Differences in the cyclometalation reactivity of bisphosphinimine-supported organo-rare earth complexes

Article abstract of DOI:10.1039/c4dt00863d

The pyrrole-based ligand N,N′-((1H-pyrrole-2,5-diyl) bis(diphenylphosphoranylylidene))bis(4-isopropylaniline) (HL_B) can be deprotonated and coordinated to yttrium and samarium ions upon reaction with their respective trialkyl precursors. In the case of yttrium, the resulting complex [L_BY(CH₂SiMe₃)₂] (1) is a Lewis base-free monomer that is remarkably resistant to cyclometalation. Conversely, the analogous samarium complex [L_BSm(CH ₂SiMe₃)₂] is dramatically more reactive and undergoes rapid orthometalation of one phosphinimine aryl substituent, generating an unusual 4-membered azasamaracyclic THF adduct [κ ⁴-L_BSm(CH₂SiMe₃)(THF)₂] (2). This species undergoes further transformation in solution to generate a new dinuclear species that features unique carbon and nitrogen bridging units [κ¹:κ²:μ²-L_BSm(THF)] ₂ (3). Alternatively, if 2 is intercepted by a second equivalent of HL_B, the doubly-ligated samarium complex [(κ⁴-L _B)L_BSm] (4) forms.

Full text of DOI:10.1039/c4dt00863d

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Differences in the cyclometalation reactivity of
bisphosphinimine-supported organo-rare earth
complexes†
Cite this: Dalton Trans., 2014, 43,
10739
Matthew T. Zamora, Kevin R. D. Johnson, Mikko M. Hänninen and Paul G. Hayes*
The pyrrole-based ligand N,N’-((1H-pyrrole-2,5-diyl)bis(diphenylphosphoranylylidene))bis(4-isopropyl-
aniline) (HLB) can be deprotonated and coordinated to yttrium and samarium ions upon reaction with
their respective trialkyl precursors. In the case of yttrium, the resulting complex [LBY(CH2SiMe3)2] (1) is a
Lewis base-free monomer that is remarkably resistant to cyclometalation. Conversely, the analogous
samarium complex [LBSm(CH2SiMe3)2] is dramatically more reactive and undergoes rapid orthometalation
of one phosphinimine aryl substituent, generating an unusual 4-membered azasamaracyclic THF adduct
[κ4-LBSm(CH2SiMe3)(THF)2] (2). This species undergoes further transformation in solution to generate a
new dinuclear species that features unique carbon and nitrogen bridging units [κ1:κ2:μ2-LBSm(THF)]2 (3).
Alternatively, if 2 is intercepted by a second equivalent of HLB, the doubly-ligated samarium complex
[(κ4-LB)LBSm] (4) forms.
Received 22nd March 2014,
Accepted 2nd May 2014
DOI: 10.1039/c4dt00863d
www.rsc.org/dalton
(Chart 1, HLA).49–53 These ligands can be attached to rare earth
metal centres via either an alkane elimination route or by a
Introduction
The organometallic chemistry of trivalent rare earth complexes salt metathesis pathway.53 Depending on the size of the metal,
has found many important applications in the field of cata- one or more molecules of coordinating solvent is occasionally
lysis, including the metal-catalyzed hydrophosphonylation of required to fill the coordination sphere of the resulting
carbonyls,1 hydrogenation,2–7 the catalytic addition of H–X complex and prevent decomposition.
to
unsaturated
moieties
(e.g.
hydroamination,8–16
We are highly interested in monitoring the metalation reac-
hydrosilylation,10,16–20 hydrophosphination,21–24 hydrobora- tivity of these bisphosphinimine-ligated complexes owing to
tion,25,26 and hydroalkoxylation27–30), and especially polymeriz- the ubiquitous nature of this process, especially in hydro-
ation chemistry.31–44 However, compared to research on elementation catalysis. Metalation is described frequently in
organotransition-metal species, the study of rare earth com- modern literature, although generally as an undesirable ligand
plexes remains much less explored. For instance, stabilization decomposition pathway. However, prior to 1960, only Group I
of rare earth organometallic complexes has traditionally metal alkyls were known to prominently convert sp2 or sp3
focused on simple carbocyclic frameworks,45 particularly cyclo- C–H bonds to a C–M bonded species. It was subsequently
pentadienyl motifs. More contemporary studies have incorpor- shown by Kaska, while under the supervision of Eisch,54,55
ated a broad array of tunable ligand scaffolds that can provide that metalative processes such as direct sp2 C–H metalation
a plethora of different steric and electronic environments.46,47
could be facilitated by other organometallic species. Further
As a result, the development of new ancillary ligands to pioneering work by Kaska demonstrated that even sp3 C–H
support stoichiometric and catalytic transformations at rare
earth metal centres is an important and intensively studied
area of organolanthanide chemistry.31,45,47,48 To this end, we
have recently reported a variety of carbazole-based bisphosphi-
nimine frameworks for supporting complexes of Sc, Y and Lu
Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University
Drive, Lethbridge, AB, Canada, T1K 3M4. E-mail: p.hayes@uleth.ca
†Electronic supplementary information (ESI) available: Atomic coordinates,
interatomic distances and angles, anisotropic thermal parameters, and hydrogen
parameters for 1, 2, 3, 4, and 5. CCDC 991840–991844 for 1–5. For crystallo-
Chart 1 Carbazole (HLA) and pyrrole (HLB) bisphosphinimine proteo
graphic data in CIF format see DOI: 10.1039/c4dt00863d
ligands.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10739–10750 | 10739

View Article Online
Paper
Dalton Transactions
metalation could be promoted by intramolecular processes were interested in exploring the stability of even larger rare
that use the entropically-driven attachment of
a pincer earth ions. Hence, we targeted the preparation of an analogous
ligand.56 Over the decades that followed, the importance of the complex of the slightly larger rare earth metal yttrium in order
C–M bond has been realized in a multitude of molecular and to establish whether C–H activation of the ligand substituents
materials applications that are intrinsically important to the would prevail.
future of sustainable catalysis. However, studies of C–H acti-
The yttrium complex 1 was prepared by reaction of
vation in the context of gaining the understanding required to [Y(CH2SiMe3)3(THF)2]60–62 with the proteo ligand HLB (R = Ph,
rationally develop metalation-resistant ligands has served as Ar = Pipp; Pipp = para-isopropylphenyl) in a pentane–THF
the impetus for our studies in this area.
solvent mixture at ambient temperature. Although [Y(CH2Si-
In our research focussed on rare earth complexes, we have Me3)3(THF)2] can be isolated, it is thermally sensitive and 1
observed that lanthanide alkyl derivatives supported by can also be obtained through an in situ alkane elimination
ligands of type A commonly undergo several C–H cyclometala- process (Scheme 2).
tion processes, however, these transformations often limit sub-
Formation of complex 1 was evident by monitoring the reac-
sequent reactivity. For example, the lutetium complex [LALu- tion by multinuclear NMR spectroscopy. Specifically, in the 31
P
(CH2SiMe3)2] rapidly decomposes via two stepwise intramole- {1H} NMR spectrum, disappearance of the signal for free
cular cyclometalation reactions at ambient temperature proteo ligand (δ −8.1) and concomitant emergence of one new
(Scheme 1).49 As a result, we initiated studies to generate a signal further downfield at δ 25.0 with diagnostic splitting (d,
new set of ligands in an attempt to prepare more thermally 2JP–Y = 7.5 Hz; 89Y = 100% abundant, I = 1/2) indicated that the
robust rare earth metal complexes.
ligand was coordinated to Y symmetrically via both phosphini-
mine donors. The 1H NMR spectrum is also consistent with
that expected for complex 1, particularly, the presence of an
upfield resonance at δ −0.04 which can be attributed to the
four –CH2SiMe3 methylene protons. The pyrrole resonance is
split into a doublet of doublets, due to coupling to two magne-
3
4
tically inequivalent 31P nuclei (dd, JH–P = 2.1 Hz, JH–P
1.2 Hz). As was observed with the previously reported Sc and
Lu analogues, complex contains no coordinated THF
=
1
Scheme 1 Cyclometalative behaviour of a carbazole-based bisphosphi-
ligands, confirmed by the absence of THF resonances in the
1H NMR spectrum; this is a notable feature considering that
base-free, 5-coordinate yttrium complexes are relatively
unusual. Finally, the 13C{1H} NMR spectrum closely matches
those of related Lu and Sc complexes,57 with the exception that
this spectrum’s CH2 resonance exhibits the expected one-bond
coupling to yttrium (d, 1JC–Y = 39 Hz).
Single crystals of 1 suitable for X-ray diffraction were readily
obtained by slow diffusion of pentane into a concentrated
toluene solution of 1 at −35 °C. The molecular structure of 1 is
depicted in Fig. 1 as a thermal displacement plot. In the solid
state, 1 is monomeric with the yttrium centre coordinated by
two trimethylsilylmethyl groups and the tridentate pyrrole
ligand bound by the three nitrogen atoms. The geometry is
best described as distorted trigonal bipyramidal with the equa-
torial plane defined by N1 and the alkyl groups, while the
bisphosphinimine donors (N2 and N3) occupy the apical sites.
The bond angles about the equatorial plane are close to the
ideal value of 120° (N1–Y1–C47 = 115.81(9)°, N1–Y1–C51 =
nimine lutetium complex.
Accordingly, we designed a new pyrrole-based bisphosphi-
nimine ligand (Chart 1, HLB) which, upon coordination to a
variety of rare earth metals, yielded substantially more stable
complexes, even when bearing substituents identical to the
carbazole-based analogue.57 More specifically, solutions of
[LBMR2] (M = Sc, Lu, Er) complexes are resistant to cyclometa-
lation at elevated temperatures (60 °C for more than 4 h),57
whereas A-type complexes immediately begin to convert into
P- and/or N-aryl cyclometalated products at ambient tempera-
ture, despite the presence of coordinating solvent.49,50 Owing
to the stabilizing effect of ligand LB, we became interested in
expanding the study of this pyrrole-based ligand to include
even larger rare earth metals. Our results are presented herein.
Results and discussion
Yttrium complex synthesis and stability
In contrast to transition metal chemistry, coordination
numbers of organolanthanide complexes are heavily governed
by the ionic radius of the encapsulated metal centre.58,59 Our
previously reported organometallic complexes of LB were of
general form [LBM(CH2SiMe3)2] (M = Sc, Lu and Er); these rare
earth metals possess small ionic radii relative to the rest of the
lanthanide series,58 with the erbium congener being the
largest of the three. Due to the fact that the erbium organo-
Scheme 2 Formation of pyrrole-based bisphosphinimine yttrium
metallic complex of LB was found to be thermally robust, we complex 1.
10740 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
complexes of LB, we were curious as to whether complexes
ligated by our new pyrrole-based bisphosphinimine ligand
would remain inert to C–H activation, or if these species
would succumb to cyclometalative transformations akin to our
prior carbazole-based systems.
Rather than the expected dialkyl species 2′, reaction of the
proteo ligand HLB with [Sm(CH2SiMe3)3(THF)3]65 resulted in
formation of the cyclometalated organosamarium product 2
(Scheme 3). Low overall reaction yields were obtained when
the trialkyl samarium reagent was isolated prior to reaction
with HLB, due to the fact that it rapidly decomposes at temp-
eratures above −35 °C. However, complex 2 was obtained in
substantially higher yields when [Sm(CH2SiMe3)3(THF)3] was
prepared in situ by reaction of [SmCl3(THF)2] with three equi-
valents of LiCH2SiMe3 in THF at −78 °C, and subsequently
reacted with a toluene solution of the proteo ligand.
The reaction leading to the formation of complex 2 most
likely proceeds through the putative non-cyclometalated THF
adduct (2′) that contains two trimethylsilylmethyl groups.
However, efforts to isolate 2′ met with little success, a fact
which was not surprising given that this species could not be
observed spectroscopically, even at temperatures as low as
−78 °C in toluene-d8. Presumably dialkyl 2′ is highly unstable
and upon formation immediately undergoes cyclometalation
of one of the ancillary ligand N-aryl rings to form the four-
membered azasamaracycle 2. Hence, the rate constant for
decomposition of 2′ is likely larger than that of its formation.
When the reaction was monitored spectroscopically in
Fig. 1 Thermal displacement plot (30% probability) of complex 1 with
hydrogen atoms and minor disordered components omitted for clarity.
Selected bond distances (Å) and angles (°): P1–N2 = 1.606(2), P2–N3 =
1.609(2), Y1–N1 = 2.345(2), Y1–N2 = 2.374(2), Y1–N3 = 2.410(2), Y1–C47
= 2.412(3), Y1–C51 = 2.394(3); C1–P1–N2 = 103.3(1), C4–P2–N3 =
104.2(1), C47–Y1–C51 = 127.3(1), N2–Y1–N3 = 141.96(7).
116.88(9)°, C47–Y1–C51 = 127.3(1)°); however, the apical
bond angle deviates significantly from 180° (N2–Y1–N3 =
141.96(7)°). As a result, the structure bears a striking resem-
blance to its Lu and Er analogues.57 Complex 1 exhibits Y–C
bond lengths (Y1–C47 = 2.412(3) Å, Y1–C51 = 2.394(3) Å) which
fall within the range for typical Y–CH2SiMe3 bonds.63
Additionally, these bond lengths are similar to those observed
in the reported Er analogue (2.375(6) Å, 2.397(5) Å), and
slightly longer than those found in the corresponding Lu
congener (2.347(4) Å, 2.355(4) Å).57
Since previously-reported rare earth (Sc, Lu, Er) dialkyl com-
plexes ligated by LB did not display evidence of intramolecular
C–H activation,57 we were interested in determining if the
larger and more reactive yttrium metal centre in complex 1
would be prone to cyclometalation at either the phosphorus or
nitrogen aryl substituents. Such a process would generate a 4-
or 5-membered heteroyttricycle, depending on the site of C–H
activation. However, similar to the previously-reported Sc, Lu
and Er species, complex 1 is surprisingly resistant to cyclo-
metalation, with no C–H activation occurring at either the
phenyl substituents on phosphorus, or the nitrogen Pipp
groups. More specifically, complex 1 showed no sign of cyclo-
metalation even after several hours in solution (benzene-d6) at
ambient temperature.
Synthesis and cyclometalation reactivity of organosamarium
derivatives
In the search for more reactive metals, we became intrigued by
the unique reactivity of other lanthanide complexes. Samarium
complexes were of particular interest, especially because of
their utility in such processes as organic synthesis,4 multifunc-
tional asymmetric catalysis,64 and hydroamination chemistry.8
Furthermore, given the even larger ionic radius of samarium(III)
Scheme 3 Synthesis
of
bisphosphinimine-ligated
samarium
relative to that of our previously described trivalent rare earth complexes.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10739–10750 | 10741

View Article Online
Paper
Dalton Transactions
benzene-d6, the 31P{1H} NMR resonance attributed to the free planarity, as indicated by the sum of the internal angles (∑∠ =
ligand HLB (δ −8.1) disappeared as two new signals emerged at 358°). As expected, the THF ligand in the equatorial plane is
δ 24.5 and 18.1 in a 1 : 1 ratio. This ligand asymmetry was also bound more strongly (Sm1–O2 = 2.480(6) Å) than the THF
observed in the 1H NMR spectrum whereby several overlapping molecule trans to the trimethylsilylmethyl group (Sm1–O1 =
resonances afforded a much more complicated spectrum than 2.588(6) Å). This difference can be rationalized either on the
would be expected for 2′. For example, the two diagnostic basis of steric repulsions between the apical THF ligand and
doublets attributed to the N-aryl groups are both more complex the adjacent aryl substituents, or by the larger trans influence
than anticipated and each signal integrates to only 1H. Fur- exerted by the alkyl ligand. The lengths of the phosphinimine
thermore, four separate signals were observed for the iPr PvN bonds (P1–N2 = 1.571(6) Å, P2–N3 = 1.593(6) Å) correlate
methyl groups, which is more likely a result of asymmetry well with other examples in the literature whereby the phosphi-
rather than restricted rotation about the Cipso–CH bond. nimine functionality exhibits considerable PvN double bond
Notably, these NMR spectra are generally devoid of the features character.49,57,67 Similar to other rare earth complexes sup-
characteristic of strongly paramagnetic complexes: most ported by this ligand, the amido group of the pyrrole moiety is
signals exhibit relatively routine chemical shifts and the reson- slightly closer (Sm1–N1 = 2.454(5) Å) to the metal centre than
ances are narrow with widths at half height (whh) in the range the phosphinimine nitrogen (Sm1–N3 = 2.582(6) Å), suggesting
of those typically observed for diamagnetic compounds (2.0 that the anionic charge is primarily localized on N1. Interest-
Hz ≤ whh ≤ 3.4 Hz); however, some minor paramagnetic ingly however, the other samarium–phosphinimine bond is
characteristics are also evident in the spectra. This phenom- significantly contracted (Sm–N2 = 2.436(6) Å), most likely due
enon can be attributed to the relatively weak magnetic to the physical constraints imposed by the cyclometalation-
moment of samarium compared to most of the other trivalent induced metallacycle that encompasses the Pipp group of N1.
lanthanides.66
Accordingly, the interatomic distance between the samarium
High quality crystals of 2 were obtained from slow diffusion centre and the ortho-carbon of the Pipp group indicates a Sm–C
of pentane into a concentrated THF solution of 2 at −35 °C. In single bond (Sm1–C34 = 2.551(7) Å). Notably, the four-mem-
the solid state, the heptacoordinate samarium complex 2 is bered azasamaracycle is highly strained with bond angles
defined by one trimethylsilylmethyl group, a tetradentate deviating considerably from ideality (Sm1–C34–C29 = 93.5(5)°,
pyrrole ligand bound by the three expected nitrogen atoms as C34–C29–N2 = 111.7(7)°, C29–N2–Sm1 = 98.2(4), N2–Sm1–C34
well as the ortho carbon of one phosphinimine N-Pipp substi- = 56.1(2)°).
tuent, and two cis-oriented THF solvent molecules (Fig. 2).
Although complex 2 is stable in toluene solution at
Complex 2 adopts a 7-coordinate geometry that is best −35 °C, it slowly converts into a different species at ambient
described as pentagonal bipyramidal, with the alkyl ligand temperature (t1/2 = 1 h). This transformation is apparent when
and one THF ligand residing in the apical sites. The equatorial monitoring complex 2 by NMR spectroscopy, especially in the
positions about samarium do not deviate substantially from 31P{1H} spectra (benzene-d6) wherein resonances at δ 24.5 and
18.1 from 2 are gradually replaced by two new signals at δ 39.4
and 24.7. As was discussed above for 2, the two different phos-
phorus environments are indicative of ligand asymmetry.
Moreover, the 1H NMR spectrum exhibits resonances that
suggest a species of increased complexity relative to that of 2.
In addition to an increase in aryl proton environments, the
extrusion of one equivalent of tetramethylsilane was evident
from the emergence of a singlet at δ 0.00. Owing to the procliv-
ity of our carbazole-based bisphosphinimine complexes to
undergo C–H activation at the ortho-position of the phos-
phorus phenyl substituents, we postulated this new compound
might be the result of another ligand cyclometalation at the
opposite phosphinimine P-aryl site (Scheme 3, 2a), since a
second N-Pipp activation would be expected to afford a sym-
metric species.
Through isolation of X-ray quality single crystals of this com-
pound (vide infra), we were able to confirm that this complex
was indeed the result of P-phenyl cyclometalation. However,
Fig. 2 Thermal displacement plot (20% probability) of complex 2 with the process appears to proceed at the phosphorus aryl group
hydrogen atoms and minor disordered components omitted for clarity.
Selected bond distances (Å) and angles (°): P1–N2 = 1.571(6), P2–N3 =
1.593(6), Sm1–N1 = 2.454(5), Sm1–N2 = 2.436(6), Sm1–N3 = 2.582(6),
Sm1–C34 = 2.551(7), Sm1–C47 = 2.47(1), Sm1–O1 = 2.588(6), Sm1–
adjacent to the azasamaracycle and intermolecularly, ejecting
an equivalent of TMS and forging several C- and N-bridging
interactions to generate complex 3, an unexpected C2-sym-
metric samarium dinuclear complex (Scheme 3, 3). Spectro-
scopically, it would be difficult to distinguish between 2a and 3.
O2 = 2.480(6); C1–P1–N2 = 102.3(3), C4–P2–N3 = 105.6(3), C47–Sm1–
O1 = 177.1(3), N1–Sm1–C34 = 122.7(2).
10742 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
i
i
As such, we were only able to ascertain the salient features of [(µ-NC6H3 Pr2-2,6)Sm(µ-NHC6H3 Pr2-2,6)(µ-Me)AlMe2]2
and
the structure of complex 3 from X-ray diffraction studies. 3.4300(4) Å for [(Et8-calix-pyrrole)Sm2(THF)2(µ-Cl)2][Li(THF)2],
Similar to 2, single crystals of 3 suitable for X-ray diffraction respectively. Although the shortest known Sm–Sm contact
were grown by slow diffusion of pentane into a benzene solu- comes from a different report by Gambarotta (3.3159(5) Å) for
tion of the complex at ambient temperature. The molecular [(OEPG)Sm2(Et2O)2] (OEPG = deprotonated octaethylporphyri-
structure of 3 is depicted in Fig. 3 as a thermal displacement nogen), each samarium atom in that complex is formally
plot. Complex 3 features a samarium centre coordinated by described as Sm(II), and as a result, is likely involved in some
one THF ligand, the three nitrogen atoms of the pyrrole degree of metal–metal bonding.77
bisphosphinimine scaffold, the carbon atom from the afore-
The dinuclear nature of 3 represents the first example of a
mentioned azasamaracycle, one phosphinimine nitrogen atom dinuclear samarium complex bridged by an η1:μ2-phenyl group
from the opposing pyrrole ligand azasamaracycle, one of the and only the third example of the whole lanthanide series with
ortho carbons from cyclometalation of the opposite pyrrole’s a similar bridging fragment.78,79 Although the NMR spectral
azasamaracycle-adjacent phosphinimine phenyl groups, as parameters of 3 indicate this complex is not strongly paramag-
well as the azasamaracyclic carbon from the opposite bispho- netic, analysis of this complex by air-sensitive SQUID measure-
sphinimine ligand. Considering these bonds are reciprocated ments to determine its magnetic properties is still of interest,
at the other samarium atom, 3 involves pairs of µ-(phenylido)- and the potential of the bridging system to evoke magnetic
1κC,2κC, μ-(phosphinimine)1κN,2κN, and μ-(phosphinimine- exchange coupling between samarium centres will be under-
phenylido)1κN,2κo-C bridging groups. This represents a rare taken in due course.
example of a dinuclear samarium complex possessing a brid-
As noted above, a variety of difficulties were encountered
ging carbon atom, with only a few other dinuclear68–72 or tri- during the attempted preparation of complex 2′. It was not
nuclear73,74 complexes known. The bridging nature of these until slow, dropwise addition of the proteo ligand HLB into a
groups is evident from the close contacts (Sm1–C39 = 2.638(2), solution of [Sm(CH2SiMe3)3(THF)3] was undertaken at low
Sm1–N3 = 2.616(2), Sm1–C27′ = 2.597(2) Å). As before, the aza- temperature that we were able to isolate a discrete complex (2).
samaracycle adopts a highly-strained 4-membered geometry In an effort to establish possible competing reactions, we pur-
(N3–Sm1–C39 = 51.93(5), Sm1–C39–C38 = 90.3(1), C39–C38–N3 posely added a solution of HLB rapidly to a cooled (–35 °C)
= 108.7(2), C38–N3–Sm1 = 91.0(1)°).
toluene solution of [Sm(CH2SiMe3)3(THF)3]. Hence, we were
At 3.3655(6) Å, the distance between the samarium centres able to identify complex 4 (Scheme 3) as the main product
is too large to suggest Sm–Sm bonding. However, the intera- under such conditions. This doubly-ligated complex
tomic distance is one of the shortest known, with reports by (vide infra) possesses two bisphosphinimine ligands, one of
Gordon75 and Gambarotta76 being only marginally shorter, which contains the azasamaracyclic moiety in 2. The two ancil-
wherein they report distances of 3.3611(10)
Å
for lary ligands are attached to Sm in an orthogonal arrangement
that appears to leave insufficient space for Lewis basic solvent
molecules to coordinate. Considering the structure of 4, it is
reasonable to postulate that 2 is an intermediate en route to
complex 4. To support this hypothesis, a sample of 2 was
treated with excess HLB, which resulted in formation of bis-
(ligated) complex 4, as well as several other minor products.
The formulation of complex 4 was supported by spectro-
scopic analysis. Three signals are evident in the 31P{1H} NMR
spectrum in a 1 : 1 : 2 ratio. The ligand asymmetry can also be
1
seen in the H NMR spectrum, whereby another complicated
array of overlapping resonances is observed. Similar to
i
i
complex 2, four separate Pr methyl and three separate Pr
methine signals were observed, but unlike either 2 or 3, no
THF resonances were visible.
The molecular structure of 4 is depicted in Fig. 4 as a
thermal displacement plot. As mentioned previously, complex
4 bears two inequivalent bisphosphinimine ligands, one with
the κ4 arrangement observed in complex 2 and one that is
bound in a prototypical tridentate fashion. This base-free
Fig. 3 Thermal displacement plot (30% probability) of complex 3 with
hydrogen atoms, benzene solvent molecules, THF backbones, Pipp pseudopentagonal bipyramidal complex bears
a strained
iPr groups, and minor disordered components omitted for clarity. Sym-
metry code ’ = 2 − x, −y, −z. Selected bond distances (Å) and angles (°):
P1–N2 = 1.600(2), P2–N3 = 1.609(1), Sm1–N1 = 2.477(2), Sm1–N2 =
2.530(2), Sm1–N3 = 2.616(2), Sm1–C39 = 2.638(2), Sm1–C27’ = 2.597(2),
Sm1–O1 = 2.575(2), Sm1⋯Sm1’ = 3.3655(6); P2–N3–C38 = 136.7(1),
N3–P2–C26 = 109.0(1).
metallacycle (Sm1–C34–C29 = 94.3(2)°, C34–C29–N2 = 111.6
(3)°, C29–N2–Sm1 = 99.3(2)°, N2–Sm1–C34 = 54.8(1)°) and
typical phosphinimine PvN (P1–N2 = 1.577(3) Å, P2–N3 =
1.597(3) Å), samarium–nitrogen (Sm1–N1 = 2.509(3) Å), and
samarium–carbon bond lengths (Sm1–C34 = 2.603(4) Å).
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10739–10750 | 10743

View Article Online
Paper
Dalton Transactions
Fig. 5 Thermal displacement plot (30% probability) of the cation of
complex 5 with hydrogen atoms, non-ipso carbons of the P-phenyl
rings, non-ipso carbons of the N12 and N13 N-aryl rings, isopropyl
groups of the N2 and N3 N-aryl rings and minor disordered components
omitted for clarity. Selected bond distances (Å) and angles (°): P1–N2 =
1.634(2), P2–N3 = 1.637(2), P11–N12 = 1.642(2), P12–N13 = 1.637(2),
Lu1–N1 = 2.340(2), Lu1–N2 = 2.324(2), Lu1–N3 = 2.326(2), Lu1–N11 =
Fig. 4 Thermal displacement plot (30% probability) of complex 4 with
hydrogen atoms, P-phenyl rings (except ipso carbons), N-aryl rings
(except ipso carbons) on the non-cyclometalated ligand and isopropyl
substituents on the cyclometalated ligand omitted for clarity. Selected
bond distances (Å) and angles (°): P1–N2 = 1.577(3), P2–N3 = 1.597(3),
Sm1–N1 = 2.509(3), Sm1–N2 = 2.471(3), Sm1–N3 = 2.615(3), Sm1–C34
= 2.603(4), Sm1–N11 = 2.466(2), Sm1–N12 = 2.469(2), Sm1–N13 =
2.472(3); N12–Sm1–N13 = 136.21(9).
2.338(2), Lu1–N12
= 2.326(2), Lu1–N13 = 2.352(2); N1–Lu1–N11 =
178.22(7), N2–Lu1–N3 = 143.90(7), N12–Lu1–N13 = 143.60(7).
Notably, the angle at samarium involving the apical phosphini-
mine nitrogens (N12–Sm1–N13 = 136.21(9)°) is dramatically
smaller than the corresponding angle in complex 2 (C47–Sm1–
O1 = 177.1(3)°), presumably because N12 and N13 belong to
the same pincer ligand.
the case of Y, the ligated complex is resistant to cyclometala-
tion, unlike its carbazole congener, and is notably free of
Lewis bases, a testament to the versatility of this pyrrole
system. Owing to the remarkable stability of these complexes,
we are in the process of using this backbone to prepare mixed
alkyl/amido yttrium complexes, which have the potential to
lead to elusive rare earth imido complexes. Investigations in
this area are on-going.
In the case of the larger, more reactive Sm, reaction of [Sm-
(CH2SiMe3)3(THF)3)] with HLB is extremely facile, however the
putative dialkyl complex [LBSm(CH2SiMe3)2] immediately
undergoes cyclometalation of an N-aryl group to form a some-
what “protected” variant [κ4-LBSm(CH2SiMe3)(THF)2] (2). This
azasamaracyclic THF adduct is reasonably stable at low temp-
erature and in the solid state, but it converts over time in solu-
tion to a unique carbon- and nitrogen-bridged dinuclear
product [κ1:κ2:μ2-LBSm(THF)]2 (3). However, if intercepted by
another equivalent of HLB, the alkyl group of 2 appears to
facilitate another alkane elimination reaction, resulting in a
doubly-ligated samarium complex [(κ4-LB)LBSm] (4).
Intriguingly, complex 4 bears a striking similarity to the cat-
ionic lutetium complex [(LB)2Lu]+[B(C6F5)4]− (5), which was
produced as
a disproportionation byproduct during the
decomposition of previously reported [LBLu(CH2SiMe3)-
(OEt2)2]+[B(C6F5)4]− (5′).57 Although only a small quantity of
crystalline 5 was isolated, rendering further characterization
impossible, it nonetheless serves as a useful crystallographic
comparison to complex 4. Both complexes feature two LB
derived ligands; however, the lutetium centre in complex 5
exhibits a distorted octahedral geometry resulting from two tri-
dentate LB moieties (Fig. 5). Furthermore, although the trans
amido groups form an angle with Lu that is relatively close to
180° (N1–Lu1–N11 = 178.22(7)°), the other angles about the
metal centre are severely strained (N2–Lu1–N3 = 143.90(7)°,
N12–Lu1–N13 = 143.60(7)°). Also of interest is the fact that the
lutetium amido bonds (Lu1–N1 = 2.340(2), Lu1–N11 = 2.338(2)
Å) are longer than the other lutetium–nitrogen distances (Lu1–
N2 = 2.324(2), Lu1–N3 = 2.326(2), Lu1–N12 = 2.326(2), Lu1–N13 =
2.352(2) Å). This is most likely an artifact of the trans influence
of these groups, which are located perfectly opposite to each
other, unlike the other “trans” groups in this species.
In many previous studies, dialkyl rare earth complexes have
often been used to form alkyl/amido complexes [LnMR(NHR)]
via the elimination of an alkyl group when a primary amine is
added. These hybrid alkyl/amido species are of interest
because a subsequent intramolecular alkane elimination reac-
tion could potentially afford elusive rare earth imido
[LnMvNR] complexes, which have been popularized in con-
temporary reports.80,81 Due to the presence of only one remain-
Conclusions
Our newly-reported pyrrole-based bisphosphinimine pincer ing –CH2SiMe3 group, complex 2 cannot be used as a direct
ligand is able to support rare earth complexes of Y and Sm. In precursor to an alkyl/amido complex. Nonetheless, we were
10744 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
interested in preparing samarium amido complexes of the Synthesis of compounds
type [κ4-LBSm(NHR)(THF)x] as such complexes could poten-
[SmCl3(THF)2]. This complex was prepared via a modified
tially be transformed into terminal imido complexes if metalla-
cycle ring-opening could be triggered intramolecularly,50,82
potentially forming a “decyclometalated” complex of the form
[LBSmvNR(THF)x]. Unfortunately, complex 2 (as well as
species 3 and 4) did not react cleanly with a variety of different
1° amines, perhaps due to the exotic nature of the cyclometa-
lated moiety. Regardless, we intend to further pursue amido
and imido studies in due course. Finally, dinuclear product 4
has interesting potential from a single-molecule magnet
(SMM) perspective, and studies in this area are ongoing, with
intentions to expand this work to more prominent SMM
metals such as dysprosium.
literature procedure.84,85
A sample of [Sm2O3] (5.211 g,
14.94 mmol) was added to a 250 mL round-bottomed flask
and dissolved in 6.0 M HCl (30 mL). The flask was attached to
a reflux condenser, and the solution was heated at reflux for
23 h. The volatiles were then removed in vacuo to afford
[SmCl3(H2O)6] as a yellow residue. A solution of SOCl2 (35 mL,
0.48 mol) in THF (90 mL) was slowly added to the residue.
After bubbling had subsided, the solution was heated at reflux
for 14 h, at which point volatiles were removed in vacuo result-
ing in a pale yellow solid. The flask was connected to a swivel
frit apparatus which was then attached to a double vacuum
manifold. A portion of Et2O (70 mL) was transferred to the
flask, the slurry was vigorously mixed, then filtered and the
1
resulting white powder was dried in vacuo (10.784 g, 90%). H
Experimental
Reagents and general procedures
NMR (benzene-d6): δ 3.58 (m, 8H, THF OCH2); 1.41 (m, 8H,
THF OCH2CH2). Accurate 13C{1H} NMR spectral data could not
be obtained because of insolubility of [SmCl3(THF)2] in
common organic solvents. Anal. Calcd (%) for C8H16Cl3O2Sm:
C, 23.97; H, 4.02. Found: C, 23.48; H, 3.71.
Manipulations of air- and moisture-sensitive materials and
reagents were carried out under an argon atmosphere using
double vacuum manifold techniques or in a glove box. Sol-
vents used for air-sensitive materials were purified using an
MBraun solvent purification system (SPS), stored in PTFE-
sealed glass vessels over sodium benzophenone ketyl (THF
and ether) or “titanocene” (pentane, benzene, and toluene),
and freshly distilled at the time of use. The deuterated solvents
benzene-d6 and toluene-d8 were dried over sodium benzophe-
none ketyl, degassed via three freeze–pump–thaw cycles, dis-
tilled in vacuo and stored over 4 Å molecular sieves in glass
bombs under argon. All NMR spectra were recorded at
ambient temperature with a Bruker Avance II NMR spectro-
meter (300.13 MHz for 1H, 75.47 MHz for 13C, and 121.48 MHz
for 31P). Chemical shifts are reported in parts per million rela-
tive to the external standards SiMe4 (1H, 13C) and 85% H3PO4
(31P); residual H-containing species in benzene-d6 (δ 7.16) were
used as internal references (1H). Assignments were aided by
the use of 13C{1H}-DEPT and 1H–13C{1H}-HSQC experiments
(s = singlet, d = doublet, t = triplet, q = quartet, sp = septet,
m = multiplet, br = broad, ov = overlapping signals). Elemental
analyses were performed using an Elementar Vario Microcube
[LBY(CH2SiMe3)2] (1). In an argon atmosphere glove box,
anhydrous yttrium chloride (168 mg, 0.86 mmol) was weighed
into a 50 mL 2-neck round-bottomed flask and 4 mL of a THF–
pentane mixture (3 : 1) was added to form an off-white slurry.
The flask was connected to a swivel frit apparatus, attached to
a double vacuum manifold, and heated at 65 °C for 1.5 h. In a
separate 50 mL 2-neck flask, LiCH2SiMe3 (246 mg, 2.61 mmol)
was dissolved in 4 mL of a THF–pentane mixture (1 : 3) and
the resulting suspension was added to the other flask dropwise
over 5 min at −94 °C via cannula. The mixture was allowed to
warm to 0 °C and stirred for an additional 3.5 h at that temp-
erature. The THF–pentane solution was removed in vacuo to
yield a white solid. In an argon atmosphere glove box, proteo
ligand HLB (595 mg, 0.85 mmol) was added to a different
50 mL two-neck round-bottom flask, dissolved in 15 mL of
toluene and attached to a double vacuum manifold. The
proteo ligand solution was added dropwise over 5 min to the
reaction mixture via cannula, and the solution was stirred at
ambient temperature for 1 h. The resulting cloudy yellow solu-
tion was filtered and the solid was washed three times with
10 mL of pentane. All solvents were evaporated under reduced
pressure yielding an oily yellow solid. The solid was then tritu-
rated once with 25 mL of pentane and dried in vacuo to yield a
light yellow powder (616 mg, 75%). 1H NMR (benzene-d6):
δ 7.71 (ddd, 3JH–P = 12.3 Hz, 3JH–H = 7.8 Hz, 4JH–H = 1.2 Hz, 8H,
o-phenyl H); 7.38 (dd, 3JH–H = 8.5 Hz, 4JH–P = 2.1 Hz, 4H, o-Pipp
instrument.83
Me3)3(THF)2],
phoranylylidene))bis(4-isopropylaniline)
(CH2SiMe3)3(THF)3],65
and
The
N,N′-((1H-pyrrole-2,5-diyl)bis(diphenylphos-
(HLB),57
[Sm-
[LBLu(CH2SiMe3)(OEt2)2]+-
reagents
[YCl3(THF)3],
[Y(CH2Si-
[B(C6F5)4]− (5′)57 were prepared according to literature
methods. Complex 5 was isolated as a minor disproportiona-
tion byproduct from decomposition of 5′ in benzene over
several days. Although reported previously by other
methods,84,85 [SmCl3(THF)2] was prepared in our laboratory
from a modified procedure as described below. A solid sample
of LiCH2SiMe3 was obtained by removal of pentane from a 1.0 M
solution purchased from Sigma-Aldrich. A sample of 6.0 M HCl
was prepared by dilution of a concentrated solution. All other
reagents were purchased from commercial sources and used as
received.
3
H); 7.06 (d, JH–H = 8.1 Hz, 4H, m-Pipp H); 7.07–6.90 (ov m,
3
4
12H, m-phenyl + p-phenyl H); 6.63 (dd, JH–P = 2.1 Hz, JH–P
=
3
i
1.2 Hz, 2H, pyrrole CH); 2.66 (sp, JH–H = 6.6 Hz, 2H, Pr CH);
3
i
1.10 (d, JH–H = 7.0 Hz, 12H, Pr CH3); 0.21 (s, 18H, YCH2Si-
(CH3)3), −0.04 (d, JH–Y = 2.8 Hz, 4H, YCH2). 13C{1H} NMR
2
1
(benzene-d6): δ 144.3 (m, 2C), 143.0 (d, 4C, JC–P = 6.0 Hz, aro-
matic ipso-C); 133.6 (d, 8C, JC–P = 10.4 Hz), 132.7 (s, 4C, aro-
3
matic CH); 131.7 (s, 2C, aromatic C); 130.5 (s, 2C, aromatic
Dalton Trans., 2014, 43, 10739–10750 | 10745
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Dalton Transactions
2
4
ipso-C); 129.1 (d, 8C, JC–P = 12.7 Hz), 128.4 (d, 4C, JC–P
=
for 3 h. The solution was then cooled to ambient temperature
2
8.2 Hz), 128.1 (s, 4C, aromatic CH); 119.4 (dd, 2C, JC–P = 28.0 and all volatiles were removed in vacuo. The solid was recrystal-
Hz, JC–P = 6.0 Hz, pyrrole CH); 34.5 (d, 2C, JC–Y = 39 Hz, lized from a concentrated solution of benzene (7 mL) layered
3
1
i
i
YCH2); 34.2, (s, 2C, Pr CH), 24.6 (s, 4C, Pr CH3), 5.1 (s, 6C, with pentane (20 mL), collected by filtration, washed with 5 ×
Si(CH3)3). 31P{1H} NMR (benzene-d6): δ 25.0 (d, JP–Y = 7.5 Hz). 2 mL pentane, and dried under vacuum to yield a dark orange
3
Anal. Calcd (%) for C54H66N3P2Si2Y: C, 67.27; H, 6.90; N, 4.36. powder. The solid was redissolved in toluene (15 mL), and the
Found: C, 66.90; H, 6.51; N, 4.53.
solution stored at −35 °C for 12 h. The resulting precipitate
[κ4-LBSm(CH2SiMe3)(THF)2] (2). In an argon atmosphere was collected by filtration, washed with pentane, and dried
1
glove box, [SmCl3(THF)2] (230 mg, 0.57 mmol) was weighed in vacuo yielding a dark orange solid (207 mg, 51%). H NMR
into a 50 mL 2-neck round-bottomed flask. A THF–pentane (benzene-d6): δ 10.61 (m, 2H), 10.16 (m, 2H), 9.13 (m, 2H), 8.92
mixture (3 : 1, 6 mL) was added to form an off-white slurry. (m, 2H), 8.80 (m, 2H), 8.64 (m, 6H), 8.50 (m, 4H), 8.38 (m, 2H),
The flask was connected to a swivel frit apparatus which was 8.15 (m, 2H), 7.00 (m, 16H), 6.63 (m, 6H), 6.47 (m, 6H, Ar);
then attached to a double vacuum manifold. In a separate 6.19 (m, 4H, pyrrole CH); 3.35 (m, 8H, THF OCH2); 2.49 (sp,
3
i
50 mL 2-neck flask, LiCH2SiMe3 (162 mg, 1.72 mmol) was dis- 3JH–H = 7.3 Hz, 2H), 1.34 (sp, JH–H = 6.9 Hz, 2H, Pr CH); 1.13
3
3
solved in 6 mL of a THF–pentane mixture (1 : 3) and cooled to (d, JH–H = 6.9 Hz, 6H), 1.09 (d, JH–H = 6.9 Hz, 6H), 0.49 (d,
0 °C. After 5 min, the resulting suspension of LiCH2SiMe3 was 3JH–H = 6.9 Hz, 6H), −0.21 (d, JH–H = 8.4 Hz, 6H, Pr CH3);
added to the other flask dropwise over 1 min at −78 °C via −0.51 (br s, 8H, THF OCH2CH2). 13C{1H} NMR (benzene-d6): δ
cannula. The mixture was stirred for an additional 3 h at that 146.9 (s, 2C), 146.8 (s, 2C), 141.2 (s, 2C), 141.1 (s, 2C), 138.2 (s,
3
i
2
temperature, wherein the solution turned bright yellow in 2C, Pipp C); 136.8 (d, 2C, JC–P = 6.6 Hz, aromatic C); 136.7 (s,
1
colour. The THF–pentane solution was removed in vacuo at 2C, aromatic CH); 135.5 (d, 2C, JC–P = 71.5 Hz, pyrrole C);
−35 °C to afford a yellow solid, at which point a portion of 135.3 (d, 4C, 1JC–P = 10.0 Hz, aromatic ipso-C); 134.3 (d, 8C, 2JC–P
2
toluene (5 mL) was added. In an argon atmosphere glove box, = 9.0 Hz), 133.9 (d, 8C, JC–P = 10.0 Hz, aromatic CH); 131.8
proteo ligand HLB (410 mg, 0.58 mmol) was added to a separ- (d, 2C, JC–P = 88.6 Hz, pyrrole C); 131.8 (d, 4C, JC–P = 9.4 Hz),
ate 50 mL two-neck round-bottom flask, dissolved in 9 mL of 131.8 (d, 4C, JC–P = 5.5 Hz), 131.3 (s, 4C), 131.2 (s, 2C, aro-
1
2
3
toluene and attached to a double vacuum manifold. The matic CH); 129.7 (om m, 2C, pyrrole CH); 129.7 (s, 4C, aro-
1
proteo ligand solution was added dropwise to the reaction matic CH); 129.3 (d, 2C, JC–P = 24.2 Hz, aromatic ipso-C); 129.2
mixture over 1 min at −78 °C via cannula, and the resulting (s, 2C, pyrrole CH); 127.5 (s, 4C, aromatic CH); 126.9 (d, 2C,
solution was allowed to gradually warm to ambient tempera- 1JC–P = 7.0 Hz, aromatic ipso-C); 126.3 (br s, 2C), 126.0 (s, 2C),
ture whereupon it was stirred for 1 h. The resulting orange 123.3 (s, 2C), 123.3 (s, 2C), 121.7 (s, 4C, aromatic CH); 105.8
i
solution was filtered and the solvent was removed under (br s, 4C, THF OCH2); 34.2 (s, 2C), 31.5 (s, 2C, Pr CH); 24.8 (s,
i
reduced pressure. The resulting orange solid was triturated 2C), 24.5 (s, 2C), 24.2 (s, 2C), 24.0 (s, 2C, Pr CH3); 21.8 (s, 4C,
with pentane and dried in vacuo. The residue was recrystallized THF OCH2CH2). 31P{1H} NMR (benzene-d6): δ 39.4 (s), 24.7 (s).
from a concentrated solution of THF (2 mL) layered with Anal. Calcd (%) for C100H100N6O2P4Sm2: C, 65.19; H, 5.47; N,
pentane (15 mL), washed with 5 × 2 mL pentane, and dried 4.56. Found: C, 66.07; H, 5.62; N, 4.39.83
in vacuo to yield a dark orange powder (444 mg, 72%). 1H NMR
[(κ4-LB)LBSm] (4). In an argon atmosphere glove box,
(benzene-d6): δ 9.42 (m, 2H), 9.21 (m, 2H), 8.29 (m, 1H), [SmCl3(THF)2] (357 mg, 0.89 mmol) was weighed into a 50 mL
8.08 (m, 1H), 7.77 (m, 2H), 7.48 (m, 7H), 7.05 (m, 4H), 6.87 2-neck round-bottomed flask and 6 mL of a THF–pentane
(m, 4H), 6.61 (m, 4H, Ar); 5.34 (m, 2H, pyrrole CH); 3.17 (br s, mixture (3 : 1) was added to form an off-white slurry. The flask
3
3
8H, THF OCH2); 1.63 (d, JH–H = 7.2 Hz, 3H), 1.59 (d, JH–H
=
was connected to a swivel frit apparatus which was then
i
i
6.9 Hz, 3H, Pr CH3); 1.30 (ov, 1H, Pr CH); 1.21 (br m, 8H, attached to a double vacuum manifold. In a separate 50 mL
THF OCH2CH2); 1.08 (ov, 1H, Pr CH); 0.89 (s, 2H, SmCH2); 2-neck flask, LiCH2SiMe3 (252 mg, 2.68 mmol) was dissolved
i
3
3
−0.14 (d, JH–H = 5.7 Hz, 3H), −0.54 (d, JH–H = 6.6 Hz, 3H, in 6 mL of a THF–pentane mixture (1 : 3) and cooled to 0 °C.
iPr CH3); −3.75 (s, 9H, SmCH2Si(CH3)3). Accurate 13C{1H} After 5 min the resulting suspension of LiCH2SiMe3 was
NMR spectral data could not be obtained because of rapid quickly added to the other flask dropwise over 1 min at −35 °C
decomposition of the sample to 3 in solution. Low temp- via cannula. The mixture was stirred for an additional 3 h at
erature studies were attempted in an effort to slow this that temperature, wherein the solution turned bright yellow in
decomposition process, but signals became sufficiently colour. The THF–pentane solution was removed in vacuo at
broad that signal-to-noise ratios were unsatisfactory. 31P{1H} −35 °C to yield a yellow solid, at which point a portion of
NMR (benzene-d6): δ 24.5 (s), 18.1 (s). Anal. Calcd (%) toluene (5 mL) was added. In an argon atmosphere glove box,
for C58H70N3O2P2SiSm2: C, 64.41; H, 6.52; N, 3.89. Found: proteo ligand HLB (625 mg, 0.89 mmol) was added to a
C, 64.77; H, 6.71; N, 4.05.
different 50 mL two-neck round-bottom flask, dissolved in
[κ1:κ2:μ2-LBSm(THF)]2 (3). In an argon atmosphere glove 18 mL of toluene and attached to a double vacuum manifold.
box, [κ4-LBSm(CH2SiMe3)(THF)2] (243 mg, 0.22 mmol) was The proteo ligand solution was quickly added to the reaction
weighed into a 50 mL round-bottomed flask and 9 mL of mixture at −35 °C via cannula, and the solution was allowed to
toluene was added to form an orange slurry. The flask was warm to ambient temperature and stirred for an additional
attached to a double vacuum manifold and warmed to 50 °C 1 h. The resulting orange solution was filtered using a swivel
10746 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
frit apparatus and the volatiles removed under vacuum. The 2C), 131.3 (ov m, 1C), 131.2 (ov m, 1C, aromatic CH); 130.5 (s,
orange solid was triturated 3 times with 3 mL of pentane and 1C), 129.8 (s, 1C), 129.2 (ov m, 4C, JC–P = 6.0 Hz, aromatic ipso-
dried in vacuo. The resulting residue was recrystallized from a C); 127.5 (s, 4C), 127.3 (s, 4C), 127.1 (s, 2C), 127.0 (s, 2C), 126.4
concentrated solution of THF (5 mL) layered with pentane (s, 1C, aromatic CH); 126.1 (1C, aromatic ipso-C); 124.7 (s, 8C),
(20 mL), washed with 5 × 2 mL pentane, and dried under 123.9 (s, 1C), 121.9 (s, 1C), 118.1 (s, 2C), 118.0 (s, 2C, aromatic
i
vacuum to yield a dark orange powder (622 mg, 45%). 1H NMR CH); 34.8 (s, 1C), 32.8 (s, 2C), 31.5 (s, 1C, Pr CH); 26.3 (s, 2C),
i
(benzene-d6): δ 10.52 (dd, 1H, JH–P = 11.1 Hz, JH–H = 7.8 Hz), 25.0 (s, 2C), 24.1 (s, 2C); 23.1 (s, 2C, Pr CH3). 31P{1H} NMR
10.28 (dd, 2H, JH–P = 3.3 Hz, JH–H = 3.3 Hz), 9.28 (br s, 1H), (benzene-d6): δ 40.8 (s, 1P), 26.3 (s, 1P), 19.3 (s, 2P). Anal.
8.98 (m, 2H, JH–H = 3.3 Hz), 8.78 (dd, 1H, JH–P = 12.9 Hz, JH–H
=
Calcd (%) for C92H87N6P4Sm: C, 71.24; H, 5.65; N, 5.42. Found:
7.2 Hz), 8.55 (m, 4H), 8.21 (m, 6H), 8.09 (m, 1H, JH–H = 3.3 Hz), C, 70.87; H, 6.05; N, 5.16.
8.00 (br s, 1H), 7.92 (br s, 1H), 7.23 (m, 4H, JH–H = 3.3 Hz), 6.89
X-ray crystallography
(m, 19H), 6.36 (m, 6H), 6.68 (d, 4H, JH–H = 7.8 Hz, aromatic
CH); 3.57 (sp, 1H, JH–H = 6.9 Hz), 2.93 (sp, 1H, JH–H = 6.9 Hz,
Single crystals suitable for X-ray diffraction were readily
obtained from slow diffusion of pentane into a concentrated
toluene (1) or THF solution (2) at −35 °C, from slow diffusion
of pentane into a concentrated benzene (3) or THF solution (4)
at ambient temperature, or from slow evaporation of a concen-
trated benzene solution (5). Crystals were coated in dry Para-
tone oil under an argon atmosphere and mounted onto a glass
fibre. Data were collected at 173 K using a Bruker SMART APEX
II diffractometer (Mo Kα radiation, λ = 0.71073 Å) outfitted
with a CCD area-detector and a KRYO-FLEX liquid nitrogen
i
iPr CH); 1.95 (d, 6H, JH–H = 6.9 Hz, Pr CH3); 1.70 (sp, 2H, JH–H
i
= 6.9 Hz, Pr CH); 1.39 (d, 6H, JH–H = 6.9 Hz), 0.46 (d, 6H, JH–H
i
= 6.9 Hz), 0.43 (d, 6H, JH–H = 6.9 Hz, Pr CH3). 13C{1H} NMR
(benzene-d6): δ 137.4 (s, 2C), 137.2 (s, 1C), 137.1 (m, 1C), 137.0
(m, 1C), 136.2 (s, 1C), 135.3 (d, 2C, JC–P = 9.0 Hz), 135.1 (d, 2C,
JC–P = 8.5 Hz, aromatic ipso-C); 134.6 (dd, 8C, JC–P = 9.3 Hz, JC–P
= 9.3 Hz), 133.9 (d, 4C, JC–P = 9.1 Hz), 133.8 (d, 8C, JC–P = 10.4
Hz), 133.4 (d, 4C, JC–P = 10.4 Hz, aromatic CH); 132.2 (s, 2C),
131.8 (m, 2C, aromatic ipso-C); 131.5 (ov m, 4C), 131.4 (ov m,
Table 1 Summary of X-ray crystallography data collection and structure refinement for compounds 1, 2, 3·2benzene, 4, and 5
1
2
3·2C6H6
4
5
Formula
C54H66N3P2SiY
964.12
Triclinic
ˉ
P1
C58H70N3O2P2SiSm C100H100N6O2P4Sm2·2C6H6 C92H87N6P4Sm
C116H88BF20LuN6P4
2255.58
Orthorhombic
Pbca
Formula weight (g mol–1
)
1081.55
Monoclinic
P21/c
1998.65
Monoclinic
P21/n
1550.90
Triclinic
P1
Crystal system
Space group
ˉ
Unit cell parameters
a (Å)
b (Å)
c (Å)
α (°)
β (°)
γ (°)
V (Å3)
Z
9.7370(7)
12.1842(9)
24.1371(18)
84.8276(9)
78.9662(9)
69.6098(9)
2633.7(3)
2
14.806(2)
14.713(2)
29.208(5)
90
101.263(2)
90
6240.0(18)
4
1.151
15.9755(13)
17.8486(15)
17.3855(14)
90
109.2030(10)
90
4681.5(7)
2
1.418
12.3595(11)
13.8452(12)
24.642(2)
104.3038(9)
101.2213(9)
95.5028(9)
3961.6(6)
2
21.649(5)
22.032(5)
45.231(10)
90
90
90
21 574(8)
8
1.389
1.060
ρcalcd (mg m–3
)
1.216
1.250
1.300
0.872
µ (mm–1
)
1.049
1.366
Crystal dimensions (mm)
Crystal colour
Crystal habit
0.30 × 0.16 × 0.16 0.43 × 0.33 × 0.30
0.31 × 0.17 × 0.11
Yellow
Prism
0.30 × 0.24 × 0.16 0.36 × 0.15 × 0.15
Colourless
Needle
Yellow
Block
Light red
Block
Colourless
Prism
θ Range (°)
Diffractometer
1.72 to 25.00
1.56 to 25.00
1.69 to 26.00
0.875 to 25.00
1.597 to 26.00
Bruker D8/APEX II CCDa
Radiation (λ [Å])
Temperature (K)
Total data collected
Mo Kα (0.71073) fine focused sealed tube source
173
173
72 436
173
173
173
25 832
9252 (0.0208)
8315
142/590
1.136
48 294
8272 (0.0219)
9183
48/587
1.061
56 897
13 911 (0.0905)
11 555
66/993
1.044
295 374
21 194 (0.0524)
17 241
51/1371
1.080
Independ reflns (Rint
)
10 985 (0.0434)
9375
Obsd reflns [Fo2 ≥ 2σ(Fo2)]
Restraints/params
Goodness-of-fit (S)b [all data]
Final R indicesc
237/719
1.202
R1 [Fo2 ≥ 2σ(Fo2)]
0.0403
0.0999
0.0460
0.1031
0.0788
0.1761
0.0899
0.1798
0.0199
0.0473
0.0244
0.0499
0.0384
0.0931
0.0487
0.0965
0.0285
0.0646
0.0406
0.0686
wR2 [Fo2 ≥ 2σ(Fo2)]
R1 [all data]
wR2 [all data]
Largest diff peak, hole (e Å–3
)
0.786 and −0.513 1.431 and −3.086
0.549 and −0.305
1.579 and −0.581 0.428 and −0.380
a Programs for diffractometer operation, data collection, data reduction, and absorption correction were those supplied by Bruker. b S = [w(Fo
−
2
Fc2)2/(n − p)]1/2 (n = number of data; p = number of parameters varied; w = [σ2(Fo2) + (0.0540P)2 + 22.8160P]–1 where P = [Max(Fo2, 0) + 2Fc2]/3).
c R1 = ∑||Fo| − |Fc||/∑|Fo|; wR2 = [∑w(Fo2 − Fc2)2/∑w(Fo4)]1/2
.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10739–10750 | 10747

View Article Online
Paper
Dalton Transactions
vapour cooling device. A data collection strategy using ω and φ
scans at 0.5° steps yielded full hemispherical data with excel-
lent intensity statistics. Unit cell parameters were determined
and refined on all observed reflections using APEX2 software.86
Data reduction and correction for Lorentz polarization were
performed using SAINT-Plus software.87 Absorption correc-
tions were applied using SADABS.88 The structures were solved
by direct methods and refined by the least squares method
on F2 using the SHELX software suite.89 All non-hydrogen
atoms were refined anisotropically. Hydrogen atom positions
were calculated and isotropically refined as riding models to
their parent atoms. Table 1 provides a summary of selected
data collection and refinement parameters.
3 G. A. Molander and J. O. Hoberg, J. Org. Chem., 1992, 57,
3266–3268.
4 G. A. Molander and J. A. C. Romero, Chem. Rev., 2002, 102,
2161–2185.
5 Y. Obora, T. Ohta, C. L. Stern and T. J. Marks, J. Am. Chem.
Soc., 1997, 119, 3745–3755.
6 M. A. Giardello, V. P. Conticello, L. Brard, M. R. Gagné and
T. J. Marks, J. Am. Chem. Soc., 1994, 116, 10241–10254.
7 T. Shima and Z. Hou, J. Am. Chem. Soc., 2006, 128, 8124–
8125.
8 S. Hong and T. J. Marks, Acc. Chem. Res., 2004, 37, 673–686.
9 T. E. Müller, K. C. Hultzsch, M. Yus, F. Foubelo and
M. Tada, Chem. Rev., 2008, 108, 3795–3892.
Special considerations were required in the refinement of 10 G. A. Molander and E. D. Dowdy, J. Org. Chem., 1998, 63,
disordered moieties in all determined structures. One of the 8983–8988.
iPr groups in 1 was found to be in two different orientations 11 V. M. Arredondo, F. E. McDonald and T. J. Marks, Organo-
which were refined with occupation ratios of 61 : 39. For 2 and metallics, 1999, 18, 1949–1960.
3, the whole non-cyclometalated Pipp phenyl ring was dis- 12 J.-S. Ryu, G. Y. Li and T. J. Marks, J. Am. Chem. Soc., 2003,
ordered in two positions which could be refined with approxi- 125, 12584–12605.
mate 57 : 43 and 56 : 44 occupancy ratios, respectively. 13 A. Motta, I. L. Fragalà and T. J. Marks, Organometallics,
Furthermore, the coordinated THF molecules of 2 were also 2006, 25, 5533–5539.
disordered in two positions with an occupation ratio of 64 : 36. 14 H. F. Yuen and T. J. Marks, Organometallics, 2009, 28,
Finally, one iPr carbon (C36) of the cyclometalated Pipp ring in
2423–2440.
2 was disordered and refined in two positions (57 : 43). In 4, 15 A. G. Trambitas, T. K. Panda, J. Jenter, P. W. Roesky,
both phenyl groups attached to phosphorus P1 were dis-
ordered over two positions and were refined to these positions
C. Daniliuc, C. G. Hrib, P. G. Jones and M. Tamm, Inorg.
Chem., 2010, 49, 2435–2446.
with occupation ratios of 56 : 44 and 51 : 49. In addition, two of 16 D. V. Gribkov, F. Hampel and K. C. Hultzsch, Eur. J. Inorg.
i
the Pr groups were disordered in two positions and refined
Chem., 2004, 4091–4101.
with the occupation ratio of 61 : 39 and 69 : 31. The disorder of 17 G. A. Molander and M. Julius, J. Org. Chem., 1992, 57,
one of the iPr groups in 5 was refined with an occupation ratio
of 76 : 24.
6347–6351.
18 P.-F. Fu, L. Brard, Y. Li and T. J. Marks, J. Am. Chem. Soc.,
1995, 117, 7157–7168.
No suitable disorder model could be found for the severely
disordered solvent molecules (THF or benzene) in 2, 4 or 5; 19 G. A. Molander and W. H. Retsch, J. Am. Chem. Soc., 1997,
hence, electron density associated with the solvent molecules 119, 8817–8825.
was treated with the solvent mask option as implemented in 20 G. A. Molander, E. D. Dowdy and B. C. Noll, Organometal-
the Olex2 program.90
lics, 1998, 17, 3754–3758.
21 M. R. Douglass and T. J. Marks, J. Am. Chem. Soc., 2000,
122, 1824–1825.
22 M. R. Douglass, C. L. Stern and T. J. Marks, J. Am. Chem.
Soc., 2001, 123, 10221–10238.
Acknowledgements
This research was financially supported by the Natural 23 A. M. Kawaoka, M. R. Douglass and T. J. Marks, Organo-
Sciences and Engineering Research Council (NSERC) of metallics, 2003, 22, 4630–4632.
Canada and the Canada Foundation for Innovation (CFI). Prof. 24 A. Motta, I. L. Fragalà and T. J. Marks, Organometallics,
Jun Okuda and RWTH Aachen University are thanked for 2005, 24, 4995–5003.
hosting PGH during the preparation of this manuscript. 25 K. N. Harrison and T. J. Marks, J. Am. Chem. Soc., 1992,
Dr Tracey Roemmele is acknowledged for aiding with elemen-
tal analysis measurements.
114, 9220–9221.
26 H. Schumann, A. Heim, J. Demtschuk and S. H. Mühle,
Organometallics, 2003, 22, 118–128.
27 S. Seo, X. Yu and T. J. Marks, J. Am. Chem. Soc., 2009, 131,
263–276.
28 X. Yu, S. Seo and T. J. Marks, J. Am. Chem. Soc., 2007, 129,
7244–7245.
References
1 S. Zhou, H. Wang, J. Ping, S. Wang, L. Zhang, X. Zhu,
Y. Wei, F. Wang, Z. Feng, X. Gu, S. Yang and H. Miao, Orga- 29 S. Seo and T. J. Marks, Chem. – Eur. J., 2010, 16, 5148–
nometallics, 2012, 31, 1696–1702. 5162.
2 G. Jeske, H. Lauke, H. Mauermann, H. Schumann and 30 A. Motta, I. L. Fragalà and T. J. Marks, Organometallics,
T. J. Marks, J. Am. Chem. Soc., 1985, 107, 8111–8118.
2010, 29, 2004–2012.
10748 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
31 H. Schumann, J. A. Meese-Marktscheffel and L. Esser, 57 K. R. D. Johnson, M. A. Hannon, J. S. Ritch and
Chem. Rev., 1995, 95, 865–986. P. G. Hayes, Dalton Trans., 2012, 41, 7873–7875.
32 S. Zhou, S. Wu, H. Zhu, S. Wang, X. Zhu, L. Zhang, 58 R. D. Shannon, Acta Crystallogr., Sect. A: Found. Crystallogr.,
G. Yang, D. Cui and H. Wang, Dalton Trans., 2011, 40,
9447–9453.
1976, 32, 751–767.
59 W. J. Evans, Inorg. Chem., 2007, 46, 3435–3449.
33 P. L. Watson and G. W. Parshall, Acc. Chem. Res., 1985, 18, 60 K. C. Hultzsch, P. Voth, K. Beckerle, T. P. Spaniol and
51–56. J. Okuda, Organometallics, 2000, 19, 228–243.
34 G. Jeske, H. Lauke, H. Mauermann, P. N. Swepston, 61 W. J. Evans, J. C. Brady and J. W. Ziller, J. Am. Chem. Soc.,
H. Schumann and T. J. Marks, J. Am. Chem. Soc., 1985, 107,
8091–8103.
35 W. J. Evans, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc.,
1990, 112, 2314.
36 M. A. Giardello, Y. Yamamoto, L. Brard and T. J. Marks,
J. Am. Chem. Soc., 1995, 117, 3276–3277.
37 M. Yamashita, Y. Takemoto, E. Ihara and H. Yasuda, Macro-
molecules, 1996, 29, 1798–1806.
2001, 123, 7711–7712.
62 M. F. Lappert and R. Pearce, J. Chem. Soc., Chem. Commun.,
1973, 126–126.
63 An analysis of 101 entries in the Cambridge Structural
Database (CSD version 5.35, updated Nov. 2013) for neutral
(trimethylsilyl)methyl yttrium complexes of the generic
form LnYIII(CH2SiMe3)x suggested an average Y–CH2SiMe3
bond length of 2.415 Å (range = 2.339–2.476 Å).
38 K. C. Hultzsch, T. P. Spaniol and J. Okuda, Organometallics, 64 M. Shibasaki and N. Yoshikawa, Chem. Rev., 2002, 102,
1997, 16, 4845–4856. 2187–2209.
39 K. C. Hultzsch, T. P. Spaniol and J. Okuda, Angew. Chem., 65 H. Schumann, D. M. M. Freckmann and S. Dechert,
Int. Ed., 1999, 38, 227–230. Z. Anorg. Allg. Chem., 2002, 628, 2422–2426.
40 Z.-M. Hou, Y.-G. Zhang, H. Tezuka, P. Xie, O. Tardif, 66 Magnetochemistry, ed. R. L. Carlin, Springer-Verlag,
T. A. Koizumi, H. Yamazaki and Y. Wakatsuki, J. Am. Chem.
Soc., 2000, 122, 10533–10543.
41 Y. Luo, J. Baldamus and Z. Hou, J. Am. Chem. Soc., 2004,
126, 13910–13911.
New York, USA, 1986, ch. 9.
67 K. D. Conroy, W. E. Piers and M. Parvez, J. Organomet.
Chem., 2008, 693, 834–846.
68 D. Barbier-Baudry, F. Bonnet, B. Domenichini, A. Dormond
and M. Visseaux, J. Organomet. Chem., 2002, 647, 167–179.
42 A.-S. Rodrigues, E. Kirillov, C. W. Lehmann, T. Roisnel,
B. Vuillemin, A. Razavi and J.-F. Carpentier, Chem. – Eur. J., 69 Y. Satoh, N. Ikitake, Y. Nakayama, S. Okuno and H. Yasuda,
2007, 13, 5548–5565. J. Organomet. Chem., 2003, 667, 42–52.
43 S. Zhou, S. Wang, E. Sheng, L. Zhang, Z. Yu, X. Xi, G. Chen, 70 D. Baudry, A. Dormond, B. Lachot, M. Visseaux and
W. Luo and Y. Li, Eur. J. Inorg. Chem., 2007, 1519–1528. G. Zucchi, J. Organomet. Chem., 1997, 547, 157–165.
44 S. Wang, X. Tang, A. Vega, J.-Y. Saillard, S. Zhou, G. Yang, 71 H. Schumann, M. R. Keitsch, J. Demtschuk and
W. Yao and Y. Wei, Organometallics, 2007, 26, 1512–1522. G. A. Molander, J. Organomet. Chem., 1999, 582, 70–82.
45 K. R. D. Johnson and P. G. Hayes, Chem. Soc. Rev., 2013, 42, 72 H. Schumann, M. R. Keitsch and S. H. Mühle, Z. Anorg.
1947–1960. Allg. Chem., 2002, 628, 1311–1318.
46 F. T. Edelmann, D. M. M. Freckmann and H. Schumann, 73 A. Venugopal, I. Kamps, D. Bojer, R. J. F. Berger, A. Mix,
Chem. Rev., 2002, 102, 1851–1896.
47 W. E. Piers and D. J. H. Emslie, Coord. Chem. Rev., 2002,
233–234, 131–155.
48 C. Pi, Z. Zhang, R. Liu, L. Weng, Z. Chen and X. Zhou,
Organometallics, 2006, 25, 5165–5172.
49 K. R. D. Johnson and P. G. Hayes, Organometallics, 2009,
28, 6352–6361.
A. Willner, B. Neumann, H.-G. Stammler and N. W. Mitzel,
Dalton Trans., 2009, 5755–5765.
74 W. J. Evans, J. M. Perotti and J. W. Ziller, J. Am. Chem. Soc.,
2005, 127, 3894–3909.
75 J. C. Gordon, G. R. Giesbrecht, D. L. Clark, P. J. Hay,
D. W. Keogh, R. Poli, B. L. Scott and J. G. Watkin, Organo-
metallics, 2002, 21, 4726–4734.
50 K. R. D. Johnson and P. G. Hayes, Organometallics, 2011, 76 T. Dubé, S. Gambarotta and G. P. A. Yap, Organometallics,
30, 58–67. 2000, 19, 817–823.
51 K. R. D. Johnson and P. G. Hayes, Organometallics, 2013, 77 J.-I. Song and S. Gambarotta, Angew. Chem., Int. Ed. Engl.,
32, 4046–4049. 1995, 34, 2141–2143.
52 K. R. D. Johnson and P. G. Hayes, Dalton Trans., 2014, 43, 78 M. N. Bochkarev, V. V. Khramenkov, Y. F. Rad’kov and
2448–2457. L. N. Zakharov, J. Organomet. Chem., 1991, 421, 29–38.
53 K. R. D. Johnson, PhD thesis, University of Lethbridge, 79 W.-X. Zhang, Z. Wang, M. Nishiura, Z. Xi and Z. Hou, J. Am.
2012. Chem. Soc., 2011, 133, 5712–5715.
54 J. J. Eisch and W. C. Kaska, J. Am. Chem. Soc., 1962, 84, 80 E. Lu, Y. Li and Y. Chen, Chem. Commun., 2010, 46, 4469–
1501–1502. 4471.
55 J. J. Eisch and W. C. Kaska, J. Organomet. Chem., 1964, 2, 81 W. Rong, J. Cheng, Z. Mou, H. Xie and D. Cui, Organometal-
184–187. lics, 2013, 32, 5523–5529.
56 H. A. Mayer and W. C. Kaska, Chem. Ber., 1990, 123, 1827– 82 B. V. Mork and T. D. Tilley, J. Am. Chem. Soc., 2004, 126,
1831.
4375–4385.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 10739–10750 | 10749

View Article Online
Paper
Dalton Transactions
83 Despite exhaustive efforts the elemental analysis data for 86 APEX2, Bruker AXS, Madison, WI, 2010.
complex 3 was not satisfactory; the best values from 87 SAINT-Plus, Bruker AXS, Madison, WI, 2009.
repeated runs are given. 88 G. M. Sheldrick, SADABS, Bruker AXS, Madison, WI, 2008.
84 S.-H. Wu, Z.-B. Ding and X.-J. Li, Polyhedron, 1994, 18, 89 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crys-
2679–2681. tallogr., 2008, 64, 112–122.
85 G. B. Deacon, T. Feng, S. Nickel, B. W. Skelton and 90 O. V. Dolomanov, J. Bourhis, R. J. Gildea,
A. H. White, J. Chem. Soc., Chem. Commun., 1993, 1328–
1329.
J. A. K. Howard and H. Puschmann, J. Appl. Crystallogr.,
2009, 42, 339–341.
10750 | Dalton Trans., 2014, 43, 10739–10750
This journal is © The Royal Society of Chemistry 2014