Syntheses of rac/meso-{PhP(3-t-Bu-C5H3)2}-Zr{RN(CH2)3NR}, structural analyses of rac-{PhP(3-t-Bu-C5H3)2}Zr{RN(CH2)3NR} (where R is SiMe3 or Ph), and meso to rac isomerization

Article abstract of DOI:10.1016/j.jorganchem.2008.09.018

Syntheses of rac/meso-{PhP(3-t-Bu-C₅H₃)₂}Zr{Me₃SiN(CH₂)₃NSiMe₃} (rac-3/meso-3) and rac/meso-{PhP(3-t-Bu-C₅H₃)₂}Zr{PhN(CH₂)₃NPh} (rac-4/meso-4) were achieved by metallation of K₂[PhP(3-t-Bu-C₅H₃)₂] · 1.3 THF (2) with Zr{RN(CH₂)₃NR}Cl₂(THF)₂ (where R = SiMe₃ or Ph, respectively) using ethereal solvent. These isomeric pairs were characterized by ¹H, ¹³C{¹H}, and ³¹P{¹H} NMR spectroscopy; rac-3 and rac-4 were also examined via single crystal X-ray crystallography. The structures of rac-3 and rac-4 are notable in the tendency of the cyclopentadienyl rings towards η³ coordination. While isolated samples of rac-3/meso-3 and rac-4/meso-4 slowly isomerize in tetrahydrofuran-d₈ to equilibrium ratios, the isomerization rate for 3 is more than 15-fold greater than that for 4. In addition, equilibrium ratios are rapidly reached when isolated samples of rac-3/meso-3 and rac-4/meso-4 are exposed to tetrabutylammonium chloride in tetrahydrofuran-d₈ solvent. We propose that a nucleophile (either chloride or the phosphine interannular linker) brings about dissociation of one cyclopentadienyl ring, thus promoting the rac/meso isomerization mechanism.

Full text of DOI:10.1016/j.jorganchem.2008.09.018

Journal of Organometallic Chemistry 693 (2008) 3741–3750
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Syntheses of rac/meso-{PhP(3-t-Bu-C5H3)2}-Zr{RN(CH2)3NR}, structural analyses of
rac-{PhP(3-t-Bu-C5H3)2}Zr{RN(CH2)3NR} (where R is SiMe3 or Ph), and meso to
rac isomerization
Jonathan C. Axtell, Susan D. Thai, Laurel A. Morton, William S. Kassel, William G. Dougherty,
*
Deanna L. Zubris
Department of Chemistry, Mendel Science Center, Villanova University, Villanova, PA 19085, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 August 2008
Received in revised form 5 September 2008
Accepted 5 September 2008
Available online 18 September 2008
Syntheses of rac/meso-{PhP(3-t-Bu-C5H3)2}Zr{Me3SiN(CH2)3NSiMe3} (rac-3/meso-3) and rac/meso-
{PhP(3-t-Bu-C5H3)2}Zr{PhN(CH2)3NPh} (rac-4/meso-4) were achieved by metallation of K2[PhP(3-t-Bu-
C5H3)2] ꢀ 1.3 THF (2) with Zr{RN(CH2)3NR}Cl2(THF)2 (where R = SiMe3 or Ph, respectively) using ethereal
solvent. These isomeric pairs were characterized by 1H, 13C{1H}, and 31P{1H} NMR spectroscopy; rac-3
and rac-4 were also examined via single crystal X-ray crystallography. The structures of rac-3 and
rac-4 are notable in the tendency of the cyclopentadienyl rings towards
g
3 coordination. While isolated
Keywords:
samples of rac-3/meso-3 and rac-4/meso-4 slowly isomerize in tetrahydrofuran-d8 to equilibrium ratios,
the isomerization rate for 3 is more than 15-fold greater than that for 4. In addition, equilibrium ratios are
rapidly reached when isolated samples of rac-3/meso-3 and rac-4/meso-4 are exposed to tetrabutylam-
monium chloride in tetrahydrofuran-d8 solvent. We propose that a nucleophile (either chloride or the
phosphine interannular linker) brings about dissociation of one cyclopentadienyl ring, thus promoting
the rac/meso isomerization mechanism.
Zirconocene
Cyclopentadienyl
Phosphine
Synthesis
X-ray structure
Ó 2008 Elsevier B.V. All rights reserved.
1. Introduction
presented (Fig. 1). The chelating bis(amide) ligand accounts for
the differing diastereomeric outcomes of these metallations. While
A wide range of group 4 metallocenes and metallocene ana-
logs have been synthesized and tested as precatalysts for poly-
the {Me3SiN- (CH2)3NSiMe3} ligand adopts a twist conformation,
best suited to the rac-isomer, the {PhN(CH2)3NPh} ligand can, in
some cases, adopt an envelope confirmation to accommodate the
steric requirements of the meso isomer; these diastereomeric pref-
erences were the subject of a recent DFT study [6e]. Solvent choice
is critical for isolation of a single diastereomer in these chelate-
controlled metallations. For example, use of the mismatched met-
allation agent/solvent pair, Zr{PhN(CH2)3NPh}Cl2(THF)2/tetrahy-
drofuran, provides a rac/meso mixture of products [6f]. Jordan
et al. propose that the solvent effect stems from differing solubility
of the lithium chloride reaction byproduct; chloride ion catalyzes
isomerization of the meso isomer to the more thermodynamically
stable rac isomer.
We sought to employ Jordan’s methodology for the preparation
of rac- and meso-ansa-zirconocenes with a phenyl-phosphine
interannular bridge. Ligands (such as ([Me2P(fluorenyl)2]ꢁ) [7]
and [R2P(2-Me-4-t-Bu-C5H2)2]ꢁ where R = Me or n-Bu [8]) and
group 4 metallocenes with a phosphorus-containing ansa-bridge
are the subject of two recent review articles [9,10]. A few of these
group 4 ansa-metallocenes have been tested as ethylene [11–13]
merization of ethylene, propylene, and other
a-olefins [1].
From this wealth of polymerization data, defined relationships
between precatalyst symmetry and polymer tacticity have
emerged [2]. Symmetry may be enforced by use of an interan-
nular bridge, or linker, between the two cyclopentadienyl rings
of these metallocenes. Ansa-metallocenes [3] differ electronically
from their unlinked counterparts; these differences were re-
cently scrutinized in the literature [4]. Choice of interannular
linker and its effect on reactivity was the subject of a recent re-
view [5].
Over the past eight years, Jordan and coworkers [6] have
described the chelate-controlled synthesis of a wide range of
ansa-zirconocenes (with substituted bis(cyclopentadienyl) or
bis(indenyl) ligands incorporating a Me2Si, or in one case, a
CH2CH2, interannular bridge). The importance of metallation agent
and solvent selection was highlighted in Ref. [6c], where the
chelate-controlled synthesis of meso- and rac-Me2Si(3-t-Bu-
C5H3)2Zr{RN(CH2)3NR} (where R is Ph or SiMe3, respectively) was
and a-olefin polymerization precatalysts [14] (Fig. 2). Other group
4 metallocene dichlorides with a phosphorus-containing ansa-
bridge have been reported with no mention of their use as
* Corresponding author. Tel.: +1 610 519 4874; fax: +1 610 519 7167.
E-mail address: deanna.zubris@villanova.edu (D.L. Zubris).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.09.018

3742
J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
Li+
Cl
Zr
Cl
R
SiMe2
N
SiMe2
N
Zr
Zr
+
solvent
N
N
N
N
THF
THF
R
R
R
R
+
SiMe2
R
Li+
meso isomer isolated
where solvent is Et2O;
R is Ph
rac isomer isolated
where solvent is THF;
R is SiMe3
Fig. 1. Choice of metallation agent, Zr{RN(CH2)3NR}Cl2(THF)2 (where R is Ph or SiMe3), and reaction solvent (Et2O or THF) allows for isolation of either meso- or rac-Me2Si(3-t-
Bu-C5H3)2Zr{RN(CH2)3NR} (where R is Ph or SiMe3, respectively) [6c].
Ph
Ph
Ph
P
Ph
P
Cl
Cl
Cl
Cl
Cl
Cl
Me
Me
Cl
Ph
P
M
t-Bu
M
M
P
Zr
Ph
P
R
P
R
Zr
Cl
P
Cl
Zr
Cl
Me
Cl
P
Cl
Me
Ph
M = Zr or Hf
Ph
Ph
meso
rac
where M = Ti, Zr, or Hf
R = Ph or i-Pr;
isolated as mixture of rac- and meso-isomers
R4
R1
R1
R2
PR2
R7
R7
PR2
R2
R7
R7
R1
R1
Ph
Cl
Ph
Cl
Ph
P
P
P
Zr
Zr
Zr
R4
Fe
Cl
Cl
Cl
Cl
Fe
R2
PR2
R2
PR2
Fig. 2. Group
precatalysts incorporating
(C5Me4)2]MCl2 (where M = Zr or Hf) [12], rac/meso-[PhP(2-Me-4-Ph-indenyl)2]ZrCl2,
rac/meso-[(i-PrP)(2-Me-4-Ph-indenyl)2]ZrCl2 [14], [t-BuP(2-indenyl)2]ZrCl2 [13],
rac-PhP(indenyl)2ZrCl2, PhP[(fluorenyl)(Cp)]ZrCl2 [14], and PhP(fluorenyl)2ZrCl2
[11,14].
4
ansa-metallocene ethylene and/or propylene polymerization
phosphine-containing interannular bridge: [PhP-
R4
meso
R4
rac
a
where R1 = H, Me, PPh2, or SiMe3;
R2, R4, and R7 = H or Me; R = Ph or i-Pr
Fig. 3. Meso/rac isomer pairs of bis(phosphoyl)zirconocenes described by Hollis
[22] and Fu [23], titanocenes described by Hollis [22], hafnocenes described by Fu
[23], and bis(phosphino-g
5-indenyl)iron(II) complexes described by Curnow [24].
polyolefin precatalysts, including [PhP(fluorenyl)2]HfCl2 [11],
[Ph(E)P(C5Me4)2]MX2 (where M = Zr or Hf; E = O, S, or Se) [12],
[PhP(2-indenyl)2]ZrCl2 [13], [PhP(C5H4)2]TiCl2 [15], [RP(C5H4)2]-
ZrCl2 (where R = Me, Et, i-Pr, or t-Bu) [16], [PhP(C5H4)2]ZrCl2 [17],
{[Me2P(C5Me4)2]MCl2}+ (where M = Zr or Hf) [18], and rac/meso-
{[R2P(2-Me-4-t-Bu-C5H2)2]ZrCl2}+ (where R = Me or n-Bu) [19].
Also, phosphorus has been used as a component of multi-atom
interannular bridges for ansa-zirconocenes, such as [{Me2P-
(C5H4)}{Cl2B(C5H4))]ZrCl2, [{Et2P(2-Me-indenyl)}{Cl2B(2-methyl-
(Fig. 3): tetrahydrofuran-assisted indenyl ring slip, phosphorus
coordination, indenide dissociation, indenide recoordination, phos-
phorus decoordination, and ring slip. Herein, we will describe the
synthesis and characterization of rac/meso-PhP(3-t-Bu-C5H3)2-
Zr{RN(CH2)3NR} (where R = SiMe3 or Ph) and will present our
observations of meso to rac isomerization in these systems. Pure
rac-PhP(3-t-Bu-C5H3)2Zr{RN(CH2)3NR} (where R = SiMe3 or Ph)
can be isolated (either via controlled isomerization or preferential
crystallization); these complexes will be employed in future work
indenyl))]ZrCl2 [20], and [PhP(CH2CH2-g
5-C5H4)2]ZrCl2 [21].
We were drawn to the phosphine interannular linker due to its
pyramidal nature and ability to enforce two different steric envi-
ronments at the rear of the metallocene wedge. This feeds into
one long-term goal of our research program; namely, the design
and testing of complexes to probe the steric requirements for reg-
directed towards a-olefin polymerization.
2. Results and discussion
iocontrol in
a-olefin polymerization. It follows that the ability to
isolate a single diastereomeric zirconocene (rac or meso) is critical
for meaningful polymerization trials in this vein.
Synthesis of our target ligand, K2[PhP(3-t-Bu-C5H3)2] ꢀ 1.3 THF
(2), was achieved via standard methods as described in Scheme
1. Treatment of a solution of Li(t-Bu-C5H4) (tetrahydrofuran/pen-
tane, 3:1, v/v) with a tetrahydrofuran solution of dichlorophenyl-
phosphine provided neutral compound, PhP(3-t-Bu-C5H4)2 (1) as
a mixture of double bond isomers [25]. Characterization via multi-
nuclear NMR spectroscopy (1H, 13C, and 31P) verified the presence
of these isomers; due to overlapping NMR signals throughout all
three spectra, individual isomer identities were not assigned. Com-
pound 1 can be doubly deprotonated by addition of 2 equiv. of
either potassium bis(trimethylsilylamide) or potassium tert-butox-
The inclusion of phosphorus in our ligand array introduces the
potential for phosphorus-mediated meso- to rac-isomerization in
addition to the possibility of chloride catalyzed rac/meso isomeri-
zation as described by Jordan (vide supra) [6f]. Hollis [22] and Fu
[23] have implicated the role of phosphorus in a ring slip-inver-
sion-ring slip mechanism of isomerization for bis(phosphoyl)tita-
nium, zirconium and hafnium(IV) complexes (Fig. 3). Curnow
[24] proposes the following stepwise mechanism for rac/meso
isomerization of bis(phosphino-g
5-indenyl)iron(II) complexes

J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
3743
tion were obtained from such a solution cooled to –35 °C for 8 days
(Fig. 4).
K+
As expected, the {Me3SiN(CH2)3NSiMe3} chelate adopts a twist
conformation, as observed previously for rac-Me2Si(3-t-Bu-
C5H3)2Zr{Me3SiN(CH2)3NSiMe3} [6c]. The following deviations in
Å of C and Si atoms from the N(1)–Zr–N(2) plane are consistent
with the twist conformation: Si(1) ꢁ0.70, C(28) +0.73, C(29)
ꢁ0.10, C(30) ꢁ0.86, Si(2) +0.60. Another similarity with rac-
Me2Si(3-t-Bu-C5H3)2Zr{Me3SiN(CH2)3NSiMe3} is the slight dis-
placement of both t-butyl substituents out of the cyclopentadienyl
ring planes, away from the zirconium. The most notable aspect of
Li+
(ii)
(i)
P
Ph
P
Ph
2
K+
· 1.3 equiv THF
1
2
rac-3 is the trend towards
g
3-coordination of the cyclopentadienyl
rings, evidenced by the 2.481–2.836 Å range of Zr–C(cyclopentadi-
enyl) bond lengths (as compared to the 2.495–2.796 Å range of
Zr–C bond lengths for dimethylsilyl-bridged rac-Me2Si(3-t-Bu-
C5H3)2Zr{Me3SiN(CH2)3NSiMe3} [6c] and the 2.435–2.608 Å range
of Zr–C bond lengths for phenylphosphine-bridged [PhP-
(C5Me4)2]ZrCl2 [12]). In fact, various angles (Fig. 5) make the
coordination mode of the cyclopentadienyl rings in rac-3 more
consonant with doubly-linked, as opposed to singly-linked, zirco-
nocenes [26]. For example, the following centroid–Zr–centroid an-
Scheme 1. Reagents and conditions: (i) chlorodiphenylphosphine, THF/pentane,
ꢁ78 °C ? 22 °C, 18 h; (ii) 2 equiv. potassium tert-butoxide, THF, 0 °C ? 22 °C, 72 h.
ide to yield K2[PhP(3-t-Bu-C5H3)2] ꢀ 1.3 THF (2) as a white powder.
We found preparation and isolation of a dipotassio salt preferable
to a dilithio salt due to persistent impurities observed upon isola-
tion of the dilithio salt; the analogous preparation of a disodium
salt was not attempted. 1H NMR analysis of 2 reveals the presence
of nonstoichiometric protio THF (1.3 equiv.); poor solubility of 2 in
THF has precluded preparative scale recrystallizations for elemen-
tal analysis purposes.
gles (c) are reported for select singly-linked zirconocenes: for rac-
Me2Si(3-t-Bu-C5H3)2Zr{Me3SiN(CH2)3NSiMe3} [6c],
c
= 122.5(9)°;
Cl2(THF)2Zr{Me3SiN(CH2)3NSiMe3}
Ph
Ph
P
Zr
P
2
Zr
+
N
N
N
N
THF
-78 °C to 22 °C
28 h
ð1Þ
SiMe3
SiMe3
Me3Si
Me3Si
rac-3
meso-3
Preparative scale metallation of K2[PhP(3-t-Bu-C5H3)2] ꢀ 1.3 THF
for [PhP(C5Me4)2]ZrCl2 [12],
c
= 125.9°; and for rac-3, c = 120.2°.
(2) with Zr{Me3SiN(CH2)3NSiMe3}Cl2(THF)2 in tetrahydrofuran sol-
vent over the course of 28 h provides both rac- and meso-PhP(3-t-
Bu-C5H3)2Zr{Me3SiN(CH2)3NSiMe3} (rac-3/meso-3) as a pale yel-
low microcrystalline solid in a 2:1 isomeric ratio (Eq. (1)). The
pyramidal geometry at phosphorus is evidenced by the number
of inequivalent 1H NMR resonances in this rac/meso mixture. For
example, rac-3 displays six cyclopentadienyl signals, two trimeth-
ylsilyl methyl signals, and two tert-butyl methyl signals. The up-
field 31P NMR chemical shifts for meso-3 and rac-3 (ꢁ43.6
(meso) and ꢁ37.1 (rac) in benzene-d6 solvent; ꢁ40.3 (meso) and
ꢁ33.8 (rac) in THF-d8 solvent) are consistent with those reported
for related phosphine-bridged zirconocenes [12,14,17]. COSY and
HETCOR methods were used to assist with NMR peak assignments.
While the pyramidal geometry at phosphorus creates the possibil-
ity of two distinct meso isomers (one with the phosphorus-phenyl
substituent on the same side of the zirconocene as the t-butyl sub-
stituents, and the other with the phosphorus-phenyl on the oppo-
site side of the zirconocene as the t-butyl substituents), only one of
these meso isomers is detected under our experimental conditions.
Based on steric arguments, one would expect the latter to be the
preferred meso isomer (as depicted in Eq. (1)). The rac-isomer pref-
erentially crystallizes from a concentrated pentane solution of rac-
3/meso-3; indeed, single crystals of rac-3 suitable for X-ray diffrac-
Further, the centroid–Zr–centroid angle, interplanar ring angle
= 73.6°), Cp(normal)–Cp(normal) dihedral angle (b = 106.4°),
and tilt angle ( = 6.9°) for rac-3 are very similar to that of dou-
bly-linked [(Me2Si)2(3,5-i-Pr2Cp)(4-i-PrCp)]ZrCl2 (where
120.3°, = 73.9°, b = 106.1°, and = 7.1°). [27] Since the success
of doubly-linked [(Me2Si)2(3,5-i-Pr2Cp)(4-i-PrCp)]ZrCl2 as an
(a
s
c
=
a
s
a-
olefin polymerization precatalyst has been, at least in part, attrib-
uted to these characteristic angles, we are encouraged that rac-3
may display similar angles upon conversion to a zirconocene
dichloride (presumed) precatalyst for a-olefin polymerization.
We have followed the course of the reaction described in Eq. (1)
by 1H and 31P NMR spectroscopy. At 22 °C, the metallation reaction
is rapid; upon thawing
a
frozen sample of
2
and
Zr{Me3SiN(CH2)3NSiMe3}Cl2(THF)2 in tetrahydrofuran-d8 solvent,
2 is immediately consumed to provide a mixture of rac-3 and
meso-3 (in a 3:2 ratio, with slightly more meso isomer than our
preparative scale metallation). Over the course of days, the propor-
tion of rac-3 increases at the expense of meso-3. In separate reac-
tion trials, we found that the presence of excess
Zr{Me3SiN(CH2)3NSiMe3}Cl2(THF)2 or substoichiometric 1 (pre-
sumably formed via partial hydrolysis of 2 during sample prepara-
tion) facilitates isomerization, and over the course of 1 day in each
case provides a sample that is 100% rac-3 (within the detection

3744
J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
Fig. 4. X-ray crystal structure of rac-3. Selected bond distances (Å) and angles (deg): Zr(1)–N(1), 2.096(2); Zr(1)–N(2), 2.107(2); N(1)–Si(1), 1.733(2); N(2)–Si(2), 1.735(2); Zr(1)–
C(7), 2.503(3); Zr(1)–C(8), 2.665(3); Zr(1)–C(9), 2.836(3); Zr(1)–C(10), 2.683(3); Zr(1)–C(11), 2.481(3); Zr(1)–C(16), 2.519(3); Zr(1)–C(17), 2.586(3); Zr(1)–C(18), 2.754(3);
Zr(1)–C(19), 2.655(3); Zr(1)–C(20), 2.523(3); Zr(1)–{C(7)–C(11) centroid}, 2.346; Zr(1)–{C(16)–C(20) centroid}, 2.314; N(1)–Zr(1)–N(2), 93.95(9); centroid–Zr–centroid, 120.2.
limit for NMR integration). The data from a representative reaction
trial (with excess Zr{Me3SiN(CH2)3NSiMe3}Cl2(THF)2 present) is gi-
α = dihedral angle between ring planes
β = Cpnormal - Cpnormal
γ = Cpcentroid - Zr - Cpcentroid
τ = tilt angle
τ
ven in Fig. 6. To further probe the role of phosphine in the isomer-
ization mechanism, triphenylphosphine (1 equiv.) was added to
one of our metallation trials. However, triphenylphosphine thwarts
the desired metallation reaction; rac-3 and meso-3 were not
observed.
γ
β
α
Zr
Fig. 5. Definition of a, b, c, and s for ansa-zirconocenes [4].
The metallation described in Eq. (1) was also tracked at low
temperature (–30 °C) in tetrahydrofuran-d8 solvent. Immediately,
upon thawing a frozen sample of 2 and Zr{Me3SiN(CH2)3NSi-
Me3}Cl2(THF)2 in tetrahydrofuran-d8 solvent, new signals appear
in both the 1H and 31P NMR spectra corresponding to unidentified
intermediate/s (the number and chemical shift range of the 31P sig-
nals suggest the intermediate/s have structural similarity to 1), and
after 45 min at ꢁ30 °C, broad signals appear for rac-3 and meso-3
(in an approximate 3:2 ratio). Upon warming to 0 °C, these signals
continue to slowly increase in intensity; further warming to 25 °C
leads to complete consumption of starting materials while main-
taining the initially observed rac/meso isomer ratio. Over the
course of days at 22 °C, the amount of rac isomer increases at the
expense of meso isomer.
We
decided
to
test
whether
or
not
the
Zr{PhN(CH2)3NPh}Cl2(THF)2/diethyl ether combination would al-
low for the preparative scale isolation of pure meso-4 (Eq. (2)).
Metallation of K2[PhP(3-t-Bu-C5H3)2] ꢀ 1.3 THF (2) with
Zr{PhN(CH2)3NPh}Cl2(THF)2 in diethyl ether at 0 °C over the course
of
2
hours provides both rac- and meso-PhP(3-t-Bu-
C5H3)2Zr{PhN(CH2)3NPh} (rac-4/meso-4) as an orange-red micro-
crystalline solid in a 0.36:1.0 isomeric ratio (Eq. (2)). Again, the
pyramidal geometry at phosphorus is evidenced by the number
of inequivalent 1H NMR resonances in this rac/meso mixture; COSY
and HETCOR methods were used to assist with NMR peak assign-
Cl2(THF)2Zr{PhN(CH2)3NPh}
Ph
Ph
P
P
Zr
2
Zr
+
N
N
N
N
ð2Þ
Ph
Ph
Et2O
-78 °C to 0 °C
2h
Ph
Ph
rac-4
meso-4

J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
3745
100
80
60
40
20
0
allowed us to obtain samples sufficiently enriched in this species to
carry out complete NMR characterization.
Similar to the rac-3/meso-3 mixture, the rac isomer preferen-
tially crystallizes from solutions of rac-4/meso-4 in various sol-
vents. A concentrated benzene-d6 solution was prepared using a
recrystallized sample of rac-4/meso-4 (1.0:0.1 isomeric ratio).
Upon standing for 5 weeks at 22 °C, single crystals suitable for X-
ray diffraction were obtained (Fig. 7).
One unusual feature of this structure is that the
{PhN(CH2)3NPh} chelate adopts a conformation intermediate be-
tween the idealized twist and envelope forms. Ref. [6e] reports
the following trends in displacement of atoms from the N(1)–Zr–
N(2) plane for idealized twist and envelope conformers of the
{PhN(CH2)3NPh} chelate: twist (ꢁ, +, 0, ꢁ, +) and envelope (ꢁ, +,
0, 0, 0), (where + is above; 0 is in; ꢁ is below the N(1)–Zr–N(2)
plane). The displacement of atoms in our system (ꢁ, ꢁ, +, ꢁ, +) is
intermediate between these two idealized conformers, with the
following deviations in Å of C atoms from the N(1)–Zr–N(2) plane:
C(34) –0.43, C(27) ꢁ0.41, C(26) +0.51, C(25) ꢁ0.94, C(28) +0.78.
Similar to rac-3, the cyclopentadienyl rings of rac-4 tend towards
0
5
10
15
20
25
30
time (h)
Fig. 6. Formation of rac-3/meso-3 mixture by combination of
2
and excess
Zr{Me3SiN(CH2)3NSiMe3}Cl2(THF)2 (time = 0) and subsequent conversion of the
meso-3/rac-3 mixture to ꢂ100% rac-3 (THF-d8, 22 °C, in the dark). rac-3 (squares,
upper curve); meso-3 (diamonds, lower curve).
g
3-coordination, evidenced by the 2.4958–2.7317 Å range of Zr–
C(cyclopentadienyl) bond lengths. The centroid–Zr–centroid angle
= 120.6°), interplanar ring angle ( = 70.5°), Cp(normal)–Cp(nor-
mal) dihedral angle (b = 109.5°), and tilt angle ( = 5.6°) for rac-4
ments. The deshielded 31P NMR chemical shifts for meso-4 and
rac-4 (ꢁ38.7 (meso) and ꢁ35.8 (rac) in benzene-d6 solvent) are
similar to those reported for meso-3/rac-3. Another species was
also detected with a 31P NMR chemical shift of ꢁ24.6 ppm, a single
t-butyl methyl 1H NMR signal, and other 1H NMR signals that over-
lap with those of meso-4/rac-4. The chemical shift difference of
this species relative to rac-4/meso-4 is not pronounced enough
to suggest oxidation of the phosphine linker (for example,
[Ph(O)P(C5Me4)2]ZrCl2 has a 31P NMR signal at +17.9 ppm in ben-
zene-d6 solvent [12]). It is conceivable that this species is the other
potential meso isomer of 4, with the phosphorus-phenyl substitu-
ent on the same side of the zirconocene as the t-butyl substituents
(not pictured in Eq. (2)), though our experiments to date have not
(c
a
s
have less tendency towards the characteristic angles of a doubly-
linked zirconocene as compared to those for rac-3. However, the
N(1)–Zr–N(2) angle of 89.45(6)° is small in comparison to that of
rac-3 (93.95(9)°), meso-Me2Si(3-t-Bu-C5H3)2Zr{PhN(CH2)3NPh}
(95.34(10)°) [6c] and rac-Me2Si(3-t-Bu-C5H3)2Zr{Me3SiN(CH2)3
NSiMe3} (92.63(8)°) [6c].
We followed the course of this metallation reaction at low tem-
perature (0 °C) using 31P NMR spectroscopy (with an unlocked
NMR sample). Upon thawing
a frozen sample of 2 and
Zr{PhN(CH2)3NPh}Cl2(THF)2 in diethyl ether solvent, peaks are
immediately observed corresponding to the meso and rac isomers
as well as another species (vide supra). Though slightly broadened,
these peaks appear in a similar relative ratio as observed in the
preparative scale metallation depicted in Eq. (2). This experiment,
along with the preferential crystallization of rac-4 from solutions
of rac-4/meso-4, suggest that the Zr{PhN(CH2)3NPh}Cl2(THF)2/
diethyl ether combination is not an appropriate choice for prepara-
tive scale isolation of pure meso-4.
Our observation that NMR samples of isolated rac-3/meso-3 in
tetrahydrofuran-d8 (initial ratio of 1.0:0.46) slowly change over the
course of weeks at 22 °C to afford a sample that is solely rac-3 led
us to undertake careful NMR isomerization studies for both rac-3/
meso-3 and rac-4/meso-4 at elevated temperature (60 °C) in tetra-
hydrofuran-d8 solvent. All of these experiments were carried out in
the dark to rule out the possibility of photochemically promoted
isomerization. In each case, the rac/meso ratio was monitored using
31P NMR spectroscopy, with frequent data sampling for the first
12 h, and daily monitoring until an equilibrium state was reached.
As described in Fig. 8, the rac-3/meso-3 ratio was monitored
with no additive, with 2 equiv. KCl, and with 2 equiv. nBu4NCl
additive present. While the presence of KCl causes an initial drop
in the percentage of meso-3, over time the rate of decay mimics
that observed in the absence of KCl. It should be noted that solid
KCl is apparent at the bottom of the NMR tube both before and
after NMR monitoring; we assume that the KCl is at best sparingly
soluble under reaction conditions. Both samples are completely
converted to 100% rac-3 over the course of 2 d at 60 °C. In contrast,
addition of nBu4NCl immediately converts the sample to 100% rac-
3. Also, a color change is immediately observed in the presence of
nBu4NCl (yellow ? orange solution). We invoke solubility and ion
pairing effects to account for the differing kinetics of rac to meso
isomerization in the presence of KCl and nBu4NCl.
Fig. 7. X-ray crystal structure of rac-4. Selected bond distances (Å) and angles (deg):
Zr(1)–N(1), 2.0956(15); Zr(1)–N(2), 2.1413(15); N(1)–C(28), 1.406(2); N(2)–C(34),
1.404(2); Zr(1)–C(1), 2.4976(17); Zr(1)–C(2), 2.5967(17); Zr(1)–C(3), 2.7211(17);
Zr(1)–C(4), 2.6290(18); Zr(1)–C(5), 2.4958(17); Zr(1)–C(10), 2.5063(17); Zr(1)–
C(11), 2.5620(17); Zr(1)–C(12), 2.7317(17); Zr(1)–C(13), 2.6337(18); Zr(1)–C(14),
2.5131(17); Zr(1)–{C(1)–C(5) centroid}, 2.291; Zr(1)–{C(10)–C(14) centroid}, 2.293;
N(1)–Zr(1)–N(2), 89.45(6); centroid–Zr–centroid, 120.6.

3746
J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
100
80
60
40
20
0
meso to rac isomerization of bis(1-diphenylphosphino-
nyl)iron(II) in THF-d8 at 23 °C [24b].
g
5-inde-
A priori, we expected chloride catalyzed rac to meso isomeriza-
tion to occur for our rac-3/meso-3 and rac-4/meso-4 mixtures, as
described by Jordan for related dimethylsilyl linked zirconocenes
[6f]. Therefore, in the presence of chloride additive, we postulate
that one operating mechanism for equilibration of our rac/meso
mixtures involves displacement of one cyclopentadienyl ligand
by chloride, P–C bond rotation, chloride loss, and cyclopentadienyl
ligand recoordination, analogous to Jordan’s proposed mechanism
(Scheme 2, where Nuꢁ is chloride). However, the fact that we ob-
serve rac to meso isomerization in the absence of an additive
raises the possibility of a second potential mechanism. We pro-
pose another nucleophile catalyzed mechanism, with the phos-
phine interannular linker acting as a nucleophile instead of
chloride. Specifically, this second mechanism involves intramolec-
ular displacement of one cyclopentadienyl ligand by the phos-
phine interannular linker (along with coordination of
tetrahydrofuran solvent), P–C bond rotation, phosphine and tetra-
hydrofuran loss, followed by cyclopentadienyl ligand recoordina-
tion (Scheme 2). We suggest that the second cyclopentadienyl
0
1
2
3
4
5
6
time (h)
Fig. 8. Conversion of meso-3 to rac-3 starting from
a 0.46:1.0 meso-3/rac-3
mixture (THF-d8, 60 °C, in the dark); percentage of rac-3 (upper curves) and meso-3
(lower curves) plotted vs. time. Run i (diamonds), no additive; run ii (squares),
2 equiv. KCl (sparingly soluble); run iii (triangles), 2 equiv. nBu4NCl.
Similar experiments using rac-4/meso-4 are depicted in Fig. 9.
In this case, the meso isomer is the major isomer at the start of
the experiments (1.0:0.24 meso-4/rac-4). Here we observed simi-
lar behavior in the both the presence and absence of KCl, with neg-
ligible isomerization over the course of 12 h at 60 °C. In fact, a full
week at 60 °C is required to obtain a 1.0:1.0 ratio of rac-4/meso-4.
In contrast, addition of nBu4NCl immediately converts the sample
to a 1.0:0.16 ratio of rac-4/meso-4, and this ratio is maintained
over the course of days at 60 °C. This appears to be the equilibrium
ratio. A subtle color change is immediately observed in the pres-
ence of nBu4NCl (orange-red ? red solution).
ligand may undergo a ring slip (g5
? g
3) to put the phosphine
interannular linker in a more appropriate location (relative to zir-
conium) for intramolecular coordination. However, we realize it is
difficult to rule out isomerization catalyzed by
a negligible
amount of adventitious chloride or water (a potential source of
hydroxide) in our solutions of rac-3/meso-3 and rac-4/meso-4.
Regardless of the operating mechanism(s) of isomerization, the
ability to isolate pure rac-3 and pure rac-4 via controlled isomer-
ization of rac/meso mixtures (with no additive required) is valu-
able for our proposed future work in the realm of
polymerization (vide supra).
a-olefin
The observed rate constant for isomerization, kobs, was deter-
mined for Fig. 8, run i, and for Fig. 9, run i. Each experiment
was treated as an approach to equilibrium system. A plot of
[ꢁln{(Xt=t ꢁ Xt=1)/(Xt=0 ꢁ Xt=1)}] versus time (in s), where X is
the mole fraction of meso isomer, affords kobs from the slope.
The kobs value for meso-4 to rac-4 isomerization
(kobs = 1.2 ꢃ 10ꢁ6 sꢁ1) is more than 15 fold smaller than that for
meso-3 to rac-3 isomerization (kobs = 2.0 ꢃ 10ꢁ5 sꢁ1). Both of
these kobs values are small in comparison to those reported by
Hollis [22], Fu [23], and Curnow [24] for other metallocenes that
incorporate phosphorus in their ligand arrays, albeit not as a
component of an interannular linker. Indeed, the kobs value for
meso-3 to rac-3 isomerization in THF-d8 at 60 °C has the same or-
der of magnitude as Curnow’s kobs value (1.59(3) ꢃ 10ꢁ5 sꢁ1) for
3. Summary
The syntheses of rac/meso-{PhP(3-t-Bu-C5H3)2}Zr{SiMe3N-
(CH2)3NSiMe3} (rac-3/meso-3, 2:1 ratio) and rac/meso-{PhP(3-t-
Bu-C5H3)2}Zr{PhN(CH2)3NPh} (rac-4/meso-4, 0.36:1.0 ratio) under-
score the influence of the interannular linker in the course of met-
allation when Jordan’s Zr{RN(CH2)3NR}Cl2(THF)2 (where R = SiMe3
or Ph, respectively) is used as a metallation agent. Both rac-3 and
rac-4 display tendencies towards
g
3-coordination of their cyclo-
pentadienyl rings, as evidenced by X-ray crystallography. Since
mixtures of rac-3/meso-3 and rac-4/meso-4 equilibrate both in
the presence and absence of chloride additive, we suggest a general
isomerization mechanism involving nucleophile catalyzed dis-
placement of one cyclopentadienyl ligand (where Nuꢁ is chloride
or the phosphine interannular linker). The ability to carry out con-
trolled isomerization of rac/meso mixtures of 3 and 4 in the ab-
sence of additive is synthetically useful, irrespective of
mechanism. Conversions of pure rac-3 and rac-4 to the corre-
100
80
60
40
20
0
sponding zirconocene dichloride precatalyst and subsequent
olefin polymerization studies are underway and will be reported
in due course.
a-
4. Experimental
4.1. General considerations
0
1
2
3
4
5
6
All chemical reactions were carried out under an atmosphere of
argon using standard Schlenk techniques [28] unless otherwise
noted. Argon gas was purified by passage over DrieriteTM. All chem-
icals were purchased from Aldrich and used as received unless
otherwise noted. Dichlorophenylphosphine was purchased from
Strem. ZrCl2(THF)2{PhN(CH2)3NPh} [6b,6c,29], ZrCl2(THF)2{Me3
time (h)
Fig. 9. Conversion of meso-4 to rac-4 starting from
a 1.0:0.24 meso-4/rac-4
mixture (THF-d8, 60 °C, in the dark); percentage of meso-3 plotted vs. time. Run i
(diamonds), no additive; run ii (squares), 2 equiv. KCl (sparingly soluble); run iii
(triangles), 2 equiv. nBu4NCl.

J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
3747
Ph
Ph
P
P
Zr
Zr
N
N
N
N
R
R
R
R
meso
rac
Nu-
- Nu-
Nu-
- Nu-
Ph
Nu
Zr
O
P
Ph
P
or
Zr
N
N
N
N
R
R
R
R
where R is SiMe3 or Ph;
where Nu is Cl-, OH-, or PR3
Scheme 2.
SiN(CH2)3NSiMe3} [30], and Li(t-Bu-C5H4) [27] were prepared
according to literature methods. Benzene-d6 and tetrahydrofuran-
d8 were purchased from Cambridge Isotope Laboratories and dried
over molecular sieves (8–12 mesh, 4 Å, activated) before use. An
MBraun Manual Solvent Purification System (MB-SPS) was used
to obtain the following anhydrous solvents: toluene, tetrahydrofu-
ran, diethyl ether, pentane, and dichloromethane; solvents were
submitted to three freeze-pump-thaw cycles before use [31]. Ben-
zene was dried over sodium/benzophenone ketyl and stored under
vacuum. Molecular sieves (8–12 Mesh, 4 Å) were purchased from
J.T. Baker and activated before use. Potassium tert-butoxide was
purchased from Aldrich and sublimed before use.
4.2. Preparation of PhP(3-t-Bu-C5H4)2 (1)
A 100-mL Schlenk flask was charged with Li(t-Bu-C5H4) (3.01 g,
23.5 mmol, 2.0 equiv.), tetrahydrofuran (20 mL), and pentane
(7 mL). A 50-mL Schlenk flask was charged with PhPCl2 (1.60 mL,
11.8 mmol, 1.0 equiv.) and tetrahydrofuran (7 mL). The Li(t-Bu-
C5H4) solution was cooled to ꢁ78 °C, followed by dropwise addi-
tion of the PhPCl2 solution via cannula transfer. The ꢁ78 °C dry
ice/acetone bath was left intact to allow the reaction mixture to
slowly warm to 22 °C over the course of 18 h. Solvent was removed
in vacuo and pentane (30 mL) was added inside the glove box to
provide an orange-brown slurry. Vacuum filtration and pentane
washes (5 ꢃ 8 mL) provided a brown filtrate, which was concen-
trated in vacuo to yield 1 as a light brown oil (3.73 g, 90.2% yield).
1H NMR (benzene-d6, 22 °C): d 7.45–7.68 (m, C6H5), 6.98–7.14 (m,
C6H5), 6.70 (m, C5H4), 6.31–6.39 (m, C5H4), 6.05 (m, C5H4), 5.99 (m,
C5H4), 4.02 (m, allylic C5H4), 3.11 (m, allylic C5H4), 3.07 (br s, allylic
C5H4), 3.06 (m, allylic C5H4), 3.01 (m, allylic C5H4), 2.98 (m, allylic
C5H4), 2.78 (m, allylic C5H4), 1.09 (s, C(CH3)), 1.08 (s, C(CH3)),
1.07 (s, C(CH3)), 1.06 (s, C(CH3)), 1.04 (s, C(CH3)), 0.99 (s, C(CH3)),
0.98 (s, C(CH3)), 0.97 (s, C(CH3)), 0.96 (s, C(CH3)). 13C{1H} NMR
(benzene-d6, 22 °C): d 141.5–143.1, 132.8–133.6, 124.8–125.2,
44.1–44.3, 43.4–43.8, 41.8–42.0, 33.7, 32.4, 31.2, 31.1, 30.1.
31P{1H} NMR (benzene-d6, 22 °C): d ꢁ18.7, ꢁ18.8, ꢁ19.1, ꢁ29.1,
ꢁ29.2, ꢁ30.5, ꢁ30.6 (relative integrations for these peaks: 0.18.
0.08, 0.08, 0.48, 1.00, 0.24, and 0.27, respectively).
4.3. Preparation of K2{PhP(3-t-Bu-C5H3)2} (2)
NMR spectra were recorded on a Varian Mercury spectrom-
A 100-mL Schlenk flask was charged with PhP(3-t-Bu-C5H4)2
(3.73 g, 10.6 mmol, 1.0 equiv.) and tetrahydrofuran (25 mL), then
cooled to 0 °C with stirring. In a separate Schlenk flask, potassium
tert-butoxide (2.58 g, 23.0 mmol, 2.16 equiv.) was dissolved in tet-
rahydrofuran (15 mL), followed by dropwise transfer of this solu-
tion to the PhP(3-t-Bu-C5H4)2 solution, with stirring at 0 °C. The
0 °C ice-water bath was left intact to slowly warm to 22 °C; After
30 min of stirring at 0 °C, precipitate began to form, yielding a pale
peach-colored slurry. After 72 h of stirring, the reaction mixture
eter at 300 MHz (1H), 75 MHz
( (
13C), and 121 MHz 31P) at
298 K. 1H and 13C chemical shifts are referenced relative to
the NMR solvent (residual protio solvent peak(s)); 31P chemical
shifts are referenced relative to phosphoric acid and triphenyl-
phosphine oxide. The following abbreviations are used for NMR
splitting patterns: pt (pseudo triplet) and br s (broad singlet).
Elemental analyses were performed at Atlantic Microlab, Inc.
in Norcross, Georgia.

3748
J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
was concentrated to dryness, followed by continued drying in va-
cuo with heating to 50 °C for 3.5 h to remove residual solvent
and tert-butanol byproduct. Inside the glove box, tetrahydrofuran
(75 mL) was added to provide a pale peach-colored slurry. This
slurry was transferred to a jar and cooled to ꢁ35 °C for 1 week.
At this point, a white solid and red colored supernatant were pres-
ent. Vacuum filtration yielded a white solid, which was transferred
to a pear-shaped flask for drying in vacuo for 1 h. Compound 2 was
isolated as a white powder (4.27 g, 70.1% yield). 1H NMR analysis
revealed the presence of nonstoichiometric protio THF (1.3 equiv.);
this residual solvent was taken into account for percent yield cal-
culations. Satisfactory elemental analysis data has not been ob-
tained to date; poor solubility in THF has precluded preparative
scale recrystallizations for elemental analysis purposes. 1H NMR
(THF-d8, 22 °C): d 7.80 (pt, J = 6.8 Hz, 2H, ortho-C6H5), 7.07 (pt,
J = 7.4 Hz, 2H, meta-C6H5), 6.91 (pt, J = 6.9 Hz, 1H, para-C6H5),
5.81 (br s, 4H, two overlapping signals, C5H3), 5.57 (br s, 2H,
C5H3), 3.62 (t, C4H8O), 1.78 (t, C4H8O), 1.15 (s, 18H, C(CH3)3).
13C{1H} NMR (THF-d8, 22 °C): d 132.2 (J = 14.9 Hz, ortho-C6H5),
126.8 (meta-C6H5), 123.8 (para-C6H5), 111.9 (J = 21.2 Hz, C5H3),
108.3 (J = 23.5 Hz, C5H3), 103.2 (J = 8.5 Hz, C5H3), 67.3 (C4H8O),
33.9 (C(CH3)3), four ipso carbon resonances not detected due to
poor solubility of 2. 31P{1H} NMR (THF-d8, 22 °C): d ꢁ31.1.
C5H3), 5.79 (dt, J = 2.6 Hz, 4.2 Hz, 1H, C5H3), 5.66 (dt, J = 2.6 Hz,
4.2 Hz, 1H, C5H3), 3.14–3.42 (m, 2H, NCH2), 2.86–3.10 (m, 2H,
NCH2), 1.27 (s, 9H, C(CH3)3), 1.13 (s, 9H, C(CH3)3), NCH2CH2 reso-
nances not detected, 0.20 (s, 9H, Si(CH3)3), 0.02 (s, 9H, Si(CH3)3).
13C{1H} NMR (benzene-d6, 22 °C): d 131.2 (J = 14 Hz, Co), 128.8
(J = 4 Hz, Cm), 128.0 (Cp), 118.1 (J = 39 Hz, Ca), 118.1 (Cb), 110.7
(J = 39 Hz, Cf), 110.4 (J = 10 Hz, Ce), 106.6 (J = 7 Hz, Cc), 104.7
(J = 7 Hz, Cd), C(CH3)3 resonances not detected, 46.3 (NCH2), 45.7
(NCH2), 31.4 (C(CH3)3), 31.0 (C(CH3)3), 28.6 (NCH2CH2), 4.1
(Si(CH3)3), 3.1 (Si(CH3)3). 13C{1H} NMR (THF-d8, 22 °C): d 131.5
(J = 14 Hz, Co), 129.1 (J = 4 Hz, Cm), 128.3 (Cp), 118.8 (Cb), 118.6
(J = 41 Hz, Ca), 111.0 (J = 40 Hz, Cf), 110.8 (J = 10 Hz, Ce), 107.1
(J = 7 Hz, Cc), 105.1 (J = 7 Hz, Cd), 46.7 (NCH2), 46.1 (NCH2),
(C(CH3)3) resonances not detected, 31.6 (C(CH3)3), 31.2 (C(CH3)3),
29.1 (NCH2CH2), 4.0 (Si(CH3)3), 3.1 (Si(CH3)3).
meso isomer: 1H NMR (benzene-d6, 22 °C): d 7.76 (m, 2H, Ho),
7.08 (m, 2H, Hm), 6.98 (m, 1H, Hp), 6.60 (m, 2H, Cp-H), 6.46 (m,
2H, Cp-H), 6.09 (m, 2H, Cp-H), 2.82–3.06 (m, 4H, NCH2), 1.24 (s,
18H, C(CH3)3), NCH2CH2 resonances not detected, 0.27 (s, 9H,
Si(CH3)3), ꢁ0.04 (s, 9H, Si(CH3)3). 1H NMR (THF-d8, 22 °C): d 7.56
(pt, J = 6.8 Hz, 2H, Ho), 7.32 (m, 2H, Hm), 7.26 (m, 1H, Hp), 6.56
(dd, J = 2.4 Hz, J = 3.2 Hz, 2H, C5H3), 6.49 (m, 2H, C5H3), 6.08 (m,
2H, C5H3), 3.14–3.42 (m, 2H, NCH2), 2.86–3.10 (m, 2H, NCH2),
1.29 (s, 18H, C(CH3)3), NCH2CH2 resonances not detected, 0.21 (s,
9H, Si(CH3)3), 0.11 (s, 9H, Si(CH3)3). 13C{1H} NMR (benzene-d6,
22 °C), selected resonances: d 39.9, 39.8, 22.9, 0.41 (presumably
two coincident peaks, Si(CH3)3). 13C{1H} NMR (THF-d8, 22 °C), se-
lected resonances: 46.7, 46.5, 45.4, 3.5 (Si(CH3)3), 1.6 (Si(CH3)3).
4.4. Preparation of rac/meso-{PhP(3-t-Bu-
C5H3)2}Zr{Me3SiN(CH2)3NSiMe3} (rac-3/meso-3)
A 250-mL Schlenk flask was charged with K2{PhP(3-t-Bu-
C5H3)2} ꢀ 1.3
THF
(1.00 g,
1.93 mmol,
1.0 equiv.)
and
ZrCl2(THF)2{Me3SiN(CH2)3NSiMe3} (0.87 g, 1.93 mmol, 1.0 equiv.),
degassed, and cooled to ꢁ78 °C. Tetrahydrofuran (50 mL) was
added by vacuum transfer, yielding a pale yellow slurry. The
ꢁ78 °C dry ice/acetone bath was replaced by a 0 °C ice-water bath,
and the reaction mixture was allowed to slowly warm to 22 °C.
Upon warming, the reaction mixture took on a lemon-yellow color
and opaque appearance. After 27.5 h, the reaction mixture was
concentrated to dryness, and then dried in vacuo for an additional
1.5 h. Inside the glove box, pentane (40 mL) was added and the
product mixture left to stir for 1 h. The product mixture was sub-
jected to vacuum filtration to yield a yellow filtrate. The filtrate
was then passed through a pad of celite, and finally though a dis-
posable pipet containing a Kimwipe plug. The resulting yellow fil-
trate was concentrated to dryness and dried in vacuo for an
additional 2.5 h to provide rac-3/meso-3 as a yellow, microcrystal-
line solid (0.89 g, 70.2% yield). Single crystals of rac-3 for X-ray dif-
fraction study were obtained from a pentane solution of rac-3/
meso-3 cooled to ꢁ35 °C for 8 days. 31P{1H} NMR (benzene-d6,
22 °C): d ꢁ43.6 (meso), ꢁ37.1 (rac). 0.46:1.00, initial meso:rac ratio.
31P{1H} NMR (THF-d8, 22 °C): d ꢁ40.3 (meso), ꢁ33.8 (rac). Anal.
Calc. for C33H53N2PSi2Zr: C, 60.41; H, 8.14; N, 4.27. Found: C,
59.46; H, 8.27; N, 4.21%.
4.5. Preparation of rac/meso-{PhP(3-t-Bu-C5H3)2}Zr{PhN(CH2)3NPh}
(rac-4/meso-4)
A 250-mL round bottom flask was charged with K2{PhP(3-t-Bu-
C5H3)2} ꢀ 1.3
THF
(0.32 g,
0.63 mmol,
1.01 equiv.)
and
ZrCl2(THF)2{PhN(CH2)3NPh} (0.33 g, 0.62 mmol, 1.00 equiv.),
topped with a 180° needle valve, removed from the glove box, de-
gassed on the Schlenk line, and cooled to ꢁ78 °C. Diethyl ether
(30 mL) was added by vacuum transfer, yielding a yellow slurry.
The ꢁ78 °C dry ice/acetone bath was replaced by a 0 °C ice-water
bath, and within 5 min the slurry became bright orange in color.
The temperature was maintained at 0 °C for 2 h, followed by re-
moval of solvent in vacuo and drying of the resulting orange resi-
due for an additional 45 min. Inside the glove box, pentane (30 mL)
was added and the product mixture was subjected to vacuum fil-
tration (using a fine porosity fritted funnel) to yield an orange fil-
trate. Pentane washes (4 ꢃ 2 mL) were used and the washes also
passed through the fritted funnel. The filtrate was concentrated
to dryness and dried in vacuo for an additional 1 h to provide
rac-4/meso-4 as an orange, microcrystalline solid (0.29 g, 66.4%
yield). 31P{1H} NMR (benzene-d6, 22 °C): d ꢁ38.7 (meso), ꢁ35.8
(rac), ꢁ24.6 (minor species). 1.00:0.36:0.24, meso:rac:minor spe-
cies ratio. 31P{1H} NMR (THF-d8, 60 °C): d ꢁ38.7 (meso), ꢁ35.6
(rac), ꢁ24.5 (minor species). Single crystals of rac-4 for X-ray dif-
fraction study were obtained from a concentrated benzene-d6 solu-
tion of rac-4/meso-4 (recrystallized material, 1.0:0.1 isomeric
ratio) that was stored for 5 weeks at 22 °C. Anal. Calc. for
C39H45N2PZr: C, 70.55; H, 6.83; N, 4.22. Found: C, 71.17; H, 7.14;
N, 3.77%. Notations for NMR peak assignments are as follows:
phosphino-phenyl is ring A, one amino-phenyl is ring B, the other
amino-phenyl is ring C, one cyclopentadienyl is ring D, and the
other cyclopentadienyl is ring E.
rac isomer: 1H NMR (benzene-d6, 22 °C): d 7.76 (pt, J = 7 Hz, 2H,
Ho), 7.18 (m, 2H, Hm), 6.98 (t, J = 6 Hz, 1H, Hp), 6.84 (t, J = 3 Hz, 1H,
Hb), 6.52 (dt, J = 3 Hz, 3 Hz, 1H, He), 6.42 (dt, J = 3 Hz, 3 Hz, 1H, Hf),
6.31 (dt, J = 3 Hz, 3 Hz, 1H, Ha), 5.87 (dt, J = 3 Hz, 3 Hz, 1H, Hd), 5.84
(dt, J = 3 Hz, 3 Hz, 1H, Hc), 3.35 (dd, J = 15.7 Hz, 4.5 Hz, 1H, NCH2),
3.23 (dd, J = 15.1 Hz, 4.9 Hz, 1H, NCH2), 3.03 (ddd, J = 15.7 Hz,
10.6 Hz, 4.8 Hz, 1H, NCH2), 2.94 (ddd, J = 15.1 Hz, 10.9 Hz, 5.4 Hz,
1H, NCH2), 1.21 (s, 9H, top ring C(CH3)3), 1.05 (s, 9H, bottom ring
C(CH3)3), 1.03 (m, 1H, NCH2CH2), 0.94 (m, 1H, NCH2CH2), 0.26 (s,
9H, top ring Si(CH3)3), 0.07 (s, 9H, bottom ring Si(CH3)3). Note:
NCH2 resonances at 3.35 ppm and 3.03 ppm are on the same car-
bon; NCH2 resonances at 3.23 ppm and 2.94 ppm are on the same
carbon. 1H NMR (THF-d8, 22 °C): d 7.68 (pt, J = 6.8 Hz, 2H, Ho), 7.40
(pt, J = 6.8 Hz, 2H, Hm), 7.32 (m, 1H, Hp), 6.83 (t, J = 2.6 Hz, 1H,
C5H3), 6.49 (m, 1H, C5H3), 6.23 (m, 2H, two overlapping resonances,
rac isomer: 1H NMR (benzene-d6, 22 °C): d 7.74 (pt, J = 7.2 Hz,
2H, Ho-ring A), 7.23 (t, J = 7.6 Hz, 2H, Hm-ring B), 7.13 (m, 2 H, Hm-
ring A), 7.05 (m, 2H, Hm-ring C), 7.03 (m, 1H, Hp-ring A), 6.89 (t,
J = 7.5 Hz, 1H, Hp-ring B), 6.86 (d, J = 7.9 Hz, 2H, Ho-ring B), 6.77 (t,
J = 7.5 Hz, 1H, Hp-ring C), 6.55 (m, 1H, Cp-H-ring D), 6.50 (m, 1H,

J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
3749
Cp-H-ring E), 6.44 (d, J = 8.2 Hz, 2H, Ho-ring C), 6.30 (m, 1H, Cp-H-
ring D), 6.12 (m, 1H, Cp-H-ring D), 5.91 (m, 1H, Cp-H-ring E), 5.87
(m, 1H, Cp-H-ring E), 3.26–3.60 (overlapping m, 4H, NCH2), 1.43–
1.72 (overlapping m, 2H, NCH2CH2), 0.97 (s, 9H, C(CH3)3), 0.94 (s,
9H, C(CH3)3). 13C{1H} NMR (benzene-d6, 22 °C): d 162.0 (ipso-ami-
no-phenyl), 160.4 (ipso-amino-phenyl), 143.2 (ipso-t-Bu-Cp,
J = 9.7 Hz), 138.4 (ipso-t-Bu-Cp, J = 6.7 Hz), ipso-phosphino-phenyl
resonance not detected, 131.2 (J = 14.2 Hz, Co-ring A), 128.9
(J = 4.4 Hz, Cm-ring A), 128.7 (Cm-ring B), 128.2 (Cp-ring A), 128.1
(Cm-ring C), 123.6 (ring D), 121.8 (Co-ring B), 120.2 (Cp-ring B),
119.8 (Co-ring C), 119.4 (Cp-ring C), 116.3 (J = 38.8 Hz, ring D),
114.9 (J = 38.8 Hz, ring E), 112.8 (J = 9.7 Hz, ring E), 108.4
(J = 8.5 Hz, ring D), 105.8 (J = 8.5 Hz, ring E), 105.1 (J = 19.1 Hz, ring
D), 101.0 (J = 19.1 Hz, ring E), 55.6 (NCH2), 55.0 (NCH2), C(CH3)3 res-
onances not detected, 31.5 (C(CH3)3), 31.2 (C(CH3)3), 26.3 (NCH2CH2).
meso isomer: 1H NMR (benzene-d6): d 7.72 (pt, J = 7 Hz, 2H, Ho-
ring A), 7.20 (pt, J = 7 Hz, 2H, Hm-ring A), 7.0 (m, 1H, Hp-ring A), 6.4–
7.8 (m, 10H, PhN(CH2)2NPh), 6.63 (m, 2H, Cp-H), 6.59 (d, J = 9.7 Hz,
2H, Ho-amino-phenyl), 6.59 (d, J = 9.7 Hz, 2H, Ho-amino-phenyl),
6.09 (m, 4H, Cp-H), 3.46 (t, J = 6 Hz, 2H, NCH2), 3.29 (t, J = 6 Hz,
2H, NCH2), 1.58 (t, J = 5 Hz, 2H, NCH2CH2), 0.88 (s, 18H, C(CH3)3).
13C{1H} NMR (benzene-d6): d 161.2, 144.9, 127–131 (overlapping
resonances), 119.8, 113.0, 108.1, 53.9, 52.4, 33.1, 31.2, 29.1.
rac-4 (17 mg, 26
ratio). To a J. Young NMR tube was added either KCl (60
2 equiv.), nBu4NCl (60 mol, 2 equiv.), or no additive. Using a
l
mol, 1.00 (meso):0.36 (rac):0.24 (minor species)
l
mol,
l
microsyringe, tetrahydrofuran-d8 solvent (0.7 mL) was added to
the small vial containing rac-3/meso-3 or rac-4/meso-4 and the
solution was immediately transferred to the J. Young NMR tube.
The J. Young NMR tube was removed from the glove box, shaken
vigorously, and inserted into a thermostated probe of a Varian
Mercury 300 MHz NMR spectrometer at 333 K (see Section 2);
overall, approximately 5 min elapsed between the addition of
NMR solvent and insertion of the sample into the NMR probe.
For 31P{1H} NMR data acquisition, 128 transients were recorded
for each experiment. The initial spectrum was recorded immedi-
ately (t = 0). Spectra were acquired every 10 min for 6 h, and then
every 15 min for an additional 6 h. The NMR tube was removed
from the instrument and placed in a 60 °C constant temperature
bath in the dark for continued heating. Subsequent spectra were
acquired daily until the rac/meso equilibrium state was reached.
4.8. Single crystal X-ray structure determination
4.8.1. General
Crystallographic data are presented in Table 1. Single crystals of
rac-3 and rac-4, respectively, were mounted using ParatoneÒ oil
4.6. NMR monitoring of metallation to form rac-3/meso-3
Table 1
Inside the glove box, a J. Young NMR tube was charged with 2
Crystal data and structure refinement for rac-3 and rac-4
(30 mg, 57
l
mol), ZrCl2(THF)2{Me3SiN(CH2)3NSiMe3} (26 mg,
mol, 1H NMR internal stan-
rac-3
rac-4
57 mol), and ferrocene (11 mg, 59
l
l
dard). The upper portion (inside) of the NMR tube was wiped clean
with a dry Kimwipe to remove residual solids and to allow the Tef-
lon cap to make an air-tight seal. Using a Schlenk line, tetrahydro-
furan-d8 solvent (1.0 mL) was added via vacuum transfer at
ꢁ196 °C. While the sample was still frozen, the NMR tube was
immediately inserted into a thermostated probe of a Varian Mer-
cury NMR spectrometer at 300 MHz at either 295 K or 243 K (see
Section 2). For 31P{1H} NMR data acquisition, 128 transients were
recorded for each experiment. Due to overlapping signals in the
1H NMR spectra, 31P{1H} NMR was used to monitor the relative ra-
tio of rac-3 and meso-3. The initial spectrum was recorded (t = 0)
immediately. Initially, spectra were acquired every 10 min; after
frequent monitoring for 2–3 h, 30 min intervals were used for
spectral acquisition for a total of 8 h. The NMR tube was removed
from the NMR spectrometer, stored at 22 °C in the dark, and sub-
sequent spectra were acquired at approximately 12 h intervals un-
til only rac-3 remained.
Empirical formula
Formula weight
Crystal color
Habit
Temperature (K)
Wavelength (Å)
Crystal system
Space group
C33H53N2PSi2Zr
656.14
Colorless
Block
C45H51N2PZr
742.07
Yellow
Blade
100(2)
100(2)
1.54178
Monoclinic
C2/c
1.54178
Triclinic
ꢀ
P1
Unit cell dimensions
a (Å)
28.6415(17)
90
17.7876(10)
126.134(2)
16.9360(10)
90
10.2741(5)
66.159(2)
14.4603(6)
84.658(2)
14.4648(6)
72.539(2)
1874.06(14)
2
a
(°)
b (Å)
b (°)
c (Å)
c
(°)
Volume (Å3)
6968.5(7)
8
Z
Density (calculated) (mg/m3) 1.251
1.315
3.053
Absorption coefficient
(mmꢁ1
F(000)
Crystal size (mm3)
range for data collection
3.842
)
2784
780
0.20 ꢃ 0.16 ꢃ 0.13
3.13–68.06°
ꢁ34 6 h 6 34,
ꢁ21 6 k 6 18,
ꢁ19 6 l 6 20
37012
0.15 ꢃ 0.15 ꢃ 0.07
4.51–69.81°
ꢁ12 6 h 6 12,
ꢁ16 6 k 6 17,
ꢁ16 6 l 6 17
13609
4.7. NMR monitoring of metallation to form rac-4/meso-4
H
Index ranges
Inside the glove box, a J. Young NMR tube was charged with 2
(33 mg, 64
lmol) and ZrCl2(THF)2{PhN(CH2)3NPh} (35 mg,
Reflections collected
Independent reflections
66 mol). Using
l
a
Schlenk line, protio diethyl ether solvent
6214 [0.0279]
6635 [0.0199]
(1.0 mL) was added via vacuum transfer at ꢁ196 °C. While the
sample was still frozen, the NMR tube was immediately inserted
into a thermostated probe of a Varian Mercury 300 MHz NMR
spectrometer at 273 K (see Section 2). The initial spectrum was re-
corded immediately (t = 0). For 31P{1H} NMR data acquisition, 128
transients were recorded for each experiment. Due to overlapping
signals in the 1H NMR spectra, 31P{1H} NMR was used to monitor
the relative ratio of rac-4 to meso-4.
[R(int)
Completeness to
Absorption correction
]
H
= 67.00°
99.1 %
Semi-empirical from
equivalents
96.3 %
Semi-empirical from
equivalents
Maximum and minimum
transmission
Refinement method
0.6310 and 0.5152
0.8147 and 0.6574
Full-matrix least-
squares on F2
6214/0/364
Full-matrix least-
squares on F2
6635/0/448
Data/restraints/parameters
Goodness-of-fit (GOF) on F2
1.050
1.031
Final R indices [I > 2
r
(I)]
R1 = 0.0385,
wR2 = 0.1009
R1 = 0.0400,
wR2 = 0.1020
1.512 and ꢁ0.468
R1 = 0.0270,
wR2 = 0.0699
R1 = 0.0284,
wR2 = 0.0709
0.583 and ꢁ0.345
4.8. NMR monitoring of isomerization in the absence or presence of
salts
R indices (all data)
Largest difference peak and
hole (e/Å3)
Inside the glove box, a small vial was charged with either meso-
3/rac-3 (20 mg, 30 lmol, 0.46 (meso):1.00 (rac) ratio) or meso-4/

3750
J.C. Axtell et al. / Journal of Organometallic Chemistry 693 (2008) 3741–3750
(c) M.D. LoCoco, R.F. Jordan, Organometallics 22 (2003) 5498–5503;
(d) M.D. LoCoco, R.F. Jordan, J. Am. Chem. Soc. 126 (2004) 13918–13919;
(e) A.R. Dunn, L.E. Sweet, D.C. Wiser, M.D. LoCoco, R.F. Jordan, Organometallics
23 (2004) 5671–5680;
(f) R.M. Buck, N. Vinayavekhin, R.F. Jordan, J. Am. Chem. Soc. 129 (2007) 3468–
3469.
onto a glass fiber and cooled to the data collection temperature of
100 K. Data were collected on a Brüker-AXS Kappa APEX II CCD dif-
fractometer with 1.54178 Å Cu K
were obtained from 90 data frames, 0.5°
a
radiation. Unit cell parameters
, from three different
U
sections of the Ewald sphere. For rac-3, the systematic absences
in the diffraction data were consistent with the centrosymmetric
monoclinic space group, C2/c. For rac-4, no higher symmetry than
triclinic was evident from the diffraction data; solution in the cen-
[7] E.D. Brady, T.P. Hanusa, M. Pink, V.G. Young Jr., Inorg. Chem. 39 (2000) 6028–
6037.
[8] N. Leyser, K. Schmidt, H.H. Brintzinger, Organometallics 17 (1998) 2155–2161.
[9] C. Freund, B. Martin-Vaca, G. Bouhadir, D. Bourissou, in: M.A. Cato (Ed.), Trends
in Organometallic Chemistry Research, Nova Science Publishers, Hauppauge,
NY, 2005, pp. 1–53.
ꢀ
trosymmetric space group option P1 yielded chemically reasonable
and computationally stable results of refinement. The data-set was
treated with SADABS absorption corrections based on redundant
multi-scan data (Sheldrick, G., Brüker-AXS, 2001) Tmax/Tmin = 1.22
(for rac-3) and Tmax/Tmin = 1.24 (for rac-4). A single molecule was
located in a general position yielding Z = 8 and Z0 = 1 for rac-3;
Z = 2 and Z0 = 1 for rac-4. All non-hydrogen atoms were refined
with anisotropic displacement parameters. All hydrogen atoms
were treated as idealized contribution. For rac-3, the largest differ-
ence peak and hole of 1.512 and ꢁ0.468 e/Å3 resulted from heavy
atom noise around the zirconium atom.
[10] D. Bourissou, C. Freund, B. Martin-Vaca, G. Bouhadir, C.R. Chimie 9 (2006)
1120–1142.
[11] H.G. Alt, M. Jung, J. Organomet. Chem. 568 (1998) 127–131.
[12] J.H. Shin, T. Hascall, G. Parkin, Organometallics 18 (1999) 6–9.
[13] D.N. Kazul’kin, A.N. Ryabov, V.V. Izmer, A.V. Churakov, I.P. Beletskaya, C.J.
Burns, A.Z. Voskoboynikov, Organometallics 24 (2005) 3024–3035.
[14] C.J. Schaverien, R. Ernst, W. Terlouw, P. Schut, O. Sudmeijer, P.H.M. Budzelaar,
J. Mol. Catal. A: Chem. 128 (1998) 245–256.
[15] H. Kopf, N. Klouras, Monatsh. Chem. 114 (1983) 243–247.
[16] M. Häp, T. Gilles, T. Kruck, K.-F. Tebbe, Z. Naturforsch. 54b (1999) 482–486.
[17] (a) G.K. Anderson, M. Lin, Inorg. Chim. Acta 142 (1988) 7–8;
(b) G.K. Anderson, M. Lin, Organometallics 7 (1988) 2285–2288.
[18] J.H. Shin, B.M. Bridgewater, G.E. Parkin, Organometallics 19 (2000) 5155–
5159.
[19] N. Leyser, K. Schmidt, H.H. Brintzinger, Organometallics 17 (1998) 2155–
2161.
[20] K.A.O. Starzewski, M.W. Kelly, A. Stumpf, D. Freitag, Angew. Chem., Int. Ed. 38
(1999) 2439–2443.
[21] (a) O.J. Curnow, G. Huttner, S.J. Smail, M.M. Turnbull, J. Organomet. Chem. 524
(1996) 267–270;
(b) J.R. Butchard, O.J. Curnow, S.J. Smail, J. Organomet. Chem. 541 (1997) 407–
416.
[22] (a) T.K. Hollis, L.-S. Wang, F. Tham, J. Am. Chem. Soc. 122 (2000) 11737–
11738;
Supplementary material
CCDC 697639 and 697640 contains the supplementary crystal-
lographic data for rac-3 and rac-4. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
(b) T.K. Hollis, Y.J. Ahn, F.S. Tham, Chem. Commun. (2002) 2996–2997;
(c) T.K. Hollis, Y.J. Ahn, F.S. Tham, Organometallics 22 (2003) 1432–1436;
(d) Y.J. Ahn, R.J. Rubio, T.K. Hollis, F.S. Tham, B. Donnadieu, Organometallics 25
(2006) 1079–1083.
Acknowledgments
Acknowledgment is made to the Donors of the American Chem-
ical Society Petroleum Research Fund and Villanova University for
support of this research. We thank Dr. Walter Boyko of Villanova
University for his assistance with NMR spectroscopy. We acknowl-
edge a reviewer’s suggestions regarding the structural assignments
for the rac-4/meso-4 reaction mixture.
[23] S. Bellemin-Laponnaz, M.M.-C. Lo, T.H. Peterson, J.M. Allen, G.C. Fu,
Organometallics 20 (2001) 3453–3458.
[24] (a) O.J. Curnow, G.M. Fern, Organometallics 21 (2002) 2827–2829;
(b) O.J. Curnow, G.M. Fern, M.L. Hamilton, A. Zahl, R. van Eldik,
Organometallics 23 (2004) 906–912;
(c) O.J. Curnow, G.M. Fern, M.L. Hamilton, E.M. Jenkins, J. Organomet. Chem.
689 (2004) 1897–1910.
[25] P. Jutzi, Chem. Rev. 86 (1986) 983–996.
References
[26] (a) B.L. Zhu, B.Q. Wang, S.S. Xu, X.Z. Zhou, Youji Huaxue 23 (2003) 1049–1057;
(b) B.L. Zhu, B.Q. Wang, Chem. Abstr. (2003) 830385;
(c) D.L. Zubris, in: L.S. Baugh, J.M. Canich (Eds.), Stereoselective
Polymerization with Single-Site Catalysts, CRC Press, Boca Raton, FL, 2008,
pp. 101–134.
[27] T.A. Herzog, Ph.D. Thesis, California Institute of Technology, Pasadena, CA,
2007.
[28] D.F. Shriver, M.A. Drezdon, The Manipulation of Air-Sensitive Compounds, 2nd
ed., Wiley, New York, 1986.
[29] J.H. Billman, L.R. Caswell, J. Org. Chem. 16 (1951) 1041–1046.
[30] (a) G. Courtois, L. Miginiac, J. Organomet. Chem. 340 (1988) 127–141;
(b) S. Friedrich, L.H. Gade, I.J. Scowen, M. McPartlin, Organometallics 14
(1995) 5344–5349.
[31] A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen, F.J. Timmers,
Organometallics 15 (1996) 1518–1520.
[1] H.H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger, R.M. Waymouth, Angew.
Chem., Int. Ed. Engl. 34 (1995) 1143–1170.
[2] G.W. Coates, Chem. Rev. 100 (2000) 1223–1252.
[3] J.A. Smith, J.V. Seyerl, G. Huttner, H.H. Brintzinger, J. Organomet. Chem. 173
(1979) 175–185.
[4] C.E. Zachmanoglou, A. Docrat, B.M. Bridgewater, G. Parkin, C.G. Brandow, J.E.
Bercaw, C.N. Jardine, M. Lyall, J.C. Green, J.B. Keister, J. Am. Chem. Soc. 124
(2002) 9525–9546.
[5] P.J. Shapiro, Coord. Chem. Rev. 231 (2002) 67–81.
[6] (a) X. Zhang, Q. Zhu, I.A. Guzei, R.F. Jordan, J. Am. Chem. Soc. 122 (2000) 8093–
8094;
(b) M.D. LoCoco, X. Zhang, R.F. Jordan, J. Am. Chem. Soc. 126 (2004) 15231–
15244;