Nucleophilically assisted and cationic ring-opening polymerization of tin-bridged [1]ferrocenophanes

Article abstract of DOI:10.1021/ja020206v

To obtain mechanistic insight, detailed studies of the intriguing "spontaneous" ambient temperature ring-opening polymerization (ROP) of tin-bridged [1]ferrocenophanes Fe(η-C₅H₄)₂SnR₂ 3a (R = t-Bu) and 3b (R = M

Full text of DOI:10.1021/ja020206v

Published on Web 08/07/2002
Nucleophilically Assisted and Cationic Ring-Opening
Polymerization of Tin-Bridged [1]Ferrocenophanes
Thomas Baumgartner,† Frieder Ja¨kle,‡ Ron Rulkens, Gernot Zech,
Alan J. Lough, and Ian Manners*
Contribution from the Department of Chemistry, UniVersity of Toronto, 80 St. George Street,
Toronto, Ontario, M5S 3H6, Canada
Received February 11, 2002
Abstract: To obtain mechanistic insight, detailed studies of the intriguing “spontaneous” ambient temperature
ring-opening polymerization (ROP) of tin-bridged [1]ferrocenophanes Fe(η-C5H4)2SnR2 3a (R ) t-Bu) and
3b (R ) Mes) in solution have been performed. The investigations explored the influence of non-nucleophilic
additives such as radicals and radical traps, neutral and anionic nucleophiles, Lewis acids, protic species,
and other cationic electrophiles. Significantly, two novel methodologies and mechanisms for the ROP of
strained [1]ferrocenophanes are proposed based on this study. First, as the addition of amine nucleophiles
such as pyridine was found to strongly accelerate the polymerization rate in solution, a new nucleophilically
assisted ROP methodology was proposed. This operates at ambient temperature in solution even in the
presence of chlorosilanes but, unlike the anionic polymerization of ferrocenophanes, does not involve
cyclopentadienyl anions. Second, the addition of small quantities of the electrophilic species H+ and
Bu3Sn+ was found to lead to a cationic ROP process. These studies suggest that the “spontaneous” ROP
of tin-bridged [1]ferrocenophanes may be a consequence of the presence of spurious, trace quantities of
Lewis basic or acidic impurities. The new ROP mechanisms reported are likely to be of general significance
for the ROP of other metallocenophanes (e.g., for thermal ROP in the melt) and for other metallacycles
containing group 14 elements.
Introduction
However, preliminary evidence based on the lack of ROP
inhibition in the presence of radical traps suggests that hetero-
Over the past decade the thermal,1 anionic,2 and transition
metal-catalyzed3 ring-opening polymerization (ROP) of strained
ring-tilted [1]- and [2]metallocenophanes has become a well-
established route to high molecular weight polymetallocenes
which possess a range of interesting properties.4 The resulting
materials are attracting attention as redox-active gels,5 sensors,6
charge dissipation coatings,7 plasma etching resists,8 self-
assembled materials,9 and precursors to magnetic ceramics over
various length scales.10 Among the various possibilities, thermally
induced ROP is currently the most general synthetic method.
Although evidence for cleavage of the silicon ipso Cp11 carbon
bond in silicon-bridged [1]ferrocenophanes has been presented,
the detailed mechanism of these reactions is still unclear.
(4) For the work of other groups on the synthesis of ring-opened polyferrocenes,
see, for example: (a) Brandt, P. F.; Rauchfuss, T. B. J. Am. Chem. Soc.
1992, 114, 1926. (b) Stanton, C. E.; Lee, T. R.; Grubbs, R. H.; Lewis, N.
S.; Pudelski, J. K.; Callstrom, M. R.; Erickson, M. S.; McLaughlin, M. L.
Macromolecules 1995, 28, 8713. (c) Mizuta, T.; Onishi, M.; Miyoshi, K.
Organometallics 2000, 19, 5005. (d) Compton, D. L.; Brandt, P. F.;
Rauchfuss, T. B.; Rosenbaum, D. F.; Zukoski, C. F. Chem. Mater. 1995,
7, 2342. (e) Heo, R. W.; Somoza, F. B.; Lee, T. R. J. Am. Chem. Soc.
1998, 120, 1621. (f) Brunner, H.; Klankermayer, J.; Zabel, M. J. Organomet.
Chem. 2000, 601, 211. (g) Herberhold, M.; Hertel, F.; Milius, W.;
Wrackmeyer, B. J. Organomet. Chem. 1999, 582, 352. (h) Antipin, M. Y.;
Vorontsov, I. I.; Dubovik, I. I.; Papkov, V.; Cervantes-Lee, F.; Pannell, K.
H. Can. J. Chem. 2000, 78, 1511. (i) Calleja, G.; Carre´, F.; Cerveau, G.;
Labbe´, P.; Coche-Gue´rente, L. Organometallics 2001, 20, 4211. (j) Papkov,
V. S.; Gerasimov, M. V.; Dubovik, I. I.; Sharma S.; Dementiev, V. V.;
Pannell, K. H. Macromolecules 2000, 33, 7107. (k) Buretea, M. A.; Tilley,
T. D. Organometallics 1997, 16, 1507. (l) Barlow, S.; Rohl, A. L. Shi, S.;
Freeman, C. M.; O’Hare, D. J. Am. Chem. Soc. 1996, 118, 7578.
(5) (a) Kulbaba, K.; MacLachlan, M. J.; Evans, C. E. B.; Manners, I.,
Macromol. Chem. Phys. 2001, 202, 1768. (b) Calleja, G.; Cerveau, G.;
Corriu, R. J. P. J. Organomet. Chem. 2001, 621, 46.
* To whom correspondence should be addressed. E-mail: imanners@
chem.utoronto.ca.
(6) Espada, L. I.; Shadaram, M.; Robillard, J.; Pannell, K. H., J. Inorg.
Organomet. Polym. 2000, 10, 169.
† Current address: Institut fu¨r Anorganische Chemie, Johannes Gutenberg
Universita¨t, Duesbergweg 10-14, D-55099 Mainz, Germany.
‡ Current address: Department of Chemistry, Rutgers University, 73
Warren St., Newark, NJ 07102-1811.
(7) Resendes, R.; Berenbaum, A.; Stojevic, G.; Ja¨kle, F.; Bartole, A.; Zamanian,
F.; Dubois, G.; Hersom. C.; Balmain, K.; Manners, I. AdV. Mater. 2000,
12, 327.
(1) (a) Foucher, D. A.; Tang, B. Z.; Manners, I. J. Am. Chem. Soc. 1992, 114,
6246. (b) Resendes, R.; Nelson, J. M.; Fischer, A.; Ja¨kle, F.; Bartole, A.;
Lough, A. J.; Manners. I. J. Am. Chem. Soc. 2001, 123, 2116.
(2) (a) Rulkens, R.; Ni, Y.; Manners, I. J. Am. Chem. Soc. 1994, 116, 12121.
(b) Ni, Y.; Rulkens, R.; Manners, I. J. Am. Chem. Soc. 1996, 118, 4102.
(3) (a) Ni, Y.; Rulkens, R.; Pudelski, J. K.; Manners, I. Macromol. Chem. Rapid
Commun. 1995, 16, 637. (b) Reddy, N. P.; Yamashita, H.; Tanaka, M.
Chem. Commun. 1995, 2263. (c) Go´mez-Elipe, P., Resendes, R., Mac-
donald, P. M., Manners, I. J. Am. Chem. Soc. 1998, 120, 8348. (d) Temple,
K.; Ja¨kle, F.; Sheridan, J. B.; Manners, I. J. Am. Chem. Soc. 2001, 123,
1355.
(8) (a) Cheng, J. Y.; Ross, C. A.; Chan, V. Z.-H.; Thomas, E. L.; Lammertink,
R. G. H.; Vancso, G. J. AdV. Mater. 2001, 13, 1174. (b) Massey, J.; Winnik,
M. A.; Manners, I.; Chan, V. C. H.; Spatz, J. P.; Ostermann, J. M.;
Enchelmaier, R.; Mo¨ller, M. J. Am. Chem. Soc. 2001, 123, 3147.
(9) (a) Massey, J.; Power, K. N.; Winnik, M. A.; Manners, I. AdV. Mater.
1998, 10, 1559. (b) Manners, I. Science 2001, 294, 1664.
(10) (a) MacLachlan, M. J.; Ginzburg, M.; Coombs, N.; Coyle, T. W.; Raju, N.
P.; Greedan, J. E.; Ozin, G. A.; Manners, I. Science 2000, 287, 1460. (b)
Kulbaba, K.; Resendes, R.; Cheng, A.; Bartole, A.; Safa-Sefat, A.; Coombs,
N.; Sto¨ver, H. D. H.; Greedan, J. E.; Ozin, G. A.; Manners, I. AdV. Mater.
2001, 13, 732.
9
10062
J. AM. CHEM. SOC. 2002, 124, 10062-10070
10.1021/ja020206v CCC: $22.00 © 2002 American Chemical Society

Polymerization of Tin-Bridged [1]Ferrocenophanes
A R T I C L E S
lytic processes may be involved in this case.12 The mechanism
of anionic ROP, on the other hand, is well-established. For
silicon-bridged [1]ferrocenophanes (e.g., 1), after initial attack
Table 1. Representative NMR Experiments with 3a (20 mg; 48
µmol; 0.10 M)a
molar ratio
n/n (3a)
approximate
relative rate
additive
1
galvinoxyl
TEMPO
BzSSBz
BHT
0.1
0.1
0.1
0.1
18
1
1
1
1
SiCl4
1
Me3SiCl
TMEDA
pyridine
8
15
26
13/8/1
1
of the nucleophile at silicon a cyclopentadienyl (Cp) anion is
generated. The latter can attack further silicon centers of other
monomer molecules in the propagation step, and the chain ends
of the resulting living anionic polymer can be capped with, for
example, SiMe3 groups by the addition of Me3SiCl (to give 2)
or used to prepare block copolymers.13
Although silicon and germanium-bridged [1]ferrocenophanes
are easily prepared from the reaction of the dilithioferrocene-
TMEDA complex with the appropriate dichloroorganosilane or
-germane,14,15 early attempts to prepare tin-bridged analogues
in a similar manner were frustrated by the formation of cyclic
and linear oligomers.16 In the mid-1990s, tin-bridged [1]ferro-
cenophanes were successfully isolated by using sterically
demanding substituents on tin.17-19 Interestingly, although the
two examples prepared in our group, 3a and 3b, were stable in
30
300
300
pyridine/Me3SiCl
a Rates for ROP are given in comparison to reaction of 3a in neat benzene
(rate ) 1) and are based on 50% conversion. The average molecular weights
and PDI show a broad distribution (Mn ) 10 000-900 000; PDI ) 1.2-
3.5) and vary from sample to sample for runs with the same combination
of reagents. Reactions performed at 25 °C. For further details, see Supporting
Information.
full details20 of our studies on the reactivity of [1]stannaferro-
cenophanes toward radicals, nucleophiles, and electrophiles and
describe the discovery two new and potentially general methods
of polymerization that involve nucleophilically assisted and
cationic mechanisms.
Results
We have previously shown that [1]stannaferrocenophane 3a
polymerizes in benzene or toluene solution at 25 °C, forming
high molecular weight polymer 4a (100% conversion, ∼6 h,
0.1 M solution, Mn > 105). For the sterically more encumbered
analogue, 3b, the ROP is much slower (∼50% conversion, 15
d, 0.1 M, Mn > 105). The rates of polymerization and the exact
molecular weights as well as the polydispersities of the resulting
polyferrocenylstannanes 4a and 4b were found to vary signifi-
cantly from sample to sample and from experiment to experi-
ment. In addition, in some cases, bimodal molecular weight
distributions resulted. However, for both 3a and 3b, high
molecular weight polymer is formed at low conversion, which
is characteristic of a chain growth process where the concentra-
tion of propagating centers is extremely low.18 Because of the
slower rate of ROP for 3b, the detection of rate enhancement
effects caused by additives was expected to be more convenient
with this species. Thus, although many experiments with
additives were performed with both 3a and 3b, most of the
following work involved the latter monomer.
the solid state, each was found to undergo “spontaneous” ROP
in solution at ambient temperature. This afforded the high
molecular weight polyferrocenylstannanes 4a and 4b, together
with, in some cases, small amounts of the corresponding cyclic
dimers 5a and 5b.18 Encouragingly, these reactions appeared
to offer the possibility of convenient mechanistic investigations
of the ROP process, which might also provide much needed
insight into the thermal ROP reactions for strained metallo-
cenophanes that proceed in the melt. In this paper, we report
(11) In this paper, the abbreviation Cp is used to denote both substituted and
unsubstituted cyclopentadienyl ligands.
1. Influence of Non-Nucleophilic Additives on the ROP
of [1]Stannaferrocenophanes 3a and 3b. (a) Influence of
Radical Initiators and Radical Traps. As a radical mechanism
for the ambient temperature ROP of 3a and 3b was suspected,
the effect of a variety of radical traps was investigated under
similar conditions (Table 1 and Table 2). The observation that
neither the traps galvinoxyl, TEMPO, PhCH2SSCH2Ph, BHT,
and 1,4-cyclohexadiene nor UV irradiation (λ ) 365 nm) had
a substantial effect on the rate of polymerization of 3a or 3b
suggested that a homolytic ROP mechanism is unlikely.21
This hypothesis was supported by experiments in which the
influence of stannyl radicals on the ROP of 3b was investigated.
(12) (a) Pudelski, J. K.; Manners, I. J. Am. Chem. Soc. 1995, 117, 7265. (b)
Pudelski, J. K.; Foucher, D. A.; Honeyman, C. H.; Macdonald, P. M.;
Manners, I.; Barlow, S.; O’Hare, D. Macromolecules 1996, 29, 1894.
(13) (a) Massey, J.; Power, K. N.; Manners, I.; Winnik, M. A. J. Am. Chem.
Soc. 1998, 120, 9533. (b) Rulkens, R.; Lough, A. J.; Manners, I.; Lovelace,
S. R.; Grant, C.; Geiger. W. E. J. Am. Chem. Soc. 1996, 118, 12683.
(14) (a) Osborne, A. G.; Whiteley, R. H. J. Organomet. Chem. 1975, 101, C27.
(b) Wrighton, R. S.; Palazzotto, M. C.; Bocarsly, A. B.; Bolts, J. M.; Fischer,
N. L. J. Am. Chem. Soc. 1978, 100, 7264.
(15) (a) Osborne, A. G.; Whiteley, R. H.; Meads, R. E. J. Organomet. Chem.
1980, 193, 345. (b) Foucher, D. A.; Edwards, M.; Burrow, R. A.; Lough,
A. J.; Manners, I. Organometallics 1994, 13, 4959. (c) Castruita, M.;
Cervantes-Lee, F.; Mahmoud, J. S.; Zhang, Y.; Pannell, K. H. J. Organomet.
Chem. 2001, 637-639, 664.
(16) (a) Stoeckli-Evans, H.; Osborne, A. G.; Whiteley, R. H.; J. Organomet.
Chem. 1980, 194, 91. (b) Seyferth, D.; Withers, H. P. Organometallics
1982, 1, 1275. (c) Clearfield, A.; Simmons, C. J.; Withers, H. P.; Seyferth,
D. Inorg. Chim. Acta 1983, 75, 139.
(17) Rulkens, R.; Lough, A. J.; Manners, I. Angew. Chem., Int. Ed. Engl. 1996,
35, 1805.
(20) Communication: Ja¨kle, F.; Rulkens, R.; Zech, G.; Massey, J. A.; Manners,
I. J. Am. Chem. Soc. 2000, 122, 4231.
(21) Homolytic cleavage of stannacycloalkanes with free radical sources has
previously been reported. See: Davies, A. G.; Roberts, B. P.; Tse, M.-W.
J. Chem. Soc., Perkin Trans. 2 1977, 1499.
(18) Ja¨kle, F.; Rulkens, R.; Zech, G.; Foucher, D. A.; Lough, A. J.; Manners,
I. Chem., Eur. J. 1998, 4, 2117.
(19) Sharma, H. K.; Cervantes-Lee, F.; Mahmoud, J. S.; Pannell, K. H.
Organometallics 1999, 18, 399.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10063

A R T I C L E S
Baumgartner et al.
Table 2. Representative NMR Experiments with 3b (20 mg; 37
µmol; 0.07 M)a
in the rate of ROP was detected on addition of amine
nucleophiles.24,25 Thus, on addition of excess pyridine, ROP of
3a (0.1 M in C6D6) was complete after less than 90 s and the
ROP rate for 3b was dramatically increased (in C6D6, 0.07 M,
∼95% conversion after 24 h compared to <3% in a control
experiment without pyridine addition). At 60 °C, strongly
accelerated polymerization of 3b was observed even with an
equimolar (95% conversion in 16 h; Mn ) 10 000, PDI ) 2.3)
or catalytic amount of pyridine (95% conversion in 16 h; Mn )
111 000, PDI ) 3.1). The use of more basic amines such as
4-(dimethylamino)pyridine led to further increases of the ROP
rate (for 3b, ∼95% conversion in 3.5 h), whereas for the more
bulky NEt3, the rate was much slower. It is interesting to note
that the highly sterically encumbered amines 2,6-di-tert-bu-
tylpyridine and 1,4-diazabicyclo[2.2.2]octane (DABCO) show
no accelerating effects at all; ROP is inhibited (Table 2). The
reaction of 3b with 2 equiv of DABCO affords only small
amounts of polymer 4b as observed by 1H NMR (<5%), even
after prolonged reaction times (1 week) at 60 °C.
25 °C
60°C
molar
ratio
n/n (3b)
molar
ratio
n/n (3b)
approx
rel rate
approx
rel rate
additive
1
1
15
15
15
15
AIBN
Bu3SnH
AIBN/Bu3SnH
BHT
1
1
5
0.5/5/1
1
1
TEMPO
Me3SiCl
pyridine
1
11
33
1
1
300
1
0.03
1000
300
4-(dimethylamino)pyridine
NEt3
33
19
1000
1
2,6-di-tert-butylpyridine
1,4-diazabicyclo[2.2.2]octane
pyridine/AIBN
pyridine/TEMPO
pyridine/1,4-cyclohexadiene
pyridine/Me3SiCl
12
2
33/1/1
33/1/1
17/6/1
17/11/1
<1
,1
300
300
300
300
(b) Influence of Amine Nucleophiles in the Presence of
Radical Traps and UV Light. Importantly, the accelerating
effect of amines was also observed in the presence of a
stoichiometric amount of the TEMPO free radical. Thus, a
nucleophilically assisted radical mechanism appeared unlikely
as treatment of 3a with amines in the presence of radical traps
(Table 2) or on irradiation with UV light (λ ) 365 nm) or
sunlight led to no significant change in the rate of ROP.
(c) Influence of Amine Nucleophiles in the Presence of
Me3SiCl as an Electrophilic Trap. If free Cp anions are
generated during the amine-promoted ROP of 3a and 3b, the
addition of electrophiles would be expected to terminate the
polymerization process. However, significantly, even the pres-
ence of a large excess of the silyl halide Me3SiCl, which is
known to cap living chain ends formed in the anionic ROP of
[1]silaferrocenophanes, was found to have no significant effect
on the rate of polymerization of 3a or 3b in the presence of
added pyridine. This important result strongly indicates that Cp
anions are not generated during the amine-catalyzed ROP
process.26
a Rates for ROP are given in comparison to reaction of 3b in neat ben-
zene (rate ) 1) and are based on 50% conversion. The average molecular
weights and PDI show a broad distribution (Mn ) 10 000-500 000; PDI
) 1.3-3.2) and vary from sample to sample for runs with the same
combination of reagents. For further details, see Supporting Information.
Scheme 1. Ring Opening of 3b with Bu3Sn• Radicals
Reaction of 3b with an excess of Bu3SnH and AIBN at 60 °C,
which is known to generate Bu3Sn• radical species,22 afforded
the ring-opened product 6 quantitatively (Scheme 1). This
process presumably involves a radical reaction, with attack of
Bu3Sn• on the ferrocenophane followed by hydrogen abstraction
from the excess Bu3SnH as the formation of 6 was not detected
when 3b was treated with Bu3SnH, alone under the same
conditions. Importantly, treatment of 3b with either an equimolar
amount of AIBN or an excess Bu3SnH at 60 °C did not signif-
icantly influence the ROP rate. Moreover, no increase in the
ROP rate was detected when 3b was treated with a deficiency
(∼15 mol %) of AIBN and excess Bu3SnH at 60 °C.
(d) Influence of Organolithium Reagents. To further inves-
tigate the possible formation and reactivity of free Cp anions
derived from 3b, we attempted to generate an anionic species
via the ring-opening addition of an organolithium reagent.
Reaction of 3b with 10 mol % BuLi in THF followed by the
addition of Me3SiCl, conditions analogous to those used for the
formation of Me3Si-capped polyferrocenylsilanes by the anionic
ROP polymerization of silicon-bridged [1]ferrocenophanes,2,13
(b) Influence of Cyclic Dimers 5a and 5b. An alternative
ROP mechanism involving sequential ring fusion via σ-bond
metathesis processes was dismissed on the basis that ring fusion
of the [1]stannaferrocenophane 3a and 3b with the cyclic dimers
5a and 5b was not detected. Thus, for example, with a 1:1
mixture of 3a and 5a in C6D6, 3a was found to form high
molecular weight polymer 4a without consuming detectable
quantities of 5a. Similarly, a mixture of 3b and 5a afforded
homopolymer 4b and no copolymer, whereas 5a remained
unchanged.23
(24) Other nucleophilically assisted or accelerated reactions with tin compounds
include hydrogenation reactions with tin hydrides (homolytic and heterolytic
reactions), the reaction of tin dihydrides to form tin-tin bonds, scrambling
reactions of Sn2R6 with Sn2R′6, and oxidative cleavage of Sn-C bonds.
See, for example: (a) Vedejs, E.; Duncan, S. M.; Haight, A. R. J. Org.
Chem. 1993, 58, 3046. (b) Clive, D. J. L.; Yang, W. J. Org. Chem. 1995,
60, 2607. (c) Bulten, E. J.; Budding, H. A.; Noltes, J. G. J. Organomet.
Chem. 1970, 22, C5. (d) Yoshida, J.; Izawa, M. J. Am. Chem. Soc. 1997,
119, 9361. (e) Suga, S.; Manabe, T.; Yoshida, J. Chem. Commun. 1999,
1237.
(25) An increased susceptibility of stannacyclopentanes and stannacyclobutanes
toward oligomerization in polar solvents had been noted previously without
mechanistic investigations or discussions. See, for example: (a) Bulten,
E. J.; Budding, H. A. J. Organomet. Chem. 1977, 137, 165. (b) Seetz, J.
W. F. L.; Schat, G.; Akkerman, O. S.; Bickelhaupt, F. J. Am. Chem. Soc.
1983, 105, 3336.
2. Influence of Neutral and Anionic Nucleophiles on the
ROP of [1]Stannaferrocenophanes 3a and 3b. (a) Influence
of Amine Nucleophiles. Next, we investigated the influence
of polar additives on the ROP of 3a and 3b. A dramatic increase
(22) Davies, A. G. Organotin Chemistry; VCH: Weinheim, 1997.
(23) For this reaction, a small amount of pyridine (see vide infra) was added so
an appreciable reaction rate was detected.
(26) In an additional control experiment, Me3SiCl was also found to have no
significant effect on the ROP rate for 3a and 3b in the absence of added
pyridine.
9
10064 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Polymerization of Tin-Bridged [1]Ferrocenophanes
A R T I C L E S
Scheme 2. Ring-Opening Reaction of 3b with MeLi: Formation of 7a and 7b
failed to generate any polyferrocenylstannane by 1H NMR and
GPC.
investigate the addition of Lewis acids, which would be expected
to act as scavengers for nucleophilic impurities.
(a) Addition of Organoaluminum Reagents. When 3b was
treated with excess AlMe3 (hexanes or toluene, 25 °C), however,
the ring-opening addition product 8a was formed as an orange
oil (Scheme 3).27 The 119Sn NMR spectrum of 8a showed a
signal at δ ) -111.6, which is in a region similar to that for
7a (δ ) -103.5) and 7b (δ ) -109.0) but at higher field than
for 3b (δ ) -128.3). In the 27Al NMR spectrum, a broad
resonance was observed for 8a at 152 ppm indicative of a
dimeric triorganoaluminum species.28 Ring-opening of 3b was
further confirmed by the fact that the 13C NMR spectrum showed
no ipso-Cp resonances in the 30-40 ppm range characteristic
of a tin-bridged [1]ferrocenophane. Instead, the 13C NMR
spectrum revealed four signals in the Cp region (δ ) 86.1, 77.3,
75.7, 73.7), of which one was strongly downfield shifted as
expected for the attachment of a Lewis acidic aluminum center
to a Cp ring. A signal at -1.8 ppm was assigned to the methyl
group transferred from the aluminum to the tin center as
supported by the appearance of tin satellites (J ) 356, 367 Hz).
To gain insight into the lack of anionic ROP for 3b under
these conditions, this species was treated with a stoichiometric
amount of MeLi in THF-d8 at -196 °C. On warming to -20
°C, analysis by 119Sn and 1H NMR provided convincing
evidence for the formation of the lithiated ring-opening addition
product 7a (Scheme 2). Thus, the 119Sn NMR spectrum of the
product showed a new singlet resonance at δ ) -103.5
consistent with the formation of a methyltriarylstannane-type
product. Attachment of a Me group to the tin center was further
confirmed by the presence of a new signal at δ ) 0.87 ppm in
the 1H NMR spectrum, which showed satellites due to coupling
to the Sn center (J (H-117/119Sn) ) 53 Hz). In addition, the
1
observation of four signals in the Cp region of the H NMR
spectrum was indicative of the generation of an unsymmetrically
disubstituted ferrocene derivative. Interestingly, two of these
resonances were found at unusually high field (δ ) 3.79, 3.62)
and were assigned to the R- and â-carbons of a lithiated Cp
ring. When the THF solution of 7a was treated with MeOH at
-30 °C, the protonated, monosubstituted ferrocene 7b was
obtained in 61% yield. The 119Sn NMR spectrum of 7b showed
a signal at δ ) -109.0, which is upfield shifted by 5.5 ppm in
comparison to the lithiated precursor 7a. The 1H NMR spectrum
of 7b also provided key characterization, displaying two
pseudotriplet resonances for the substituted Cp ring (δ ) 4.24
and 4.17) and a singlet resonance for an unsubstituted Cp ring
(δ ) 3.94). Significantly, when the reaction mixture containing
7a was not quenched at -30 °C, but was allowed to warm to
room temperature, fast decomposition to a range of uncharac-
terized products occurred. This may be a consequence of
transmetalation reactions between the lithium and tin centers.
The thermal instability of the lithiated ring-opened species 7a
and the inability of this compound to induce anionic ROP of
3b was apparent from the reaction of the latter with 0.1 equiv
of MeLi in THF. Analysis of the products formed on slow
warming from - 78 to 25 °C by 1H NMR showed the generation
of ∼10% of unidentified species in addition to ∼90% of
unreacted 3b. Significantly, no polymer 4b was detected. These
experiments therefore indicate that 3b does not undergo anionic
ROP with organolithium reagents as the intermediates with free
Cp anions, such as 7a, are unstable at room temperature and
do not appear to participate in a ring-opening propagation step
with 3b at the low temperatures where such species are stable.
These results provide additional support for the assertion that
amine-promoted ROP reactions cannot proceed by a mechanism
involving Cp anions as intermediates.
1
A singlet displaying tin satellites was also observed in the H
NMR spectrum at 0.82 ppm (J ) 52 Hz). Importantly, a broad
signal with an intensity of ∼15 protons at -0.32 ppm indicated
that 8a was an adduct with AlMe3, which is consistent with the
structures of previously reported ferrocenylalanes.29
Scheme 3. Ring-Opening Reaction of 3b with AlMe3: Formation
of 8a, 8b, and 8c
When 8a was heated under vacuum at 50 °C, 1 equiv of
AlMe3 was removed according to 1H NMR and 8b was formed
(Scheme 3). Interestingly, the 1H and 13C NMR chemical shifts
for the methyl groups changed significantly (8a, δ ) -0.32/-
5.8; 8b, δ ) -0.38/-4.5). The dimeric nature of 8b was
corroborated by the broad 27Al NMR signal at δ ) 154 ppm,
which is comparable to the signal observed for 8a. In both
species, the Cp ring acts as a bridging ligand between two
aluminum centers bearing additional methyl substituents. The
3. Influence of Lewis Acids on the ROP of [1]Stannafer-
rocenophane 3b. The discovery that amines have a dramatic
effect on the ROP rate for tin-bridged [1]ferrocenophanes
suggests that such species may be responsible for the “spontane-
ous” ROP of the latter at room temperature. Indeed, amines are
likely trace impurities in the monomers formed in their synthesis
in the presence of TMEDA catalyst. This prompted us to
(27) A similar reaction was detected with the more encumbered species AliBu3.
See: Supporting Information.
(28) NMR of Newly Accessible Nuclei; Laszlo, P., Ed.; Academic Press: New
York, 1993; Vol. 2, p 166.
(29) (a) Atwood, J. L.; Shoemaker, A. L. Chem. Commun. 1976, 536. (b) Rogers,
R. D.; Cook, W. J.; Atwood, J. L. Inorg. Chem. 1979, 18, 279.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10065

A R T I C L E S
Baumgartner et al.
Scheme 4. Homopolymerization versus Addition-Oligomerization
in the Reaction of 3b with Me3SnCl
preference of aryl groups over alkyl groups for the bridging
positions in organoaluminum dimers was previously reported
for the analogous species Al2Me5Ph and [PhAlMe2]2.30,31
When 8b was reacted with 2 equiv of pyridine, the mono-
meric adduct 8c was formed. The existence of a monomeric
species was indicated by the 27Al NMR signal at δ ) 167 ppm,
which is similar to the previously reported shift of the adduct
AlMe3‚py (δ ) 167 ppm).28 Due to the reduced Lewis acidity
of the aluminum center in 8c, no strongly downfield shifted Cp
carbon atom was observed by 13C NMR (δ ) 76.7, 74.2, 74.1,
72.1, 71.1, 68.6), and the resonance of the methyl groups bound
to aluminum was high-field shifted (δ ) -8.9).
Figure 1. 1H NMR spectra (in C6D6) after 14 d for reactions of 3b with
varying amounts of Me3SnCl. Only the region for the methyl groups on
the mesityl substituents is shown.
(b) Addition of Tin Halides. Although silyl halides had no
effect on the ROP rate for 3a or 3b (see above), organotin
halides such as Me3SnCl would be expected to be more effective
scavengers for nucleophilic impurities. However, as with the
case of organoaluminum reagents, Me3SnCl was found to react
directly with 3b to afford ring-opening addition products. Thus,
reaction of 1 equiv of Me3SnCl with 3b (25 °C, C6D6) gave
approximately equal amounts of 9a and the dimer 101. The latter
was presumably formed by the further reaction of 9a with 3b
(Scheme 4).32 With a 2:1 ratio of Me3SnCl to 3b, the addition
product 9a was obtained almost exclusively.
nances and two sets of signals for the mesityl groups and in
the mass spectrum of 101 the molecular ion peak was observed
(m/z ) 1282).
When the reaction of Me3SnCl and 3b was performed in ratios
of 1:2-1:10 (C6D6, 14 d), increasing amounts of higher
oligomers 10x (x g 2) were formed (see Figure 1). However,
significantly, unreacted 3b was also detected.33 This sug-
gested that the reactivity of 9a and the oligomers 10x toward
3b is lower than that of Me3SnCl. This would be expected
based on steric effects. When Me3SnCl and 3b were reacted in
C6D6 in a 1:100 ratio under the same conditions, the major
products were polyferrocenylstannane 4b (GPC, Mn > 105)
together with small amounts of 10x (x g 2; GPC, Mn < 103)
and significant quantities of unreacted monomer 3b were de-
tected (Figure 1, top). These results, together with the observa-
tion that the conversion of 3b strongly decreased with decreasing
amounts of Me3SnCl (see Figure 1), suggest that 4b and 10x
are formed by different competing mechanisms in such reac-
tions. The formation of 4b appears typical of the “spontaneous”
ROP of 3b via a chain growth process whereas 10x is formed
by a sluggish ring-opening polycondensation process in which
3b successively reacts with the moderately reactive SnMes2Cl
group of 9a and subsequently 10x. The results of these
experiments therefore indicate that the “spontaneous” ROP
of 3b is unaffected by the Lewis acidic tin centers present in
10x.
Formation of 9a and 101 was confirmed by multinuclear NMR
spectroscopy and by mass spectrometry. The 119Sn NMR
spectrum of 9a showed two signals at -6.2 and -27.9 ppm,
which were assigned to the SnMe3 and SnMes2Cl substituents,
respectively. Formation of a product bearing two different
1
substituents on the Cp rings was apparent in the H NMR
spectrum, which showed four pseudotriplets at δ ) 4.47, 4.42,
4.30, and 4.11. A signal at 0.19 ppm was assigned to the
trimethylstannyl group. The 13C NMR spectrum of 9a shows
four signals for the Cp-H resonances at 75.4 (48 Hz), 73.9
(81 Hz), 72.3 (38 Hz), 71.6 (55 Hz) ppm, two of which show
strong coupling to tin. For the higher homologue 101, additional
signals in the NMR spectra for an internal -Mes2SnfcSnMes2-
fragment would be expected in a region similar to those for
polyferrocenylstannane 4b. Indeed, the 119Sn NMR spectrum
for 101 showed signals at -6.0 and -27.9 ppm for the terminal
stannyl groups and a resonance at -128.1 ppm assigned to the
interior tin center (cf. 9a, δ ) -6.2, -27.9; 4b, δ ) -127.0).
4. Reactivity of [1]Stannaferrocenophane 3b toward Protic
Species. Studies of the reactivity of silicon-bridged [1]ferro-
cenophanes have shown that ring-opening addition processes
occur in the presence of protic agents such as MeOH, H2O,
and HCl.34,35 To explore whether similar chemistry is possible
1
Furthermore, the H NMR spectrum displays eight Cp reso-
(30) Mole, E. A. J.; Saunders: J. K. Chem. Commun. 1967, 697.
(31) Malone, J. F.; McDonald, W. S. J. Chem. Soc., Dalton Trans. 1972, 2649.
(32) Reaction of 3b with 9a results in a mixture of higher oligomers 10x and
polymer 4b (see Supporting Information).
(33) Similar results and reaction rates were observed when the experiments
were performed in CH2Cl2.
9
10066 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Polymerization of Tin-Bridged [1]Ferrocenophanes
A R T I C L E S
with the tin-bridged [1]ferrocenophane system, we performed
a set of reactions that involve protic species.
(a) Reaction with HCl. The reaction of 3b with 0.1 equiv
of HCl (1 M in Et2O) in either C6D6 or CD2Cl2 at ambient
temperature led to chemistry similar to that observed in the case
of Me3SnCl. Thus, HCl reacted directly with 3b to afford the
ring-opened addition product 9b. However, the formation of
oligomeric/polymeric products was very sluggish, presumably
for the same reasons as in the case of Me3SnCl, and complete
reaction required prolonged reaction times and slightly elevated
reaction temperatures (50-60 °C). A more facile formation of
oligomeric/polymeric products (Mn ) 6,450, PDI ) 1.4) was
observed in the case of monomer 3a with 0.1 equiv of HCl (1
M in Et2O) in benzene, which is consistent with the lower kinetic
stability of 3a compared to 3b. In this case, the reaction is
complete after 2 d at room temperature.
On the basis of these and the aforementioned studies with
Me3SnCl, we chose triflate (OTf) as an alternative counteranion
based on the expectation that the resulting SnMes2OTf chain
end should provide a more reactive site for the induction of
ring-opening processes of the type identified in the formation
of oligomers 10x. We anticipated that this might allow poly-
ferrocenylstannanes of appreciable molecular weight to be
rapidly formed.
Figure 2. Molecular Structure of 12‚C6D6.
allowed to slowly warm to ambient temperature. Formation of
(b) Reaction with HOTf. (i) Reaction of 3b with 1 Equiv
of HOTf: Synthesis of Ring-Opening Addition Product 11a
and the Pyridine Adduct 12. When a solution of 3b in
CH2Cl2 was treated with 1 equiv of HOTf at ambient temper-
ature, the reaction mixture immediately turned dark green,
indicating that a significant amount of oxidation of the ferrocene
units had occurred. However, when the addition was performed
at low temperature (-78 °C) and the reaction mixture was
slowly allowed to warm to 25 °C, only a small amount of
oxidation was detected in the form of a slight darkening of the
solution. The ring-opening addition product 11a was subse-
quently isolated as a yellow oil in ∼90% purity (Scheme 5).36
the adduct 12 was confirmed by multinuclear NMR spectroscopy
and a single-crystal X-ray diffraction study (Figure 2, Tables 3
and 4). Complexation of 11a is apparent from a significant
upfield shift of the 119Sn NMR signal (δ ) -66.1). Signals at
1
δ ) 8.42, 6.95, and 6.59 in the H NMR spectrum and at δ )
149.4, 136.4, and 123.8 in the 13C NMR spectrum were assigned
to the coordinated pyridine moiety. Orange crystals of 12‚C6D6
for the X-ray study formed over a period of 2 days at ambient
temperature and were isolated by decanting the solvent.
The crystal structure of 12‚C6D6 confirmed the formation of
a monosubstituted ferrocene derivative via addition of HOTf
across one of the ipso-C-Sn bonds of 3b. Compound 12
represents the first structurally characterized triorganotin triflate
containing an additional donor ligand.37 A distorted trigonal
bipyramidal geometry around the tin center with a triflate and
a pyridine ligand in axial positions is apparent from the N(31)-
Sn(1)-O(1) angle of 172.6°. The angles between axial and
equatorial ligands are in the range of 86.0-99.5°, and the angles
between the equatorial ligands were found to be 116.8, 120.1,
1
This species was characterized by multinuclear NMR. The H
NMR spectrum was consistent with the assigned structure as a
monosubstituted ferrocene derivative, with two triplets at 4.41
and 4.20 ppm and a singlet at 4.13 ppm in the Cp region.
Similarly, the Cp region of the 13C NMR spectrum of 11a
showed two Cp-H signals displaying satellites due to coup-
ling to the tin center (δ ) 74.7, J ) 79 Hz; δ ) 72.4, J ) 60
Hz) and a singlet for an unsubstituted Cp ring (δ ) 69.5). The
119Sn NMR spectrum showed a strongly downfield shifted signal
at δ ) 36.9.
Scheme 5. Ring-Opening Addition Reaction of 3b with ROTfa
(34) (a) Fischer, A. B.; Kinney, R. H.; Staley, R. H.; Wrighton, M. S. J. Am.
Chem. Soc. 1979, 101, 6501. (b) O’Brien, S.; Tudor, J.; Barlow, S.; Drewitt,
M. J.; Heyes, S. J.; O’Hare, D. Chem. Commun. 1997, 641.
(35) (a) MacLachlan, M. J.; Ginzburg, M.; Zheng, J.; Kno¨ll, O.; Lough, A. J.,
Manners, I. New J. Chem. 1998, 22, 1409. (b) MacLachlan, M. J.; Bourke,
S. C.; Lough, A. J.; Manners, I. J. Am. Chem. Soc. 2000, 122, 2126.
(36) We found that formation of 11a can also be achieved at ambient temperature
when 3b and HOTf are reacted in benzene.
(37) For related structures of tin(IV) sulfonates and sulfates, see: (a) Harrison,
P. G.; Phillips, R. C.; Richards, J. A. J. Organomet. Chem. 1976, 114, 47.
(b) Molloy, K. C.; Quill, K.; Cunningham, D.; McArdle, P.; Higgins, T. J.
Chem. Soc., Dalton Trans. 1989, 267. (c) Hiemisch, O.; Henschel, D.; Jones,
P. G.; Blaschette, A. Z. Anorg. Allg. Chem. 1997, 623, 147. (d) Wester-
hausen, M.; Schwarz, W. Main Group Met. Chem. 1997, 20, 351. (e)
Nakazawa, H.; Yamaguchi, Y.; Kawamura, K.; Miyoshi, K. Organome-
tallics 1997, 16, 4626.
a Conditions: 11a: R ) H, -78 ˚C, CH2Cl2. 11b: R ) nBu3Sn. Ambient
Temperature, C6D6. OTf ) OSO2CF3.
As the ring-opening addition product 11a could not be
obtained completely pure, an attempt was made to isolate an
adduct of this species with pyridine for further characterization.
Species 11a was generated from the reaction of 1 equiv of HOTf
and 3b at -78 °C in CH2Cl2; the reaction mixture was
subsequently treated with neat pyridine and the solution was
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10067

A R T I C L E S
Baumgartner et al.
Table 3. Selected Bond Lengths (Å) and Angles (deg) for 12‚C6D6
Sn(1)-C(1)
Sn(1)-C(11)
Sn(1)-C(21)
Sn(1)-N(31)
Sn(1)-O(1)
S(1)-O(1)
2.117(10)
2.135(12)
2.160(13)
2.332(11)
2.340(8)
1.445(9)
1.456(9)
1.421(9)
C(1)-Sn(1)-C(11)
C(1)-Sn(1)-C(21)
C(11)-Sn(1)-C(21)
C(1)-Sn(1)-N(31)
C(11)-Sn(1)-N(31)
C(21)-Sn(1)-N(31)
C(1)-Sn(1)-O(1)
C(11)-Sn(1)-O(1)
116.8(4)
123.0(4)
120.1(4)
88.0(4)
99.5(4)
86.2(4)
86.0(4)
87.1(4)
C(21)-Sn(1)-O(1)
N(31)-Sn(1)-O(1)
O(1)-S(1)-O(2)
O(1)-S(1)-O(3)
O(2)-S(1)-O(3)
Sn(1)-O(1)-S(1)
93.5(4)
172.6(3)
115.1(5)
113.4(5)
116.2(6)
166.6(6)
S(1)-O(2)
S(1)-O(3)
Table 4. Crystal Data and Structure Refinement for 12‚C6D6
Scheme 6. Ring-Opening Addition Reaction of 3b with MeOH
formula
Mr
C40H42F3FeNO3SSn
848.35
T, K
100.0(1)
wavelength, Å
crystal system
space group
a, Å
λ(Mo KR) ) 0.710 73
monoclinic
P21/n
12.989(1)
9.922(1)
b, Å
in the presence of catalytic amounts of HOTf in CH2Cl2,
c, Å
28.220(2)
90
91.463(5)
90
857.7(2)
4
1.550
1.198
1728
0.27 × 0.25 × 0.20
Kappa-CCD
φ, 1
4.11-25.40
0 ) h ) 15, 0 ) k ) 11,
-34 ) l ) 34
11331
6321 (Rint ) 0.098)
empirical, multiscan/
0.7956, - 0.7380
6321/457
1.092
0.0806
0.2218
0.0005(5)
2.504/-0.861
presumably via the reaction of 11a with the monomer.40
R, deg
â, deg
γ, deg
V, Å3
(c) Reaction with MeOH. To establish the generality of this
chemistry, we also investigated the reaction of 3b with the weak
proton source MeOH. However, only when a large excess of
MeOH was added to a solution of 3b in C6D6 at ambient
temperature was slow conversion to the ring-opening addition
Z
F
calc, g cm-3
µ(Mo KR), mm-1
F(000)
1
product 13 detected by H NMR spectroscopy (C6D6, 6 d)
crystal size, mm
diffractometer
scan type, range
θ range, deg
(Scheme 6). Formation of 13 was confirmed by multinuclear
NMR spectroscopy and mass spectrometry. Ring-opening of 3b
is apparent from a significant downfield shift of the 119Sn NMR
signal, which is consistent with the attachment of the electroneg-
ative methoxy substituent (δ ) -35.5). The 1H NMR spectrum
of 13 was also consistent with the assigned structure and showed
two pseudotriplets and a singlet in the Cp region. Furthermore,
signals at δ ) 3.73 (J (H-117/119Sn) ) 42 Hz) in the 1H NMR
limiting indices
reflns collected
independent reflns
abs correction/
min and max transmn coeff
data/parameters
GoF on F2
spectrum and at δ ) 53.9 (J (C-117/119Sn) ) 22 Hz) in the 13
C
R1a (I > 2σ(I))
NMR spectrum showed Sn satellites and were assigned to the
methoxy substituent on tin.
wR2b (all data)
ext coefficient
The sluggish formation of 13 is consistent with the low acidity
of MeOH compared to HCl and triflic acid, and the resulting
SnMes2OMe group is insufficiently reactive for further reaction
with 3b. The latter is corroborated by the reactions of 3b with
smaller amounts of MeOH (0.1 and 0.01 equiv), which show
the generation of the ring-opened product 13 (10 or 1%, respec-
tively), which does not react with the remaining 3b even after
prolonged reactions times (3 weeks) at ambient temperature.
peak/hole (eÅ-3
)
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) {∑[w(Fo2 - Fc ) ]/∑[w(Fo ) ]}1/2
.
2 2
2 2
and 123.0°, respectively. The Sn-N bond length of 2.33(1) Å
is typical for pyridine adducts with tin(IV) centers, which range
from 2.24 to 2.44 Å.38 The Sn-O bond length of 2.340(8)
Å is similar to that in trimethyltin benzenesulfonate hydrate
(2.37(1) Å).37,39 As a result of the bulky substituents, the tin
atom is bent out of the Cp plane by an angle of -11(1)° and
the tilt-angle between the Cp planes was found to be -9(1)°.
(ii) Reaction of 3b with 0.05 Equiv of HOTf. To explore
whether adduct 11a was capable of further reaction with 3b, a
small quantity (0.05 equiv) of HOTf was added dropwise to a
solution of 3b in CH2Cl2 at -78 °C. After 30 min, the solution
was slowly warmed to 25 °C and an amber precipitate of
polyferrocenylstannane 4b formed. The formation of the latter
was confirmed by 1H and 119Sn NMR analysis in C6D6 and GPC
(in THF) indicated that the material was of high molecular
weight (Mn ) 385 000; PDI ) 2.2).
5. Reactivity of [1]Stannaferrocenophane 3b toward other
Electrophilic Cations. (a) Reaction with MeOTf. To further
investigate whether other electrophilic cations are capable of
inducing the ROP of 3b in a manner similar to HOTf, we
explored the reactions of 3b with methyl triflate, a widely used
source of methyl electrophiles.
When an equimolar amount of neat MeOTf was added
dropwise to a solution of 3b in CH2Cl2 at ambient temperature,
a yellow precipitate gradually formed over a period of 1-2 h.
1H and 119Sn NMR analysis in C6D6 indicated the formation of
high molecular weight polymer 4b, which was confirmed by
GPC (Mn ) 116 900, PDI ) 2.4). However, no evidence for
the ring-opening addition product 14 was apparent. Similar
observations were made during the reaction of 0.1 equiv of
MeOTf with 3b in CH2Cl2 at 25 °C, except that in this case,
These results indicate that HOTf reacts at -78 °C with 3b
to give the ring-opened species 11a, which can be isolated as
the pyridine adduct 12. Rapid polymerization of 3b is induced
(38) A CSD search gave eight structures of tin(IV) pyridine adducts.
(39) Harrison, P. G.; Phillips, R. C.; Richards, J. A. J. Organomet. Chem. 1976,
114, 47.
(40) Reaction of 3b with ∼0.1 equiv of 11a also results in the formation of
polymer 4b (see Supporting Information).
9
10068 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Polymerization of Tin-Bridged [1]Ferrocenophanes
A R T I C L E S
Scheme 7. Proposed Mechanism for the Nucleophilically Assisted
ROP of [1]Stannaferrocenophanes
the yellow precipitate of high molecular weight 4b formed more
slowly over a period of 5-6 h (Mn ) 222 700, PDI ) 3.0).
The observation that no ring-opening addition product was
formed despite the fact that polymer 4b formed relatively slowly
suggests that the ROP is induced by the small concentration of
HOTf impurities present in the MeOTf. This assumption is
further supported by the observation that the addition of 1 equiv
of the proton scavenger 2,6-di-tert-butylpyridine slowed the
polymerization process significantly (95% conversion after 10-
12 h). The formation of the polymer 4b even under these
conditions might be explained by the presence of an equilibrium
between free HOTf and the corresponding pyridinium triflate
species in the reaction mixture, which could still allow catalytic
quantities of HOTf to initiate the ROP.
(b) Reaction with nBu3SnOTf. (i) Reaction of 3b with 1
Equiv of nBu3SnOTf. To probe the possibility that a source
of highly electrophilic Sn centers could induce ROP of
tin-bridged [1]ferrocenophanes, we investigated the reactivity
of nBu3SnOTf toward 3b at ambient temperature. The 1:1
reaction between these compounds in C6D6 led to the formation
of ring-opening addition species 11b as the main product
according to NMR analysis of the reaction mixture after 1 d
(Scheme 5). The 1H NMR spectrum of 11b showed four
pseudotriplets at 4.62, 4.56, 4.42, and 4.13 ppm in the Cp region,
which suggested the formation of a disubstituted ferrocene
compound. This feature was also apparent in the Cp region of
the 13C NMR spectrum, which displayed four signals with tin
satellites for the Cp-H resonances (δ ) 75.8, J ) 40 Hz; 74.4,
J ) 79 Hz; 72.6, J ) 60 Hz; δ ) 72.3, J ) 34 Hz). The signals
at 74.4 and 72.6 ppm showed strong coupling to tin and were
therefore assigned to the electron-deficient tin center bearing a
triflate substituent (cf. 1,1′-fc(SnBu3)2, δ ) 74.7, J ) 43 Hz;
71.1, J ) 34 Hz). Formation of a new tin triflate species was
also apparent from the 119Sn NMR spectrum, which showed a
strongly downfield-shifted signal at δ ) 34.7 and a signal at
-20.5 ppm, which was assigned to the tributylstannyl group
by comparison with that for 1,1′-fc(SnBu3)2 (δ ) -17.4).
Furthermore, the 19F NMR spectrum gave rise to a signal at δ
) -77.9, which was assigned to the triflate group.
results for the cases of the ROP of 3b initiated by nBu3SnOTf
and HOTf are clearly indicative of an inhibiting effect for excess
triflate anion. This strongly suggests that a cationic or “cation-
like” tin center functions as the propagating species in the ROP
mechanism.
Discussion
Two New Polymerization Mechanisms for [1]Ferro-
cenophanes: Nucleophilically-Assisted ROP and Cationic
ROP. Our studies of the influence of additives on the
“spontaneous” ROP of tin-bridged [1]ferrocenophanes have led
to the discovery of two new polymerization mechanisms.
The rate enhancement detected for 3a and 3b in the presence
of amines points to a mechanism (Scheme 7), in which amine
coordination to tin increases the nucleophilicity of the Cp carbon
bonded to tin without generating a free anion. In the initiation
step, a small equilibrium concentration of a pentacoordinate tin
species 15 may form.24,41 Although examples of amine coor-
dination to tetraorganotin compounds are still rare, evidence
for the formation of stable pentacoordinate tetraorganostannanes
has been reported.42 Moreover, intramolecular coordination of
amines is known to lead to elongated tin-carbon bonds trans
to the incoming nucleophile, which display enhanced reactiv-
ity.43 The increased nucleophilicity of the Cp carbon bonded to
tin in 15 may allow for attack at the tin center of another
monomer molecule of 3 (Scheme 7). Crucially, heterolytic
cleavage of the Sn-C bond trans to the coordinating amine does
(ii) Reaction of 3b with 0.3 Equiv of nBu3SnOTf. With
the analogous reaction involving 30 mol % nBu3SnOTf in C6D6,
initial formation of 11b was detected followed by ROP to give
polymer 4b over 14 d. Polymer 4b was isolated as a yellow
solid after precipitation into MeOH, and the molecular weight
was found to be high by GPC (Mn ) 58 100, PDI ) 2.2). The
1
analogous process in CH2Cl2 was complete within 3 d by H
NMR (GPC: Mn ) 244 000; PDI ) 2.3). These results suggest
that the reactivity of the SnMes2OTf end group is dependent
on the ability of the solvent to promote ionization to form a
cationic tin center.
(41) (a) We have isolated a 5-coordinate silicon-bridged [1]ferrocenophane with
an intramolecular Si-N interaction; in the solid state, Fe(η-C5H4)2SiMe-
(2-C6H4NMe2) possesses a trigonal bipyramidal Si center with an elongated
Cp-Si bond trans to the amine substituent. Ja¨kle, F.; Vejzovic, E.: Power-
Billard, K. N.; MacLachlan, M. J.; Lough, A. J.; Manners, I. Organome-
tallics 2000, 19, 2826. (b) The formation of 5-coordinate [1]ferrocenophanes
is favored by the contracted Cp-E-Cp bond angle (E ) bridging element)
relative to analogous acyclic species.
(42) See, for example: (a) Tzschach, A.; Jurkschat, K. Pure Appl. Chem. 1986,
58, 639. (b) Jurkschat, K.; Tzschach, A.; Meunier-Piret, J. J. Organomet.
Chem. 1986, 315, 45. (c) Kumar-Das, V. G.; Mun, L. K.; Wei, C.; Blunden,
S. J.; Mak, T. C. W. J. Organomet. Chem. 1987, 322, 163. (d) Kumar-
Das, V. G.; Mun, L. K.; Wei, C.; Mak, T. C. W. Organometallics 1987, 6,
10.
6. Attempted Inhibition of Cationic ROP: Reaction of 3b
with HOTf or nBu3SnOTf in the Presence of Excess [NBu4]-
OTf. If stannyl cations are generated during the polymerization
of 3b with HOTf or nBu3SnOTf, then the presence of a large
concentration of triflate anions would be expected to terminate
or at least hinder the ROP process. Indeed, on treatment of 3b
with catalytic amounts of either HOTf or Bu3SnOTf in the
presence of 100 equiv of [NBu4]OTf no polymer formation was
detected. 1H NMR spectroscopy and GPC revealed the formation
of only oligomeric products in both cases (Mn < 1000). These
(43) See for example: (a) Yoshida, J.; Izawa, M. J. Am. Chem. Soc. 1997, 119,
9361. (b) Steenwinkel, P.; Jastrzebski, J. T. B. H.; Deelman, B.-J.; Grove,
D. M.; Kooijman, H.; Veldman, N.; Smeets, W. J. J.; Spek, A. L.; van
Koten, G. Organometallics 1997, 16, 5486.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10069

A R T I C L E S
Baumgartner et al.
Scheme 8. Proposed Mechanism for the Cationic ROP of
[1]Stannaferrocenophanes
proposed to be the key step. By contrast, the proposed
mechanism for the cationic ROP process discovered for 3a and
3b, the first for a [1]ferrocenophane, appears to involve a
propagating cationic or cation-like tin site.
Although the mechanism for the spontaneous ROP of tin-
bridged [1]ferrocenophanes is still not known with certainty,
the results reported clearly indicate that trace quantities of bases
or perhaps strong protic acids may explain this phenomenon.
For example, minute amounts of amines are likely impurities
in the monomers, which are prepared from dilithioferrocene-
TMEDA complex, and catalytic quantities of acids are possible
as impurities in many solvents and reagents. Indeed, the noted
variation of the rate of ROP and molecular weight from sample
to sample is also consistent with this explanation.
Summary
Our investigations of the reactivity of tin-bridged [1]ferro-
cenophanes toward a variety of reagents including radicals and
radical traps, nucleophiles, and electrophiles have led to the
discovery of two new ROP mechanisms for [1]ferrocenophanes.
One involves nucleophilic assistance by amines and the other
proceeds via a cationic process.
The observed “spontaneous” ambient temperature ROP of
tin-bridged [1]ferrocenophanes in solution can be tentatively
explained in terms of the presence of trace quantities of amine
(or other nucleophilic) impurities or electrophilic species, such
as protons. Significantly, similar impurities might also be
anticipated in other, less labile metallocenophane monomers
such as silicon-bridged [1]ferrocenophanes, and as a conse-
quence, analogous mechanisms need to be considered for the
thermal ROP of these species in the melt as well. Future
experiments will test this possibility and will also explore the
generality of these new ROP methods, which should be
applicable to many other strained rings containing group 14
elements.
not generate a free anion but results in a tin ate complex (16)
as an intermediate. Pentacoordinate stannate complexes are
known to show decreased reactivity toward electrophiles in
comparison to the corresponding lithium carbanions,44 which
may account for the fact that silicon halides have no significant
influence on the polymerization process. The formation of cyclic
dimer 5 can be explained by backbiting and the regeneration
of the amine catalyst and would be expected to occur in
situations of lower monomer concentration.45
The surprising inhibition of the ROP process by the highly
sterically emcumbered, strongly basic amines 2,6-di-tert-bu-
tylpyridine and 1,4-diazabicyclo[2.2.2]octane can be attributed
to two effects. First, the inacessability of the tin center of 3 due
to the steric bulk of the amine nucleophile, which hinders the
formation of key intermediate 15, and second, their ability to
act as scavengers for electrophiles (e.g., protons). The latter
suppresses a possible ROP process through a cationic mecha-
nism induced by, for example, acid impurities present in the
reaction mixture.
Our studies indicate that a cationic mechanism operates in
the case of the ROP of 3b in the presence of small quantities
of HOTf or nBu3SnOTf. An initial rapid reaction of 3b with
these species forms the ring-opened tin triflate species 11a or
11b, which reacts further with 3b via a chain growth ROP
process. The observation that the reaction is faster in more polar
solvents (CH2Cl2 vs C6D6) and the observed inhibition by excess
triflate anions supports the assertion that the propagating tin
center is either cationic or cation-like (Scheme 8). The ROP of
3a in the presence of HCl as initiator is presumed to occur by
an analogous mechanism.
Acknowledgment. We thank the donors of the Petroleum
Research Fund (PRF) administered by the American Chemical
Society, for financial support of this work. We also wish to
acknowledge the DFG for Fellowships for T.B. (research) and
for F.J. (postdoctoral), and the Ontario Government for an
Ontario Graduate Scholarship for R.R. I.M. thanks the NSERC
for an E. W. R. Steacie Fellowship (1997-1999), the University
of Toronto for a McLean Fellowship (1997-2003), the Ontario
Government for a PREA Award (1999-), and the Canadian
Government for a Canada Research Chair (2001-) for the
duration of the work described. In addition we thank S.C.
Bourke, C.A. Jaska, Prof. S. A. Mabury, D. Ellis and Prof. A.
J. Poe¨ for experimental assistance and helpful discussions. We
also thank the reviewers for helpful comments.
We recently found that MeOTf and BF3‚OEt2 induce the ROP
of a [2]carbathiaferrocenophane.1b Methyl cations and (hydro-
lytically derived) protons were shown to be the likely initiators,
respectively, and attack at the sulfur lone pair was tentatively
Supporting Information Available: Full experimental pro-
cedures and 1H, 13C, and 119Sn NMR, MS, and GPC data (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
(44) Reich, H. J.; Phillips, N. H. J. Am. Chem. Soc. 1986, 108, 2102.
(45) In the case of polymer formation, we anticipate that the coordinated amine
and ferrocenophane chain termini remain intact until workup where
hydrolysis would likely generate SnR2OH and SnR2(η-C5H4)Fe(η-C5H5)
end groups.
JA020206V
9
10070 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002