Kinetic and mechanistic investigation of metallacycle ring opening in an ortho-metalated lutetium aryl complex

Article abstract of DOI:10.1021/om100814h

Reactivity involving metallacycle ring opening of ortho-metalated bis(phosphinimine)carbazolide complexes (L^Ph-κ³N, κ²C^P-Ph)Lu(THF) (1a) and (L^Pipp- κ³N,κ²C^P-Ph)Lu(THF) (1b); L = [1,8-(Ph₂P=NAr)₂dmc]; Ar = Ph (L^Ph), p-isopropylphenyl (L^Pipp); dmc = 3,6-dimethylcarbazolide) is described. Reaction of 1a,b with bulky anilines (2,4,6-trimethylaniline, MesNH₂; 2,4,6-triisopropylaniline, TripNH₂) promoted metallacycle ring opening of two ortho-metalated P-phenyl groups to liberate the bis(anilide) products (L^Ph-κ³N)Lu(NHMes) ₂ (2) and (L^Pipp-κ³N)Lu(NHTrip) ₂ (3). Regardless of the quantity of TripNH₂ or MesNH ₂ utilized, double ring opening always prevailed to afford the bis(anilide) product, rather than the mono(anilide). In contrast, reaction of complex 1b with the bulkier reagent 2,4,6-tri-tert-butylaniline (Mes NH ₂) only caused metallacycle ring opening of one ortho-metalated P-phenyl group, affording the mono(anilide) complex (L^Pipp- κ³N,κC^P-Ph)Lu(NHMes) (4) exclusively. Complex 4 rapidly undergoes an intramolecular rearrangement whereby metalation of an N-aryl group promotes metallacycle ring opening of the ligated P-phenyl moiety to give (L^Pipp-κ³N,κC^N-Pipp) Lu(NHMes) (5) as the structural isomer. Through deuterium labeling and kinetic studies it was established that the thermal rearrangement of 4 does not proceed through an imido intermediate. Compounds 2, 3, and 5 were characterized by single-crystal X-ray diffraction studies.

Full text of DOI:10.1021/om100814h

58 Organometallics 2011, 30, 58–67
DOI: 10.1021/om100814h
Kinetic and Mechanistic Investigation of Metallacycle Ring Opening
in an Ortho-Metalated Lutetium Aryl Complex
Kevin R. D. Johnson and Paul G. Hayes*
Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive,
Lethbridge, Alberta, Canada T1K 3M4
Received August 19, 2010
Reactivity involving metallacycle ring opening of ortho-metalated bis(phosphinimine)carbazolide
complexes (LPh-κ3N,κ2CP-Ph)Lu(THF) (1a) and (LPipp-κ3N,κ2CP-Ph)Lu(THF) (1b); L=[1,8-(Ph2Pd
NAr)2dmc]; Ar=Ph (LPh), p-isopropylphenyl (LPipp); dmc=3,6-dimethylcarbazolide) is described.
Reaction of 1a,b with bulky anilines (2,4,6-trimethylaniline, MesNH2; 2,4,6-triisopropylaniline,
TripNH2) promoted metallacycle ring opening of two ortho-metalated P-phenyl groups to liberate
the bis(anilide) products (LPh-κ3N)Lu(NHMes)2 (2) and (LPipp-κ3N)Lu(NHTrip)2 (3). Regardless
of the quantity of TripNH2 or MesNH2 utilized, double ring opening always prevailed to afford
the bis(anilide) product, rather than the mono(anilide). In contrast, reaction of complex 1b with the
bulkier reagent 2,4,6-tri-tert-butylaniline (Mes*NH2) only caused metallacycle ring opening of one
ortho-metalated P-phenyl group, affording the mono(anilide) complex (LPipp-κ3N,κCP-Ph)Lu(NHMes*)
(4) exclusively. Complex 4 rapidly undergoes an intramolecular rearrangement whereby meta-
lation of an N-aryl group promotes metallacycle ring opening of the ligated P-phenyl moiety to give
(LPipp-κ3N,κCN-Pipp)Lu(NHMes*) (5) as the structural isomer. Through deuterium labeling and kinetic
studies it was established that the thermal rearrangement of 4 does not proceed through an imido
intermediate. Compounds 2, 3, and 5 were characterized by single-crystal X-ray diffraction studies.
Introduction
anilido-phosphinimine ligands,4 in addition to many other
scaffolds.
A decomposition route often encountered in organo-
lanthanide complexes is ligand cyclometalation via intra-
molecular C-H bond activation. Such pathways have been
well-documented in highly reactive alkyl and hydrido rare-
earth complexes supported by Cp* and Cp0 (Cp0 =substituted
cyclopentadienyl),1 β-diketiminate,2 amido-pyridinate,3 and
From a synthetic perspective, a ligand metalation process
can have diverse consequences. For example, in the context
of an olefin polymerization catalyst, the cyclometalative
C-H bond activation often results in catalyst deactivation
and deprivation of any living polymerization processes.5
Furthermore, ligand metalation may occur through numer-
ous competing intramolecular C-H bond activation path-
ways. If multiple products are generated, it often proves
difficult to separate or characterize the mixture.
*To whom correspondence should be addressed. E-mail: p.hayes@
uleth.ca.
(1) (a) Evans, W. J.; Ulibarri, T. A.; Ziller, J. W. Organometallics
1991, 10, 134–142. (b) Evans, W. J.; Perotti, J. M.; Ziller, J. W. Inorg. Chem.
2005, 44, 5820–5825. (c) Evans, W. J.; Champagne, T. M.; Ziller, J. W.
J. Am. Chem. Soc. 2006, 128, 14270–14271. (d) den Haan, K. H.; Teuben, J. H.
J. Chem. Soc., Chem. Commun. 1986, 682–683. (e) Booij, M.; Meetsma, A.;
Teuben, J. H. Organometallics 1991, 10, 3246–3252. (f) Booij, M.; Deelman,
B.-J.; Duchateau, R.; Postma, D. S.; Meetsma, A.; Teuben, J. H. Organometallics
1993, 12, 3531–3540. (g) Thompson, M. E.; Bercaw, J. E. Pure Appl. Chem.
1984,56, 1–11. (h) Takenaka, Y.; Hou, Z. Organometallics 2009, 28, 5196–5203.
(2) (a) Conroy, K. D.; Hayes, P. G.; Piers, W. E.; Parvez, M. Organo-
metallics 2007, 26, 4464–4470. (b) Hayes, P. G.; Piers, W. E.; Lee, L. W. M.;
Knight, L. K.; Parvez, M.; Elsegood, M. R. J.; Clegg, W. Organometallics
2001, 20, 2533–2544. (c) Hayes, P. G.; Piers, W. E.; Parvez, M. Organo-
metallics 2005, 24, 1173–1183. (d) Knight, L. K.; Piers, W. E.; Fleurat-
Lessard, P.; Parvez, M.; McDonald, R. Organometallics 2004, 23, 2087–
2094. (e) Knight, L. K.; Piers, W. E.; McDonald, R. Organometallics 2006,
25, 3289–3292. (f) Kenward, A. L.; Piers, W. E.; Parvez, M. Organometallics
2009, 28, 3012–3020.
While frequently unfavorable, ligand metalation processes
can sometimes be exploited to achieve a desired form of
reactivity.6 A recent example reported by Waterman et al.
involves a cyclometalated zirconium triamidoamine species
that exhibits catalytic reactivity for the selective dehydro-
coupling of phosphines and arsines (Scheme 1),7 in addition
to catalytic hydrophosphination of terminal alkynes.8
(4) (a) Conroy, K.; Piers, W.; Parvez, M. J. Organomet. Chem. 2008,
693, 834–846. (b) Liu, B.; Cui, D.; Ma, J.; Chen, X.; Jing, X. Chem. Eur. J.
2007, 13, 834–845. (c) Liu, B.; Liu, X.; Cui, D.; Liu, L. Organometallics
2009, 28, 1453–1460.
(5) (a) Schrock, R. R.; Bonitatebus, P. J., Jr.; Schrodi, Y. Organo-
metallics 2001, 20, 1056–1058. (b) Schrodi, Y.; Schrock, R. R.; Bonitatebus,
P. J., Jr. Organometallics 2001, 20, 3560–3573.
(6) (a) Mork, B. V.; Tilley, T. D. J. Am. Chem. Soc. 2001, 123, 9702–9703.
(b) Mork, B. V.; Tilley, T. D. J. Am. Chem. Soc. 2004, 126, 4375–4385.
(7) (a) Roering, A. J.; Davidson, J. J.; MacMillan, S. N.; Tanski,
J. M.; Waterman, R. Dalton Trans. 2008, 4488–4498. (b) Waterman, R.
Organometallics 2007, 26, 2492–2494.
(3) (a) Luconi, L.; Lyubov, D. M.; Bianchini, C.; Rossin, A.; Faggi,
C.; Fukin, G. K.; Cherkasov, A. V.; Shavyrin, A. S.; Trifonov, A. A.;
Giambastiani, G. Eur. J. Inorg. Chem. 2010, 608–620. (b) Qayyum, S.;
€
Skvortsov, G. G.; Fukin, G. K.; Trifonov, A. A.; Kretschmer, W. P.; Doring,
C.; Kempe, R. Eur. J. Inorg. Chem. 2010, 248–257. (c) Skvortsov, G.;
€
Fukin, G.; Trifonov, A.; Noor, A.; Doring, C.; Kempe, R. Organometallics
(8) Roering, A. J.; Leshinski, S. E.; Chan, S. M.; Shalumova, T.;
Macmillan, S. N.; Tanski, J. M.; Waterman, R. Organometallics 2010,
29, 2557–2565.
2007, 26, 5770–5773. (d) Zimmermann, M.; Estler, F.; Herdtweck, E.;
Tornroos, K. W.; Anwander, R. Organometallics 2007, 26, 6029–6041.
€
r
pubs.acs.org/Organometallics Published on Web 12/08/2010
2010 American Chemical Society

Article
Scheme 1. Mechanism of Pnictogen Dehydrocoupling7
Organometallics, Vol. 30, No. 1, 2011 59
Recently, we described the synthesis of a bis(phosphinimine)-
carbazolide pincer ligand, L (L = [1,8-(Ph2PdNAr)2dmc];
Ar=Ph (LPh), p-isopropylphenyl (LPipp); dmc=3,6-dimethyl-
carbazolide), and its use in the preparation of well-defined
organolutetium complexes.9 It was found that dialkyl lutetium
complexes of L were thermally unstable and rapidly underwent
two sequential intramolecular metalative alkane elimination
processes. The final product of this transformation contained
the ligand bound in a κ5 manner through the three nitrogen
atoms of the ligand framework and two ortho-metalated P-phenyl
rings (LAr-κ3N,κ2CP-Ph)Lu(THF) (Ar=Ph (1a), p-isopropyl-
phenyl (1b)). Herein, we report an investigation into the
reactivity patterns of these ortho-metalated organolutetium
complexes through the process of metallacycle ring opening.
In particular, we prepared bis(anilide) lutetium complexes
supported by L as well as a mixed aryl/anilide lutetium
complex. The latter was assessed for its potential to liberate
a lutetium imido complex (LLudNR) by thermolysis. As a
result of this study, novel reactivity patterns were uncovered
in conjunction with the clean formation of complexes that
exhibit unique bonding modes and structures.
Figure 1. Thermal ellipsoid plot (50% probability) of 2 with
hydrogen atoms (except H1N and H2N) and solvent molecules
of crystallization omitted for clarity.
Scheme 2. Metallacycle Ring-Opening Reaction of Complex 1a
with MesNH2
of 2 is illustrated in Figure 1 as a thermal ellipsoid plot. In the
solid state, complex 2 is defined by coordination of two 2,4,6-
trimethylanilide ligands and the ancillary pincer ligand bound
in a κ3 fashion through its three nitrogen atoms. The five-
coordinate lutetium center exhibits a distorted-trigonal-
bipyramidal geometry with the anilide ligands (N1 and N2)
and N4 of the ancillary in the equatorial positions. The phos-
phinimine nitrogen donors of the pincer ligand (N3 and N5)
occupy the apical sites. The metal center sits above the plane
Results and Discussion
Metallacycle Ring-Opening Reactivity. The ortho-metalated
lutetium aryl complex 1 can be reacted with various anilines in
toluene solution at ambient temperature to give the metallacycle
ring-opened product. For example, treatment of 1a with 2 equiv
of 2,4,6-trimethylaniline (MesNH2) resulted in an immediate
reaction whereby ring opening of the metalated P-phenyl rings
liberated the bis(anilide) complex 2, (LPh-κ3N)Lu(NHMes)2
(Scheme 2). Similar reactivity has been previously documented
in rare-earth complexes supported by an anilido-phosphinimine
ligand.4b,c
˚
of the dimethylcarbazole backbone by 0.770 A. The lutetium-
anilide bond lengths fall within the normal range at 2.1777(18) A
˚
˚
(Lu1-N1) and 2.1749(19) A (Lu1-N2) (Table 1). Similarly,
the ancillary ligand, LPh, coordinates to lutetium with bond
˚
˚
lengths of 2.3586(17) A (Lu1-N3), 2.3595(16) A (Lu1-N4),
˚
and 2.3586(17) A (Lu1-N5), which correspond well with
previously reported values.9
The bis(anilide) lutetium complex 2 is C2v symmetric in
solution, as depicted by a sharp singlet (δ 30.55) in its 31P{1H}
NMR spectrum (benzene-d6). The H NMR spectrum for
Of particular interest to us was the installation of only one
anilide group on lutetium so as to afford a mixed aryl/anilide
species. The impetus behind this goal stemmed from the idea
that thermolysis of a mixed aryl/anilide complex may pro-
mote intramolecular metallacycle ring opening to yield a
terminal lutetium imido complex.10 To this end, we explored
the reaction of complex 1a with only 1 equiv of MesNH2 in
the prospect of generating the mono(anilide) congener of 2.
Unfortunately, repeated attempts of this reaction were
1
complex 2 exhibited the expected signals for the ancillary
ligand, in addition to a set of resonances corresponding to
two mesityl anilide ligands. In particular, the NH anilide
protons of complex 2 gave rise to a singlet at δ 3.97 with an
integration of 2H (benzene-d6).
Single crystals of complex 2 suitable for an X-ray diffraction
study were readily obtained from a benzene solution layered
with pentane at ambient temperature. The molecular structure
(10) (a) Scott, J.; Basuli, F.; Fout, A. R.; Huffman, J. C.; Mindiola,
D. J. Angew. Chem., Int. Ed. 2008, 47, 8502–8505. (b) Basuli, F.;
Tomaszewski, J.; Huffman, J. C.; Mindiola, D. J. Organometallics 2003,
22, 4705–4714.
(9) Johnson, K. R. D.; Hayes, P. G. Organometallics 2009, 28, 6352–
6361.

60 Organometallics, Vol. 30, No. 1, 2011
Johnson and Hayes
˚
Table 1. Selected Bond Distances (A) and Angles (deg)
for Compounds 2 and 3
2
3
Lu1-N1
Lu1-N2
Lu1-N3
Lu1-N4
Lu1-N5
P1-N3
2.1777(18)
2.1749(19)
2.3586(17)
2.3595(16)
2.3586(17)
1.5943(18)
1.6097(17)
2.2030(72)
2.1698(81)
2.3309(76)
2.3225(71)
2.3394(69)
1.5764(83)
1.6000(72)
P2-N5
N3-Lu1-N5
N3-Lu1-N4
N4-Lu1-N5
N1-Lu1-N2
Lu1-N1-C51a
Lu1-N2-C60a
Lu1-N1-C57b
Lu1-N2-C72b
168.76(6)
86.38(6)
84.44(6)
118.11(7)
144.02(15)
151.34(17)
170.75(27)
84.91(26)
86.35(26)
110.07(28)
146.53(66)
143.77(65)
a The listed angle pertains only to complex 2. b The listed angle pertains
only to complex 3.
Scheme 3. Metallacycle Ring-Opening Reaction of Complex 1b
with TripNH2
Figure 2. Thermal ellipsoid plot (30% probability) of 3 with
hydrogen atoms (except H1N and H2N) and solvent molecules
of crystallization omitted for clarity.
Similar to that observed in 2, the lutetium center in complex 3
adopts a trigonal-bipyramidal geometry with two Trip anilide
ligands and the ancillary pincer bound in a κ3 fashion through its
three nitrogen atoms. Likewise, the anilide ligands (N1 and
N2) and N4 of the ancillary ligand occupy the equatorial
positions, while N3 and N5 define the apical sites. The
lutetium-anilide bond lengths in complex 3 are comparable
˚
to that of 2 with distances of 2.2030(72) and 2.1689(81) A for
Lu1-N1 and Lu1-N2, respectively (Table 1). In addition,
the Lu-N-C anilide bond angles in both complexes 2 and 3
are similar with values ranging from 143.77(65) to 151.34(17)°.
In contrast to the reactivity observed upon reaction of 1
with MesNH2 and TripNH2, addition of 1 equiv of 2,4,6-tri-
tert-butylaniline (Mes*NH2) to 1b only promoted metalla-
cycle ring opening of a single ortho-metalated P-phenyl group,
generating the desired mono(anilide) complex (LPipp-κ3N,
κCP-Ph)Lu(NHMes*) (4) (Scheme 4). Even under forcing
conditions (100 °C for 24 h) with multiple equivalents
of Mes*NH2, it was found that double substitution of 1b
(to make (LPipp-κ3N)Lu(NHMes*)2) was not possible.
Interestingly, complex 4 was highly unstable toward a
thermally induced intramolecular rearrangement to the struc-
tural isomer (LPipp-κ3N,κCN-Pipp)Lu(NHMes*) (5) (Scheme 4).
Unfortunately, the high thermal instability of 4 precluded its
isolation as a solid. Complex 4 could, however, be readily
observed in situ by 31P{1H} NMR spectroscopy throughout
the transformation from 1b to 5. The 31P{1H} NMR spec-
trum of 4 revealed a marked difference from that observed
for 2 and 3. In the solution state, complex 4 exhibited low
symmetry (C1), as demonstrated by two singlets of equal
intensity in the 31P{1H} NMR spectrum at δ 31.77 and 22.64
(benzene-d6), corresponding to the chemically inequivalent
phosphinimine groups. Attempts to fully characterize 4 in
situ by other NMR nuclei (1H or 13C{1H}) were unsuccessful
due to the severity of overlapping signals in the 1H or 13C{1H}
NMR spectra corresponding to complexes 1b, 4, and 5.
The thermal transformation of 4 to 5 liberated a structural
isomer whereby the ancillary ligand is ortho-metalated via an
N-aryl ring in 5, as compared to a P-phenyl ring in 4. The
31P{1H} NMR spectrum of complex 5 contains two resonances
hampered by Schlenk-type ligand redistribution processes
whereby only the bis(anilide) complex 2 could be isolated.
Similar reactivity has previously been documented in the
preparation of other mono(anilide) rare-earth complexes.2d
In a further effort to prepare a mono(anilide) lutetium com-
plex, we pursued the possibility of reacting 1 with anilines of
steric bulk even greater than that of MesNH2. The premise
behind this approach was to install a sufficiently bulky anilide
ligand to inhibit further intermolecular metallacycle ring opening.
Thus, we performed reactions of complex 1with various anilines
of gradually increasing steric bulk; specifically, the reagents
2,4,6-triisopropylaniline (TripNH2) and 2,4,6-tri-tert-butyl-
aniline (Mes*NH2) were utilized.
Reaction of complex 1b with TripNH2 afforded the double
metallacycle ring-opening product (LPipp-κ3N)Lu(NHTrip)2
(3), analogous to the bis(anilide) 2 (Scheme 3). Similar to the
case for 2, complex 3 exhibited C2v symmetry in solution on
the NMR time scale. In the 31P{1H} NMR spectrum of 3, a
sharp singlet resonating at δ 30.57 (benzene-d6) was observed.
This shift was highly comparable to the 31P{1H} NMR signal
for complex 2 (δ 30.55). The 1H NMR spectrum consisted of
the expected resonances for the ancillary ligand as well as signals
corresponding to two Trip anilide ligands. Remarkably, the
two NH protons for complex 3 were found to resonate with
the same chemical shift as complex 2 at δ 3.97 (benzene-d6).
Single crystals of complex 3 were obtained by recrystalli-
zation from a concentrated pentane solution at ambient tem-
perature. The solid-state structure of 3, as determined from
an X-ray diffraction experiment, is depicted in Figure 2.

Article
Organometallics, Vol. 30, No. 1, 2011 61
Scheme 4. Metallacycle Ring-Opening Reaction of 1 with 1 Equiv of Mes*NH2
˚
Table 2. Selected Bond Distances (A) and Angles (deg)
for Compound 5
Lu1-N1
Lu1-N2
Lu1-N3
Lu1-N4
Lu1-C40
2.1632(32)
2.3072(30)
2.3123(30)
2.3122(30)
2.3368(42)
C39-C40
C39-N2
P1-N2
1.4083(50)
1.4436(46)
1.5840(33)
1.6142(31)
P2-N4
N2-Lu1-N4
N2-Lu1-N3
N3-Lu1-N4
N1-Lu1-N3
Lu1-N1-C57
142.46(11)
81.80(11)
86.50(10)
125.19(12)
164.15(29)
Lu1-N2-C39
N2-Lu1-C40
Lu1-C40-C39
N2-C39-C40
92.88(21)
61.31(12)
92.59(25)
112.25(32)
N2, C39, and C40. In the solid state the metallacycle takes on
a kite-shaped geometry defined by two long bonds (Lu1-N2,
˚
˚
2.3072(30) A; Lu-C40, 2.3368(42) A) and two short bonds
˚
˚
(C39-N2, 1.4436(46) A; C39-C40, 1.4083(50) A) (Table 2).
The sum of the angles within the metallacycle is 359.03°,
indicating a nearly planar conformation. The Lu-N1-C57
anilide bond angle in 5 (164.15(29)°) is substantially more
linear than that observed in complexes 2 and 3 (which range
from 143.77(65) to 151.34(17)°). This difference is likely due
to the increased steric bulk of the 2,4,6-tri-tert-butylanilide
ligand.
Figure 3. Thermal ellipsoid plot (50% probability) of 5 with
hydrogen atoms (except H1N) and solvent molecules of crystal-
lization omitted for clarity.
(δ 29.96 and 11.83 (benzene-d6)) slightly upfield of those
observed for 4. The 1H and 13C{1H} NMR spectra of 5 were
found to be extremely complicated, especially in the aromatic
regions, due to the low symmetry (C1) of the complex. The 1H
and 13C{1H} NMR spectra exhibited the expected reso-
nances for the ancillary ligand and one 2,4,6-tri-tert-butyl-
anilide moiety. In particular, the NH anilide proton of
Kinetic Analysis. Due to its rapid rate of decomposition,
complex 4 could be neither isolated nor fully characterized by
1H or 13C{1H} NMR spectroscopy. However, the formation
of 4 from 1b, followed by its decomposition to complex 5
(eq 1), was quantitatively monitored by 31P{1H} NMR spectros-
copy. The progress of reaction at 296.9 K (from t=185 s to
t = 157 000 s) is portrayed in Figure 4 as a stacked plot of
31P{1H} NMR spectra (toluene-d8) recorded at predefined
time intervals. Over the course of the reaction, the decreasing
concentration of 1b (δ 29.7) is accompanied by the forma-
tion of asymmetric intermediate 4, depicted by two signals
resonating at δ 31.7 and 22.4. Within two days at this tem-
perature, complex 4 gradually undergoes an intramolecular
metalation exchange to afford exclusively product 5 (δ 29.7
and 11.4).
1
complex 5 gave rise to a singlet in the H NMR spectrum
at δ 4.88 with an integration of 1H (benzene-d6).
In order to unambiguously establish the connectivity of 5,
a single-crystal X-ray diffraction study was performed. Com-
plex 5 was found to be highly crystalline in nature, and single
crystals were readily obtained from a concentrated toluene
solution layered with pentane at -35 °C. As depicted in
the molecular structure of 5 (Figure 3), the low symmetry of
the complex in the solid state (C1) matched that observed in
solution. The complex adopts a five-coordinate geometry
with one site occupied by a 2,4,6-tri-tert-butylanilide ligand.
The remaining four coordination sites are defined by the
ancillary ligand bound in a κ4 fashion through the three
nitrogen atoms and the ortho carbon of one Pipp group. At
k1
s
f
k2
s
f
ꢀ
1b þ Mes NH2
4
5
ð1Þ
The observed rate constant (k1(obsd)) for the formation of
complex 4 was obtained from a second-order plot of the
reaction of 1b with Mes*NH2. The reaction was monitored
over a broad range of temperatures (296.9-349.1 K), with
observed t1/2 values ranging from 18 500 to 198 s (Table 3).
An Eyring plot was constructed that allowed for extraction
of the activation parameters ΔHq = 73.5(2) kJ mol-1 and
ΔSq=-50.3(5) J K-1 mol-1 for this transformation (Figure 5a).
˚
1.554 A, the lutetium center sits substantially out of the plane
of the dimethylcarbazole ligand backbone, presumably due
to the extremely sterically demanding nature of the ligands
coordinated to it. Of particular interest in complex 5 is the
unusual four-membered metallacycle constituted by Lu1,

62 Organometallics, Vol. 30, No. 1, 2011
Johnson and Hayes
Figure 4. Stacked 31P{1H} NMR spectra depicting the metallacycle ring-opening reaction of complex 1b to 4 followed by
transformation to complex 5.
Table 3. Observed and Calculated Rate Constants for the
Metallacycle Ring-Opening Reaction of Complex 1b with
Mes*NH2 at Temperatures Ranging from 296.9 to 349.1 K
Table 4. Calculated Rate Constants for the Intramolecular
Rearrangement of Complex 4 to Complex 5 at Temperatures
Ranging from 296.9 to 349.1 K
T/K k1(obsd)/M-1 s-1 t1/2(obsd)/s k1(calcd)/M-1 s-1 t1/2(calcd)/s
T/K
k2(calcd)/s-1
t1/2(calcd)/s
3
296.9
304.6
315.7
326.8
338.0
349.1
1.81 ꢁ 10-3
3.43 ꢁ 10-3
9.96 ꢁ 10-3
3.07 ꢁ 10-2
7.44 ꢁ 10-2
1.66 ꢁ 10-1
18500
9660
3330
1070
445
1.69 ꢁ 10-3
3.22 ꢁ 10-3
9.44 ꢁ 10-3
2.68 ꢁ 10-2
6.42 ꢁ 10-2
1.49 ꢁ 10-1
19900
10300
3510
1230
517
296.9
304.6
315.7
326.8
338.0
349.1
3.17 ꢁ 10-5
8.46 ꢁ 10-5
2.54 ꢁ 10-4
6.84 ꢁ 10-4
2.02 ꢁ 10-3
5.14 ꢁ 10-3
21900
8200
2730
1010
342
198
222
135
Table 5. Transition State Activation Parameters
for the Transformation of Complex 1b to 5
rate const
ΔHq/kJ mol-1
ΔSq/J K-1 mol-1
k1(obsd)
k1(calcd)
k2(calcd)
73.5(2)
72.3(1)
80.3(1)
-50.3(5)
-55.0(4)
-60.0(4)
In contrast to the second-order reaction which converted
complex 1b to 4, the transformation from 4 to 5 involved
significantly more complicated kinetic behavior. No simple
mathematical rate law could be derived for the expression of
k2 due to the complexity of the consecutive reactions. Thus,
no values for k2(obsd) could be determined from the experi-
mental data. However, using the kinetic simulation software
COPASI,11 we were able to model the two-step process from
1b to 5. As such, the modeled data set allowed for calculation
of the simulated rate constants, k1(calcd) and k2(calcd), for
the consecutive reactions; these values are listed in Tables 3
and 4, respectively. The k1(calcd) values agree fairly well with
the k1(obsd) values; however, it should be noted that the
calculated rate constants were consistently slightly slower
(by 5-14%) than the observed rate constants. Due to this
observation, it is reasonable to assume that the calculated
rate constants for k2 (Table 4) may also be slow by a similar
margin of error. However, a visual inspection of the simu-
lated reaction progress over time indicated good agreement
with the experimental reaction plots.
As with the observed rate constant k1(obsd), Eyring plots
were constructed to express the temperature dependence of
the simulated rate constants k1(calcd) and k2(calcd). From these
plots, the activation parameters of ΔHq = 72.3(1) kJ mol-1
Figure 5. (a) Eyring plots for (a) the metallacycle ring-opening
reaction of 1b to 4 (derived from k1(obsd)) and (b) the intramo-
lecular rearrangement of 4 to 5 (derived from k2(calcd)).
The large negative entropy of activation suggests a highly ordered
transition state, consistent with the expected σ-bond metathesis
mechanism.
(11) Hoops, S.; Sahle, S.; Gauges, R.; Lee, C.; Pahle, J.; Simus, N.;
Singhal, M.; Xu, L.; Mendes, P.; Kummer, U. Bioinformatics 2006, 22,
3067–3074.

Article
Organometallics, Vol. 30, No. 1, 2011 63
Scheme 5. Deuterium Labeling Experiment 1: Reaction of Complex 1b with Mes*ND2
and ΔSq=-55.0(4) J K-1 mol-1 and ΔHq=80.3(1) kJ mol-1
and ΔSq =-60.0(4) J K-1 mol-1 were extracted for k1(calcd)
and k2(calcd), respectively (Table 5). The parameters obtained
for k1(calcd) agree very well with those obtained from k1(obsd).
For the activation parameters obtained from the k2(calcd)
rate constants, both the enthalpic barrier and entropy of
activation remained similar to that for k1.
of 5 from 1b, we performed two independent deuterium
labeling experiments (Schemes 5 and 6).15
The first deuterium labeling experiment involved the reac-
tion of complex 1b with Mes*ND2. As outlined in Scheme 5,
if pathway 1 was operative, the labeled anilide formed upon
initial reaction (4-N-d1-ring-d1) would retain a deuterium atom
on the anilide nitrogen throughout the transformation to give
the final product 5-N-d1-ring-d1, with a deuterium-labeled
anilide nitrogen. Conversely, if pathway 2 was operative, the
deuterium on the anilide nitrogen of 4-N-d1-ring-d1 would
become scrambled into the P-phenyl rings upon imido forma-
tion. This would be followed by remetalation of an N-Pipp
group, thus installing a proton onto the anilide nitrogen
atom of the final putative product, 5-ring-d2. When this
Deuterium Labeling and Mechanism. The structure of com-
plex 5, confirmed by solution multinuclear NMR spectros-
copy and solid-state X-ray diffraction analysis, suggests that
an unusual reaction mechanism is operative in its formation
from starting material 1b. It is evident that the mechanism
for the generation of 5 from 1b requires multiple steps, due to
the intermediacy of 4, as observed by multinuclear NMR
spectroscopy. Several pathways for this transformation can
be envisioned, the two most plausible of which will be
described in depth. The first mechanism (pathway 1) involves
the metallacycle ring-opening reaction of complex 1b with
Mes*NH2 to give mono(anilide) 4, followed by direct meta-
lation exchange of the aryl rings between P-Ph and N-Pipp
groups of the ancillary ligand to afford complex 5 as the final
product. An alternative mechanism (pathway 2) could involve
the formation of 4 as in pathway 1. Following this, intramo-
lecular metallacycle ring opening of complex 4 could give
rise to a transient lutetium imido complex, whereby remeta-
lation of an N-Pipp group would afford the final product 5.
Although there have been no terminal, unconstrained lute-
tium imides reported to date,12 there are several examples of
rare-earth-metal complexes that are formed via a transient
terminal imido intermediate.10,13 More recently, the isola-
tion of a terminal scandium imide has been realized,14 thus
suggesting that the paucity of such rare-earth species in the
literature is not due to thermodynamic limitations. In order
to establish which mechanism is operative in the formation
1
reaction was followed on an NMR tube scale by H NMR
spectroscopy, it was determined that the final product of the
transformation contained a deuterium atom on the anilide
nitrogen, thus suggesting that pathway 1, rather than path-
way 2, was operative. This conclusion was supported by
the lack of a resonance at δ 4.88 in the 1H NMR spectrum of
(15) Two additional mechanisms (pathways 3 and 4) have also been
suggested to us. Pathway 3 involves the reaction of complex 4 with a
second equivalent of Mes*NH2 to give the bis(anilide) complex analo-
gous to 2 and 3. From such a species, loss of 2,4,6-tri-tert-butylaniline
with concomitant metalation of an N-Pipp group would result in 5.
Pathway 3 is not considered to be a probable mechanism on the grounds
of steric hindrance. It does not appear to be possible to fit two 2,4,6-tri-
tert-butylanilide groups into the coordination pocket defined by the
ancillary ligand because of too much steric crowding. Furthermore,
operation of pathway 3 was nullified through the aforementioned
deuterium labeling studies. For example, in deuterium labeling experi-
ment 1, the reaction of 1b with 2 equiv of Mes*ND2 would result in the
intermediate (LPipp-κ3N-ring-d2)Lu(NDMes*)2 if pathway 3 were op-
erative. This species would then undergo loss of Mes*NHD to give the
final product 5-N-d1-ring-d2. It can be expected then that, in subsequent
reactions, competition between 1b reacting with Mes*ND2 or
Mes*NHD would occur. As a result, deuterium incorporation on the
anilide nitrogen of complex 5 would not be either 100% or 0% (as for
pathways 1 and 2, respectively) but, rather, a statistical mixture. In
pathway 4, the anilide ligand in complex 4 could serve to shuttle an H
atom from the N-Ar group to the metalated P-Ph moiety via the
intermediate (L-κ3N,κCP-Ph,κCN-Ar)Lu(NH2Mes*). This mechanism
was disproven by both deuterium labeling experiments, as Mes*NHD,
which would afford a statistical mixture of D and H incorporation on the
anilide nitrogen of complex 5, would be produced in both cases.
(12) (a) Giesbrecht, G. R.; Gordon, J. C. Dalton Trans. 2004, 2387–
2393. (b) Scott, J.; Mindiola, D. J. Dalton Trans. 2009, 8463–8472.
(13) (a) Beetstra, D. J.; Meetsma, A.; Hessen, B.; Teuben, J. H.
Organometallics 2003, 22, 4372–4374. (b) Conroy, K. D.; Piers, W. E.;
Parvez, M. Organometallics 2009, 28, 6228–6233.
(14) Lu, E.; Li, Y.; Chen, Y. Chem. Commun. 2010, 46, 4469–4471.

64 Organometallics, Vol. 30, No. 1, 2011
Johnson and Hayes
Scheme 6. Deuterium Labeling Experiment 2: Reaction of Complex 1a-ring-d10 with Mes*NH2
5-N-d1-ring-d1 and an aromatic region that integrated for
one less proton than for 5. Other than these details, the 1H
NMR spectrum of 5-N-d1-ring-d1 was identical with that of the
proteo control, 5, where the NH signal can be clearly
observed at δ 4.88.
Conclusion
The process of metallacycle ring opening has been probed
in detail using a doubly ortho-metalated lutetium aryl complex.
While reaction of (LAr-κ3N,κ2CP-Ph)Lu(THF) with bulky
anilines (MesNH2, TripNH2) resulted in double metallacycle
ring opening to generate the corresponding bis(anilide) lutetium
complexes, utilization of the extremely sterically demanding
Mes*NH2 promoted single metallacycle ring opening to afford
the mono(anilide) complex (LPipp-κ3N,κCP-Ph)Lu(NHMes*) (4)
exclusively. The latter product was found to be highly thermally
sensitive and rapidly underwent an unusual metalation exchange
process to give (LPipp-κ3N,κCN-Pipp)Lu(NHMes*) (5) in high
yield. Through various deuterium labeling and kinetic stud-
ies it was determined that complex 5 forms through direct
metalation exchange, with no evidence of a transient imido
intermediate.
In an effort to access elusive LudE functionalities future
work will explore the reactions of complex 1 with the heavier
group 15 analogues of Mes*NH2. These larger congeners
may exhibit significantly different reactivity patterns, where-
by a complex possessing a terminal lutetium-main-group
multiple bond may be realized through a metallacycle ring-
opening pathway.
The second deuterium labeling experiment involved the
reaction of fully protonated Mes*NH2 with a deuterium-
labeled lutetium analogue to 1b, 1a-ring-d10 (Scheme 6). The
starting material 1-ring-d10 contained fully deuterated
N-phenyl groups as opposed to the proton-containing 4-iso-
propylphenyl groups in 1b. Despite the lack of an isopropyl
group in the para position of the N-aryl ring, we were
confident that 1a-ring-d10 would react in a manner identical
with that for 1b, excluding any kinetic isotope effects.
The identical reactivity patterns and kinetic behavior of 1a
(the protonated version of 1a-ring-d10) and 1b have been
previously documented.9 As depicted in Scheme 6, pathway
1 dictates that the reaction of 1a-ring-d10 with Mes*NH2
would result in the products 40-ring-d10 and 50-ring-d10,
whereby a proton is retained on the anilide nitrogen through-
out the entire process. Conversely, pathway 2 would result in
loss of the anilide proton upon imido formation, followed by
remetalation of a deuterium-labeled N-aryl ring, thus instal-
ling a deuterium atom on the anilide nitrogen. When the
transformation was followed by 1H NMR spectroscopy, the
final product of the reaction of 1a-ring-d10 with Mes*NH2
was observed to be 50-ring-d10, with a proton bound to the
anilide nitrogen. Thus, the deuterium labeling experiment 2
corroborated the results from experiment 1, in that pathway
1 rather than pathway 2 appears to be operative.
The mechanistic work presented herein suggests that the
formation of complex 5 occurs via two sequential metalla-
cycle ring-opening reactions. The first ring opening occurs
during the reaction of 1b with Mes*NH2 to give complex 4,
which possesses a metalated P-phenyl ring. Complex 4 then
undergoes a thermal rearrangement via a rare direct metala-
tion exchange between an N-aryl ring and the metalated
P-phenyl ring to yield the structural isomer 5. The results
from deuterium labeling experiments argue against the pos-
sibility of a transient lutetium imido species being formed as
an intermediate in this transformation.
Experimental Section
General Procedures. Unless otherwise specified, all reactions
were carried out under an argon atmosphere with the rigorous
exclusion of oxygen and water using standard glovebox (MBraun)
or high-vacuum-line techniques. The solvents pentane, benzene,
and toluene were dried and purified using a solvent purification
system (MBraun) and stored in evacuated 500 mL bombs over a
“titanocene” indicator. Deuterated solvents (benzene-d6 and
toluene-d8) were dried over sodium benzophenone ketyl, de-
gassed via three freeze-pump-thaw cycles, distilled under
vacuum, and stored in glass bombs under argon. Unless other-
wise specified, all solvents required for air-sensitive manipula-
tions were introduced directly into the reaction flasks by vacuum
transfer with condensation at -78 °C. For air-stable manipula-
tions, the solvents THF, diethyl ether, and hexanes were purchased
from Fisher Scientific and used without further purification.
Samples for NMR spectroscopy were recorded on a 300 MHz

Article
Organometallics, Vol. 30, No. 1, 2011 65
Bruker Avance II (Ultrashield) spectrometer (1H 300.13 MHz,
13C{1H} 75.47 MHz, 31P{1H} 121.49 MHz) and referenced rela-
tive to either SiMe4 through the residual solvent resonance(s) for
1H and 13C{1H} or external 85% H3PO4 for 31P{1H}. All NMR
spectra were recorded at ambient temperature (295 K) unless
specified otherwise. FT-IR spectra were recorded on a Bruker
ALPHA FT infrared spectrometer with Platinum ATR sam-
pling. Elemental analyses were performed using an Elementar
Americas Vario MicroCube instrument. The reagent 2,4,6-tri-
tert-butylaniline was purchased from Frinton Laboratories and
used as received. Mes*ND2 was prepared via the exchange reac-
tion of Mes*NH2 with D2O under the presence of a catalytic
amount of anhydrous HCl in diethyl ether. TripNH2,16 1a,9 and
1b9 were prepared according to literature procedures. All deu-
terated solvents and aniline-ring-d5 were purchased from Cam-
bridge Isotope Laboratories. All other reagents were obtained
from Aldrich Chemicals or Alfa Aesar and used as received.
(LPh-K3N)Lu(NHMes)2 (2). MesNH2 (0.151 mL, 1.07 mmol)
was added to a solution of 1a (0.531 g, 0.537 mmol) in toluene
(20 mL) at ambient temperature. The resulting orange solution
was stirred for 1 h, following which all volatiles were removed
under reduced pressure. In a glovebox, the oily residue was
washed with pentane (2ꢁ2 mL) and then dried under vacuum.
The solid was taken up in hot benzene, filtered, and cooled to
ambient temperature, where it was left to crystallize. After 2 days
the mother liquor was decanted off, leaving a yellow crystalline
solid that was washed with pentane (5 mL) and thoroughly dried
in vacuo. Yield: 0.376 g (59.0%). 1H NMR (benzene-d6): δ 8.12
(s, 2H, 4,5-Cz CH), 7.71 (dd, 3JHP=11.1 Hz, 3JHH=7.3 Hz, 8H,
PPh o-CH), 7.42 (d, 2H, J=17.1 Hz, 2,7-Cz CH), 6.94-6.82 (m,
20H, aromatic H), 6.70 (m, 6H, aromatic H), 3.97 (s, 2H, NH),
2.37 (s, 6H, mesityl p-CH3), 2.22 (s, 6H, Cz CH3), 2.20 (s, 12H,
mesityl o-CH3). 13C{1H} NMR (benzene-d6): δ 154.2 (s, aro-
matic ipso-C), 151.0 (d, JCP = 3.3 Hz, aromatic ipso-C), 145.9
(d, JCP =8.2 Hz, aromatic ipso-C), 134.2 (d, JCP =9.4 Hz, PPh
o-CH), 133.3 (d, JCP=13.2 Hz, 2,7-Cz CH), 132.1 (d, JCP=2.5Hz,
aromatic CH), 131.6 (d, JCP = 6.5 Hz, aromatic CH), 131.1 (s,
ipso-C), 129.8 (s, ipso-C), 129.2 (s, aromatic CH), 128.6 (d,
partially obscured by solvent, JCP = 11.9 Hz, aromatic CH),
127.7 (d, obscured by solvent, JCP = 3.5 Hz, aromatic CH),
(s, aromatic ipso-C), 151.5 (d, JCP = 3.3 Hz, aromatic ipso-C),
143.5 (d, JCP=4.3 Hz, aromatic ipso-C), 143.1 (d, JCP=8.7 Hz,
aromatic ipso-C), 134.9 (d, JCP =9.5 Hz, PPh o-CH), 134.3 (d,
J
CP =13.1 Hz, 2,7-Cz CH), 132.6 (s, aromatic ipso-C), 132.3 (s,
aromatic ipso-C), 131.9 (s, PPh p-CH), 131.9 (d, JCP =10.4 Hz,
Pipp o-CH), 130.7 (d, JCP=92.0 Hz, aromatic ipso-C), 128.1 (d,
JCP=11.7 Hz, PPh m-CH), 125.9 (d, JCP=3.6 Hz, Pipp m-CH),
125.4 (d, JCP = 14.1 Hz, aromatic ipso-C), 125.1 (d, JCP = 3.7
Hz, 4,5-Cz CH), 120.6 (s, Trip m-CH), 120.5 (s, aromatic ipso-
C), 107.6 (d, JCP = 115.6 Hz, aromatic ipso-C), 34.7 (s, Trip
p-CH(CH3)2), 33.5 (s, Pipp p-CH(CH3)2), 30.2 (br s, Trip
o-CH(CH3)2), 25.4 (s, Trip p-CH(CH3)2), 24.3 (s, Trip o-CH-
(CH3)2), 24.1 (s, Pipp p-CH(CH3)2), 21.0 (s, Cz CH3). 31P{1H}
NMR (benzene-d6): δ 30.57. Anal. Calcd for C91H112LuN5P2
(3 (pentane)): C, 72.25; H, 7.46; N, 4.63. Found: C, 72.33; H,
3
7.82; N, 4.85.
(LPipp-K3N,KCN-Pipp)Lu(NHMes*) (5). Toluene (40 mL) was
added to a bomb charged with an intimate mixture of 1b and
Mes*NH2 to give an orange solution. The reaction mixture was
heated to 100 °C for 3 h, following which it was cooled to ambient
temperature and the volume concentrated under vacuum to
∼5 mL. Upon standing for 5 min the product crystallized out of
solution as a solid orange mass. In a glovebox, the crystals were
redissolved in 5 mL of hot toluene to give a dark red solution.
After it was cooled to ambient temperature, the toluene solution
was layered with pentane (5 mL) and left for 16 h to crystallize.
Matted needles of the product were collected by filtration,
washed with pentane (2 ꢁ 2 mL), and thoroughly dried under
reduced pressure. Yield: 0.716 g (77.3%). 1H{31P} NMR
(benzene-d6): δ 8.02 (s, 1H, 4-Cz CH), 7.96 (ov d, 3JHH =8.1 Hz,
2H, PPh o-CH), 7.94 (ov s, 1H, 5-Cz CH), 7.84 (d, 3JHH=7.4 Hz,
2H, PPh o-CH), 7.81-7.78 (m, 2H, aromatic H), 7.56 (d, 3JHH
=
8.2 Hz, 2H, PPh o-CH), 7.84 (d, 3JHH=8.3 Hz, 2H, PPh o-CH),
7.39 (s, 2H, Mes* m-CH), 7.20 (s, 1H, 2-Cz CH), 7.14 (s,
obscured by solvent, 1H, 7-Cz CH), 7.09-7.06 (m, 2H, aromatic
H), 7.02-7.00 (m, 4H, aromatic H), 6.94-6.82 (m, 5H, aromatic
H), 6.73 (d, 2H, aromatic H), 6.69 (d, 2H, aromatic H), 6.53 (m,
2H, PPh m-CH), 4.88 (s, 1H, NH), 2.83 (sp, 3JHH =7.0 Hz, 1H,
Pipp0 CH(CH3)2), 2.67 (sp, 3JHH=6.8 Hz, 1H, Pipp CH(CH3)2),
2.31 (s, 3H, 3-Cz CH3), 2.17 (s, 3H, 6-Cz CH3), 1.40 (s, 9H,
3
p-tBu), 1.36 (s, 18H, o-tBu), 1.30 (d, JHH = 7.0 Hz, 3H, Pipp0
125.5 (d, JCP = 13.9 Hz, aromatic ipso-C), 125.0 (d, JCP
2.1 Hz, 4,5-Cz CH), 123.4 (d, JCP=3.8 Hz, aromatic CH), 121.4
(s, aromatic ipso-C), 120.0 (s, aromatic ipso-C), 109.2 (d, JCP
=
CH(CH3)(CH3)0), 1.27 (d, 3JHH =7.0 Hz, 3H, Pipp0 CH(CH3)-
(CH3)0), 1.18 (d, 3JHH=6.8 Hz, 3H, Pipp CH(CH3)(CH3)0), 1.16
(d, 3JHH = 6.8 Hz, 3H, Pipp CH(CH3)(CH3)0). 13C{1H} NMR
(benzene-d6): δ 182.8 (d, JCP = 21.7 Hz, C-Lu), 154.1 (s,
aromatic ipso-C), 151.6 (d, JCP = 2.6 Hz, aromatic ipso-C),
151.0 (d, JCP = 6.9 Hz, aromatic ipso-C), 150.5 (d, JCP = 3.6
Hz, aromatic ipso-C), 143.0 (d, JCP =6.2 Hz, aromatic ipso-C),
142.6 (d, JCP=1.9 Hz, aromatic ipso-C), 140.0 (s, aromatic ipso-
C), 136.5 (d, JCP=4.2 Hz, aromatic CH), 134.5 (d, JCP=9.1 Hz,
aromatic CH), 134.2 (d, JCP =10.1 Hz, aromatic CH), 133.8 (s,
aromatic ipso-C), 133.7 (s, aromatic ipso-C), 133.2 (s, aromatic
CH), 133.1 (s, aromatic CH), 133.0 (s, aromatic CH), 132.9 (s,
aromatic CH), 132.6 (d, JCP =2.9 Hz, aromatic CH), 131.7 (d,
=
115.7 Hz, aromatic ipso-C), 21.1 (s, mesityl p-CH3), 21.0 (s,
mesityl o-CH3), 20.3 (s, Cz CH3). 31P{1H} NMR (benzene-d6):
δ 30.55. Anal. Calcd for C68H64LuN5P2: C, 68.74; H, 5.43; N,
5.89. Found: C, 68.35; H, 5.41; N, 5.40.
(LPipp-K3N)Lu(NHTrip)2 (3). TripNH2 (0.214 g 0.974 mmol)
was added via syringe to a solution of 1b (0.502 g, 0.467 mmol) in
toluene at ambient temperature. The orange solution was stirred
for 30 min, after which all volatiles were removed under reduced
pressure. In a glovebox, the residue was washed with pentane
(2 ꢁ 2 mL) and then dried under vacuum. The solid was
reconstituted in toluene (2 mL), layered with pentane, and left
at -35 °C for 16 h to crystallize. The crystalline material was
collected by filtration, washed with pentane, and thoroughly
dried in vacuo. Yield: 0.295 g (43.9%). 1H NMR (benzene-d6): δ
8.13 (s, 2H, 4,5-Cz CH), 7.52 (dd, 3JHP = 11.2 Hz, 3JHH = 7.1
Hz, 8H, PPh o-CH), 7.44 (d, 3JHP = 17.7 Hz, 2H, 2,7-Cz CH),
7.11 (s, 4H, Trip m-CH), 6.97 (ov m, 3JHH = 7.2 Hz, 4H, PPh
J
CP = 2.8 Hz, aromatic CH), 131.2 (d, JCP = 8.6 Hz, aromatic
CH), 131.1 (s, aromatic CH), 130.4 (s, aromatic ipso-C), 130.1
(d, JCP = 34.9 Hz, aromatic ipso-C), 129.1 (d, JCP = 10.3 Hz,
aromatic CH), 129.0 (d, JCP=10.7 Hz, aromatic CH), 128.6 (d,
JCP =12.3 Hz, aromatic CH), 128.4 (s, aromatic CH), 127.9 (s,
aromatic ipso-C), 127.6 (s, aromatic CH), 127.5 (d, JCP=1.1 Hz,
aromatic ipso-C), 127.4 (d, JCP=1.1 Hz, aromatic ipso-C), 126.8
(d, JCP = 11.3 Hz, aromatic CH), 126.1 (d, JCP = 47.9 Hz,
aromatic ipso-C), 125.9 (d, JCP = 47.1 Hz, aromatic ipso-C),
125.6 (d, JCP = 2.5 Hz, aromatic CH), 125.4 (d, JCP = 2.6 Hz,
aromatic CH), 124.8 (s, aromatic CH), 124.6 (d, JCP =90.7 Hz,
aromatic ipso-C), 121.3 (s, Mes* m-CH), 115.9 (d, JCP=5.4 Hz,
aromatic CH), 113.2 (d, JCP=104.4 Hz, aromatic ipso-C), 111.6
(d, JCP = 112.3 Hz, aromatic ipso-C), 35.0 (s, Mes* C(CH3)3),
34.7 (s, Pipp0 CH(CH3)2), 34.5 (s, Mes* C(CH3)3), 33.7 (s, Pipp
CH(CH3)2), 32.4 (s, Mes* p-C(CH3)3), 30.2 (s, Mes* o-C(CH3)3),
p-CH), 6.89 (ov m, 12H, PPh m-CH þ Pipp o-CH), 6.56 (d, 3JHH
=
7.2 Hz, 4H, Pipp m-CH), 3.97 (s, 2H, NH), 3.26 (br sp, 4H, Trip
o-CH(CH3)2), 3.01 (sp, 3JHH =6.9 Hz, 2H, Trip p-CH(CH3)2),
2.54 (sp, 3JHH =6.9 Hz, 2H, Pipp p-CH(CH3)2), 2.20 (s, 6H, Cz
CH3), 1.43 (d, 3JHH =6.9 Hz, 12H, Trip p-CH(CH3)2), 1.11 (d,
3JHH=6.6 Hz, 24H, Trip o-CH(CH3)2), 1.02 (d, 3JHH=6.9 Hz,
12H, Pipp p-CH(CH3)2). 13C{1H} NMR (benzene-d6): δ 152.7
€
(16) Grubert, L.; Jacobi, D.; Buck, K.; Abraham, W.; Mugge, C.;
Krause, E. Eur. J. Org. Chem. 2001, 3921–3932.

66 Organometallics, Vol. 30, No. 1, 2011
Johnson and Hayes
Table 6. Summary of X-ray Crystallography Data Collection and Structure Refinement for Compounds 2, 3, and 5
2a
3b
5c
formulad
fwe
C74H70LuN5P2
1266.26
triclinic
C91H112LuN5P2
1512.77
triclinic
P1
14.357(2)
16.959(3)
17.593(3)
85.737(2)
74.590(2)
77.178(2)
4026.1(11)
2
C74H81LuN4P2
1263.34
monoclinic
C2/c
32.990(3)
23.689(2)
23.446(2)
90
104.0580(10)
90
17774(3)
8
0.944
cryst syst
space group
P1
˚
a/A
12.6755(9)
13.9958(10)
17.6268(13)
98.9330(10)
98.6410(10)
90.7270(10)
3052.0(4)
2
˚
b/A
˚
c/A
R/deg
β/deg
γ/deg
3
˚
V/A
Z
Dcalcdd/Mg m-3
1.378
1.718
1.248
1.313
μd/mm-1
F000
1.179
1.179
1300
0.34 ꢁ 0.29 ꢁ 0.09
1588
0.25 ꢁ 0.21 ꢁ 0.17
cryst size/mm
cryst color
cryst habit
0.38 ꢁ 0.13 ꢁ 0.07
yellow
plate
yellow
prism
yellow
prism
θ range/deg
N
Nind
1.63-27.10
43225
13380
1.73-25.03
47761
14118
1.79-27.10
123715
19565
completeness to θ = 27.10°/%
Tmax; Tmin
99.4
99.3
99.8
0.8677; 0.5887
13 380/0/755
1.046
0.0219
0.0523
0.7456; 0.5673
14 118/0/910
0.948
0.0731
0.1486
0.9210; 0.6628
19 565/1/745
[0.942]
[0.0449]
[0.1119]
[2.391; -0.668]
no. of data/restraints/params
GOF on F2
R1e (I > 2σ(I))
wR2f (I > 2σ(I))
ΔFmax and ΔFmin/e A
-3
˚
0.947; -0.363
4.515; -1.416
a Crystallized with one molecule of benzene in the asymmetric unit. b Crystallized with one molecule of pentane in the asymmetric unit. c Crystallized
with two highly disordered molecules of pentane in the asymmetric unit, which were removed from the reflection file using the SQUEEZE subroutine
P
P
of PLATON; statistics following treatment of data with SQUEEZE are listed in brackets. d For non-SQUEEZED data. e R1 = ||Fo| - |Fc||/ |Fo|.
P
P
f wR2 = { [w(Fo2 - Fc2)2]/ [w(Fo2)2]1/2
.
25.3 (s, Pipp0 CH(CH3)(CH3)0), 24.8 (s, Pipp0 CH(CH3)(CH3)0),
24.4 (s, Pipp CH(CH3)(CH3)0), 24.3 (s, Pipp CH(CH3)(CH3)0),
21.4 (s, Cz CH3). 21.3 (s, Cz CH3). 31P{1H} NMR (benzene-d6):
δ 30.01 (s, 1P, PippNdPPh2), 11.85 (s, 1P, Pipp0NdPPh2). Anal.
Calcd for C74H81LuN4P2: C, 70.35; H, 6.46; N, 4.43. Found: C,
70.17; H, 7.12; N, 4.43.
with the exception that no resonances were observed for the
deuterated N-aryl groups.
(LPh-K3N,K2CP-Ph)Lu(THF)-d10 (1a-ring-d10). This com-
pound was prepared in a manner identical with that previously
described for 1a,9 with the exception that HLPh-d10 was used in
place of HLPh. The 1H and 31P{1H} NMR spectra matched
those of 1a, with the exception that no resonances were observed
for the deuterated N-aryl groups.
Phenyl-d5 Azide. Aqueous 8 M HCl (30 mL) was added drop-
wise in air to a clear yellow solution of aniline-ring-d5 (2.52 g,
25.7 mmol) in THF (100 mL) at 0 °C. The pale yellow solution
was stirred for 15 min, following which a solution of NaNO2
(1.95 g, 28.3 mmol) in H2O (16.5 mL) was added dropwise over
10 min. Urea (0.253 g, 4.21 mmol) was added as a solid to
remove excess nitrous acid. A solution of NaN3 (1.85 g, 28.4
mmol) in H2O (15 mL) was added over 30 min at 0 °C, after
which the cloudy white solution was stirred at this temperature
for a further 1.75 h. The product was extracted into hexanes
(3ꢁ50 mL), and the combined organic layers were washed with
1 ꢁ 50 mL of 1 M HCl, dried over MgSO4, and concentrated in
vacuo to give a yellow liquid. The product was purified by
passage through a silica column (20 cm), with hexanes as eluent.
The solvent was removed from the eluent by rotational evapora-
tion, leaving the product as a canary yellow liquid. Yield: 2.66 g
(83.5%). 13C NMR (CDCl3): δ 140.0 (s), 129.4 (t, 1JCD = 24.5
NMR Kinetics. The rate constants k1 and k2 were determined
by monitoring the 31P{1H} NMR resonance(s) over the course
of the reaction (to at least 3 half-lives) at a given temperature. In
a typical experiment, 1b (0.0163 g, 0.0152 mmol) and 2,4,6-tri-
tert-butylaniline (0.0040 g, 0.0152 mmol) were added to a
Wilmad NMR tube that was then sealed with a rubber septum
(Sigma-Aldrich) and Parafilm. The tube was cooled to 0 °C, and
0.5 mL of toluene-d8 was injected via syringe. The tube was removed
from the cold bath and shaken briefly to mix the reagents. The
tube was then immediately inserted into the NMR probe, which
was pre-equilibrated to the appropriate temperature. The sam-
ple was allowed to equilibrate at the set temperature over the
course of shimming the tube in the magnet. 31P{1H} NMR
spectra were recorded at preset time intervals until the reaction
had progressed to at least 3 half-lives. The extent of reaction at
each time interval was determined by integration of the peak
intensity of the starting material relative to that of the inter-
mediate and final product. An appropriately long delay between
scans was utilized to ensure that integration was quantitative
and not affected by the T1 relaxation times of the reacting
species. The observed rate constant k1(obsd) was determined
according to the law of mass action. The simulated rate constants
k1(calcd) andk2(calcd) were deduced using the program COPASI.11
A summary of the observed and calculated rate constants and
half-lives are given in Tables 3 and 4 for k1 and k2, respectively.
X-ray Crystallography. Recrystallization of compound 2
from a concentrated benzene solution layered with pentane at
295 K, 3 from a concentrated pentane solution at 295 K, and 5
from a concentrated toluene solution layered with pentane at
1
1
Hz), 124.5 (t, JCD = 24.4 Hz), 118.7 (t, JCD = 24.3 Hz). IR:
ν (cm-1) 2276 (vw), 2109 (s), 2094 (s), 2034 (w), 1560 (m), 1409
(vw), 1370 (s), 1302 (vw), 1260 (s), 1098 (w), 1068 (vw), 1040
(vw), 958 (vw), 876 (vw), 841 (vw), 818 (w), 775 (vw), 753 (w),
650 (m), 625 (m), 590 (vw), 547 (s), 530 (m), 425 (s). The spectros-
copic analysis of this compound agrees with previously published
data for phenyl-d5 azide.17
HLPh-d10. This compound was prepared in a manner identical
with that previously described for HLPh,9 with the exception
that phenyl-d5 azide was used in place of phenyl azide. The 1H
and 31P{1H} NMR spectra matched that previously described,
(17) El-Shahawy, A. Spectrochim. Acta, Part A 1983, 39A, 115–117.

Article
Organometallics, Vol. 30, No. 1, 2011 67
238 K afforded single crystals suitable for X-ray diffraction.
Crystals were coated in dry Paratone oil under an argon atmo-
sphere and mounted onto a glass fiber. Data were collected at
173 K using a Bruker SMART APEX II diffractometer (Mo KR
refined freely. No decomposition was observed during data
collection. Table 6 provides a summary of selected data collec-
tion and refinement parameters. Note: In the refinement of 5,
disordered solvent molecules were removed from the reflection file
using the SQUEEZE subroutine of the PLATON program.22
Reduced residuals were observed in the final SQUEEZED struc-
ture, confirming that the uncertainty in the model was a result of
the disordered solvent.23
˚
radiation, λ = 0.710 73 A) outfitted with a CCD area detector
and a KRYO-FLEX liquid nitrogen vapor cooling device. A
data collection strategy using ω-2θ scans at 0.5° steps yielded
full hemispherical data with excellent intensity statistics. Unit
cell parameters were determined and refined on all observed
reflections using APEX2 software.18 Data reduction and correc-
tion for Lorentz-polarization were performed using SAINT-
Plus software.19 Absorption corrections were applied using
SADABS.20 The structures were solved by direct (2) or Patter-
son (3, 5) methods and refined by the least-squares method on F2
using the SHELXTL software suite.21 All non-hydrogen atoms
were refined anisotropically. Hydrogen atom positions were
calculated and isotropically refined as riding models to their parent
atoms, with the exception of the anilide protons in 2 (H1N and
H2N) and 5 (H1N), which were located on the Fourier map and
Acknowledgment. P.G.H. acknowledges financial sup-
port from the Natural Sciences and Engineering Research
Council (NSERC) of Canada for a Discovery Grant and
the Canada Foundation for Innovation for a Leaders
Opportunity Grant. We wish to thank Dr. Marc R. Roussel
for useful discussions regarding the kinetic analysis,
Mr. Tony Montina for expert technical assistance, and
Mr. Craig A. Wheaton for performing elemental analyses.
Supporting Information Available: CIF files giving X-ray
crystallographic data for 2, 3, and 5. This material is available
free of charge via the Internet at http://pubs.acs.org.
(18) APEX2, version 2.1-4; Bruker AXS, Madison, WI, 2006.
(19) SAINT-Plus, version 7.23a; Bruker AXS, Madison, WI, 2004.
(20) Sheldrick, G. M. SADABS, version 2004/1; Bruker AXS, Madison,
WI, 2004.
(22) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7–13.
(23) Sudik, A. C.; Millward, A. R.; Ockwig, N. W.; Cote, A. P.; Kim,
^ ꢀ
(21) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2007, A64, 112–122.
J.; Yaghi, O. M. J. Am. Chem. Soc. 2005, 127, 7110–7118.