
TpMe2 239°, TpiPr,4Br 243°, and TptBu,Me 243°, implying similar
size for TpiPr2 and TptBu,Me ligands. However, in TlTpiPr,4Br the
iPr methyl substituents point toward Tl and not away, as it is
more commonly observed. Hence, not surprisingly, the cone
angle is the same as in the TptBu,Me ligand. There are
independent geometrical features and reactivity behavior that
indicate that the steric profile of the iPr-substituted Tp ligand is
smaller than that of the corresponding tBu-substituted ligand,35
and we support this conclusion. What is clear is that the low
barriers to alkyl group rotation in Tp-type ligands make it
appropriate that estimates of their cone angles should be
accompanied by much larger error bars than the rather
simplistic figures typically quoted suggest.]
Pushing the steric congestion even further, increasing the
bulk in the 3-position of the pyrazolyl rings and using the
superbulky TptBu,Me ligand led to a change in structure. In the
homoleptic Ln(TptBu,Me)2 (Ln = Sm, Yb)32 molecules, one
scorpionate switched from the classical κ3-N3 bonding mode to
being coordinated by two nitrogens and one “agostic” B−H
bond, i.e., a κ3-N2H TptBu,Me ligand.32 This interesting structural
1
motif is maintained in solution, as shown by VT H NMR
studies, which also revealed fascinating two-stage dynamic
behavior: a low-energy process involving the κ3-N2H ligand,
and high-energy process which results in exchange of the
coordination mode of the two distinct scorpionates.
In this article we report the synthesis and solid-state structure
of Ln(TpiPr2)2 (Ln = Sm, 3Sm; Tm, 3Tm), highly sterically
congested complexes that have been found to adopt an
unexpected “bent-sandwich-like” geometry. We have used DFT
to probe the intramolecular origins of this origins of this
unusual distortion. The solid-state structures of the oxidized
Questions therefore arise regarding the structural preferences
of Ln(TpR,R′)2 complexes for ligands with a steric demand and
cone angle intermediate between those of TpMe2 and TptBu,Me
.
analogue of 3Tm, [Tm(TpiPr2)2]I (3Tm+), and Tm(TpMe2,4Et
(4) are also reported.
)
2
The TpiPr2 ligand appeared to fit the bill, since, with the typical
12067
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
RESULTS AND DISCUSSION
■
Synthesis and Characterization of Ln(TpiPr2
) (Ln =
2
Sm, Tm). Addition of two equivalents of KTpiPr2, dissolved in
THF, to a solution of LnI2 in the same solvent resulted in an
immediate color change from blue (Sm) or green (Tm) to
green and reddish-brown, respectively, and the precipitation of
KI, eq 1. After workup the Ln(TpiPr2)2 (Ln = Sm, 3Sm; Tm,
3Tm) compounds were obtained as green (Sm) and plum-red
(Tm) fluffy solids, sufficiently pure for further reactions.
LnI2 + 2KTpiPr2 → Ln(TpiPr2)2 + 2KI( ↓ )
(1)
The Sm compound is moderately stable at room temperature
and can be kept at −35 °C for months but then shows signs of
decomposition. The Tm compound, however, is significantly
less thermally stable and starts to decompose even at −35 °C
after a week, giving insoluble white powders of unknown
composition.
The compounds are extremely soluble in ether and
hydrocarbon solvents, and crystallization from these solvents
proved challenging (see Experimental Section), but eventually
led to crystals suitable for X-ray analysis and NMR studies.
The 1H NMR spectra of 3Sm and 3Tm are shown in Figure
1. The spectra establish the purity of these delicate and highly
air sensitive compounds. The spectra exhibit one set of
pyrazolyl group resonances with the expected intensity ratio.
The signals are paramagnetically shifted, and the two spectra
provide a nice illustration of the greater shifts induced by
thulium with its greater magnetic moment.36,37 The simple
spectra show no broadening of the resonances down to −80
°C, implying either a symmetrical structure, similar to the
previously reported Ln(TpR,R′)2,26−28 or a time-averaged
structure due to rapid fluxionality. To resolve the ambiguity,
the solid-state structures of 3Sm and 3Tm were determined by
single-crystal X-ray diffraction.
Solid-State Structure of Ln(TpiPr2)2 (Ln = 3Sm, 3Tm).
The structures of 3Sm and 3Tm, with their respective
numbering schemes, are shown in Figure 2; relevant metrical
parameters are listed in Table 1.
Figure 2. Molecular structures of (a, above) 3Sm, viewed
perpendicular to the B1SmB2 plane (side view), and (b, below)
3Tm, viewed perpendicular to the N12N32N53 plane (front view);
showing the “bent sandwich” geometry (above) and the close to Cs
symmetry (below). The atoms are drawn with 20% probability
ellipsoids. Hydrogen atoms have been omitted for clarity.
Close examination of Figure 2 reveals that the molecular
geometry of both of these complexes is different both from the
26−28
linear sandwich-like structures of Ln(TpMe2
)
and from
2
that of the highly congested Ln(TptBu,Me)2 that feature a κ3-
N2H-TptBu,Me ligand.32 The TpiPr2 ligands in complexes 3
display the classical κ3-bonding mode, but the B1−Ln−B2
angle is not 180°. Instead it is significantly reduced to 150.1°
(3Sm) and 152.2° (3Tm), a structural motif that has not been
observed before in the absence of either an additional ligand
(e.g., Sm(TpMe2)2X)28 or a stereochemically active lone pair in
the coordination sphere (as in M(TpMe2)2, M = Sn, Pb).38,39
The distortion of the metal’s coordination sphere is also
reflected in the ca. 30° angle between the TpiPr2 ligand’s
coordination plane (i.e., the angle between planes
N12N22N32/N42N52N62). The “bent” geometry is reminis-
cent of the bent-sandwich structure of [Sm(C5Me5)2].6 Indeed
the molecular symmetry is “close” to Cs, the mirror plane
running through B1LnB2, with N51N52 deviating only slightly,
but N21N22 more so, from this plane (see Table SI 1).
The bending of the TpiPr2 ligands along the BNB line and the
molecular symmetry of complexes 3 are also distinct from that
of seven-coordinate Sm(TpMe2)2X (X = F, Cl).28 In the latter,
the cavity accommodating the X ligand is achieved not just by
simple bending of the BSmB angle but also by twisting and
bending of the two TpMe2 ligands around the B−N bond,
resulting in an approximate C2-symmetric structure in the solid
state. Nevertheless, symmetrization of both Cs and C2 structures
is a very low activation process, as seen by the simple and
1
temperature-invariant H NMR spectra of these complexes.
The “bent-sandwich” geometry of complexes 3 has as its
result an expansion of the N12N32N52 and contraction of the
N22N42N62 triangular faces (“front” and “back” in Figure 2b),
as seen by the much larger N−Ln−N(inter ligand)cis angles:
N12−Ln−N52/N32−Ln−N52, 122.5°av compared to N22−
Ln−N42/N22−Ln−N62, 86.0°av. Viewing it this way, the
coordination geometry could be described as “capped
octahedral” with the capping position on the N12N32N52
face missing. Indeed, in spite of the bending along the B1LnB2
line, the two triangular faces, N12N32N52/N22N42N62, are
close to being parallel (angles: 3Sm 12°, 3Tm 13°).
The opposite change in the degree of congestion of these
two faces is also reflected in the LnNNB torsional and
12068
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
Table 1. Interatomic Distances (Å) and Angles (deg) for Sm(TpiPr2)2 (3Sm), Tm(TpiPr2)2 (3Tm), and [Tm(TpiPr2)2]I (3Tm+)
a
a
a
a
3Sm(A)
3Sm(B)
3Tm(A)
3Tm(B)
3Tm+
Ln−N12
Ln−N22
Ln−N32
Ln−N42
Ln−N52
Ln−N62
Ln−Nrange
2.638(2)
2.668(2)
2.645(2)
2.587(2)
2.680(2)
2.605(2)
2.59−2.68
2.664(2)
2.639(2)
2.664(2)
2.624(2)
2.643(2)
2.616(2)
2.62−2.66
2.534(2)
2.543(2)
2.553(2)
2.500(2)
2.599(2)
2.488(2)
2.538(2)
2.536(2)
2.534(2)
2.518(2)
2.554(2)
2.503(2)
2.426(2)
2.413(2)
2.379(2)
2.376(2)
2.438(2)
2.378(2)
2.38−2.44
2.49−2.60
2.50−2.55
N−Ln−N(intra)
N12−Ln−N22
N12−Ln−N32
N22−Ln−N32
N42−Ln−N52
N42−Ln−N62
N52−Ln−N62
N−Ln−N(inter)cis
N12−Ln−N52
N12−Ln−N62
N22−Ln−N42
N22−Ln−N62
N32−Ln−N42
N32−Ln−N52
N−Ln−N(inter)trans
N12−Ln−N42
N22−Ln−N52
N32−Ln−N62
N−Ln−N(inter)trans
N12−Ln−N42
N22−Ln−N52
N32−Ln−N62
Torsion Angles
Ln−N12N11−B1
Ln−N22−N21−B1
Ln−N32−N31−B1
Ln−N42−N41−B2
Ln−N52−N51−B2
Ln−N62−N61−B2
B1−Ln−B
78.68(5)
69.08(5)
77.52(5)
69.71(5)
81.45(5)
72.96(5)
77.34(5)
70.85(5)
78.85
81.80(7)
71.42(7)
81.87(7)
76.12(7)
86.03(5)
72.61(7)
80.59(7)
71.41(7)
83.53(7)
76.45(7)
87.69(7)
72.54(7)
88.27(6)
74.32(6)
82.93(6)
76.16(5)
88.26(5)
80.84(5)
73.80(5)
83.05(5)
68.86(5)
127.86(5)
101.05(5)
84.17(5)
84.93(5)
104.24(5)
125.83(5)
122.71(6)
102.59(5)
87.48(5)
87.93(5)
100.09(5)
124.92(6)
119.32(7)
99.84(7)
84.94(7)
84.33(7)
99.42(7)
122.61(7
118.87(7)
100.62(7)
86.31(7)
87.59(7)
98.32(7)
117.86(8)
111.87(6)
97.22(6)
85.61(5)
78.70(5)
98.14(6)
119.03(6)
162.34(5)
147.69(5)
161.18(6)
163.49(5)
151.43(5)
166.20(6)
164.4(7)
150.88(7)
164.60(7)
164.10(7)
153.94(7)
168.91(8)
170.84(6)
152.90(5)
160.03(6)
162.34(5)
147.69(5)
161.18(6)
163.49(5)
151.43(5)
166.20(6)
164.4(7)
150.88(7)
164.60(7)
164.10(7)
153.94(7)
168.91(8)
170.84(6)
152.90(5)
160.03(6)
8.5(2)
33.3(2)
28.7(2)
17.2(2)
7.4(2)
32.5(3)
22.3(3)
6.6(3)
30.9(3)
24.1(3)
3.9(3)
0.8(2)
23.5(2)
32.8(2)
22.8(2)
1.3(2)
26.4(2)
26.8(2)
22.6(2)
3.3(2)
9.1(3)
11.6(3)
0.6(3)
2.5(2)
2.3(3)
7.1(2)
30.7(2)
151.79(5)
22.8(3)
151.13(7)
28.2(3)
153.40(7)
3.2(2)
148.34(5)
152.59(5)
a
There are two independent molecules per asymmetric unit.
intraligand angles. Thus, in the expanded N12N32N52 face, the
LnN52N51B2 torsional angles are close to zero and result in
easy nestling of the N51N52 pyrazolyl ring and its 3-iPr
substituent between the N12/N32 pyrazolyl groups; these
retain the typical, small intraligand N12−Ln−N32 angle of ca.
70°. On the other hand, in the contracted N22N42N62 face,
the Ln−N22−N21−B1 torsional angles are greater than 20°
and the intraligand N42−Ln−N62 angles have opened up to
over 80° to accommodate the 3-iPr substituent of the N21N22
pyrazolyl ring. As expected, the N42−Tm−N62 angle (86.4°av)
has opened up more than in 3Sm because of the smaller and
more congested Tm(II) center.
complexes,41 an indication of the ability of the TpR,R′ ligands
to interlock efficiently.
As often found in complexes with TpiPr2 ligands,33,34 most of
i
the Pr substituents are arranged such that the C−H bonds
point toward either the B−H bond or the Ln center, to
minimize steric crowding. However, some of the iPr
substituents around the contracted triangular face exhibit subtly
different conformational orientation, and possible reasons for
this will be discussed in the Computational Studies subsection.
These structural studies revealed that both complexes 3Sm
and 3Tm crystallize with two independent molecules per
asymmetric unit. Although the distances and angles in Table 1
show some small variations, by and large the two independent
molecules are very similar, clearly indicating that the observed
“bent-sandwich” geometry is not the result of packing forces in
the crystal, but it is an inherent feature of complexes 3.
It is noteworthy that, in spite of the distortion, the Sm−N
distances in 3Sm remain similar to those observed in the
related Sm(TpR,R′)2 complexes.26−28 For 3Tm, the decrease of
0.1 Å in the Ln−N distance compared with 3Sm is in line with
the reduction in the six-coordinate ionic radius from Sm(II) to
Tm(II).40 However, the Tm−N distances are shorter than in
the four-coordinate Tm(TptBu,Me)N(SiMe3)2/CH(SiMe3)2
Synthesis and Solid-State Structure of Tm(TpMe2,4Et
)
2
(4Tm). To complete the picture and to rule out the possibility
that the “bent” geometry was an inherent feature of Tm(TpR,R′
)
complexes, a less crowded complex was contemplated. In view
of the thermal sensitivity of Tm(TpiPr2)2 and the anticipated
12069
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
insolubility of Tm(TpMe2)2, attention was focused on the
soluble compound Tm(TpMe2,4‑Et)2.
distances between Sm(TpMe2,4Et
)
(Sm−N, 2.615 Å) and
2
Tm(TpMe2,4Et)2 is exactly what is expected on the basis of the
difference between the respective ionic radii.40 The similarity
extends to the intra- and interligand N−Ln−N angles between
the above complexes; in particular the small Tm−N−N−B
torsional angles (average 1.4(9)°, range 0.6−2.0°) mirror those
seen in Sm(TpMe2,4Et)2 (average 5.6(18)°, range 2.2−8.0°).
Also, just as in the latter complex, the ethyl groups are rotated
approximately perpendicular to the linear B−Tm−B axis, with
two oriented toward each other and one away.
The reaction of TmI2 with KTpMe2,4Et was carried out in
DME, a solvent in which Tm(II) compounds are known to be
more stable.42 This approach gave good-quality Tm(TpMe2,4Et
)
2
(4Tm), after crystallization. However, even pure samples of
4Tm in DME or THF at RT start to deposit white solids of
unknown composition after a few hours, and the same happens
when solid 4Tm, stored at −35 °C for a few days, is redissolved
in the same solvents. Nevertheless, the compound is stable
1
Computational Studies on Tm(TpMe2
)
Tm(TpMe2,4Et)2,
enough to give a H NMR spectrum in C6D6 at RT. The
2,
and Ln(TpiPr2)2 (Ln = Sm, Tm). Taking into account the
above experimental data, DFT calculations were carried out in
order to establish the origin of the unexpected change from the
symmetrical sandwich structure of Ln(TpMe2)2 (Ln = Sm, Yb)
spectrum shows a simple pattern, one signal each for the 3- and
5-Me groups and CH2 and CH3 of the 4-Et substituent,
consistent with a symmetrical sandwich-like or a time-averaged
symmetrical solution structure. The solid-state X-ray structure
confirms the former.
Compound 4Tm is isomorphous to the Tm analogue
Sm(TpMe2,4Et)2,28 with the Sm atom on a center of inversion,
Figure 3. The metrical parameters, Table 2, are very similar to
and Ln(TpMe2,4Et
) (Ln = Sm, Tm) to the unusual bent
2
geometry of Ln(TpiPr2)2 (Ln = Sm, Tm). In order to establish
the suitability of the chosen computational method for the
subsequent study of the various structural models of the
Ln(TpiPr2
) compounds, we first compared the optimized
2
computed structures of Ln(TpMe2)2, Ln(TpMe2,4Et)2, 3Sm, and
3Tm with those determined by X-ray crystallography.
The calculated structures of Ln(TpMe2)2, Ln(TpMe2,4Et)2, and
Ln(TpiPr2)2 (Ln = Tm) are shown in Figures SI 3 and SI 4,
respectively, and the key metrical parameters are summarized in
Tables SI 3 and SI 4. There is good agreement between the
experimental and DFT-optimized structures; the calculated
distances are on average 0.08 Å longer and the bond angles
between 2° and 4° larger than in the X-ray crystal structures.
These differences are consistent with the use of large-core
relativistic effective core potentials and the lack of core−valence
correlation effects.43 This agreement with experiment gives us
confidence that our computational methods are appropriate for
the study of these electronically complex molecular systems.
To understand the peculiar bent geometry observed with the
TpiPr2 ligands, we optimized three additional structures: (a) a
linear sandwich structure, analogous to that obtained with the
TpMe2 ligand, (b) an unsymmetrical structure, similar to that
obtained with the TptBu,Me ligand, with one normal Tp ligand
and the other N-bidentate with a single B−H···Ln agostic
interaction, and (c) a (N2H)2-TpiPr2 structure with two
identical TpiPr2 ligands coordinated to the metal center via
two nitrogen donor atoms and an agostic B−H interaction. The
optimized structures are shown in Figures SI 5 and SI 6,
respectively.
Figure 3. Molecular structure of 4Tm viewed down the B---Tm---B
axis, showing the disposition of the ethyl groups. The atoms are drawn
with 20% probability ellipsoids. Hydrogen atoms have been omitted
for clarity. Primed atoms are related to unprimed ones via the
crystallographic inversion center (0, 0, 0) on which the Tm atom is
located.
Table 2. Interatomic Distances (Å) and Angles (deg) for
Tm(TpMe2,4Et)2 (4Tm)
It is noteworthy that for the Ln(TpiPr2)2 compounds the
optimized linear sandwich structure has its minimum energy
with the 3-iPr substituents having a conformation that orients
one isopropyl methyl group toward the metal center; the
alternate orientation, with the isopropyl C−H bond pointing
toward the metal, would result in prohibitive steric congestion
between the methyl groups lying in the equatorial plane.
Nevertheless, the linear structure is strongly destabilized with
respect to the bent geometry by +51.8 kcal mol−1 (Tm) and
+42.7 kcal mol−1 (Sm).
The two structures involving either one (b) or two (c)
agostic B−H interactions are also unstable with respect to the
bent structure of Ln(TpiPr2)2 by 7.6 and 20.9 kcal mol−1 for Sm
and by 7.0 and 17.5 kcal mol−1 for Tm, respectively. For the
Ln(TpiPr2)2 complexes, and unlike for Ln(TptBu,Me)2, the NBO
analysis indicates that the stabilization due to the decrease in
steric hindrance around the metal center and the formation of
Tm−N12
Tm−N22
2.482(5)
2.473(6)
Tm−N32
2.476(6)
N−Tm−N(intra)
torsion angles
N12−Tm−22
N12−Tm−32
N22−Tm−32
78.27(19)
76.87(18)
78.19(19)
Tm−N12 N11−B
Tm−N22 N21−B
Tm−N32 N31−B
2.0(8)
1.6(8)
0.6(9)
N−Tm−N(inter)cis
N−Tm−N(inter)trans
N12−Tm−N22′
N12−Tm−N32′
N22−Tm−N32′
101.73(19)
103.16(18)
101.81(19)
N12−Tm−N12′
N22−Tm−N22′
N32−Tm−N32′
180.0
180.00(18)
180.0
the almost identically sized Yb(II) complex Yb(TpMe2)2. The
average Ln−N distances are Tm−N, 2.477(6) Å, and Yb−N,
2.483(4) Å, respectively. The difference of 0.14 Å in the Ln−N
12070
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
one agostic B−H bond does not offset the energetic cost of
breaking one Ln−N interaction.
driven by release of steric hindrance or whether the
conformational flexibility of the Pr group makes possible
i
In order to identify clearly the origin of the bent structure,
additional interactions. We have already seen that the
symmetrical linear structure, with the Me groups of the
proximal-3iPr pointing toward Ln, is a very high energy
structure. However, close examination of the solid-state
structures of 3Sm and 3Tm and the computed structures
reveals that the conformations of the proximal-3iPr groups in
the expanded, N12N32N52, and contracted, N22N42N52,
faces are different. While those in the expanded face have the
3-iPr C−H bond pointing toward the metal, those on the
contracted face have the C−H pointing toward a nitrogen atom
from another TpiPr2 ligand, with which they are in close contact,
Figure 5. The relevant atoms and distances (in parentheses are
the X-ray values) in complex 3Tm are N22iPrC−H···N62 2.68
(2.61) Å, N42iPrC−H···N32 2.81 (2.62) Å, and N62iPrC−H···
N12 2.89 (2.68) Å. The same is seen in the computed
structures of Ln(TpiPr,Me)2 (Figure 4b) with C−H···N distances
in the same range, N22iPrC−H···N62 2.64 Å, N42iPrC−H···
N32 2.75 Å, and N62iPrC−H···N12 2.97 Å. These C−H···N
contacts are in the range of values normally associated with C−
H···N hydrogen bonds.44 Hence these hydrogen-bonding
interactions provide additional stabilization in the bent-
sandwich structures of Ln(TpiPr2)2 and Ln(TpiPr,Me)2 over and
above the release of steric congestion. This further demon-
strates the influence of the proximal 3-iPr substituents in the
two isomeric compounds with less bulky ligands, Ln(TpMe,iPr
)
2
and Ln(TpiPr,Me)2, were considered. With respect to Ln-
(TpiPr2)2, in the Ln(TpMe,iPr)2 complex, it is the 3-iPr groups
in the vicinity of the metal center (proximal) that are replaced
by methyl groups, while for the Ln(TpiPr,Me)2 complex, it is the
5-iPr groups around the boron atom that are substituted by
methyl groups (distal). In our calculations, the substitution of
the proximal 3-iPr ligands leads to a fully linear structure with a
B−Ln−B angle of 180°, both for Sm and Tm (Figure 4a). On
i
pyrazole ring. Attempts to corroborate the presence of PrC−
H···N H-bonding by IR spectroscopy proved inconclusive.
Although the IR spectra (Figures SI 1and SI 2) show weak
peaks at the lower frequency side of the main C−H stretching
region of the TpiPR2 ligand, it is not clear that these belong to
the, expectedly red-shifted, C−H···N stretches.
Figure 4. Optimized computed structures (side views) of (a)
Ln(TpMe,iPr
) ) compounds. The molecules
and (b) Ln(TpiPr,Me
2
2
shown are for Ln = Tm. The Sm complexes adopt similar geometries.
In (a) Ln(TpMe,iPr)2, the proximal 3-Me groups are close to Ln, and in
(b) Ln(TpiPr,Me)2, the distal 5-Me groups are close to the B atoms.
Hydrogen atoms have been omitted for clarity.
To highlight the crucial influence of the two effects (steric
and electronic), two virtual, model complexes were considered.
In the first one, the three proximal 3-iPr groups in the expanded
N12N32N52 face of Ln(TpiPr,Me)2, not involved in any
hydrogen bonds, were replaced by methyl ligands (Figure
6a), while, in the second complex, the three proximal 3-iPr
groups in the contracted N22N42N62 face, involved in the
formation of the hydrogen bonds, were substituted by methyl
groups (Figure 6b). Of course, these model complexes do not
the other hand, with substitution of the distal 5-iPr ligands by
methyl groups the bent geometry is favored; the B−Ln−B
angle does not change, remaining approximately 154° for both
metals (Figure 4b).
Thus, the origin of the bent structure can be traced to the
presence of the proximal-3iPr groups, since decreasing the steric
hindrance of the ligands by replacing these substituents by
methyl groups leads, in our calculations, to the linear sandwich
structure. Although at first sight trivial, the question remains
whether the distortion toward the bent structure is purely
i
represent the optimal distribution of three Pr groups on a Tp
ligand, but they allow us systematically to investigate the role of
both electronic and steric effects on the metal coordination
Figure 5. Enlarged side view of the optimized computed structure of Tm(TpiPr2)2 (3Tm), showing the close 3-iPrC−H---N contacts on the
contracted N22N42N62 face; N22iPrC−H···N62, N42iPrC−H···N32, and N62iPrC−H···N12 (all other H atoms have been omitted for clarity).
The Sm(TpiPr2)2 (3Sm) compound exhibits similar close contacts.
12071
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
Figure 6. Computed structures (side views) of two model compounds. (a) The 3-iPr groups on the expanded N12N31N52 face and (b) the 3-iPr
groups on the contracted N22N42N62 face have been replaced by 3-Me substituents, respectively. Also shown are the close 3-iPrC−H---N contacts
in the former (N22iPrC−H···N62, N42iPrC−H···N32, and N62iPrC−H···N12) and steric repulsion between the 3-iPr groups in the later; all other
H atoms have been omitted for clarity.
geometry and to probe the origin of the unusual B−Ln−B
angle. Remarkably, in both cases, the structures remain bent
with a B−Ln−B angle around 162° and 166° respectively for
Sm and Tm. The bent geometry of the first structure is
associated with the presence of the close C−H···N contacts
(Figure 6a)a nonclassical hydrogen bondwhereas the
second is the result of steric repulsion between the proximal
3-iPr substituent on each Tp ligand that are facing each other
(Figure 6b). Therefore, the “bent sandwich-like” coordination
geometry of Ln(TpiPr2)2 is intrinsically linked to the nature and
i
conformational flexibility of the Pr groups, which induce
sufficient steric repulsion between the two TpiPr2 ligands to
destabilize the symmetrical linear geometry but also allow the
formation of 3-iPrC−H···N hydrogen bonds, leading to further
stabilization of the bent geometry.
The identification of secondary H-bonding, revealed from
the computational work, as a stabilizing influence of the unusual
coordination geometry of Ln(TpiPr2)2 (Ln = Sm, Tm) is yet
Figure 7. Molecular structure of 3Sm+, view perpendicular to the
another demonstration of the pervasive presence and
importance of H-bonds in just about all aspects of chemistry.45
N12N32N52 plane (front view), as in Figure 2b. The atoms are drawn
with 20% probability ellipsoids. Hydrogen atoms have been omitted
for clarity.
Of course other nonbonding interactions are of import also,
and properly accounting for these effects remains a challenge
for computational methods. There is much current effort
directed to remedy this situation.46,47
Solid-State Structure of [Tm(TpiPr2)2]I (3Tm+). From
B1TmB2 plane perpendicular to the page, illustrates this and
can be compared with the similar view of 3Tm shown in Figure
some preparations of Tm(TpiPr2)2 (3Tm), colorless crystals
were also isolated. To establish the nature of the compound,
single-crystal X-ray diffraction was carried out, which showed
that it was the oxidized complex [Tm(TpiPr2)2]I (3Tm+). The
solid-state structure consists of well-separated [Tm(TpiPr2)2]+
cations and iodide counterions with no short contacts between
iodide and the Tm(III) center.
2b. As with the 3Ln complexes, the bent geometry results in an
expansion of the “front” N12N32N52 and contraction of the
“back” N22N42N62 faces, but this also is not as regular as with
3Ln. This is shown by the respective N−Tm−N(interligand)cis
angles: while the angles N12−Tm−N52/N32−Tm−N52 in
the expanded and N22−Tm−N42/N22−Tm−N62 in the
contracted faces of 3Tm are very similar, the same pair of
angles are some 7° different in 3Tm+: 111.87(6)/119.03(6)°
and 85.61(5)/78.70(5)°, respectively.
A major reason for the differences in bending and the more
irregular geometry is undoubtedly the more congested nature
of 3Tm+ compared to 3Tm, due to the smaller size of Tm(III)
and the shorter Tm−N distances in the former. Indeed the 0.15
Å decrease in the Tm−N distances from 3Tm to 3Tm+
corresponds exactly to the expected change in six-coordinate
ionic radius between the two ions.40
The structure of 3Tm+ is shown in Figure 7, while the
metrical parameters are also listed in Table 1. The molecular
geometry is again of the “bent-sandwich” type, with a B1−Tm−
B2 angle of 152.59(5)°, being close to the value seen in 3Tm.
Though bent, the geometry shows more significant deviation
from the approximate Cs symmetry of complexes 3Ln. Thus,
both N52 and N22 deviate more than 0.3 Å from the B1TmB2
plane, Table SI 1, and the deviation is in the opposite direction;
the distortion tends toward C2 symmetry, as seen with
Sm(TpMe2)2X complexes.28 The view in Figure 7, with the
12072
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
pyrazole was purchased from TCI America, and KTpiPr2 was prepared
by the published procedure of Kitajima,51 with the following
modification: instead of crystallization, excess diisopropylpyrazole
was removed by careful, high-vacuum sublimation, resulting in a higher
The computational study, carried out on the cationic system,
again corroborated the bent geometry (see Table SI 4) and also
revealed close 3-iPrC−H···N contacts in the contracted
N22N42N62 face (X-ray distances in parentheses):
N22iPrC−H···N42 2.47 (2.70) Å, N42iPrC−H···N32 2.81
(2.77) Å, and N62iPrC−H···N12 2.47 (2.82) Å. We conclude
that in 3Tm+ also the bending can be ascribed to the synergistic
effect of steric repulsions between the iPr substituents
(increased with respect to the neutral system by the shrinkage
of the metal coordination sphere in the cationic system) and
the presence of hydrogen bonds.
yield of KTpiPr2
.
NMR spectra were recorded on Varian Inova 300, 400, or 500 MHz
instruments, with shifts reported relative to residual solvent peaks. IR
spectra of Ln(TpiPr2)2 (3Sm, 3Tm) were recorded as cast film from
pentane, on a Thermo Scientific, Nicolet 8700 FT IR instrument, with
KCl plates protected by being sandwiched between O-rings. Elemental
microanalyses were performed on a Carlo Erba (Thermo Fisher
Scientific) CHNS-O EA1108 elemental analyzer by the staff of the
Analytical and Instrumentation Laboratory, University of Alberta.
Synthetic Procedures. Sm(TpiPr2
) (3Sm). A THF solution (6
2
CONCLUSIONS
mL) of KTpiPr2 (1.009 g, 2.00 mmol) was added dropwise to a freshly
prepared THF solution (25 mL) of SmI2 (1.00 mmol) at room
temperature. The initial blue-green color of SmI2 quickly turned very
dark green, accompanied by the formation of a white precipitate of KI.
The mixture was stirred for 1 h and centrifuged to remove the KI
precipitate. Removal of the THF solvent under vacuum resulted in the
formation of a very viscous, dark green oil, which upon further drying
produced a dark green, foamy substance; 1.04 g (95% yield of the
crude Sm(TpiPr2)2 (3Sm)). The crude product was dissolved in 6 mL
of pentane, filtered through a plug of glass microfiber filter, and stored
in a −35 °C refrigerator for 2 days. The green supernatant was
pipetted off, and the dark green precipitate dried under vacuum,
yielding only 107 mg of still somewhat sticky, solid 3Sm. Obtaining
better quality solid 3Sm proved difficult due to the very high solubility
of 3Sm even in hydrocarbon solvents. Eventually, dissolution of the
green solid, after removing pentane from the above supernatant, in 2
mL of ether, followed by addition of 1 mL of hexamethyldisiloxane
(HMDSO), concentration at room temperature, and a further 2 days
in a −35 °C refrigerator, resulted in the precipitation of dark green,
solid 3Sm (470 mg, 43% yield), with satisfactory EA. From one such
crystallization, pure crystals of 3Sm sufficient for X-ray crystallography
and NMR characterization were also obtained. Dark green 3Sm is
stable at room temperature for a few hours and at −35 °C for weeks,
but shows some decomposition to a white solid of unknown
composition.
■
We have shown that it is possible to prepare bis-scorpionate
complexes of the highly reducing thulium(II) ion. Disappoint-
ingly, these complexes have proved to be thermally unstable.
Nevertheless we have obtained structural, spectroscopic, and
computational data that reveal unexpected features, in particular
the bent sandwich-like geometry of the Ln(TpiPr2)2 (3Sm,
3Tm) complexes and the occurrence of an unexpected internal
C−H···N interaction. The effect of the latter is to lend
additional stabilization to this, previously unseen, distortion in
the coordination sphere of bis-scorpionate metal complexes.
This distortion does not arise from metal−ligand bonding, but
rather is driven by the interplay of repulsive clashes between
bulky substituents and the conformational flexibility of the 3-iPr
i
pyrazolyl substituents that allows for favorable PrC-H···N
interactions to develop. The conformational flexibility of some
3-R substituents calls into question estimation of steric bulk of
scorpionate ligands based simply on accepted cone angle values.
Subtle trade-offs between conformations of substituents can
render such judgments suspect.
The success of the calculations in reproducing some of the
more subtle features of the coordination sphere is a testament
to the maturity of DFT calculations in coping with such
complex electronic systems. The results presented here are a
further illustration of the remarkable flexibility of the
pyrazolylborate ligands. It is this flexibility that allows for a
far richer variety of coordination geometries for homoleptic bis-
scorpionate metal complexes compared to the related MX2 and
1H NMR (400 MHz, C7D8, 25 °C, δ ppm): −2.11 (br s, 2H, BH),
−0.85 (d, 36H, 5-iPrCH3), −0.08 (s, 6H, 5-iPrCH), 3.35 (s, 6H, 4-
pzCH), 5.41 (d, 36H, 3-iPrCH3), 14.15 (s, 6H, 3-iPrCH). 13C{1H}
NMR (125.7 MHz, C7D8, 25C, δ ppm): 19.65 (6C, 5-iPrCH), 26.03
(12C, 5-iPrCH3), 41.54 (12C 3-iPrCH3), 53.57 (6C, 4-pzCH), 70.50
(6C, 3-iPrCH), 144.74 (6C, 5-pzC), 184.12 (6C, 3-pzC). 11B{1H}
NMR (128.3 MHz, C7D8, −80 °C, δ ppm): −79.6 (br s). IR, ν(B−H):
2553 and 2492 cm−1. Anal. Calcd for C54H92B2N12Sm: C, 59.98; H,
8.58; N, 15.54. Found: C, 59.50; H, 8.44; N, 14.92.
i
M(C5R5)2 compounds (X = halide, R = Me, Pr, Ph, and 4-
nBuPh; M = alkaline earth and Eu, Sm, Yb). Furthermore, it is
noteworthy that the combination of steric bulk/secondary H-
bonding that, in the case of the planar cyclopentadienyl ligands,
resulted in a change from bent- to parallel-sandwich geometry
produced the opposite effect with the tripodal TpiPr2 ligand.
This vividly demonstrates how the subtle interplay between
nonbonding and secondary H-bonding can have a profound
effect on the geometry of metal compounds and, indeed, on
their structures in general.
Tm(TpiPr2)2 (3Tm). Solid KTpiPr2 (2.86 g, 5.66 mmol) was added in
small portions at room temperature to freshly prepared TmI2 (2.83
mmol), slurried in 15 mL of THF. The mixture immediately turned
from deep green to red-brown and to red. The mixture was stirred for
10−30 min, and the THF removed under vacuum. The sticky product
was extracted with 3 × 10 mL of pentane and centrifuged. The plum-
red supernatant was concentrated to ca. 5 mL and kept in the −35 °C
refrigerator for several days. The precipitated plum-red solid was
isolated by pipetting off the supernatant liquid and drying the solid
under dynamic vacuum; 1.32 g, 42% yield. The resulting Tm(TpiPr2
)
EXPERIMENTAL SECTION
■
2
(3Tm) is sufficiently pure for further reactions. Small amounts of
crystalline 3Tm can be obtained from either concentrated pentane or
pentane/HMDSO solutions kept at −35 °C. These crystallizations
were also accompanied by the formation of a small amount of white
solid material of unknown composition, giving evidence of the
thermally sensitive nature of 3Tm. Solid 3Tm can be kept at −35 °C
for at least a week without noticeable change in color; solutions in
benzene are stable enough for NMR analysis, but with time and even
at −35 °C the compound decomposes to a white solid of unknown
composition. Several attempts were made to obtain elemental analysis
of 3Tm, on different samples, but were frustrated by the extreme
sensitivity of the compound.
General Procedures. All preparations and subsequent manipu-
lations were carried out in a Vacuum Atmospheres glovebox, model
HE-553-2, in an atmosphere of helium/argon. THF, diethyl ether,
toluene, hexane, and pentane were dried by standard methods and
degassed before use. Deuterated solvents, benzene-d6 and toluene-d8,
were dried over Na or Na/K alloy and distilled before use. Samarium
and thulium metals were purchased from HEFA Rare Earth Canada,
Co. Ltd., and fresh filings were used for the synthesis of THF solutions
48
of SmI2 and TmI2.49 KTpMe2,4Et was prepared as previously
described28 with the following modification: 3-ethylpentane-2,4-
dione was prepared by alkylation of pentane-2,4-dione with ethyl
iodide using anhydrous sodium carbonate in acetone.50 Diisopropyl-
12073
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
1H NMR (500 MHz, C6D6, 25 °C, δ ppm): −34.79 (br s, 2H, BH),
−12.55 (s, 6H, 5-iPrCH), −5.08 (s, 36H, 5-iPrCH3), 4.75 (s, 6H, 4-
pzCH), 16.64 (s, 36H, 3-iPrCH3), 53.18 (s, 6H, 3-iPrCH). 11B{1H}
NMR (159.8 MHz, C6D6, 25 °C, δ ppm): −102.1 (br s). IR, ν(B−H):
2555 and 2468 cm−1 Anal. Calcd for C54H92B2N12Sm: C, 58.96; H,
8.43; N, 15.28.
#Lee Carter: Faculty of Law, University of Calgary, Murray
Fraser Hall, 2500 University Drive NW, Calgary, AB T2N 1N4,
Canada.
△Steph
́ anie Labouille: School of Chemistry and Centre for
Computational Chemistry, University of Bristol, Cantock’s
Close, Bristol BS8 1TS, United Kingdom.
Tm(TpMe2,4Et
)
(4Tm). Solid KTpMe2,Et (0.62 g, 1.46 mmol) was
2
added in small portions to a solution of freshly prepared TmI2 (0.73
mmol) in 40 mL of DME at ca. −40 °C. The dark green solution was
allowed to warm slowly to room temperature and stirred for another
0.5 h. During that time, some white solid precipitated. The reaction
mixture was filtered, and the filtrate dried under vacuum, to afford 4 as
dark green solid (0.60 g, 0.64 mmol, 88% yield). Complex 4 is
thermally unstable, even at −35 °C. When stirred in DME at room
temperature, a white solid of unknown composition forms, and the
same happens with solid 4Tm, kept at −35 °C when redissolved in
DME or THF. Single crystals of 4 suitable of X-ray analysis were
grown from a solution of DME and toluene (4:1) at −35 °C. 1H NMR
(C6D6, 27 °C, δ ppm): 44.55 (br, 18H, 3-Me-TpMe2,4Et), 5.27 (br, 12H,
4−CH2CH3-TpMe2,4Et), 3.82 (br, 18H, 4-CH2CH3-TpMe2,4Et), −14.17
(br, 18H, 5-Me-TpMe2,4Et). Anal. Calcd for C46H78B2N12O2Tm: C,
54.07; H, 7.69; N, 16.45. Found: C, 53.88; H, 7.49; N, 16.61.
X-ray Crystallographic Studies. Crystals suitable for single-crystal
X-ray diffraction studies were obtained as described in the
Experimental Section. The crystals were manipulated in the glovebox,
coated with Paratone-N oil, and transferred to a cold gas stream on the
diffractometer. Data were collected on a Bruker D8/APEX II CCD
diffractometer. (Programs for diffractometer operation, data collection,
data reduction, and absorption correction were those supplied by
Bruker.) The data were corrected for absorption by the Gaussian
integration (face-index) method. See Table SI 2 in the Supporting
Information for summaries of crystal data.
Author Contributions
The experimental work was carried out by A.M., L.C., and Y.Y.
A.M. was cosupervised by J.T. and F.N., and L.C. and Y.Y. by
J.T. The calculations and analysis were done by S.L. and I.d.R.;
L.M. supervised S.L. and discussed the results with I.d.R. The
crystal structures were determined by R.McD. The theoretical
part of the manuscript was written by L.M. and I.d.R., and the
rest by A.S. and J.T., with contributions from the other authors.
We thank the Reviewers for their careful work and perceptive
comments.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
J.T. is grateful to NSERC Canada (Discovery Grant) and the
University of Alberta for financial support. L.M. is a member of
the Institute Universitaire de France and also acknowledges the
Humboldt Fundation for a grant for experienced researchers.
REFERENCES
■
(1) Gillespie, R. J.; Hargittai, I. The VSEPR Model of Molecular
Geometry; Allyn & Bacon: Boston, 1991.
The structures of compounds 3Sm, 3Tm, and 3Tm+ were solved by
a Patterson/structure expansion (DIRDIF-2008),52 and that of 4Tm
by direct methods.53 Refinement was completed by full-matrix least-
squares on F2 using the program SHELXL-97.53 Complex 4
crystallized with one molecule of dimethoxyethane. Attempts to refine
peaks of residual electron density as disordered or partial-occupancy
solvent dimethoxyethane oxygen or carbon were unsuccessful. The
data were corrected for disordered electron density through the use of
the SQUEEZE procedure54 as implemented in PLATON.55,56 A total
solvent-accessible void volume of 542 Å3 with a total electron count of
110 (consistent with two molecules of solvent dimethoxyethane or one
molecule per formula unit of 4) was found in the unit cell.
(2) Wharton, L.; Berg, R. A.; Klemperer, W. J. Chem. Phys. 1963, 39,
2023.
(3) Hargittai, M. Chem. Rev. 2000, 100, 2233.
(4) Donald, K. J.; Hoffmann, R. J. Am. Chem. Soc. 2006, 128, 11236.
(5) Evans, W. J.; Bloom, I.; Hunter, W. E.; Atwood, J. L. J. Am. Chem.
Soc. 1981, 103, 6508.
(6) Evans, W. J.; Hughes, L. A.; Hanusa, T. P. J. Am. Chem. Soc. 1984,
106, 4270.
(7) Evans, W. J.; Hughes, L. A.; Hanusa, T. P. Organometallics 1986,
5, 1285.
(8) Andersen, R. A.; Boncella, J. M.; Burns, C. J.; Blom, R.; Haaland,
A.; Volden, H. V. J. Organomet. Chem. 1986, 312, C49.
(9) Andersen, R. A.; Blom, R.; Burns, C. J.; Volden, H. V. J. Chem.
Soc., Chem. Commun. 1987, 768.
(10) Williams, R. A.; Hanusa, T. P.; Huffmann, J. C. Organometallics
1990, 9, 1128.
(11) Kaltsoyannis, N.; Russo, M. R. J. Nucl. Sci. Technol. 2002, 393.
(12) Bytheway, I.; Popelier, P. L. A.; Gillespie, R. J. Can. J. Chem.
1996, 74, 1059.
(13) Timofeeva, T. V.; Lii, J. H.; Allinger, N. L. J. Am. Chem. Soc.
1995, 117, 7452.
ASSOCIATED CONTENT
* Supporting Information
■
S
Simplified diagrams of complexes and ligands discussed in the
paper. Tables of plane calculations, X-ray data, and CIF files for
complexes 3Sm, 3Tm, 3Tm+, IR spectra of Ln(TpiPr2)2, and
4Tm. Computational details, tables of computed distances and
angles, figures of IR spectra of Ln(TpiPr2)2, and optimized
computed structures of Ln(TpMe2)2, Ln(TpMe2,4Et)2, Ln(TpiPr2
)
(14) Hollis, T. K.; Burdett, J. K.; Bosnich, B. Organometallics 1993,
12, 3385.
2
(bent and linear) and the two structures involving formation of
one and two B−H···Ln agostic interactions (Ln = Sm, Tm),
and energies and cartesian coordinates for all computed
structures. This material is available free of charge via the
Internet at http://pubs.acs.org.
(15) Kaupp, M.; Schleyer, P. R.; Dolg, M.; Stoll, H. J. Am. Chem. Soc.
1992, 114, 8202.
(16) Labouille, S.; Clavaguer
1265.
(17) Sitzmann, H.; Dezember, T.; Schmitt, O.; Weber, F.;
́
a, C.; Nief, F. Organometallics 2013, 32,
Wolmershauser, G. Z. Anorg. Allg. Chem. 2000, 626, 2241.
(18) Sitzmann, H.; Dezember, T.; Ruck, M. Angew. Chem., Int. Ed.
1998, 37, 3114.
(19) Deacon, G. B.; Forsyth, C. M.; Jaroschik, F.; Junk, P. C.; Kay, D.
L.; Maschmeyer, T.; Masters, A. F.; Wang, J.; Field, L. D.
Organometallics 2008, 27, 4772.
(20) Ruspic, C.; Moss, J. R.; Schurmann, M.; Harder, S. Angew.
Chem., Int. Ed. 2008, 47, 2121.
(21) Orzechowski, L.; Piesik, D. F. J.; Ruspic, C.; Harder, S. Dalton
Trans. 2008, 4742.
̈
AUTHOR INFORMATION
Corresponding Authors
*E-mail: a.sella@ucl.ac.uk (A.S.).
*E-mail: laurent.maron@irsamc.ups-tlse.fr (L.M.).
*E-mail: joe.takats@ualberta.ca (J.T.).
■
Present Addresses
∥Aurel
́
ien Momin: Lycee
́ ́
Baimbridge, Boulevard des Heros, F-
97139 Les Abymes, Guadeloupe.
12074
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075

Inorganic Chemistry
Article
(22) Harder, S.; Naglav, D.; Ruspic, C.; Wickleder, C.; Adlung, M.;
(56) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7−13.
Hermes, W.; Eul, M.; Pottgen, R.; Rego, D. B.; Poineau, F.;
̈
Czerwinski, K. R.; Herber, R. H.; Nowik, I. Chem.Eur. J. 2013, 19,
12272.
(23) Harder, S.; Naglav, D.; Schwerdtfeger, P.; Nowik, I.; Herber, R.
H. Inorg. Chem. 2014, 53, 2188.
(24) Trofimenko, S. The Coordination Chemistry of Scorpionates-
Polypyrazolylborate Ligands; Imperial College Press: London, 1999.
(25) Pettinari, C. Scorpionates II: Chelating Borate Ligands; Imperial
College Press: London, 2008.
(26) Takats, J.; Zhang, X. W.; Day, V. W.; Eberspacher, T. A.
Organometallics 1993, 12, 4286.
(27) Maunder, G. H.; Sella, A.; Tocher, D. A. J. Chem. Soc., Chem.
Commun. 1994, 885.
(28) Hillier, A. C.; Zhang; Maunder, G. H.; Liu, S. Y.; Eberspacher,
̂
T. A.; Metz, M. V.; McDonald, R.; Domingos, A.; Marques, N.; Day,
V. W.; Sella, A.; Takats, J. Inorg. Chem. 2001, 40, 5106.
(29) Marques, N.; Sella, A.; Takats, J. Chem. Rev. 2002, 102, 2137.
(30) Davies, C. E.; Gardiner, I. M.; Green, J. C.; Green, M. L. H.;
Hazel, N. J.; Grebenik, P. D.; Mtetwa, V. S. B.; Prout, K. J. Chem. Soc.,
Dalton Trans. 1985, 669.
̂
(31) Domingos, A.; Lopes, I.; Waerenborgh, J. C.; Marques, N.; Lin,
G. Y.; Zhang, X. W.; Takats, J.; McDonald, R.; Hillier, A. C.; Sella, A.;
Elsegood, M. R. J.; Day, V. W. Inorg. Chem. 2007, 46, 9415.
(32) Zhang, X. W.; Takats, J.; McDonald, R. New J. Chem. 1995, 19,
573.
(33) Fujisawa, K.; Ono, T.; Ishikawa, Y.; Amir, N.; Miyashita, Y.;
Okamoto, K.-I.; Lehnert, N. Inorg. Chem. 2006, 45, 1698.
(34) Cheng, J.; Ferguson, M. J.; Takats, J. J. Am. Chem. Soc. 2010,
132, 2.
(35) Reinaud, O. M.; Theopold, K. H. J. Am. Chem. Soc. 1994, 116,
6979.
(36) Bleaney, B. R. J. Magn. Reson. 1972, 8, 91.
(37) Piguet, C.; Geraldes, C. F. G. C. In Handbook of Chemistry and
Physics of the Rare Earths; Gschneider, K., Jr.; Bunzli, J.-C.; Pecharsky,
V. K., Eds.; Elsevier BV: North-Holland, 2003; Vol. 33, pp 353−463.
(38) Reger, D. L.; Huff, M. F.; Rheingold, A. L.; Haggerty, B. S. J. Am.
Chem. Soc. 1992, 114, 579.
(39) Cowley, A. H.; Geerts, R. L.; Nunn, C. M.; Carrano, C. J. J.
Organomet. Chem. 1988, 341, C27.
(40) Shannon, R. D. Acta Crystallogr. 1976, A32, 751.
(41) Cheng, J.; Takats, J.; Ferguson, M. J.; McDonald, R. J. Am.
Chem. Soc. 2008, 130, 1544.
(42) Nief, F. In Handbook on the Physics and Chemistry of Rare Earths;
Gschneider, K., Jr.; Bunzli, J.-C.; Pecharsky, V. K., Eds.; Handbook on
the Physics and Chemistry of Rare Earths; Elsevier BV: North-
Holland, 2010; Vol. 40, pp 241−300.
(43) Maron, L.; Eisenstein, O. J. Phys. Chem. A 2000, 104, 7140.
(44) Taylor, R.; Kennard, O. J. Am. Chem. Soc. 1982, 104, 5063.
(45) Arunan, E.; Desiraju, G. R.; Klein, R. A.; Sadlej, J.; Scheiner, S.;
Alkorta, I.; Clary, D. C.; Crabtree, R. H.; Dannenberg, J. J.; Hobza, P.;
Kjaergaard, H. G.; Legon, A. C.; Mennucci, B.; Nesbitt, D. J. Pure Appl.
Chem. 2011, 83, 1619.
(46) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
(47) Corminboeuf, C. Acc. Chem. Res. 2014, 140321125824001.
(48) Namy, J. L.; Kagan, H. B.; Caro, P. E. New J. Chem. 1981, 5, 479.
(49) Bochkarev, M. N.; Fagin, A. A. Chem.Eur. J. 1999, 5, 2990.
(50) Johnson, A. W.; Markham, E.; Price, R. Organic Syntheses; 1973;
Collect. Vol. V, p 785
(51) Kitajima, N.; Fujisawa, K.; Fujimoto, C.; Morooka, Y.;
Hashimoto, S.; Kitagawa, T.; Toriumi, K.; Tatsumi, K.; Nakamura,
A. J. Am. Chem. Soc. 1992, 114, 1277.
(52) Beurskens, P. T.; Beurskens, G.; de Gelder, R.; Smits, J. M. M.;
Garcia-Granda, S.; Gould, R. O. The DIRDIF-2008 program system
Crystallography Laboratory, University of Nijmegen: The Nether-
lands, 2008. http://www.xtal.science.ru.nl/dirdif/software/dirdif.html.
(53) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(54) van der Sluis, P.; Spek, A. L. Acta Crystallogr. 1990, A46, 194.
(55) Spek, A. L. Acta Crystallogr. 1990, A46, C36.
12075
dx.doi.org/10.1021/ic501816v | Inorg. Chem. 2014, 53, 12066−12075