Organolutetium-mediated dearomatization and functionalization of pyrimidine rings

Article abstract of DOI:10.1021/om400413e

The synthesis of a pentadentate NNNNN bis(phosphinimine)carbazole pincer ligand bearing two pyrimidine N-aryl rings is reported. A dialkyl lutetium complex of the ligand, prepared by alkane elimination with Lu(CH ₂SiMe₃)₃(THF)₂, was found to undergo an unusual transformation involving a double alkyl shift whereby the -CH ₂SiMe₃ groups on the metal migrated to both pyrimidine rings. This migration afforded a final product that featured alkyl-functionalized, dearomatized ligand pyrimidine rings.

Full text of DOI:10.1021/om400413e

Communication
pubs.acs.org/Organometallics
Organolutetium-Mediated Dearomatization and Functionalization of
Pyrimidine Rings
Kevin R. D. Johnson and Paul G. Hayes*
Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive, Lethbridge, AB, Canada, T1K 3M4
S
* Supporting Information
ABSTRACT: The synthesis of a pentadentate NNNNN
bis(phosphinimine)carbazole pincer ligand bearing two
pyrimidine N-aryl rings is reported. A dialkyl lutetium complex
of the ligand, prepared by alkane elimination with Lu-
(CH2SiMe3)3(THF)2, was found to undergo an unusual
transformation involving a double alkyl shift whereby the
−CH2SiMe3 groups on the metal migrated to both pyrimidine
rings. This migration afforded a final product that featured
alkyl-functionalized, dearomatized ligand pyrimidine rings.
n recent years, there has been a surge in the development of
non-carbocyclic ligands for use in supporting highly reactive
Chart 1. Cyclometalation of a Bis(phosphinimine)carbazole
Ligand
I
rare earth complexes.1−4 This “post-metallocene era” has
witnessed a diverse array of new ligand scaffolds that can
support these metal ions by providing a wide variety of different
steric and electronic environments. As a result, unique
coordination modes have often been encountered in lanthanide
complexes supported by non-Cp ligands, a trend that remains
largely unparalleled by traditional carbocyclic ligands.
To this end, we recently reported two carbazole-based
bis(phosphinimine) pincer ligands for use in stabilizing rare
earth metals.5 These ancillary ligands were designed to impart a
unique coordination environment on metal complexes and,
thus, act as a platform for obtaining new bonding modes and
reaction behavior. Unfortunately, initial screening of the
scaffolds, as the corresponding lutetium dialkyl complexes,
revealed a high degree of steric crowding at the metal center,
resulting in cyclometalative decomposition.6 Specifically, the P-
phenyl rings of the phosphinimine functionalities underwent
cyclometalation reactions with concomitant loss of alkane to
afford doubly cyclometalated lutetium complexes that were
coordinated by the ligand in a κ5 bonding mode via three
nitrogen atoms and two ortho-metalated P-phenyl rings
(complex i, Chart 1). We have also demonstrated that the
bis(phosphinimine)carbazole ligand is capable of cyclometala-
tive reactivity at the N-aryl site of the phosphinimine
functionality (complex ii, Chart 1).7 Accordingly, we were
inclined to modify the ligand framework so as to prevent such
decomposition pathways. Since previous iterations of this
ancillary possessing phenyl and 4-isopropylphenyl rings at the
N-aryl site of the phosphinimine functionality were prone to
C−H bond activation at the ortho positions, these moieties
were replaced with pyrimidine groups that lack ortho C−H
bonds. Notably, such pyrimidine rings incorporate additional
hard nitrogen donors into the scaffold, thus increasing its
potential denticity and capacity to stabilize sterically and
electronically unsaturated rare earth complexes. Herein, we
report the effect of pyrimidine ring incorporation into the
phosphinimine functionality of our carbazole pincer ligand and
its resultant ability to stabilize rare earth metal complexes, using
lutetium as a model system. Significantly, our investigation has
revealed unique reactivity involving a double alkyl migration
reaction that resulted in dearomatization and functionalization
of the ligand pyrimidine rings.
Installation of N-aryl groups into the ligand framework can
be readily achieved by the reaction of an aryl azide (2 equiv)
with 1,8-bis(diphenylphosphino)-3,6-dimethylcarbazole.5 In
this case, we aimed to use 2-azidopyrimidine as the pyrimidine
source (Scheme 1). In solution, 2-azidopyrimidine is
susceptible to valence tautomerism and exists in equilibrium
with its tetrazole form (tetrazolo[1,5-a]pyrimidine). This
equilibrium is influenced by choice of solvent and temperature,
but typically favors the tetrazole tautomer.8,9 Nonetheless, the
desired Staudinger reaction between 1,8-bis-
(diphenylphosphino)-3,6-dimethylcarbazole and 2-azidopyrimi-
dine proceeds cleanly to afford the anticipated bis-
Received: May 10, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om400413e | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
rotated significantly away from the carbazole N−H (N3−P2−
C8−C9 torsion angle of 159.1(3)°). The phosphinimine PN
bond lengths in 1 (1.597(3) and 1.592(3) Å) are comparable to
those found in related structures.5,7,10
Scheme 1. Synthesis of Pyrimidine-Substituted
Bis(phosphinimine)carbazole Scaffold (1)
In order to assess the ability of bis(phosphinimine) ligand 1
to support well-defined organolanthanide complexes, a dialkyl
lutetium species was targeted. Proteo ligand 1 reacted readily
with Lu(CH2SiMe3)3(THF)2 to give the respective dialkyl
lutetium complex via alkane elimination. When this reaction
was monitored in situ on an NMR tube scale in toluene-d8,
rapid formation of the desired dialkyl metal complex,
(L)Lu(CH2SiMe3)2 (2), one equivalent of SiMe4, and two
equivalents of liberated THF were observed.
In contrast to previously reported bis(phosphinimine)-
carbazole ligands,5 ligand 1 presents a chelation environment
defined by one nitrogen atom from each pyrimidine ring, in
addition to the two phosphinimine donors and the carbazole
nitrogen. As such, the alkane elimination reaction of
Lu(CH2SiMe3)3(THF)2 with 1 resulted in dialkyl lutetium
complex (L)Lu(CH2SiMe3)2 (2), whereby the ligand was
bound κ5 to the metal center through five nitrogen atoms
(Scheme 2). In complex 2, the presence of two extra nitrogen
(phosphinimine) ligand (1) in excellent (97.3%) yield,
although the rate of reaction is significantly slower than that
observed for other aryl azides, such as phenyl azide or para-
isopropylphenyl azide.5
Scheme 2. Synthesis of κ5 Dialkyl Lutetium Complex 2 and
Reactivity via Double −CH2SiMe3 Migration
Proteo ligand 1 is C2v symmetric on the NMR time scale and
exhibits a sharp singlet at δ 18.5 (benzene-d6) in its 31P{1H}
1
NMR spectrum. The H NMR spectrum of 1 (benzene-d6)
displays a single carbazole methyl resonance at δ 2.16, a broad
NH peak at δ 12.2, and the expected aromatic signals.
Diagnostically, the pyrimidine rings on 1 give rise to a triplet
(2H) at δ 5.97 and a doublet (4H) at δ 8.02 corresponding to
the para and meta protons, respectively.
In addition to characterization by multinuclear NMR
spectroscopy, the ligand structure was unambiguously estab-
lished by a single-crystal X-ray diffraction study. Proteo ligand 1
crystallized from a concentrated chloroform-d solution in the
space group P21/c with two solvent molecules. A rendering of
molecule 1 is depicted in Figure 1 as a thermal ellipsoid plot. In
the structure, one phosphinimine donor (N1) participates in a
hydrogen bond interaction with the carbazole N−H (d(N···N)
= 2.838(5) Å; however, the group lies substantially out of the
dimethylcarbazole plane (N1−P1−C1−C12 torsion angle of
33.6(4)°). The other phosphinimine moiety (P2−N3) is
donors provides enhanced electron donation to the metal
center and, as a result, enhanced thermal stability. Compared to
previously described lutetium dialkyl species of related
bis(phosphinimine)carbazole ligands, which exhibited half-
lives on the order of only 20 min at ambient temperature,5
complex 2 can be left in solution (benzene-d6/THF) at ambient
temperature for over 5 h with only minimal signs of
decomposition. Eventually, 2 does decompose to a single
new product (vide infra), the process of which is rapidly
accelerated at elevated temperature. Unfortunately, attempts to
isolate 2 as an analytically pure solid were unsuccessful and
always resulted in mixtures of 2 and its decomposition product.
However, 2 can be quantitatively generated in situ at low
temperature and used in this form to further investigate
reactivity.
Figure 1. Thermal ellipsoid plot (50% probability) of HL (1) with
hydrogen atoms (except H1) and two chloroform-d solvent molecules
omitted for clarity. Selected bond distances (Å), angles (deg), and
torsion angles (deg): P1−N1 = 1.597(3), P2−N3 = 1.592(3), N2···N1
= 2.838(5), N1−P1−C1 = 103.4(2), N3−P2−C8 = 114.7(2), N1−
P1−C1−C12 = 33.6(4), N3−P2−C8−C9 = 159.1(3).
In the 31P{1H} NMR spectrum (toluene-d8, 263.2 K), a
downfield shift of the phosphinimine resonance was observed
1
for (L)Lu(CH2SiMe3)2 (2) at δ 27.2. As expected, the H
B
dx.doi.org/10.1021/om400413e | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
NMR spectrum revealed diagnostic methylene (δ −1.22) and
trimethylsilyl signals (δ −0.33) as sharp singlets, integrating to
4H and 18H, respectively. In the aromatic region of the
spectrum, three separate signals corresponding to pyrimidine
protons were observed at δ 8.28 (br m, 2H), 7.74 (m, 2H), and
6.08 (dd, 2H). Thus, it was concluded that both pyrimidine
rings were bound to the lutetium center, affording a metal
complex where the ancillary ligand was coordinated in the
anticipated κ5N mode.
Gratifyingly, complex 2 does not suffer from intramolecular
cyclometalation pathways previously reported for related
bis(phosphinimine)carbazole lutetium complexes.6 Interest-
ingly, though, 2 undergoes an unusual double alkyl shift,
whereby both of the alkyl groups on the metal migrate to the
pyrimidine rings, resulting in a final product with trimethylsi-
lylmethyl groups para to the nitrogen atom coordinated to
lutetium. A result of the double alkyl migration reaction is
dearomatization of the pyrimidine rings and, consequently,
formation of a formal negative charge on the coordinated
pyrimidinyl nitrogen atoms. The product of this alkyl transfer is
an asymmetric bimetallic complex, 3, wherein the ligand is κ5-
coordinated to lutetium through five nitrogen atoms (three
anionic nitrogens and two neutral phosphinimine donors). To
the best of our knowledge, this is the first example of
dearomatization and alkyl group functionalization of pyrimidine
rings by a rare earth metal.
found to dimerize via bridging nitrogen atoms (N5 and N13) of
the dearomatized pyrimidinyl rings. As each bridging
pyrimidinyl nitrogen atom bears a negative charge, it can be
viewed as formally bonding to one lutetium as an anionic ligand
and the other metal center as a neutral Lewis base.
Correspondingly, N5 exhibits short (2.315(6) Å) and long
(2.493(6) Å) bonds to Lu1 and Lu2, respectively. Likewise,
N13 is bonded to Lu1 and Lu2 by similar long and short
interactions (2.436(5) and (2.354(5) Å), respectively). These
contact distances suggest that the bimetallic complex is not held
together by Lewis acid−base interactions, but rather by the
anionic ligand-to-metal bonds.
In the complex, one lutetium center (Lu2) is seven-
coordinate, while the other metal center (Lu1) is hexacoordi-
nate. The geometry at Lu2 is best described as distorted
pentagonal bipyramidal with five nitrogen atoms from one
ligand subunit (N1, N2, N3, N5, and N6) occupying the
equatorial plane (average N−Lu−N angle = 72.0°). The apical
sites of the pentagonal bipyramid are defined by coordination
of one molecule of THF (O1) and a bridging pyrimidine
nitrogen atom (N13). At 170.2(4)°, the N13−Lu2−O1 angle
approaches linearity. In contrast to Lu2, the molecular
geometry at Lu1 is not easily assigned. Upon initial inspection,
a distorted trigonal prismatic geometry with the trigonal faces
defined by N9, N8, and N5 and N10, N12, and N13 was
considered. The N−N−N bond angles measured on the N9,
N8, and N5 trigonal face range from 50.0(2)° to 65.6(2)° and
are in relatively good agreement with the expected value of 60°.
However, the N−N−N bond angles measured on the N10,
N12, and N13 trigonal face are largely distorted (ranging from
38.4(2)° to 80.2(2)°), the result of which generates a trigonal
prismatic geometry with significant twist. Perhaps a better
description of the geometry at Lu1 is a bicapped tetrahedron,11
whereby the tetrahedron is defined by N5, N9, N10, and N12,
with N8 and N13 serving as the capping atoms. The average
tetrahedral angle at Lu1 was 107.8°, which is slightly less than
the ideal value of 109.5°. Notably, a large angle of 138.6(2)°
was measured between the capping atoms of the tetrahedron
(N8−Lu1−N13), suggesting distortion from this geometry as
well.
Due to the complexity of the reaction product, and the
broad, uninformative spectral features in its NMR spectra
between −60 and 90 °C, the structure of complex 3 was
unambiguously ascertained by a single-crystal X-ray diffraction
experiment. The solid-state structure of 3 is depicted in Figure
2 as a thermal ellipsoid plot. In the solid state, the complex was
Alkyl migration involving rare earth metals has been
previously documented in complexes containing different N-
heterocyclic ligands. For example, a 1,3-alkyl migration was
reported by Kiplinger et al. in lutetium alkyl complexes
supported by 2,2′:6′,2″-terpyridine (terpy) and 4,4′,4″-tri-tert-
butyl-2,2′:6′,2″-terpyridine ligands.12 With the terpy-based
scaffolds, migration was limited to a single alkyl group, thus
converting the neutral terpy ligand into a monoanionic
ancillary; this ligand has since proven useful in supporting a
wide range of organometallic lutetium complexes.13 Likewise,
alkyl migration was also documented in group 3 benzyl
complexes of a ferrocene diamide ligand coordinated to
heterocycles such as isoquinoline, pyridine, and 2,2′-bipyr-
idine.14,15 Notably, in the systems involving pyridine and 2,2′-
bipyridine, initial 1,3-alkyl transfer yielded the corresponding
1,2-dihydroheterocycle as the kinetic product. This was then
followed by subsequent isomerization to the 1,4-dihydrohetero-
cycle as the thermodynamic product. Comparable behavior
involving pyridine and pyridine-based ligands has also been
reported in rare earth hydride,4,16−22 alkaline earth hydride,23,24
and transition metal complexes.25−27
Figure 2. Thermal ellipsoid plot (30% probability) of 3 with hydrogen
atoms and P-phenyl rings (except for ipso carbons) omitted for clarity.
Positionally disordered atoms are depicted as spheres of arbitrary
radius. Selected bond distances (Å) and angles (deg): Lu2−N2 =
2.412(5), Lu2−N3 = 2.297(6), Lu2−N1 = 2.321(5), Lu2−N5 =
2.493(6), Lu2−N13 = 2.354(5), Lu2−N6 = 2.389(16), Lu1−N13 =
2.436(5), Lu1−N10 = 2.264(6), Lu1−N5 = 2.315(6), Lu1−N9 =
2.327(5), Lu1−N8 = 2.305(6), Lu1−N12 = 2.240(6), O1−Lu2 =
2.341(11), N13−Lu2−O1 = 170.2(4), N3−Lu2−N2 = 81.4(2), N2−
Lu2−N1 = 81.0(2), N1−Lu2−N5 = 56.4(2), N5−Lu2−N6 = 87.2(5),
N6−Lu2−N3 = 53.8(5), N12−Lu1−N5 = 129.0(2), N12−Lu1−N10
= 105.1(2), N12−Lu1−N9 = 118.1(2), N9−Lu1−N10 = 81.1(2),
N10−Lu1−N5 = 110.7(2), N9−Lu1−N5 = 102.6(2), N8−Lu1−N13
= 138.6(2).
With our ligand system, the final product of the double
migration reaction, 3, possessed 4-trimethylsilylmethyl-1,4-
C
dx.doi.org/10.1021/om400413e | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Communication
Notes
dihydropyrimidin-1-yl rings bound to each phosphinimine
group of the ligand. In these dearomatized pyrimidinyl rings,
the alkyl group is situated para to the nitrogen atom bridging
the lutetium centers. It is highly probable that dearomatization
and functionalization of the pyrimidine rings occur via a
mechanism related to those discussed above for other N-
heterocycles. Accordingly, we have reasoned that the operative
pathway likely proceeds from dialkyl 2 via an initial 1,3-alkyl
migration followed by subsequent isomerization, a second 1,3-
migration, isomerization, and finally dimerization to afford the
final product 3, which is supported by a remarkable trianionic
pentadentate ancillary that possesses two 4-trimethylsilylmeth-
yl-1,4-dihydropyrimidin-1-yl rings.
We were interested in investigating the generality of this
migration process and the possibility of extending the observed
reactivity to other group 3 metals. Specifically, we utilized the
reagents Sc(CH2SiMe3)3(THF)2 and Y(CH2SiMe3)3(THF)2 to
prepare dialkyl analogues of complex 2 of the general form
(L)Ln(CH2SiMe3)2, Ln = Sc and Y. Unfortunately, these
derivatives did not cleanly undergo alkyl migration/dearoma-
tization of the ligand pyrimidine rings. This is notable
considering that the lutetium derivative 2 exclusively afforded
species 3 in high yield. With Sc and Y, however, a complicated
blend of products was observed; these intractable mixtures
likely formed via competing decomposition routes, and we have
thus far been unable to separate the various compounds.
A fine balance is clearly required when tuning the properties
of an ancillary ligand for use in rare earth metal chemistry.
While incorporation of pyrimidine rings into the ligand scaffold
afforded a lutetium complex that was resistant to cyclo-
metalative alkane elimination reactivity, it introduced an
alternative route for complex decomposition, namely, pyrimi-
dine ring dearomatization and functionalization. We will
mitigate this issue by judiciously selecting non-N-heterocyclic
ligand R groups in future iterations of our carbazole scaffold. As
a different approach to a “cyclometalation-resistant” bis-
(phosphinimine)carbazole ligand, we are currently replacing
the phenyl rings at the phosphinimine phosphorus atoms with
less bulky and geometrically constrained moieties, such as
methyl and phospholane groups.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was financially supported by the Natural Sciences
and Engineering Research Council (NSERC) of Canada and
the Canada Foundation for Innovation (CFI). The authors
wish to thank Ms. Breanne Kamenz for preparing the reagent 2-
azidopyrimidine.
REFERENCES
■
(1) Edelmann, F. T.; Freckmann, D. M. M.; Schumann, H. Chem.
Rev. 2002, 102, 1851−1896.
(2) Piers, W. E.; Emslie, D. J. H. Coord. Chem. Rev. 2002, 233−234,
131−155.
(3) Mountford, P.; Ward, B. D. Chem. Commun. 2003, 1797−1803.
(4) Konkol, M.; Okuda, J. Coord. Chem. Rev. 2008, 252, 1577−1591.
(5) Johnson, K. R. D.; Hayes, P. G. Organometallics 2009, 28, 6352−
6361.
(6) Johnson, K. R. D.; Hayes, P. G. Chem. Soc. Rev. 2013, 42, 1947−
1960.
(7) Johnson, K. R. D.; Hayes, P. G. Organometallics 2011, 30, 58−67.
(8) Temple, C., Jr.; Thorpe, M. C.; Coburn, W. C., Jr.; Montgomery,
J. A. J. Org. Chem. 1966, 31, 935−938.
(9) Temple, C., Jr.; Montgomery, J. A. J. Org. Chem. 1965, 30, 826−
829.
(10) Johnson, K. R. D.; Hannon, M. A.; Ritch, J. S.; Hayes, P. G.
Dalton Trans. 2012, 41, 7873−7875.
(11) Fryzuk, M. D.; Carter, A.; Rettig, S. J. Organometallics 1992, 11,
469−472.
(12) Jantunen, K. C.; Scott, B. L.; Hay, P. J.; Gordon, J. C.; Kiplinger,
J. L. J. Am. Chem. Soc. 2006, 128, 6322−6323.
(13) Masuda, J. D.; Jantunen, K. C.; Scott, B. L.; Kiplinger, J. L.
Organometallics 2008, 27, 803−806.
(14) Diaconescu, P. L. Acc. Chem. Res. 2010, 43, 1352−1363.
(15) Miller, K. L.; Williams, B. N.; Benitez, D.; Carver, C. T.; Ogilby,
K. R.; Tkatchouk, E.; Goddard, W. A., , III; Diaconescu, P. L. J. Am.
Chem. Soc. 2010, 132, 342−355.
(16) Kirillov, E.; Lehmann, C. W.; Razavi, A.; Carpentier, J.-F. Eur. J.
Inorg. Chem. 2004, 943−945.
(17) Robert, D.; Voth, P.; Spaniol, T. P.; Okuda, J. Eur. J. Inorg.
Chem. 2008, 2810−2819.
(18) Evans, W. J.; Meadows, J. H.; Hunter, W. E.; Atwood, J. L. J. Am.
Chem. Soc. 1984, 106, 1291−1300.
In summary, the new pyrimidine-substituted ligand 1 was
utilized to successfully prepare a dialkyl lutetium complex that
was resistant to ortho cyclometalative decomposition pathways
that plagued previous generations of this framework. However,
this species proved to be prone to an unusual double alkyl shift
that resulted in functionalization and dearomatization of the
ligand pyrimidine rings. Despite the unexpected nature of the
reactivity of the developed compounds, these well-defined
lutetium complexes serve as valuable models for studying the
reactivity patterns of potentially useful aromatic ring function-
alization processes. Accordingly, future studies will aim to fully
explore the potentially rich small-molecule reaction chemistry
of complexes 2 and 3.
(19) Deelman, B.-J.; Stevels, W. M.; Teuben, J. H.; Lakin, M. T.;
Spek, A. L. Organometallics 1994, 13, 3881−3891.
(20) Gountchev, T. I.; Tilley, T. D. Organometallics 1999, 18, 2896−
2905.
(21) Duchateau, R.; van Wee, C. T.; Teuben, J. H. Organometallics
1996, 15, 2291−2302.
(22) Konkol, M.; Spaniol, T. P.; Kondracka, M.; Okuda, J. Dalton
Trans. 2007, 4095−4102.
(23) Hill, M. S.; Kociok-Kohn, G.; MacDougall, D. J.; Mahon, M. F.;
̈
Weetman, C. Dalton Trans. 2011, 40, 12500−12509.
(24) Hill, M. S.; MacDougall, D. J.; Mahon, M. F. Dalton Trans.
2010, 39, 11129−11131.
(25) Scott, J.; Gambarotta, S.; Korobkov, I.; Budzelaar, P. H. M. J.
Am. Chem. Soc. 2005, 127, 13019−13029.
(26) Sugiyama, H.; Aharonian, G.; Gambarotta, S.; Yap, G. P. A.;
Budzelaar, P. H. M. J. Am. Chem. Soc. 2002, 124, 12268−12274.
(27) Reardon, D.; Conan, F.; Gambarotta, S.; Yap, G.; Wang, Q. J.
Am. Chem. Soc. 1999, 121, 9318−9325.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details in PDF format and X-ray crystallographic
details in CIF format are available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: p.hayes@uleth.ca.
■
D
dx.doi.org/10.1021/om400413e | Organometallics XXXX, XXX, XXX−XXX