Goldilocks effect of a distal substituent on living Ziegler-Natta polymerization activity and stereoselectivity within a class of zirconium amidinate-based initiators

Article abstract of DOI:10.1021/om049835z

A new series of cationic zirconium amidinates, [(η⁵-C ₅Me₅)ZrMe{N(Et)C(R³)N(^tBu)}] [B(C₆F₅)₄] (R³ = Me, H, Ph, and ^tBu) (1a-d), were synthesized and their ability to function as initiators for the stereoselective living Ziegler-Natta polymerization of 1-hexene evaluated. Whereas 1a is highly active for the isospecific living polymerization of 1-hexene as previously reported, polymerizations conducted with 1b and 1c both display a significant loss of stereocontrol, and in the case of 1b, the polymerization is no longer living. Further, 1d was found to be inactive for polymerization. The Golidlocks effect observed for the distal R³ substituent in this series, i.e., 1b too small , 1c too large , 1a just right , appears to be steric in origin. Solution and solid state structural studies suggest, however, that two different mechanisms are most likely operative for the loss of stereocontrol observed for 1b and 1c: a low barrier to metal-centered epimerization in the case of 1b and a lack of steric discrimination at the metal center for olefin binding in the case of 1c.

Full text of DOI:10.1021/om049835z

3512
Organometallics 2004, 23, 3512-3520
Gold ilock s Effect of a Dista l Su bstitu en t on Livin g
Ziegler -Na tta P olym er iza tion Activity a n d
Ster eoselectivity w ith in a Cla ss of Zir con iu m
Am id in a te-Ba sed In itia tor s
Yonghui Zhang, Erin K. Reeder, Richard J . Keaton, and Lawrence R. Sita*
Department of Chemistry and Biochemistry, University of Maryland,
College Park, Maryland 20742
Received March 5, 2004
A new series of cationic zirconium amidinates, [(η5-C5Me5)ZrMe{N(Et)C(R3)N(tBu)}]-
t
[B(C6F5)4] (R3 ) Me, H, Ph, and Bu) (1a -d ), were synthesized and their ability to function
as initiators for the stereoselective living Ziegler-Natta polymerization of 1-hexene evaluated.
Whereas 1a is highly active for the isospecific living polymerization of 1-hexene as previously
reported, polymerizations conducted with 1b and 1c both display a significant loss of
stereocontrol, and in the case of 1b, the polymerization is no longer living. Further, 1d was
found to be inactive for polymerization. The Golidlocks effect observed for the distal R3
substituent in this series, i.e., 1b “too small”, 1c “too large”, 1a “just right”, appears to be
steric in origin. Solution and solid state structural studies suggest, however, that two different
mechanisms are most likely operative for the loss of stereocontrol observed for 1b and 1c:
a low barrier to metal-centered epimerization in the case of 1b and a lack of steric
discrimination at the metal center for olefin binding in the case of 1c.
In tr od u ction
naturally been of interest to probe the steric and
electronic factors that govern both the reactivity and
stereoselectivity displayed by these various classes of
initiators through manipulation of their ligand frame-
works. In this regard, we originally reported that
cationic zirconium amidinates of the general structure
of 1 are capable of serving as highly active initiators
Since the first important discovery by Doi and co-
workers1 of a well-defined metal complex that could
serve as a homogeneous initiator for the living Ziegler-
Natta polymerization of R-olefins below -65 °C, the past
decade has witnessed the successful demonstration of
this same process by a small number of other nonmet-
allocene-based initiators that can operate at much
higher temperatures, and in some cases, with a high
degree of stereocontrol.2-9 With this success, it has
(1) (a) Doi, Y.; Ueki, S.; Keii, T. Macromolecules 1979, 12, 814-
819. (b) Doi, Y.; Ueki, S.; Keii, T. Makromol. Chem. Macromol. Chem.
Phys. 1979, 180, 1359-1361.
(2) For a recent review, see: Coates, G. W.; Hustad, P. D.; Reinartz,
S. Angew. Chem., Int. Ed. 2002, 41, 2236-2257.
(3) (a) Brookhart, M.; DeSimone, J . M.; Grant, B. E.; Tanner, M. J .
Macromolecules 1995, 28, 5378-5380. (b) Killian, C. M.; Tempel, D.
J .; J ohnson, L. K.; Brookhart, M. J . Am. Chem. Soc. 1996, 118, 11664-
11665.
for the living polymerization of R-olefins and noncon-
jugated dienes.7 Further, in the specific case where R1
t
) Et, R2 ) Bu, and R3 ) R ) Me (1a ), the polymeri-
zation of R-olefins proceeds in a stereospecific manner
to provide isotactic polyolefins through propagation
involving strict 1,2-olefin insertion.7a,b Subsequent in-
vestigations of this living system have focused on
delineating structure/activity/stereoselectivity relation-
ships by varying components of the ligand sphere that
are in close proximity to the presumed olefin binding
site. Importantly, these studies have served to show how
sensitive polymerization activity and stereoselectivity
are to nonbonded steric interactions presented by the
ligand environment to the approaching olefin substrate.
(4) Scollard, J . D.; McConville, D. H. J . Am. Chem. Soc. 1996, 118,
10008-10009.
(5) (a) Baumann, R.; Davis, W. M.; Schrock, R. R. J . Am. Chem.
Soc. 1997, 119, 3830-3831. (b) Mehrkhodavandi, P.; Bonitatebus, P.
J .; Schrock, R. R. J . Am. Chem. Soc. 2000, 122, 7481-7842. (c)
Mehrkhodavandi, P.; Schrock, R. R. J . Am. Chem. Soc. 2001, 123,
10746-10747.
(6) (a) Matsui, S.; Tohi, Y.; Mitani, M.; Saito, J .; Makio, H.; Tanaka,
H.; Nitabaru, M.; Nakano, T.; Fujita, T. Chem. Lett. 1999, 1065-1066.
(b) Saito, J .; Mitani, M.; Mohri, J .; Yoshida, Y.; Matsui, S.; Ishii, S.;
Kojoh, S.; Kashiwa, N.; Fujita, T. Angew. Chem., Int. Ed. 2001, 40,
2918-2920. (c) Mitani, M.; Mohri, J .; Yoshida, Y.; Saito, J .; Ishii, S.;
Tsuru, K.; Matsui, S.; Furuyama, R.; Nakano, T.; Tanaka, H.; Kojoh,
S.; Matsugi, T.; Kashiwa, N.; Fujita, T. J . Am. Chem. Soc. 2002, 124,
3327-3336.
(7) (a) J ayaratne, K. C.; Sita, L. R. J . Am. Chem. Soc. 2000, 122,
958-959. (b) J ayaratne, K. C.; Keaton, R. J .; Henningsen, D. A.; Sita,
L. R. J . Am. Chem. Soc. 2000, 122, 10490-10491. (c) Keaton, R. J .;
J ayaratne, K. C.; Henningsen, D. A.; Koterwas, L. A.; Sita, L. R. J .
Am. Chem. Soc. 2001, 123, 6197-6198. (d) Kissounko, D. A.; Fettinger,
J . C.; Sita, L. R. Inorg. Chim. Acta 2003, 345, 121-129. (e) Zhang, Y.;
Sita, L. R. J . Chem. Soc., Chem. Commun. 2003, 2358-2359.
(8) (a) Tshuva, E. Y.; Goldberg, I.; Kol, M. J . Am. Chem. Soc. 2000,
122, 10706-10707. (b) Tshuva, E. Y.; Goldberg, I.; Kol, M.; Gold-
schmidt, Z. J . Chem. Soc., Chem. Commun. 2001, 2120-2121.
(9) (a) Tian, J .; Coates, G. W. Angew. Chem., Int. Ed. 2000, 39,
3626-3629. (b) Tian, J .; Hustad, P. D.; Coates, G. W. J . Am. Chem.
Soc. 2001, 123, 5134-5135.
10.1021/om049835z CCC: $27.50 © 2004 American Chemical Society
Publication on Web 06/08/2004

Zirconium Amidinate-Based Initiators
Organometallics, Vol. 23, No. 14, 2004 3513
Sch em e 1
Sch em e 2
To briefly summarize what has currently been mapped
out, the larger steric bulk provided by the η5-C5Me5
(Cp*) vs the η5-C5H5 (Cp) group (i.e., R ) Me and H in
1, respectively) is critical for directing olefin stereoface
selectivity, thereby providing a highly isotactic polymer
microstructure (mmmm g99%) in the former case, but
nearly atactic material in the latter.7c Of greater
surprise, however, was the observed sensitivity of
activity and stereoselectivity to replacement of the N-Et
group of the amidinate ligand in 1a with N-CH2X
substituents, where X varied in size from isopropyl to
phenyl to tert-butyl.7d More specifically, with a N-Et f
N-iBu substitution, polymerization activity decreased
dramatically, and a loss of stereocontrol was observed
to provide a nearly atactic microstructure. In contrast,
the N-Et f N-CH2Ph substitution yielded an initiator
that retained both the high activity and stereoselectivity
displayed by 1a . In both of these substitutions, the living
character of the polymerization was also preserved.
Finally, the N-Et f N-CH2CMe3 substitution provided
an inactive cationic complex, presumably due to sub-
stantial crowding about the metal center that prevents
productive olefin binding.
Zr-CMe bond of in situ generated Cp*ZrMe3,7 to vary
the nature of the R3 substituent, different synthetic
strategies were employed to prepare compounds 2a -d
t
where R3 ) Me, H, Ph, and Bu, respectively. To begin,
following the synthetic route to 2a published previ-
ously,11 compounds 2c and 2d were prepared by first
generating the corresponding lithium amidinate salts
through addition of the appropriate organolithium
reagent to commercially available 1-tert-butyl-3-ethyl-
carbodiimide. As shown in Scheme 1, these lithium
reagents were then used to produce 2c and 2d in high
yield from Cp*ZrCl3 through standard procedures.
Synthesis of the formamidinate complex 2b (i.e., R3
) H) required a slightly different path. As shown in
Scheme 2, the required lithium formamidinate was
prepared through deprotonation of the corresponding
amidine, which in turn was obtained through hydrosi-
lyation of the carbodiimide followed by methanolysis of
the silyl group according to the procedure of Ojima and
co-workers.12 Following the same procedure, the Cs-
symmetric compound 3 was prepared as well since a
greater range of structure/property relationships were
of interest. In this regard, we have previously observed
t
for 1a that replacement of the R1 ) Et, R2 ) Bu
With the high degree of sensitivity to proximal groups
established for 1, attention was next turned to deter-
mining what influence distal groups within the ligand
environment might have on living character, activity,
and stereoselectivity. In this report, the results of
varying the steric bulk of the Me group in the amidinate
fragment of 1a (i.e., R3 ) Me in 1) are presented. More
specifically, a comparison of living character, activity,
and stereoselectivity in the Ziegler-Natta polymeriza-
i
combination by the R1 ) R2 ) Pr substituent pattern
greatly attenuates polymerization activity, presumably
due to a greater steric shielding of the metal center in
the latter case.
Spectroscopic and chemical analyses were all consis-
tent with the structures depicted for compounds 2a -d
and 3. Importantly, 1H NMR spectroscopy revealed that
a low barrier for racemization of the metal center
through amidinate ring-flipping13 exists for these com-
pounds, regardless of the steric bulk of the distal
amidinate substituent, such that this racemization
proceeds readily at room temperature on the NMR time
scale. This result is somewhat surprising in that we
have previously determined that monoalkylation of 2a
provides chloro, alkyl complexes, such as 4 (X ) Cl, R
t
tion of 1-hexene for the R3 ) Me f H, Me, Ph, and Bu
replacements reveals how sensitive these parameters
can also be to the influence of distal parts of the ligand
environment. Perhaps more importantly, these observa-
tions show, once again, how difficult it is to “rationally”
design an optimum initiator based on a new ligand from
first principles.10
(10) See also for instance: (a) Busico, V.; Van Axel Castelli, V.;
Aprea, P.; Cipullo, R.; Segre, A.; Talarico, G.; Vacatello, M. J . Am.
Chem. Soc. 2003, 125, 5451-5460. (b) Talarico, G.; Busico, V.; Cavallo,
L. J . Am. Chem. Soc. 2003, 125, 7172-7173.
Resu lts a n d Discu ssion
(a ) Syn th esis of Cp *Zr Cl2[N(R1)C(R3)N(R2)] (R1
(11) Keaton, R. J .; Koterwas, L. A.; Fettinger, J . C.; Sita, L. R. J .
Am. Chem. Soc. 2002, 124, 5932-5933.
t
t
) Et, R2 ) Bu , R3 ) Me, H, P h , Bu (2a -d ) a n d R1
) R2 ) P r , R3 ) H (3)). Although a large number of
i
(12) Ojima, I.; Inaba, S. J . Organomet. Chem. 1977, 140, 97-111.
(13) (a) Sita, L. R.; Babcock, J . R. Organometallics 1998, 17, 5228-
5230. (b) Koterwas, L. A.; Fettinger, J . C.; Sita, L. R. Organometallics
1999, 18, 4183-4190.
derivatives of 1 have now been conveniently prepared
in one step by utilizing carbodiimide insertion into a

3514 Organometallics, Vol. 23, No. 14, 2004
Zhang et al.
F igu r e 1. Molecular structures (30% thermal ellipsoids) of (a) 2b, (b) 2c, and (c) 2d . Hydrogen atoms, except for that on
C(11) of 2b (drawn as a sphere of arbitrary size), have been removed for the sake of clarity.
Ta ble 1. Selected Bon d Len gth s a n d Bon d An gles for th e Molecu la r Str u ctu r es of 2b-d , 6a , 6b, 6d , a n d 1d
2b
2c
2d
6b
6a a
6d
1d
bond lengths (Å)
Zr(1)-N(1)
Zr(1)-N(2)
N(1)-C(11)
N(2)-C(11)
2.2446(11)
2.2218(17)
1.327(2)
2.2645(2)
2.176(2)
1.331(4)
1.334(4)
2.177(3)
2.278(3)
1.3197(18)
1.3268(19)
2.3029(9)
2.2770(9)
1.3238(14)
1.3219(15)
2.265(2)
2.251(3)
1.332(4)
1.323(4)
2.215(4)
2.333(4)
1.367(6)
1.320(6)
2.123(3)
2.157(3)
1.349(4)
1.338(4)
1.321(2)
bond angles (deg)
N(1)-Zr(1)-N(2)
Zr(1)-N(1)-C(11)
60.01(5)
59.60(9)
92.20(17)
140.6(2)
125.4(2)
96.13(17)
140.4(2)
123.3(3)
111.9(2)
58.95(12)
92.5(2)
136.6(2)
130.4(3)
89.2(2)
131.4(2)
127.0(3)
108.3(3)
58.94(3)
89.86(6)
58.40(9)
92.76(17)
142.5(16)
122(2)
93.62(18)
136.4(17)
123.5(16)
112.2(2)
57.50(13)
93.4(3)
135.1(3)
128.5(4)
89.5(3)
132.7(3)
126.5(4)
109.2(4)
61.33(10)
95.57(19)
117.95(19)
141.6(3)
94.36(19)
136.3(2)
129.0(3)
108.7(3)
91.59(10)
146.40(9)
118.88(12)
92.77(11)
145.64(16)
119.93(17)
115.05(15)
t
Zr(1)-N(1)-C Bu
147.30(7)
118.01(9)
91.03(6)
141.20(8)
117.65(10)
116.79(10)
t
C(11)-N(1)-C Bu
Zr(1)-N(2)-C(11)
Zr(1)-N(2)-CEt
C(11)-N(2)-CEt
N(1)-C(11)-N(2)
∑θN(1) (deg)b
∑θN(2) (deg)c
φ (deg)d
356.87
358.34
7.8
358.20
359.83
3.8
359.5
347.6
34.7
355.17
349.88
18.6
357.26
353.52
18.3
357.0
348.7
33.6
355.12
359.66
3.1
a
b
c Sum of angles: Zr(1)-N(2)-C(11),
t
t
From ref 19. Sum of angles: Zr(1)-N(1)-C(11), Zr(1)-N(1)-C Bu
,
C(11)-N(1)-C Bu
.
d
Zr(1)-N(2)-CEt, C(11)-N(2)-CEt
.
Angle between the mean planes defined by the following: Zr(1)-N(1)-C(11) and Zr(1)-N(2)-C(11).
) isobutyl), that are configurationally stable, even at
elevated temperatures up to 80 °C, whereas dialkylation
of 2a provides dialkyl complexes, such as 5 (X ) Me, R
) isobutyl), that undergo metal-centered racemization
even at subambient temperatures.11,14,15 On the basis
of solid-state structures for different derivatives of 4 and
5, we speculated that the greater electronegativity of
Figure 1 presents the molecular structures of com-
pounds 2b-d as obtained from single-crystal X-ray
analysis; and Table 1 gives some selected bond lengths
and bond angles. As anticipated, having two chloro
groups on the zirconium center does indeed appear to
reduce zirconium-nitrogen bonds further relative to
either 4 or 5. For example, in compound 2b, the Zr-N
distances are 2.2218(17) and 2.2446(11) Å, and in the
structure of 3, which has also been determined,17 they
are 2.2159(12) and 2.2229(11) Å. In contrast, the cor-
responding Zr-N bond lengths in a chloro, isopropyl
zirconium acetamidinate complex (i.e., X ) Cl, R ) iPr
in 4) are 2.2419(11) and 2.2528(11) Å, while for a
i
methyl, isopropyl derivative (i.e., X ) Me, R ) Pr in
5), these values are 2.2594(9) and 2.2809(9) Å.15,16
However, it is now clear that given the configurational
instability of 2, this reduction in Zr-N bond lengths
relative to 4 and 5 is not sufficient by itself to provide
a significant barrier to racemization. Accordingly, con-
figurational stability is most likely determined by a
subtle combination of both electronic and steric effects
provided by the level of chloro vs alkyl substitution at
the metal center. Regardless of the influence of these
proximal substituents, it certainly does appear true that
the chloro group results in a strengthening of zirconium-
nitrogen bonding interactions that, in turn, raise the
barrier to amidinate ring-flipping.16 Given that com-
pounds 2a -d and 3 possess two chloro substituents, it
was naturally of interest to test this hypothesis further
by obtaining solid-state structural information for them.
(14) Zhang, Y.; Keaton, R. J .; Sita, L. R. J . Am. Chem. Soc. 2003,
125, 9062-9069.
(15) Zhang, Y.; Keaton, R. J .; Sita, L. R. J . Am. Chem. Soc. 2003,
125, 8746-8747.
(16) Harney, H. B.; Keaton, R. J .; Sita, L. R. Manuscript in
preparation.
(17) Detailed information is provided in the Supporting Information.

Zirconium Amidinate-Based Initiators
Organometallics, Vol. 23, No. 14, 2004 3515
F igu r e 2. Molecular structures (30% thermal ellipsoids) of (a) 6b, (b) 6a , and (c) 6d . Hydrogen atoms, except for that on
C(11) of 6b (drawn as a sphere of arbitrary size), have been removed for the sake of clarity.
Sch em e 3
the steric bulk of the distal amidinate substituent plays
no or at best a very small role in controlling the barrier
to racemization in 2.
The strongest influence of distal effects within the
series of 2b-d can be seen in the value of the bond
angles about the nitrogen atoms. Of greatest signifi-
cance is the reduction in the Zr-N-Cexo bond angles
as the steric bulk of the amidinate substituent increases
in going from 2b [146.40(9) and 145.64(16)°] to 2c
[140.6(2) and 140.4(2)°] to 2d [136.6(2) and 131.4(2)°].
This reduction is accompanied by a concomitant in-
crease in the Cexo-N-C(11) bond angles [cf. 2b, 118.88-
(12)° and 119.93(17)°; 2c, 125.4(2)° and 123.3(3)°; 2d ,
136.6(2)° and 131.4(2)°]. Thus, as a result of “buttress-
ing” effects18 that arise with the increasing steric bulk
of the C-amidinate substituent (i.e., R3 in 1), the metal
center becomes increasingly more shielded in the order
2b < 2c < 2d . A final geometric perturbation of note is
the extent of puckering of the four-membered amidinate
ring as quantified by the angle between the two mean
planes defined by Zr(1)-N(1)-C(11) and Zr(1)-N(2)-
C(11). As can be seen in Table 1, the amidinate ring is
nearly planar for compounds 2b and 2c (cf. 7.8° and 3.8°,
respectively), whereas in compound 2d it is significantly
puckered (cf. φ ) 34.7°). Closer inspection of the mo-
lecular structure of 2d , including analysis of a space-
filling representation, reveals that the strong nonbonded
steric interactions prevent all three amidinate substit-
uents from lying in the same plane. Oddly, however, it
is the Zr-N bond distance involving the nitrogen with
the N-Et substituent that is the longest, and it is this
same nitrogen atom, N(2), that becomes significantly
more pyramidalized than the one bearing the N-tBu
group (cf. ∑θN(2) ) 347.6° vs ∑θN(1) ) 359.5°).
with the formulation of these compounds as presented,
and in keeping with all other dialkyl derivatives pre-
pared to date, 6a -e were found to be configurationally
unstable on the 1H NMR time scale with respect to
metal-centered racemization. Finally, to aid in the
further development of structure/property relationships,
the molecular structures of 6a , 6b , and 6d were
determined by single-crystal X-ray analysis and Figure
2 and Table 1 present the molecular structures and
selected geometric parameters for these compounds,
respectively.17,19 As expected, the Zr-N bond distances
in these compounds are significantly longer than those
in their corresponding dichloro derivatives [cf. 2.2446-
(11) and 2.2218(17) Å for 2b vs 2.2770(9) and 2.3029(9)
Å for 6b]. With respect to the influence of the distal
amidinate substituent, the trends observed previously
for the angles about the nitrogen atoms are observed
here as well. More specifically, as one goes from 6b (R3
) H) to 6a (R3 ) Me) to 6d (R3 ) tBu), there is a steady
reduction in the Zr-N-Cexo bond angles [cf. 147.30(7)°
and 141.20(8)° for 6b; 142.5(16)° and 136.4(17)° for 6b;
135.1(3)° and 132.7(3)° for 6d ] and a steady increase in
the Cexo-N-C(11) bond angle [cf. 118.01(9)° and 117.65-
(10)° for 6b; 122(2)° and 123.50(16)° for 6b; 128.5(4)°
and 126.5(4)° for 6d ]. Thus, buttressing effects again
serve to increasingly shield the metal center as the
steric bulk of the distal R3 ligand increases. One
surprising feature for some of these dimethyl structures,
however, is that the amidinate fragment is much more
nonplanar than in the corresponding dichloro analogues.
For instance, the value of the angle, φ, is ∼18.5° for both
6a and 6b, whereas both of the respective dichloro
derivatives have nearly planar four-membered rings
(vide supra). Given the similar sizes of chloro and
methyl groups, it is possible that this geometric differ-
(b) Syn th esis of Cp *Zr Me2[N(R1)C(R3)N(R2)] (R1
t
t
) Et, R2 ) Bu , R3 ) Me, H, P h , Bu (6a -d ) a n d R1
) R2 ) iP r , R3 ) H (6e)). Following standard published
procedures, methylation of 2a -d and 3 with 2 equiv of
methyllithium proceeded smoothly in diethyl ether to
provide the corresponding dimethyl derivatives 6a -e
in high yield according to Scheme 3. Once again,
spectroscopic and chemical analyses were consistent
(18) For a discussion of similar buttressing effects in group 13
amidinate complexes, see: (a) Dagorne, S.; J ordan, R. F.; Young, V.
G., J r. Organometallics 1999, 18, 4619-4623. (b) Dagorne, S.; Guzei,
I. A.; Coles, M. P.; J ordan, R. F. J . Am. Chem. Soc. 2000, 122, 274-
289.
(19) Keaton, R. J .; J ayaratne, K. C.; Fettinger, J . C.; Sita, L. R. J .
Am. Chem. Soc. 2000, 122, 12909-12910.

3516 Organometallics, Vol. 23, No. 14, 2004
Zhang et al.
Sch em e 4
Ta ble 2. Da ta for P olym er iza tion of 1-Hexen e by
1a -ea ,b
initiator
yield (%)
Mn
Mw/Mn
1a
1b
1c
1d
1e
95
45
90
19 800
20 100
19 300
1.03
1.59
1.02
no activity
98
25 100
1.23
a
Polymerizations were conducted for 2 h in chlorobenzene at
b
-10 °C, using 200 equiv of 1-hexene and 2.5 µmol of 6. Mn and
Mw/ Mn values were obtained from GPC and are reported relative
to polystyrene standards.
ence between the two series is largely electronic in
origin. On the other hand, the φ value of 33.6° for 6d is
quite similar to that found for 2d , thus suggesting that
strong nonbonded interactions between the amidinate
substituents can override any electronic effects at the
metal center.
(c) In Situ Gen er a tion of [Cp *Zr Me{N(R1)C-
(R3)N(R2)}][B(C6F 5)4] a n d P olym er iza tion Stu d ies
w ith 1-Hexen e. Polymerizations of 200 equiv of 1-hex-
ene were conducted according to Scheme 4 by generating
the cationic initiators 1a -e in chlorobenzene at -10 °C
through reaction of the dimethyl precursors 6a -e with
1 equiv of the borate, [PhNHMe2][B(C6F5)4]. After a
fixed polymerization time of 2 h, the polymerizations
were quenched and the poly(1-hexene) material isolated
1
and purified by precipitation. H NMR was then used
F igu r e 3. 13C{1H} NMR (100 MHz, chloroform-d, 25 °C)
spectra of poly(1-hexene) obtained from, top to bottom, 1a ,
1c, 1b, and 1e.
to assess the presence of vinylic end groups that might
arise from termination during propagation via â-hydride
elimination, while 13C NMR was used to determine the
tacticity of the polymer microstructure. Table 2 and
Figure 3 present relevant data obtained from these
polymerization studies. To begin, both the Mn and Mw/
Mn values recorded for poly(1-hexene) obtained from 1a
are similar to previously published results7 and they are
in agreement with those expected for a living polymer-
ization proceeding with quantitative conversion of the
monomer. Further, the 13C NMR spectrum shown in
Figure 3 confirms that the microstructure of this mate-
rial is perfectly isotactic. For initiator 1c, similar Mn,
Mw/Mn, and yield of polymer were obtained; however, a
13C NMR spectrum now revealed that a significant loss
of stereocontrol had occurred to provide a microstructure
that is only slightly iso-rich as presented in Figure 3.
On the other hand, polymerizations with 1b proceeded
in poor yield for the same period of time and the
polymerization was not living as indicated by the
significantly larger polydispersity that is observed (i.e,
Mw/Mn ) 1.58). The 13C NMR spectrum of the poly(1-
hexene) obtained from 1b shown in Figure 3 reveals it
to be more iso-rich than that obtained from 1c, but a
loss of stereocontrol relative to 1a is definitely seen. For
the Cs-symmetric derivative 1e, poly(1-hexene) was
obtained in high yield and with a narrower polydisper-
sity (cf. Mw/Mn ) 1.23); however, the higher Mn and Mw/
Mn values relative to 1a still suggest that this polym-
erization was not strictly living in character. The 13C
NMR spectrum of the polymer obtained from 1e was
expectedly much more nearly atactic as revealed by
Figure 3 with the small degree of isoselectivity possibly
arising from chain-end control.7c Finally, the cationic
complex 1d displayed no catalytic activity even at room
temperature presumably due to a large degree of steric
crowding about the metal center that prevents produc-
tive olefin binding.
The polymerization data shown in Table 2 and Figure
3 present a clear example of a “Goldilocks” effect for the
relationship between the steric bulk of the R3 group and
several key parameters for 1-hexene polymerization by
the initiators 1a -d , including activity, stereocontrol,
and living character. Restating this in more general
terms, it would appear that while R3 ) H (1b) is “too
t
small” and R3 ) Ph (1c) and Bu (1d ) are “too large”, R3
) Me (1a ) is “just right”. A closer look at the solution
structures of 1a -d and the solid-state structure of 1d
shed more light on the steric origins of this Goldilocks
effect of the distal amidinate substituent. To begin, we
have previously demonstrated through the use of vari-

Zirconium Amidinate-Based Initiators
Organometallics, Vol. 23, No. 14, 2004 3517
F igu r e 5. Molecular structures (30% thermal ellipsoids)
of 7. Hydrogen atoms, except those on C(11) and C(11A),
and the borate counterions have been removed for the sake
of clarity.17
F igu r e 4. Molecular structures (30% thermal ellipsoids)
of 1d . Hydrogen atoms and the borate counterion have been
removed for the sake of clarity.
able-temperature NMR spectroscopy that while neutral
6a undergoes facile metal-centered racemization even
at low temperatures, cationic 1a is configurationally
stable up to temperatures of at least 25 °C.14,20 Using
this same technique, it was surprisingly discovered that
initiator 1b is not configurationally stable at -10 °C,
whereas both 1c and 1d are. Thus, it becomes more
obvious that the buttressing effects initiated by the R3
amidinate group that were observed in the solid-
structures of neutral 2 and 6 are playing a critical role
in controlling the barrier to amidinate ring-flipping in
cationic 1. In the case of 1b, the rate of epimerization
of the metal center of the propagating center must be
close to the rate of propagation, which then gives rise
to the partial loss of stereocontrol observed for this
initiator. In the case of 1c, we have previously noted
that to obtain a high degree of stereocontrol in the
polymerization of 1-hexene, there must be sufficient
steric discrimination being presented to the incoming
monomer by the two N-substituents of the amidinate
group (vide supra).7d It would thus appear likely that
this steric discrimination is lost in the case where R3 )
Ph due to the buttressing effects that serve to push both
of the N-substituents forward. Importantly, this analy-
sis concludes that the steric origin of decreased stereo-
control observed for 1c is different than that observed
for 1b. Finally, it was of interest to determine at what
point does this buttressing effect of the amidinate group
prevent productive olefin binding, and to this end, the
solid-state structure of 1d was obtained. Figure 4
presents the molecular structure of 1d , and Table 1
gives some selected bond lengths and bond angles. To
begin, 1d was found to be monomeric in the solid state
as opposed to the dimeric dication structure obtained
for 1a ,19 and this observation already suggests that
significant steric crowding must be present about the
metal to prevent such dimerization from occurring. By
now looking at the same set of structural parameters
discussed previously for 6d , some notable differences
can be seen in the molecular structure of cationic 1d ,
and in particular, in the geometry of the amidinate
fragment. More specifically, although the Zr-N bonds
of 1d are shorter than those in 6d , the amidinate four-
membered ring is now nearly planar (cf. a φ value of
3.1° for 1d vs 33.6° for 6d ) and both of the nitrogen
atoms are more trigonal coplanar. Further, the Zr-N-
C
exo and Cexo-N-C(11) bond angles of 117.94(19)° and
141.6(3)° in 1d are opposite in magnitude to the corre-
sponding values seen for 6d . Thus, altogether, these
differences suggest that formation of the cationic species
1d through demethylation of 6d can be likened to
releasing a spring under compression in which the steric
interactions pent up in 6d are removed by allowing the
N-tBu substituent to swing into the position previously
occupied by the departing methyl group. It is then likely
that this structural displacement of the N-tBu group is
responsible for preventing the remaining methyl group
from exiting the olefin binding pocket, which is pre-
sumed to be on the N-Et side of the amidinate group.
One answer to the question of why polymerizations
using the formamidinate initiators 1b and 1e are not
living comes from analysis of the orange-red crystalline
material obtained from chlorobenzene solutions of 1e
left for a few hours at 25 °C. As Figure 5 reveals, this
compound proved to be the µ-Cl dimeric dication 7 that
presumably arises from chloride abstraction from the
solvent according to Scheme 5, although the ultimate
fate of the methyl group is not presently known. Similar
µ-chloro dicationic dizirconium species have been ob-
served by Green and co-workers21 and Collins and co-
workers22 from demethylations of other zirconium com-
plexes. We have also obtained the analogous compound,
{Cp*Zr(µ-Cl)[N(tBu)C(Me)N(Et)]}2{[B(C6F5)4]}2 (8), from
chlorobenzene solutions of 1a ; however, in this case,
much longer reaction times were required. If this
difference in reactivity is associated with a lesser degree
of steric crowding about the metal center in 1e than in
1a , then it is safe to assume that given the even greater
metal accessibility of 1b, propagating species derived
from it will be even more highly susceptible to reaction
with the solvent. Thus, the increase in living character
in the order 1b < 1e < 1a is readily explained.
(21) Gomez, R.; Green, M. L. H.; Haggit, J . L. J . Chem. Soc., Dalton
Trans. 1996, 939-946.
(22) Vollmerhaus, R.; Rahim, M.; Tomaszewski, R.; Xin, S.; Taylor,
N. J .; Collins, S. Organometallics 2000, 19, 2161-2169.
(20) J ayaratne, K. C.; Sita, L. R. J . Am. Chem. Soc. 2001, 123,
10754-10755.

3518 Organometallics, Vol. 23, No. 14, 2004
Zhang et al.
Sch em e 5
tBu NdC(H)-N(Et)SiEt3. A mixture of 10.1 g (80.0 mmol)
Con clu d in g Rem a r k s
of 1-tert-butyl-3-ethylcarbodiimide, 11.1 g (95.4 mmol) of
triethylsilane, and 0.24 g (1.35 mmol) of PdCl2 was sealed in
a Schlenk tube and heated at 150 °C overnight. After the
mixture was cooled to room temperature, the crude product
was vacuum distilled at 85 °C (4 mmHg) to provide the desired
The results described herein demonstrate how sensi-
tive key parameters for the homogeneous metal-
catalyzed Ziegler-Natta polymerization of R-olefins,
including activity, stereoselectivity, and living character,
can be to the steric bulk of distal parts of the ligand
sphere that are transmitted to the metal center of the
propagating species through buttressing effects. In the
examples shown, a distinct Goldilocks effect was ob-
served where a fine balance between nonliving character
and no activity is controlled by the steric bulk of the
distal amidinate substituent. Importantly, with respect
to the loss of stereocontrol observed for 1b and 1c, it
was determined that two different mechanisms can be
operative depending upon the steric bulk of the distal
amidinate substituent: a low barrier to metal-centered
epimerization in the case of 1b and a lack of steric
discrimination at the metal center for olefin binding in
the case of 1c. Taken together, the present study once
again points out the difficulties in obtaining a highly
active initiator for living and/or stereoselective Ziegler-
Natta polymerization from first principles based on a
new ligand design, and only through fortuitous circum-
stance might one achieve success the first time, or after
only a few iterative modifications of the ligand frame-
work. In this regard, the introduction of high through-
put screening strategies that can more easily explore a
greater range of ligand structural diversity in a shorter
period of time has much to offer.23
1
material as a light yellow liquid (11.6 g, 61% yield). H NMR
3
(400 MHz, benzene-d6, 25 °C) δ 7.57 (s, 1H), 3.26 (q, J ) 6.8
3
3
Hz, 2H), 1.28 (s, 9H), 1.18 (t, J ) 6.8 Hz, 3H), 0.90 (t, J )
3
7.6 Hz, 9H), 0.59 (q, J ) 7.6 Hz, 6H).
tBu NdC(H)-NHEt. To 11.6 g (47.8 mmol) of freshly dis-
tilled BuNdC(H)-NH(Et)SiEt3 was added 1.8 g (56 mmol) of
t
methanol. After 1 h, vacuum distillation provided 5.0 g (81%
yield) of the desired product as a colorless liquid. 1H NMR (400
3
MHz, CDCl3, 25 °C) δ 7.40 (s, 1H), 3.16 (q, J ) 7.2 Hz, 2H),
3
1.18 (s, 9H), 1.10 (t, J ) 7.2 Hz, 3H).
Com p ou n d 2b. To a solution of 1.03 g (8.00 mmol) of
tBuNdC(H)-NH(Et) in 100 mL of Et2O, cooled to -30 °C, was
added 2.81 mL of n-BuLi (2.90 M in hexane, 8.16 mmol) via a
syringe. The reaction was allowed to warm to room temper-
ature within 3 h and then the clear solution was transferred
via cannula into a flask containing 2.66 g (8.00 mmol) of
Cp*ZrCl3 in 250 mL of Et2O cooled to -78 °C. The mixture
was slowly warmed to room temperature and stirred for 12 h
whereupon the volatiles were removed in vacuo. The residue
was extracted with toluene and filtered through a short pad
of Celite to afford a yellow solution, which upon concentration
and cooling to -35° C provided yellow crystals of 2b (2.49 g,
73% yield). For 2b: 1H NMR (400 MHz, benzene-d6) δ 7.98 (s,
1H), 3.06 (q, 3J ) 7.2 Hz, 2H), 2.03 (s, 15H), 1.12 (s, 15H),
0.96 (d, 3J ) 7.2 Hz, 3H). Anal. Calcd for C17H30Cl2N2Zr: C
48.09, H 7.12, N 6.60. Found: C 48.19, H 7.02, N 6.36.
Com p ou n d 2c. To a solution of 0.25 g (2.00 mmol) of 1-tert-
butyl-3-ethylcarbodiimide in 50 mL of Et2O, cooled to 0 °C,
was added 1.02 mL of PhLi (2.0 M in cyclopentane/Et2O 7:3
mixture, 2.05 mmol) and the resulting mixture was stirred for
3 h. The clear solution was then transferred via cannula into
a flask containing a solution of 0.67 g (2.00 mmol) of Cp*ZrCl3
in 40 mL of Et2O at -78 °C. After addition, the mixture was
slowly warmed to room temperature and stirred for 12 h
whereupon the volatiles were removed in vacuo. The residue
was extracted with toluene and filtered through a short pad
of Celite to afford a yellow solution, which upon concentration
and cooling to -35 °C provided yellow crystals (0.78 g, 78%
yield). For 2c: 1H NMR (400 MHz, benzene-d6) δ 6.93 (d, 2H),
6.88 (m, 3H), 2.91(q, J ) 7.2 Hz, 2H, CH2CH3), 2.13 (s, 15H,
C5Me5), 1.15 (s, 9H, CMe3), 0.87 (t, J ) 7.2 Hz, 3H, CH2CH3).
Anal. Calcd for C23H34Cl2N2Zr: C 55.17, H 6.79, N 5.60.
Found: C 54.99, H 6.53, N 5.81.
Exp er im en ta l Section
All manipulations were performed under an inert atmo-
sphere of dinitrogen, using standard Schlenk techniques or a
Vacuum Atmospheres glovebox. Dry, oxygen-free solvents were
employed throughout. Diethyl ether (Et2O) and pentane were
distilled from NaK. MeLi (in Et2O), PhLi (in cyclopentane/
t
Et2O), and BuLi (in pentane) were purchased from Aldrich
and used as received. Cp*ZrCl3 was obtained from Strem and
[PhNHMe2][B(C6F5)4] was obtained from Boulder Scientific.
Compounds 2a and 6a were prepared according to previously
reported procedures.7,11 Chlorobenzene was distilled from
calcium hydride and 1-hexene was vacuum transferred from
NaK prior to being used for polymerizations. GPC analyses
were performed with a Waters GPC system equipped with a
column oven and a differential refractometer both maintained
at 40 °C and four columns (Waters Ultrastyragel 500 Å, Waters
Styragel HR3, Waters Styragel HR4, and Shodex K-806M) also
maintained at 40 °C. THF was used as the eluant at a flow
rate of 1.1 mL /min. Mn and Mw/Mn values were obtained with
the Waters GPC software and seven different polystyrene
standards (Polymer Laboratories). For NMR, benzene-d6 was
Com p ou n d 2d . The same procedure as for 2c was followed
t
except BuLi was used in place of PhLi. For 2d : 1H NMR (400
MHz, benzene-d6) δ 3.39 (q, J ) 7.2 Hz, 2H, CH2CH3), 2.04 (s,
15H, C5Me5), 1.37 (s, 9H), 1.20 (s, 9H). 1.09 (t, J ) 7.2 Hz, 3H,
CH2CH3). Anal. Calcd for C21H38Cl2N2Zr: C 51.04, H 7.76, N
5.55. Found: C 51.23, H 7.77, N 5.43.
iP r NdC(H)-NH(iP r )SiEt3. A mixture of 10.1 g (80.0 mmol)
of 1,3-diisopropylcarbodiimide, 11.1 g (95.4 mmol) of triethyl-
silane, and 0.24 g (1.35 mmol) of PdCl2 was sealed in a Schlenk
tube and heated at 150 °C overnight. After the mixture was
cooled to room temperature, the crude product was vacuum
distilled (90 °C (7 mmHg)) to provide a light yellow liquid (12.6
1
vacuum transferred from NaK prior to use. H NMR and 13C-
{1H}NMR spectra were recorded at 400 and 100 MHz, respec-
tively, using chloroform-d or benzene-d6 as the solvents.
(23) For a recent review, see: Reetz, M. T. Combinatorial Methods
in Catalysis by Metal Complexes. In Comprehensive Coordination
Chemistry II 2004, 9, 509-548.

Zirconium Amidinate-Based Initiators
Organometallics, Vol. 23, No. 14, 2004 3519

3520 Organometallics, Vol. 23, No. 14, 2004
Zhang et al.
g, 66% yield). 1H NMR (400 MHz, benzene-d6, 25 °C) δ 7.53
Cr ysta llogr a p h y. The crystal structure and refinement
data for compounds 1d , 2b-d , 6b, 6d , and 7 are presented in
Table 3. The crystallographic analysis of compound 6a has
been previously reported.19 As a general procedure, for 2b, a
colorless block with approximate orthogonal dimensions 0.40
× 0.38 × 0.38 mm3 was placed and optically centered on the
Bruker SMART CCD system at -70 °C. The initial unit cell
was indexed by using a least-squares analysis of a random
set of reflections collected from three series of 0.3° wide
ω-scans, 10 s per frame, and 25 frames per series that were
well distributed in reciprocal space. Data frames were collected
[Mo KR] with 0.3° wide ω-scans, 14 s per frame, and 606
frames per series. Five complete series were collected at
varying æ angles (æ ) 0°, 72°, 144°, 216°, 288°). An additional
200 frames, a repeat of the first series for redundancy and
decay purposes, were also collected. The crystal-to-detector
distance was 4.356 cm, thus providing a complete sphere of
data to 2θmax ) 60.0°. A total of 40872 reflections were collected
and corrected for Lorentz and polarization effects and absorp-
tion, using Blessing’s method as incorporated into the program
SADABS24,25 with 6178 unique data [R(int) ) 0.0197].
3
3
(s, 1H), 3.52 (septet, J ) 6.8 Hz, 1H), 3.11 (septet, J ) 6.8
3
3
Hz, 1H), 1.26 (d, J ) 6.8 Hz, 1H), 1.26 (d, J ) 6.8 Hz, 6H),
3
3
3
1.05 (d, J ) 6.8 Hz, 6H), 0.92 (t, J ) 7.6 Hz, 9H), 0.71 (q, J
) 7.6 Hz, 6H).
iP r NdC(H)-NH(iP r ). To 11.6 g (47.8 mmol) of freshly
distilled PrNdC(H)-NH(iPr)SiEt3 was added 1.8 g (56 mmol)
i
of methanol. After 1 h, the crude product was vacuum distilled
to provide 5.0 g (81%) of a colorless liquid. 1H NMR (400 MHz,
CDCl3, 25 °C) δ 7.35 (s, 1H), 3.42 (septet, 3J ) 6.4 Hz, 2H),
3.09 (br s, 1H), 1.12 (d, 3J ) 6.4 Hz, 12H). 13C NMR (100 MHz,
CDCl3, 25 °C) δ 150.2, 24.9, 7.2, 6.8.
Com p ou n d 3. The same procedure as that for 2b was
t
followed, except iPrNdCH-NH(iPr) was used instead of -
BuNdCH-NH(Et). Yield: 75%. For 3: 1H NMR (400 MHz,
benzene-d6) δ 8.14 (s, 1H), 3.46 (septet, 3J ) 6.8 Hz, 2H), 2.01
(s, 15H), 1.00 (d, 3J ) 6.8 Hz, 12H). 13C NMR (100 MHz,
benzene-d6, 25 °C) δ 159.9, 125.4, 50.9, 24.2, 12.4. Anal. Calcd
for C17H30Cl2N2Zr: C 48.09, H 7.12, N 6.60. Found: C 48.35,
H 7.17, N 6.23.
Com p ou n d s 6b-e. To a solution of 0.30 g (0.71 mmol) of
2b in 20 mL of Et2O, cooled to -78 °C, was added 1.10 mL
(14.3 mmol) of MeLi in Et2O. The mixture was allowed to warm
to room temperature over a period of 1 h and then stirred at
this temperature for an additional hour before being quenching
by the addition of an excess of chlorotrimethylsilane and the
volatiles rwere emoved in vacuo. Extraction of the residue into
pentane and filtration through a thin pad of Celite afforded a
light-yellow solution, which upon concentration and cooling
to -35° C provided off-white crystals of 6b (0.19 g, 70%). For
6b: 1H NMR (400 MHz, benzene-d6) δ 8.25 (s, 1H), 2.91 (q, 3J
All crystallographic calculations were performed on a Per-
sonal computer (PC) with a Pentium 1.80 GHz processor and
512 MB of extended memory. The SHELXTL26 program
package was implemented to determine the probable space
group and set up the initial files. System symmetry, systematic
absences, and intensity statistics indicated the unique centric
monoclinic space group P21/c (no. 14). The 40872 data collected
were merged based upon identical indices yielding 22800 data
[R(int) ) 0.0152], which were further merged during least-
squares refinement to 5952 unique data [R(int) ) 0.0145]. The
structure was determined by direct methods with the success-
ful location of the all non-hydrogen atoms using the program
XS.27 The structure was refined with XL.28 An additional least-
squares difference Fourier cycle was required to locate the
hydrogen atoms. All full-occupancy non-hydrogen atoms were
refined anisotropically. Disorder was modeled within the main
molecule for part of the amidinate ligand and terminal ethyl
group. Hydrogen atoms were allowed to refine freely. A
centroid, C(X), was calculated for the pentamethylcyclodienyl
ligand. The final structure was refined to convergence [∆/σ e
0.001] with R(F) ) 2.76%, wR(F2) ) 6.41%, GOF ) 1.074 for
all 5952 unique reflections [R(F) ) 2.33%, wR(F2) ) 6.11% for
those 5347 data with Fo > 4σ(Fo)]. The final difference Fourier
map was featureless, indicating that the structure is both
correct and complete. The function minimized during the full-
3
) 7.2 Hz, 2H), 1.99 (s, 15H), 1.07 (s, 9H), 0.94 (t, J ) 7.2 Hz,
3H), 0.26 (s, 6H). 13C NMR (100 MHz, benzene-d6, 25 °C) 165.2,
120.5, 46.1, 32.1, 23.0, 18.7, 14.9, 12.3. Anal. Calcd for
C
19H36N2Zr: C 59.46, H 9.48, N 7.30. Found: C 59.20, H 9.33,
N 7.21.
Compounds 6c-e were prepared in a similar manner from
2c, 2d , and 3, respectively.
For 6c: 79% yield.1H NMR (400 MHz, benzene-d6) δ 7.10
(d, 2H), 7.01 (m, 3H), 2.68(q, J ) 7.2 Hz, 2H, CH2CH3), 2.07
(s, 15H, C5Me5), 1.06 (s, 9H, CMe3), 0.81 (t, J ) 7.2 Hz, 3H,
CH2CH3), 0.38 (s, 6H, ZrMe2). Anal. Calcd for C25H40N2Zr: C
5.29, H 8.79, N 6.09. Found: C 65.20, H 8.62, N 6.14.
1
For 6d : 70% yield. H NMR (400 MHz, benzene-d6) δ 3.18
(q, J ) 7.2 Hz, 2H, CH2CH3), 1.99 (s, 15H, C5Me5), 1.35 (s,
3H, CMe3), 1.25 (s, 9H, CMe3), 1.05 (t, J ) 7.2 Hz, 3H,
CH2CH3), 0.08 (s, 6H, ZrMe2). 13C {1H} NMR (benzene-d6) δ
180.1, 119.3, 55.8, 43.7, 40.7, 40.5, 34.4, 31.4, 19.1, 12.1. Anal.
Calcd for C23H44N2Zr: C 62.79, H 10.10, N 6.37. Found: C
62.89, H 10.41, N 6.12.
2
matrix least-squares refinement was ∑w(Fo - Fc2), where w
) 1/[σ2(Fo2) + (0.0328P)2 + 0.6232P] and P ) (max(Fo2,0) +
2Fc2)/3. An empirical correction for extinction was also at-
tempted but found to be negative and therefore not applied.
1
For 6e: 75% yield. H NMR (400 MHz, benzene-d6) δ 8.31
3
(s, 1H), 3.17 (septet, J ) 6.8 Hz, 2H), 1.98 (s, 15H), 1.01 (d,
Ack n ow led gm en t. Funding for this work was pro-
vided by the NSF (CHE-0092493) for which we are
grateful.
3J ) 6.8 Hz, 12H), 0.24 (s, 6H). 13C NMR (100 MHz, benzene-
d6, 25 °C) 161.9, 96.8, 49.0, 43.9, 23.1, 9.9. Anal. Calcd for
C
19H36N2Zr: C 59.47, H 9.48, N 7.30. Found: C 59.33, H 9.39,
N 7.18.
Gen er a l P r oced u r e for P olym er iza tion of 1-Hexen e.
Su p p or tin g In for m a tion Ava ila ble: Details of the struc-
ture determinations and crystallographic data for complexes
1d , 2b-d , 3, 6b, 6d , and 7. This material is available free of
charge via the Internet at http://pubs.acs.org.
To a solution of 20 mg (25 µmol) of [PhNHMe2][B(C6F5)4] in 8
mL of chlorobenzene, cooled to -10 °C, was added all at once
a solution of 11.5 mg (25 µmol) of 6c in 2 mL of chlorobenzene,
also cooled to -10 °C. After 5 min, 0.421 g (5.0 mmol) of
1-hexene, precooled to -10 °C, was added all at once and the
resulting mixture was allowed to stir for 2 h at -10 °C, after
which time it was rapidly quenched by the addition of
methanol. The volatiles were removed in vacuo, and the crude
material was purified through precipitation of a toluene
solution into a large volume of acidic methanol. The final pure
poly(1-hexene) (0.37 g, Mn ) 19 600, PDI ) 1.02) was collected
and dried overnight at 60 °C (0.01 mmHg).
OM049835Z
(24) Blessing, R. H. Acta Crystallogr. 1995, A51, 33-38.
(25) Sheldrick, G. M. SADABS; Siemens Area Detector Absorption
Correction; University of Go¨ttingen: Go¨ttingen, Germany, 1996.
(26) Sheldrick, G. M. SHELXTL/ PC, Version 5.03; Siemens Ana-
lytical X-ray Instruments Inc.: Madison, WI, 1994.
(27) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467-473.
(28) Sheldrick, G. M. Shelxl93, Program for the Refinement of
Crystal Structures; University of Go¨ttingen: Go¨ttingen, Germany,
1993.