Living organotitanium(IV)-catalyzed polymerizations of isocyanates

Article abstract of DOI:10.1021/ja9534516

An organotitanium(IV) compound, TiCl₃OCH₂CF₃, 1, was found to polymerize n-hexyl isocyanate to high yields and without the formation of cyclic trimer. CpTiCl₂L (L = -OCH₂CF₃, -N(CH₃)₂, -CH₃), 2-4, respectively, likewise polymerized n-hexyl isocyanate but also polymerized isocyanates in the presence of donor solvents and isocyanates possessing donor functional groups, activated olefins, and strained olefins. The activity of the organotitanium(IV) catalysts decreased with increasing steric bulk about the metal center and increasing electron donation to the metal center from the ligands. The polymerization of n-hexyl isocyanate using organotitanium(IV) compounds is living. The PDIs of PHIC synthesized using catalysts 1-4 were found to range from 1.05 to 1.2. The molecular weight of the polymer formed in polymerizations of n-hexyl isocyanate using catalysts 1-4 varied linearly as a function of the monomer-to-initiator ratio and the percent conversion of the polymerization. Polymerizations using 2 can be endcapped quantitatively, and well-defined block copolymers can be synthesized using catalysts 1-4. The kinetics for polymerizations using catalysts 1 and 2 are first-order in both monomer and catalyst (k₁ = 8.5 x 10^-4 mol L^-1 s^-1, k_-1 = 3.8 x 10^-4 s^-1). The active endgroup of a polymerization using 3 was observed using IR spectroscopy, and the frequency of the IR stretch (1548 cm^-1) was consistent with an η²-amidate endgroup structure. Finally, the kinetic data for the polymerization of n-hexyl isocyanate and the known chemistry of CpTiCl₂L compounds were found to be consistent with a propagation step that occurs via a bifunctional activation mechanism.

Full text of DOI:10.1021/ja9534516

1
906
J. Am. Chem. Soc. 1996, 118, 1906-1916
Living Organotitanium(IV)-Catalyzed Polymerizations of
Isocyanates
1
,2
Timothy E. Patten and Bruce M. Novak*
Contribution from the Department of Polymer Science and Engineering and the Materials
Research Science and Engineering Center, UniVersity of Massachusetts, Amherst,
Amherst, Massachusetts 01003, and Department of Chemistry, UniVersity of California, Berkeley,
Berkeley, California 94720
X
ReceiVed October 13, 1995
Abstract: An organotitanium(IV) compound, TiCl3OCH2CF3, 1, was found to polymerize n-hexyl isocyanate to
high yields and without the formation of cyclic trimer. CpTiCl2L (L ) -OCH2CF3, -N(CH3)2, -CH3), 2-4,
respectively, likewise polymerized n-hexyl isocyanate but also polymerized isocyanates in the presence of donor
solvents and isocyanates possessing donor functional groups, activated olefins, and strained olefins. The activity of
the organotitanium(IV) catalysts decreased with increasing steric bulk about the metal center and increasing electron
donation to the metal center from the ligands. The polymerization of n-hexyl isocyanate using organotitanium(IV)
compounds is living. The PDIs of PHIC synthesized using catalysts 1-4 were found to range from 1.05 to 1.2.
The molecular weight of the polymer formed in polymerizations of n-hexyl isocyanate using catalysts 1-4 varied
linearly as a function of the monomer-to-initiator ratio and the percent conversion of the polymerization.
Polymerizations using 2 can be endcapped quantitatively, and well-defined block copolymers can be synthesized
using catalysts 1-4. The kinetics for polymerizations using catalysts 1 and 2 are first-order in both monomer and
-
4
-1 -1
-4 -1
catalyst (k1 ) 8.5 × 10 mol L s , k-1 ) 3.8 × 10 s ). The active endgroup of a polymerization using 3 was
-
1
2
observed using IR spectroscopy, and the frequency of the IR stretch (1548 cm ) was consistent with an η -amidate
endgroup structure. Finally, the kinetic data for the polymerization of n-hexyl isocyanate and the known chemistry
of CpTiCl2L compounds were found to be consistent with a propagation step that occurs via a bifunctional activation
mechanism.
9
10
Introduction
dielectric, mechanically- or electrically-induced birefringence,
1
1
and viscometric measurements, demonstrated that poly(alkyl
isocyanates) are stiff-chain polymers whose hydrodynamic
properties are highly temperature-, solvent-, and molecular-
weight-dependent. At lower molecular weights these polymers
behave as true rigid rods, while at very high molecular weights
they behave as semiflexible chains or “worm-like” polymers.
The dynamic stereochemical properties of polyisocyanates have
also been well-characterized. Because the helix is a chiral
topological form and occurs with either a left-handed (M) or
right-handed (P) screw sense, the backbone of polyisocyanates
is inherently chiral, as observed in the X-ray fiber diffraction
One of the most important classes of polymerizations is the
living polymerization, which is a chain growth polymerization
3
that proceeds in the absence of the kinetics steps of chain
termination and chain transfer. In other words, during the time
scale of the polymerization the only reaction pathway available
to the active chain end is propagation. Living polymerizations
are highly valued, because the molecular weight of the polymer
formed can be established simply by varying the monomer-to-
initiator ratio and because the molecular weight distribution of
the resulting sample is very narrow. These types of polymeriza-
tions also permit the synthesis of well-defined block copolymers,
since after completion of the polymerization the active chain
ends persist until terminated or until an amount of the same or
5
study of PBIC. In several studies, remarkably large optical
activities, associated with the polymer backbone chromophore,
have been observed for polyisocyanate homo- and copolymers
4
another monomer is added to the polymerization medium.
1
2
of optically active isocyanates.
Polyisocyanates are an unusual class of macromolecules,
because they are one of the few types of synthetic polymers
that adopt a stable helical conformation in solution as well as
the solid state. Fiber diffraction experiments demonstrated that
poly(n-butyl isocyanate) [PBIC] adopts an 8/3 helical conforma-
In 1960, Shashoua and co-workers reported the anionic
polymerization of a variety of isocyanates using NaCN as the
initiator and polar solvents, such as dimethylformamide (eq 1).13
(6) (a) Schneider, N. S.; Furusaki, S.; Lenz, R. W. J. Polym. Sci., Part
5
A: Polym. Chem. 1965, 3, 933. (b) Murakami, H.; Norisuye, T.; Fujita, H.
Macromolecules 1980, 13, 346. (c) Fetters, L. J.; Yu, H. Macromolecules
tion in the solid state. Static characterization methods, such
6
7
as light-scattering, osmotic pressure and electric dichroism
1
971, 4, 385.
(7) Reference 6c.
8
measurements, and dynamic characterization methods, such as
(8) (a) Troxell, T. C.; Scheraga, H. A. Macromolecules 1971, 4, 519.
X
Abstract published in AdVance ACS Abstracts, February 1, 1996.
(b) Troxell, T. C.; Scheraga, H. A. Macromolecules 1971, 4, 528. (c)
Milstien, J. B.; Charney, E. Macromolecules 1969, 2, 678.
(9) (a) Yu, H.; Bur, A. J.; Fetters, L. J. J. Chem. Phys. 1966, 44, 2568.
(b) Bur, A. J.; Roberts, D. E. J. Chem. Phys. 1969, 51, 406. (c) Bur, A. J.
J. Chem. Phys. 1970, 52, 3813.
(1) Current address: Department of Chemistry, Carnegie Mellon Uni-
versity, 4400 Fifth Avenue, Pittsburgh, PA 15213.
(
2) University of Massachusetts, Amherst.
(3) (a) Flory, P. J. J. Am. Chem. Soc. 1940, 62, 1561. (b) Szwarc, N.;
Levy, M.; Milkovich, R. J. Am. Chem. Soc. 1956, 78, 2657. (c) Szwarc,
M. Nature 1956, 178, 1168. (d) Brown, W. B.; Szwarc, M. J. Chem. Phys.
(10) (a) Tsvetkov, V. N.; Shtennikova, I. N.; Rjumtsev, E. I.; Getman-
chuk, Y. P. Eur. Polym. J. 1971, 7, 805. (b) Jennings, B. R.; Brown, B. L.
Eur. Polym. J. 1971, 7, 805.
1
957, 27, 416. (d) Gold, L. J. Chem. Phys. 1958, 28, 91.
(
(
4) Webster, O. W. Science 1991, 251, 887 and references therein.
5) Shmueli, U.; Traub, W.; Rosenheck, K. J. Polym. Sci., Polym. Phys.
(11) (a) Bur, A. J.; Fetters, L. J. Macromolecules 1973, 6, 874. (b)
Rodriguez, R. T.; Romo, A. U.; Olayo, R. G. J. Macromol. Sci., Macromol.
Phys. 1991, B30, 75.
1
969, 7, 515.
0
002-7863/96/1518-1906$12.00/0 © 1996 American Chemical Society

Organotitanium(IV)-Catalyzed Polymerizations of Isocyanates
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1907
biting and spirotetramerization, significantly obstruct molecular
weight control in the anionic polymerization of isocyanates.
Furthermore, another side reaction, cyclic trimerization, can
degrade the active polymer chains when polymerizations are
conducted above a certain maximum temperature.
The authors examined polymerization temperatures ranging from
Some work has been done on alternative methods (cationic,
free-radical, electrochemical, irradiative, and a few transition
metal) of polymerizing isocyanates with little success.19 Re-
cently we communicated the use of organotitanium(IV) com-
pounds as catalysts for the living polymerization of isocyan-
-
20 °C to -100 °C and found that the optimum temperature
for polymerization to proceed in high yield was between -50
and -70 °C. Initiation of the anionic polymerization of
isocyanates is proposed to occur by attack of the cyanide anion
onto the electrophillic carbon of the isocyanate group to form
an amidate anion. This anion acts as the propagating species
and can add further isocyanate monomers or be terminated by
added or adventitious protic sources.
20
ates. In contrast to the anionic polymerization of isocyanates,
these polymerizations exhibit excellent control over the molec-
ular weight and polydispersity of the resulting polymer. We
report here a complete account of the living polymerization of
isocyanates using TiCl3OCH2CF3, 1, and CpTiCl2L (L )
The anionic polymerization of isocyanates is not living.
Although no individual, comprehensive study of molecular
weight control in anionic isocyanate polymerizations has been
reported, experiments from several published reports can be
assembled to demonstrate a lack of well-controlled, living
behavior. Various authors have published results on the control
of molecular weight as a function of the monomer-to-initiator
ratio in anionic isocyanate polymerizations. Both Owadh et
-
OCH2CF3, -N(CH3)2, -CH3), 2-4, respectively, as catalysts
including a complete assessment of the living behavior of the
polymerizations along with kinetic and mechanistic investiga-
tions.
1
4
13
al. and Shashoua et al. observed that the molecular weight
of anionically polymerized n-alkyl isocyanates showed little or
no dependence upon the monomer-to-initiator ratio. By care-
fully performing n-butyl isocyanate polymerizations with high
monomer-to-initiator ratios, Ahmed et al. did observe a relation-
ship between the molecular weight (Mv) of the polymer and
the monomer-to-initiator ratio.15 However, the reported absolute
molecular weights, on average, were greater than the expected
molecular weights by an order of magnitude, and the observed
relationship between Mv and the monomer-to-initiator ratio was
curved and not linear. Natta and co-workers observed that the
Results and Discussion
Overview of the Polymerizations. When n-hexyl isocyanate
is added to a concentrated solution of catalyst 1, an orange
solution is formed which then increases in viscosity until it
eventually solidifies. A waxy, colorless solid is isolated upon
workup, typically in 85-95% yields (eq 2). Cyclic trimer has
inherent viscosity of anionic PBIC increased with increasing
time and percent yield;16 however, a plot of the authors’ data
as molecular weight (Mv) versus the percent yield of the
polymerization reveals that the linear-like increase in molecular
weight did not begin until the polymerization was at about 20%
completion or greater. Moreover, the experimental molecular
weights were too high by one-to-two orders of magnitude. The
polydispersities [PDIs] of anionically polymerized n-butyl
isocyanate and n-hexyl isocyanate were reported in only a few
instances. Fetters and Yu measured Mn (membrane osmometry)
and Mw (static light-scattering) values for a wide range of PHIC
and PBIC samples prepared using alkyllithium and NaCN
initiators and found that the PDIs varied from 1.1-9.5, with
the average value being 3.7.17 Okamoto et al., using a gel
permeation chromatography [GPC] apparatus calibrated with
their own fractionated poly(n-hexyl isocyanate) [PHIC] stan-
dards, found that the PDIs of their PHIC samples prepared using
NaCN initiator varied from 1.2 to 1.8.18 These results indicate
that side reactions early in the polymerization, such as end-
not been isolated from or detected in these polymerizations.
1
13
1
GPC, elemental analysis, H NMR spectroscopy, and C{ H}
NMR spectroscopy together confirm the formulation of this
product as PHIC. In a series of polymerizations, a plot of the
molecular weight of the polymer as a function of the monomer-
to-initiator ratio can be fitted with a straight line with a slope
near that of the molecular weight of the monomer repeat unit
(Figure 1). When polymer samples are taken from a polym-
erization at various times, a plot of the samples’ molecular
weights as a function of percent conversion can be fitted with
2
0
a straight line. The polydispersities of these polymerizations
typically fall within the range of 1.1-1.2. The bis-THF
complex of 1, TiCl3(OCH2CF3)(THF)2, behaves identically as
1 in isocyanate polymerizations.
(12) (a) Goodman, M.; Chen, S. Macromolecules 1970, 3, 398. (b)
Goodman, M.; Chen, S. Macromolecules 1971, 4, 625. (c) Green, M. M.;
Lifson, S.; Teramoto, A. Chirality 1991, 3, 285. (d) Green, M. M.; Andreola,
C.; Mu n˜ oz, B.; Reidy, M. P.; Zero, K. J. Am. Chem. Soc. 1988, 110, 4063.
(e) Lifson, S.; Andreola, C.; Peterson, N. C.; Green, M. M. J. Am. Chem.
Soc. 1989, 111, 8850. (f) Green, M. M.; Reidy, M. P.; Johnson, R. J.;
Darling, G.; O’Leary, D. J.; Wilson, G. J. Am. Chem. Soc. 1989, 111, 6452.
(19) (a) See ref 6b for free-radical and cationic initiation attempts. (b)
Sobue, Y.; Tabata, M.; Hiraoka, M.; Oshima, K. J. Polym. Sci., Polym.
Lett. 1963, 2, 943. (c) Suguwara, I.; Marchal, E.; Kudoi, H.; Tabata, Y.;
Oshima, K. J. Macromol. Sci., Macromol. Chem. 1974, 8, 995. (d)
Suguwara, I.; Marchal, E.; Kudoi, H.; Tabata, Y.; Oshima, K. Radiat. Res.
1974, 59, 199. (e) Tidswell, B. M.; Gilchrist, S. C. Polymer 1980, 21, 805.
(f) Tidswell, B. M.; Gilchrist, S. C. Polymer 1980, 21, 812. (g) Kashiwagi,
T.; Hidai, M.; Uchida, Y.; Misono, A. J. Polym. Sci., Polym. Lett. 1970, 8,
173. (h) Graham, J. C.; Xu, X.; Orticochea, M. Catal. Lett. 1989, 3, 413.
(i) Graham, J. C.; Xu, X.; Jones, L.; Orticochea, M. J. Polym. Sci., Polym.
Chem. 1990, 28, 1179. (j) Yilmaz, O.; Usanmaz, A.; Aly u¨ r u¨ k, K. J. Polym.
Sci., Polym. Lett. 1990, 28, 341.
(
13) (a) Shashoua, V. E. J. Am. Chem. Soc. 1959, 81, 3156. (b) Shashoua,
V. E.; Sweeny, W.; Tietz, R. J. Am. Chem. Soc. 1960, 82, 866.
14) Owadh, A. A.; Parsons, I. W.; Hay, J. N.; Haward, R. N. Polymer
978, 19, 386.
15) Ahmed, M. S.; Parsons, I. W.; Haward, R. N. J. Polym. Sci., Polym.
Chem. 1980, 18, 371.
16) Natta, G.; DiPietro, J.; Cambini, M. Makromol. Chem. 1962, 56,
00.
(
1
(
(
2
(
(
17) Reference 6c.
18) Okamoto, Y.; Nagamura, Y.; Hatada, K.; Khatri, C.; Green, M. M.
Macromolecules, 1992, 25, 5536.
(20) Patten, T. E.; Novak, B. M. J. Am. Chem. Soc. 1991, 113, 5065.

1
908 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Patten and NoVak
Table 1. Data for the Oligomerization of n-Hexyl Isocyanate
These polymerizations are necessarily conducted in either
Using Catalysts 5 and 6
bulk or concentrated solutions due to the strong dependence of
the yield of the polymerization upon the initial concentration
of the monomer. When a polymerization of n-hexyl isocyanate
was monitored using IR spectroscopy, the monomer concentra-
tion decreased throughout the course of the polymerization, but
the concentration asymptotically approached a nonzero value
oligomer
endgroup
contribution
(deg)
c
d
yield
(%)
[R]
D
catalyst
MWb
PDIb
(deg)
5
6
60a
91
4850
3340
1.19
1.16
33
21
0.056
0.018
(Figure 2). The aforementioned molecular weight control
a
Catalyst had a limited solubility in the polymerization medium.
b
c
experiments preclude the possibility that this nonzero limit is
indicative of termination processes. When more monomer was
added to a polymerization in this state, propagation of the
polymerization ensued until most of the added monomer was
consumed. A comparison of GPC chromatograms of the
polymer before and after this addition revealed that the
additional monomer was incorporated into existing polymer
chains and that no new polymer chains had formed. The GPC
peak for the polymer after the addition was shifted into a higher
molecular weight region that was well-resolved from the initial
region, and there was no signal corresponding to the molecular
weight of the polymer before the addition. Furthermore, when
the polymer from such polymerizations was isolated with the
active endgroup intact and redissolved in new solvent, the
monomer concentration in the solution increased over time due
to depolymerization (Figure 3). The limit that the depolymer-
ization curve approached was essentially the identical concentra-
tion observed for the polymerization curve. Thus, the value
that the curve in Figure 2 approached was the equilibrium
monomer concentration for the polymerization, and the polym-
erization of n-hexyl isocyanate using organotitanium(IV) cata-
lysts is fully reversible between monomer and polymer.
Determined by GPC calibrated relative to polystyrene standards. In
d
3
CHCl . Optical rotation expected if all of the initiating ligand were
present as the free alcohol.
catalyst decomposes to a brown suspension), and isocyanates
with enolizeable protons. As with polymerizations using 1,
polymerizations using catalysts 2-4 display a linear cor-
respondence between molecular weight and the percent conver-
2
3
sion and monomer-to-initiator ratio (Molecular weight plots,
as shown in Figure 1 for catalyst 1, were obtained for the
polymerization of n-hexyl isocyanate using catalysts 2-4). The
polydispersities of PHIC synthesized using catalysts 2-4 are
typically lower than those using 1 (1.05-1.15 vs 1.1-1.2).
With both classes of catalysts, the initiating ligand subse-
quently can be observed on the end of the polymer chain. For
polymerizations using catalysts 1 and 2, the 2,2,2-trifluoroethoxy
1
9
1
endgroups are observed in the F{ H} NMR spectra of the
resulting polymer samples (δ -74.8 ppm), with the intensity
of the fluorine NMR signal being proportional to the molecular
weight of the polymer. In addition to using fluorine labeling
to demonstrate the presence of the initiating ligand on the end
of the polymer chain, we labeled the initiating ligand with a
chiral center. CpTiCl3OR derivatives with two chiral alcohols,
The occurrence of a monomer-polymer equilibrium explains
the strong dependence of yield upon the initial monomer
concentration of these polymerizations, because as the concen-
tration of monomer approaches the equilibrium concentration
for the polymerization, the amount of monomer that can be
incorporated into polymer diminishes to zero. If the initial
concentration of monomer is equal to the equilibrium monomer
concentration, then no polymerization will occur. Several
authors have suggested that polyisocyanates have a low ceiling
temperature, but ceiling temperature effects have never been
observed directly in anionic polymerizations due to the pre-
ponderance of trimerization at higher reaction temperatures.
Because trimerization does not occur in the organotitanium-
(
(
R)-(+)-R-methyl-2-naphthalene methanol, 5, and (1S,2R,5S)-
+)-menthol, 6, were prepared and used to oligomerize n-hexyl
isocyanate. As shown in Table 1, the resulting oligomers were
2
1
obtained in high yields, with narrow molecular weight distribu-
tions, and with sizable optical rotations. The optical rotations
were in excess of what would be expected if all of the initiating
alkoxide ligand were present as the free alcohol. Therefore,
the optical rotations must be due to the initiating ligand, as an
endgroup, exerting a stereochemical bias upon the conformation
of the oligomer chain. Okamoto and co-workers recently
demonstrated that optically active amido endgroups had a far
greater stereogenic influence upon the conformation of the
polymer chain than alkoxide endgroups. Because monoamido
derivatives of CpTiCl3 are as equally easy to prepare as
monoalkoxide derivatives, the use of CpTiCl2L catalysts
represents an efficient method for the preparation of well-
defined, end-functionalized, optically active isocyanate oligo-
mers.
In fact, a polyisocyanate chain can be end-functionalized with
just about any molecular structure imaginable, with the only
stipulation being that the molecule contains no functional groups
that will react with the titanium center. In addition to the
aforementioned chiral alkoxides, we prepared CpTiCl3 deriva-
tives of two polymerizable alcohols, methyl R-(hydroxymethyl)-
(IV)-catalyzed polymerizations, a ceiling temperature for poly-
isocyanates could be observed directly for the first time. The
reversible nature of the polymerizations allowed us to measure
the thermodynamic polymerization parameters for PHIC: ∆Hp
-
1
20
)
-8.5 ( 0.5 kcal mol ; ∆Sp ) -27 ( 1 eu; Tc ) 43 °C.
The Lewis acidity of the titanium center can be lessened by
2
4
the replacement of a chloride ligand with a more electron
5
donating and sterically bulkier ligand such as η -cyclopentadi-
enyl (Cp). Polymerizations of n-hexyl isocyanate using catalysts
2-4 proceed in an analogous manner to those using 1, although
they are noticeably slower. The nature of the initiating ligand
on the titanium center in catalysts 2-4 generally had no
discernible effect upon its activity. The notable exception is
that titanium phenoxides will not initiate polymerization. These
catalysts, unlike 1, can polymerize monomers possessing donor
2
2
functional groups, such as 2-isocyanatoethyl methacrylate. In
general, we have found that catalysts 2-4 will polymerize most
isocyanates, the exceptions being sterically bulky isocyanates
(secondary and tertiary isocyanates), aryl isocyanates (the
(
21) (a) Reference 13. (b) Eromosele, I. C.; Pepper, D. C. J. Polym. Sci.,
Polym. Chem. 1987, 25, 3499. (c) Ivin K. J. Angew. Chem., Int. Ed. Engl.
973, 12, 487.
22) Patten, T. E.; Novak, B. M. Macromolecules 1993, 26, 436.
(23) Hoff, S. M.; Novak, B. M. Macromolecules 1993, 26, 4067.
(24) Okamoto, Y.; Matsuda, M.; Nakano, T.; Yashima, E. Polym. J.
(Tokyo) 1993, 25, 391.
1
(

Organotitanium(IV)-Catalyzed Polymerizations of Isocyanates
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1909
acrylate, 7, and endo/exo-5-norbornene-2-methanol, 8. Both of
these compounds are active catalysts for the polymerization of
isocyanates and have been used to synthesize PHIC mac-
romonomers. These titanium compounds might also be stitched
together to form a macromolecule with pendant catalyst centers.
Polyisocyanate chains could then be grown from the polymer
to any desired length.
Living Character of the Polymerizations. The microscopic
(i.e., kinetic) behavior of a living polymerization is manifested
in measurable, macroscopic characteristics of the polymerization
and resulting polymer sample.3 These include the following:
25
(
1) qualitative conversion of the monomer into polymer with a
Figure 1. Plots of molecular weight, determined using various
techniques, versus the monomer-to-initiator ratio for the polymerization
of n-hexyl isocyanate using 1. Legend: (A) Determined using GPC
calibrated relative to polystyrene standards (since PHIC is a rigid
polymer, it is expected that these molecular weight values will be
inflated relative to the absolute molecular weights). (B) Determined
using Ubbelhode viscometry using the Mark-Houwink constants
measured by Teremoto, Aharoni, and co-workers in ref 44. (C)
Determined using static light-scattering. (D) Determined using Ubbel-
hode viscometry using the Mark-Houwink constants measured by
Berger and Tidswell in ref 45. (E) Expected molecular weights.
first order dependence of the rate upon both catalyst and
monomer concentration; (2) a linear correlation between the
molecular weight of the polymer and the percent conversion of
the polymerization; (3) a linear correlation between the molec-
ular weight of the polymer and the monomer-to-initiator ratio
of the polymerization; (4) the resulting polymer has a narrow
polydispersity; (5) the living chain ends can be quantitatively
functionalized; and (6) the living chain ends can be employed
in the synthesis of block copolymers. Demonstration of one
of these criteria, by itself, is not sufficient to establish the living
character of the polymerization, because each criterion has its
own particular caveats. However, demonstration of the group
as a whole represents a strong case for a living polymerization.
For the organotitanium(IV)-catalyzed polymerization of iso-
cyanates see the following:
(1) The polymerizations are quantitative because they proceed
until they reach the equilibrium concentration of monomer.
When polymerizations are conducted in bulk, isolated yields
as high as 95% have been obtained. Also, the rate of
polymerization displays a first order dependence upon both the
catalyst concentration and monomer concentration (Vide infra).
(2) The molecular weight of the growing polymer chains in
polymerizations using catalysts 1-4 is linearly dependent upon
the percent conversion of the polymerizations.
(3) As shown in Figures 1-4, the molecular weight of the
isolated polymer is linearly dependent upon the monomer-to-
initiator ratio.
(
4) The polydispersity of polymers synthesized using catalysts
-4 range from 1.05 to 1.2.
5) Polymerizations using catalysts 2-4 can be endcapped
Figure 2. A plot of monomer concentration versus time for the
polymerization of n-hexyl isocyanate using 1.
1
(
quantitatively using carboxylic acid anhydrides. For example,
when the initiating end of PHIC is fluorine labeled using a
catalyst with a fluorine-containing initiating ligand (such as 2)
and then the polymerization is treated with trifluoroacetic
anhydride, the F{ H} NMR spectrum of the resulting isolated,
purified polymer shows two signals in a one-to-one integration
ratio. One signal, with a chemical shift of -74.8 ppm,
corresponds to that found for the 2,2,2-trifluoroethoxy initiating
(
6) We have shown in GPC experiments that addition of a
different isocyanate to a living polyisocyanate increases the
molecular weight of the final copolymer relative to the
intermediate polymer and that the intermediate homopolymer
19
1
2
0
is not observed in the final copolymer.
We have also
synthesized and characterized well-defined block copolymers
between monoisocyanates and diisocyanates (Table 2).27
Kinetics and Mechanism of Titanium(IV)-Catalyzed Po-
lymerizations of Isocyanates. Because these polymerizations
are well-behaved and the polymerization technique, itself, is
novel, we investigated the kinetics and mechanism of organ-
otitanium(IV)-catalyzed isocyanate polymerizations. For these
studies we chose to work primarily with catalysts 2-4, because
ligand, and the other, with a chemical shift of -70.5 ppm,
corresponds to the trifluoroacetamido endcap.26
(25) For recent discussion of experimental criteria for living polymeriza-
tions, see: (a) Matyjaszewski, K. Macromolecules 1993, 26, 1787. (b)
Penczek, S.; Kubisa, P.; Szymanski, R. Makromol. Chem., Rapid Commun.
1
991, 12, 77. (c) Quirk, R. P.; Lee, B. Polym. Int. 1992, 27, 359.
26) Recently, Hatada and co-workers investigated endcapping reactions
(
of n-butyl isocyanate polymerizations where 1 was used as the catalyst:
Ute, K.; Asai, T.; Fukunishi, Y.; Hatada, K. Polym. J. (Tokyo) 1995, 27,
(27) A complete account of the synthesis and polymerization of 1,2-
diisocyanates will appear elsewhere: Patten, T. E.; Novak, B. M. Manuscript
in preparation.
4
45.

1
910 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Patten and NoVak
(
eq 3). Similar insertions of isocyanates into metal-ligand
bonds of small molecule titanium(IV) and zirconium(IV)
complexes have been observed as well.
2
8,29
An IR spectrum of an active oligomerization of n-hexyl
isocyanate using catalyst 3 showed, in addition to other peaks,
-
1
a medium-intensity band at 1548 cm
(Figure 4). The
monomer, starting catalyst, solvent, and the resulting polymer
showed no absorptions in the region in which this band appears,
-
1
1
684-1468 cm , so this band was due to a new species formed
in the polymerization. Upon addition of methanol to the solution
this band disappeared, and a new band at 1524 cm appeared
Figure 4). The intensites of the 1548 cm band and the 1524
cm band varied inversely with the initial monomer-to-initiator
ratio of the polymerization. From this evidence we concluded
that these IR absorptions were due to the active and terminated
endgroups of the polymerization.
Figure 3. A plot of monomer concentration versus time for the
depolymerization of active PHIC polymerized using 1.
-
1
-
1
(
-1
Confirmation of this assignment and information concerning
the structure of the endgroup was obtained from literature reports
-
1
of titanium-isocyanate insertion compounds. The 1548 cm
band corresponds to a region of IR stretching frequencies found
for titanium amidates. For example, the IR spectrum of the
tetrainsertion compound of tetrakisdimethylamidotitanium with
29a
-1
phenyl isocyanate shows an absorption at 1555 cm assigned
to the stretching frequency of the O-C-N unit of the
2
η -ureidate ligands. The IR spectra of the monoinsertion
3
29d
29c
derivatives of Cp2Ti(η -allyl) and Cp2TiCH3 with phenyl
-1
isocyanate showed absorptions at 1536 and 1550 cm , respec-
tively, assigned to the vibrations of the O-C-N unit of the
2
η -amidate ligand. With each of these three compounds, the
ligands could be protonated off the titanium center using H2O
or an alcohol to give essentially quantitative yields of the
corresponding urea or amide derivatives. Because the IR stretch
observed for the active oligomer is very similar to the data
reported for these literature compounds, we believe that the bis-
5
2
Figure 4. Regions (1700-1400 cm-1) of the IR spectra taken of the
chloro-η -cyclopentadienyltitanium(IV)-η -amidato structure, as
shown in eq 4, is an accurate representation of the structure of
the active propagating endgroup in these polymerizations.
active oligomerization of n-hexyl isocyanate using 3. Left: before
3 3
quenching with CH OH. Right: after quenching with CH OH.
Table 2. Experimental Data for the Block Copolymerization of
n-Hexyl Isocyanate with 1,2-Diisocyanatopropane Using Catalyst 3
after addition of
after addition of
n-hexyl isocyanate 1,2-diisocyanatopropane
characterization
(GPC)
/M (GPC)
to compd 3
to the active PHIC blockb
M
M
n
6790
1.1
yes
9960
1.2
no
w
n
The 1524 cm- band of the quenched oligomer is within the
1
a
c
solubility in hexanes
a
region normally found for the amide II band of secondary
solubility in CHCl
3
yes
yes
-
1 30
amides in dilute solution, 1550-1510 cm . For example,
IR spectra of dilute chloroform solutions of N-methyl acetamide
and trimethylurea showed amide II bands at 1534 and 1539
a
3
PHIC is soluble in hexanes and CHCl , whereas poly(1,2-
b
diisocyanatopropane) is insoluble in both of these solvents. Neither
a shoulder nor a peak corresponding to the molecular weight of the
PHIC block was observed in the GPC chromatogram of the final
product. The molecular weight distribution was unimodal. Extraction
of the final block copolymer with hexanes yielded no PHIC.
-
1
cm , respectively. Thus, termination of the polymerization,
c
(
28) Gambarotta, S.; Strologo, S.; Floriani, C.; Chiesi-Villa, A.; Guastini,
C. Inorg. Chem. 1985, 24, 654.
29) (a) Chandra, G.; Jenkins, A. D.; Lappert, M. F.; Srivastava, R. C.
J. Chem. Soc. (A) 1970, 2550. (b) Klei, E.; Teuben, J. H. J. Organomet.
Chem. 1981, 222, 79. (c) Klei, E.; Telgen, J. H.; Teuben, J. H. J. Organomet.
Chem. 1981, 209, 297. (d) Klei, E.; Teuben, J. H.; De Liefde Meijer, H. J.;
Kwak, E. J.; Bruins, A. P. J. Organomet. Chem. 1982, 224, 327.
30) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric
Identification of Organic Compounds; John Wiley and Sons: New York,
1981; pp 124-127.
(
polymerizations using these compounds occur at a rate that is
convenient for analysis using IR techniques. In each of these
polymerizations the initiating ligand is subsequently observed
on the end of the purified polymer chains (Vide supra), which
indicates that initiation of the polymerization occurs by the
insertion of an isocyanate into the titanium-initiating ligand bond
(

Organotitanium(IV)-Catalyzed Polymerizations of Isocyanates
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1911
through addition of an alcohol, occurs via protonation of the
amidate endgroup from the metal center to yield the corre-
sponding secondary amide endgroup and titanium alkoxide
compound (eq 5). The excess alcohol used in the termination
of these polymerizations as well as any adventitious water can
react further with the free titanium alkoxide compound to form
various other titanium species, so the catalyst utilized in these
polymerizations is neither recovered nor recycled.
The reaction of particular interest is the propagation step of
the polymerization, because there is little knowledge concerning
the mechanism of multiple insertions of isocyanates into metal-
ligand bonds. Equation 4 shows the general propagation step
of the polymerization where the forward and reverse processes
can occur. In this step, either the carbon-nitrogen or carbon-
oxygen double bond of an isocyanate adds accross the titanium-
2
nitrogen bond of the η -amidate endgroup. This particular
addition regiochemistry is required in order to obtain the amide-
repeat-unit structure of the polymer. The resulting species must
2
then rearrange in a fast process to yield the observed η -amidate
endgroup structure. The rate law for this reaction is shown in
eq 6. This expression can be simplified and then integrated
after using the equilibrium expression for the polymerization
(Keq ) k1/k-1 ) 1/[monomer]eq) and assuming that (1) the active
Figure 5. Plots of the kinetic data from Figures 2 and 3 after applying
eq 7. Above: polymerization. Below: depolymerization.
chain ends on polymer molecules of differing chain length show
the same reactivity31 and (2) the concentration of active chain
ends equals the initial concentration of catalyst, which is true
for a living polymerization (eq 7).
Table 3. Kinetic Data for the Polymerization and
Depolymerization of n-Hexyl Isocyanate Using 1
reaction
[catalyst] (M)
0.0825
k
obs (M s-1)
2
1
k )
(M s-1
MWx
polymerization
1.3 × 10-3
1.6 × 10-2
-
d[RNCO]
-
3
-2
0
0
0
0
.0853
.0841
.0504
.0304
1.1 × 10
9.0 × 10
7.8 × 10
9.4 × 10
5.2 × 10
4.3 × 10
1.3 × 10
)
[RNCO](
k [CpTiCl -P ]) -
n 2 n
∑
-4
-2
-2
-2
-2
-3
1.1 × 10
1.5 × 10
3.1 × 10
1.7 × 10
9.0 × 10
dt
n)MW
1
-4
MWx
-
-
-
4
4
4
(
k [CpTiCl -P ]) (6)
depolymerization
0.0306
0.0484
∑
-n
2
n
n)MW1
[
RNCO] - [RNCO]eq
kinetics of several polymerizations of n-hexyl isocyanate using
2 at various concentrations were monitored, and a plot of the
observed rate constants as a function of catalyst concentration
was fitted with a least-squares line. The slope of the plot yielded
ln
) -k [catalyst]t
(7)
1
(
)
[RNCO] - [RNCO]
0
eq
If this integrated rate law is consistent with the actual
-
4
-1 -1
mechanism of propagation, a plot of the logarithmic concentra-
tion function of the experimental data versus time should fit a
straight line with a negative slope of the magnitude of the
product of the catalyst concentration and the forward macro-
scopic rate constant. The observed rate of the polymerization
should also exhibit a first-order dependence upon catalyst
concentration. Furthermore, eq 7 should also hold true for the
kinetics of depolymerization, because the derived rate law for
depolymerization is identical to this equation. We took the
complete set of polymerization and depolymerization data shown
in Figures 2 and 3 and applied eq 7. The resulting plots are
shown in Figure 5, and the data in both plots could be fitted to
least-squares lines with R values of 0.986 and 0.997 for the
full 4-5 measured half-lives. The observed rate constants for
these and several other polymerizations/depolymerizations using
are shown in Table 3. When the concentration of catalyst is
factored out, the forward rate constants obtained for each
experiment were, within experimental error, the same. The
a forward rate constant of 8.5 × 10 mol L
s
(Figure 6).
-
4
-1
The depolymerization rate constant, 3.8 × 10
s
was
calculated using the equilibrium expression for the polymeri-
zation. The above kinetics measurements illustrate that these
polymerizations are well-behaved, living systems and are
consistent with the derived rate law.
To quantify the effect of the ligand environment upon the
activity of the catalyst, several polymerizations were conducted
using identical concentrations of several different catalysts: 1;
5
2
9
; Cp*TiCl2OCH2CF3 [Cp* ) η -pentamethylcyclopentadienyl],
; and Cp2TiCl2OCH2CF3, 10. There was an order-of-
magnitude decrease in the rate of polymerization when one of
the ligands on the titanium center was varied from Cl to Cp to
2
-4
-5
Cp* (kobs decreased from 7.8 × 10 to 3.8 × 10 to 3.3 ×
-6
2
-2 -1
1
0
mol L s , respectively). With 10 the rate of polym-
1
erization was too slow for an accurate measurement. These
data confirm the qualitative observations discussed earlier that
increasing steric bulk about the metal center and increasing
electron donation to the metal center from the ligands reduces
the activity of the titanium catalyst.
(
31) Szwarc, M. AdVances in Polymer Science; Springer-Verlag: New
York, 1983; Vol. 49, pp 15-30 and 79-81.

1
912 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Patten and NoVak
Scheme 1
reluctance of these compounds to coordinate donor ligands
suggests that the interaction between the titanium center and
the isocyanate molecule during propagation must be weak.
Instead of through formal coordination, the isocyanate molecule
may insert into the metal-ligand bond via a bifunctional
2
8,34
activation mechanism.
In this process, an empty orbital on
the highly electron-deficient titanium center can accept electron
density from the isocyanate group and polarize the carbon-
oxygen or carbon-nitrogen double bond. The metal-ligand
bond of interest is polarized in the opposite sense, and therefore
nucleophilic addition of the ligand onto the isocyanate group
can be viewed as a concerted addition (Scheme 2, assuming
Scheme 2
Figure 6. A plot of kobs versus catalyst concentration for polymeriza-
tions of n-hexyl isocyanate using 2.
One issue that remains open about the polymerization
mechanism is whether or not a discreet monomer-catalyst
complex exists along the reaction coordinate. A propagation
mechanism can be written where precoordination of the
monomer occurs before the insertion step (Scheme 1, assuming
oxygen coordination). The simplified rate expression is shown
in eq 8. Manual integration of this equation does not yield a
oxygen coordination). Instead of the deep potential well of a
formal coordination complex, such a species would be a
transition state or a very shallow potential well along the reaction
coordinate. Carbon dioxide, which is isoelectronic with an
isocyanate, has been shown to undergo this type of bifunctional
activation addition in its reactions with simple bases and
3
4
inorganic complexes. This type of addition would be con-
sistent with either mechanism and kinetic rate law.
-
d[RNCO]
k
1
k
2
[RNCO] - k-1
k
-2
)
[catalyst]
(8)
The possible insertion mechanisms discussed here can be
viewed as a spectrum where at one end a formal coordination
complex intermediate is formed, and at the other end the
isocyanate undergoes a concerted [2 + 2] addition across the
metal-ligand bond with no prerequisite interaction with the
metal center. In light of the reluctance of CpTiCl2L complexes
to coordinate Lewis bases and the effects of steric bulk and
electrophillicity of the metal center upon the rate of polymer-
dt
k-1 + k + k + k [RNCO]
2
-2
1
simple rate law that can be plotted, so numeric integration would
be necessary to fit the data. An equation with four variables
could most likely be minimized to fit the data, so other methods
are necessary to distinguish between the two possible mecha-
nisms. The chemical behavior of CpTiCl2L complexes was used
to gain insights into which of the two mechanisms is more likely.
If a coordination intermediate occurs along the mechanistic
pathway, then the isolation of CpTiCl2L adducts with Lewis
bases should be possible. However, no monodentate ether,
amine, or phosphine complex of CpTiCl3 has ever been
prepared, and stable Lewis base adducts have only been isolated
and characterized using very strong chelating donor ligands such
(
32) Clark, R. J. H.; Stockwell, J. A.; Wilkins, J. D. J. Chem. Soc., Dalton
Trans. 1976, 120.
33) Hughes D. L.; Leigh, G. J.; Walker, D. G. J. Organomet. Chem.
(
1988, 355, 113.
(34) (a) Chisholm, M. H.; Extine, M. W. J. Am. Chem. Soc. 1977, 99,
7
92. (b) Dennard, A. E.; Williams, R. J. P. Transition Metal Chemistry;
Carlin, R. L., Ed.; Marcel Dekker: New York, 1966; Vol. 2, pp 123-164.
c) Gambarotta, S.; Arena, F.; Floriani, C.; Zanazzi, P. F. J. Am. Chem.
Soc. 1982, 104, 5082.
(
3
2
33
as 2,2′-bipyridyl and bisdimethylphosphinoethane.
The

Organotitanium(IV)-Catalyzed Polymerizations of Isocyanates
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1913
1
ization, it seems most probable that the actual mechanism lies
closer to the latter side of the spectrum.
Chemical shifts for H spectra are reported in units of δ (ppm), positive
values indicating shifts downfield of tetramethylsilane, and are
referenced to a selected residual proton signal of the solvent: CDCl
3
,
7
.25 ppm, singlet; toluene-d , 2.09 ppm, quintet; benzene-d , 7.15 ppm,
8
6
Conclusion
1
broad. Significant H NMR data are tabulated in order: multiplicity
(s ) singlet, d ) doublet, t ) triplet, q ) quartet, m ) multiplet, br )
broadened signal), coupling constant(s) in Hertz, and number of protons.
Chemical shifts for 13C{ H} spectra are reported in units of δ (ppm),
positive values indicating shifts downfield of tetramethylsilane, and
are referenced to a selected carbon resonance of the solvent em-
A living organotitanium(IV)-catalyzed polymerization of
isocyanates was developed. A titanium trichloride alkoxide,
1
1
, was found to polymerize n-hexyl isocyanate to high yields
and without the formation of cyclic trimer; however, polymer-
izations using 1 were inhibited by the presence of donor solvents
or monomers possessing donor groups. Compounds 2-4 were
found to polymerize isocyanates in the presence of donor
solvents and isocyanates possessing donor functional groups,
activated olefins, and strained olefins. The activity of the
organotitanium(IV) catalysts decreased with increasing steric
bulk about the metal center and increasing electron donation to
the metal center from the ligands, and only polymerizations
using catalysts 1-4 gave high yields of polymer within an
experimentally practical time period. The nature of the initiating
ligand on the titanium center in catalysts 2-4 had no detectable
effect upon the activity of the catalyst, and both fluorine-labeled
and chiral initiating ligands were detected on the end of the
polymer chain using 19F NMR spectroscopy and polarimetry,
respectively.
ployed: CDCl
, 128.0 ppm, triplet. Chemical shifts for 19F{ H} NMR spectra are
reported in units of δ (ppm), negative values indicating shifts upfield
of CFCl , and are referenced to R,R,R-trifluorotoluene internal stan-
3 8
, 77.0 ppm, triplet; toluene-d , 20.4 ppm, septet; benzene-
1
d
6
3
dard: -63.72 ppm. Infrared spectra were determined using a Perkin-
Elmer 1600 series FTIR. Polarimetry was performed using a Perkin-
Elmer 141 polarimeter (λ ) 589 nm). Melting points were determined
using an Electrothermal capillary melting point apparatus and are
uncorrected. Elemental analyses were performed by the Microanalytical
Laboratory operated by the College of Chemistry, University of
California, Berkeley, CA.
Polymer Characterizations. Gel permeation chromatography
(GPC) was performed using a Hewlett/Packard 1050 series liquid
chromatograph pump equipped with a HP Model 1047 refractive index
detector and a HP model 3396A integrator. Separations were effected
by 10 , 10 , 10 , and 500 Å Waters Ultrastyragel columns and molecular
weights were calibrated relative to polystyrene standards (Scientific
Polymer Products, Inc.): Mobile phase: CHCl or THF; flow rate, 1
mL min . Light-scattering in the tandem GPC/light-scattering experi-
ments was performed using a Wyatt Technology Dawn Model F
Multiangle Photometer. Measurements of (δn/δc)T,P were performed
using a Chromatics KMX-16 Refractometer. Static light-scattering
5
4
3
The polymerization of n-hexyl isocyanate using organotita-
nium(IV) compounds is living. Several experiments support
this conclusion. The PDIs of PHIC synthesized using catalysts
3
-1
1
-4 were found to range from 1.05 to 1.2. The molecular
weight of the polymer formed in polymerizations of n-hexyl
isocyanate using catalysts 1-4 varied linearly as a function of
the monomer-to-initiator ratio and the percent conversion of the
polymerization. Polymerizations using 2 can be endcapped
quantitatively, and well-defined block copolymers can be
synthesized using catalysts 1-4.
(
(
SLS) measurements were performed using a Brookhaven Instruments-
BI)-9000 correlator and BI-200SM goniometer driven by BI-ZP
software for automated acquisition and analysis of SLS data. Calibra-
tion to absolute intensities was performed using toluene standards.
Thermal gravimetric analysis (TGA) was performed using a Perkin-
Elmer PC Series TGA7: scanning rate ) 10 °C min , atmosphere )
nitrogen. Differential scanning calorimetry (DSC) was performed using
a Perkin-Elmer PC series DSC.
Polymerizations of isocyanates using organotitanium(IV)
compounds were found to be completely reversible between
monomer and polymer, enabling the determination of the
thermodynamic parameters of PHIC. The kinetics of polymer-
izations using a wide variety of organotitanium compounds were
measured. For the series of catalysts 1, 2, and 9, the observed
rate constants for polymerizations conducted under identical
conditions decreased by an order of magnitude corresponding
with each increase in the steric bulk and electron-donating
capability of the spectator ligand. The kinetics for polymeriza-
tions using catalysts 1 and 2 were first-order in both monomer
and catalyst. The active endgroup of a polymerization using 3
was observed using IR spectroscopy, and the frequency of the
-1
Reagents. Solvents were purified as follows: toluene, benzene,
hexanes, and tetrahydrofuran (THF) were distilled from Na/benzophe-
none. Isocyanate monomers were distilled from CaH
an argon atmosphere. Methylene chloride (CH Cl ) and triethylamine
Et N) were distilled from CaH . endo/exo-5-Norbornene-2-methanol
was distilled from CaH under vacuum and stored under an argon
atmosphere. (R)-(+)-R-Methyl-2-naphthalene methanol and (1S,2R,5S)-
+)-menthol were sublimed under vacuum and stored in a drybox. 2,2,2-
Trifluoroethanol was vacuum transferred from MgSO that had been
2
and stored under
2
2
(
3
2
2
(
4
dehydrated by heating at 150 °C under a high dynamic vacuum for 2
days. (CH ) SiN(CH ) was obtained from Huls America, fractionally
3
3
3 2
2
IR stretch was consistent with an η -amidate endgroup structure.
distilled, and stored under an argon atmosphere. Titanium(IV) chloride
was stirred over copper turnings for 24 h and distilled through an all-
glass apparatus. Catalyst 1 was prepared according to the procedure
of Paul and co-workers.35 Methyl R-(hydroxymethyl)acrylate was
Finally, the kinetic data for the polymerization of n-hexyl
isocyanate and the known chemistry of CpTiCl2L compounds
were found to be consistent with a propagation step that occurs
via a bifunctional activation mechanism.
3
6
prepared according to the procedure of Mathias and co-workers.
3
Cp*TiCl was prepared according to the procedure of Llin a´ s and co-
workers.37 Unless specified otherwise, all other reagents were obtained
from Aldrich Chemical Co. and used without further purification.
Experimental Section
General Procedures and Characterizations. All manipulations
involving air- and moisture-sensitive compounds were carried out under
argon atmospheres using standard Schlenk techniques. Solids and
monomers were transferred in an argon-filled Vacuum Atmospheres
HE533 Dri Lab with attached HE493 Dri Train.
Experimental Procedures and Characterizations. TiCl OCH CF -
3
2
3
(THF) . To a 25 mL Schlenk flask containing 5 mL of CH Cl was
2
2
2
added via syringe 1.80 mL (3.11 g, 16.4 mmol) of TiCl . The flask
4
was placed in an ice/water bath, and then 1.00 mL (1.37 g, 13.7 mmol)
of HOCH CF was added via syringe. The solution was stirred for 1
2
3
1
13
1
H and C{ H} NMR spectra were determined at 400 and 100 MHz,
h, and then the flask was fitted with a reflux condenser. The solution
respectively, using a Bruker AM-400 or AMX-400 FT NMR spec-
trometer or at 500 and 125 MHz, respectively, using a Bruker AM-
(35) Paul, R. C.; Sharma, P.; Gupta, P. K.; Chadha, S. L. Inorg. Chim.
Acta 1976, 20, 7.
1
5
00 FT NMR spectrometer. H NMR spectra were determined at 250
(
36) (a) Mathias, L. J.; Kusefoglu, S. H. Macromolecules 1987, 20, 2041.
b) Mathias, L. J.; Kusefoglu, S. H.; Kress, A. O. Macromolecules 1987,
0, 2327.
37) Llin a´ s, G. H.; Mena, M.; Palacios, F.; Royo, P.; Serrano, R. J.
Organomet. Chem. 1988, 340, 37.
MHz on a superconducting spectrometer constructed in the UC Berkeley
NMR Laboratory, equipped with a Cryomagnets, Inc. magnet and a
(
2
19
1
Nicolet Model 1280 data collection system. F{ H} NMR spectra were
(
determined at 376 MHz using the Bruker AM-400 spectrometer.

1
914 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Patten and NoVak
was stirred and heated at reflux under argon overnight. The next day
deep red and was stirred for 24 h. THF and volatile products were
removed under vacuum, and the product (0.615 g, 79%) was transferred
to a sublimation apparatus. Sublimation at 90 °C and 0.01 Torr afforded
volatile materials were removed under vacuum to leave 1.40 g (40%
1
yield) of a white powder: H NMR (250 MHz, toluene-d
8
) δ 3.91 (br).
1
Next, 1.31 g of this powder was weighed into a 100 mL Schlenk
flask. Via cannula, 25 mL of CH Cl was added, and the flask was
fitted with a pressure-equalizing addition funnel. The addition funnel
was charged with 25 mL of CH Cl and 2.00 mL of THF. The flask
0.521 g (66% yield) of orange crystals: mp ) 121-123 °C; H NMR
(400 MHz, CDCl
(100 MHz, CDCl
3
) δ (ppm) 6.61 (s, 5H), 3.82 (s, 6H); 13C{ H} NMR
3
1
2
2
) δ (ppm) 188.44, 52.11; IR (thin film) 3125 (m),
3101 (m), 3004 (w), 2961 (m), 2903 (m), 2878 (m), 2838 (w), 2797
(w), 2787 (w), 1863 (vw), 1786 (vw), 1684 (vw), 1448 (sh), 1438 (m),
1427 (m), 1406 (w), 1366 (w), 1244 (w), 1133 (w), 1072 (vw), 1044
2
2
was cooled to 0 °C using and ice/water bath, and the THF solution
was added dropwise to the solution of 1. After complete addition, the
amount of solvent was reduced to approximately 10 mL under vacuum.
Via cannula 80 mL of hexanes was layered onto the solution, and the
flask was tightly closed and placed in a -40 °C freezer. Over a 2 day
period colorless crystals formed. The supernatant was removed, and
-
1
(w), 1028 (m), 1020 (m), 913 (s), 832 (s), 784 (sh), 707 (vw) cm
Anal. Calcd for C NTi: C, 36.88; H, 4.86; N, 6.14. Found:
C, 36.49; H, 4.69; N, 6.02.
CpTiCl CH , 4. This compound was prepared using a modification
.
7 2
H11Cl
2
3
of the procedure of Erskine et al.41 In a 50 mL Schlenk flask, CpTiCl3
(0.503 g, 2.29 mmol) was dissolved in 25 mL of THF. The flask was
cooled to 0 °C with an ice bath, and then 0.760 mL of a 3.0 M solution
of MeMgCl in THF was added dropwise via syringe over a 1 h period.
THF was then removed under vacuum. The solid was extracted with
25 mL of benzene for 1 h and filtered. Benzene was removed under
vacuum, and the product (0.276 g, 60%) was transferred to a sublimation
apparatus. Sublimation at 50 °C and 0.01 Torr onto a dry ice/isopropyl
alcohol cold finger afforded 0.157 g (34% yield) of orange crystals:
solvents were removed from the crystals under vacuum leaving 1.67 g
1
(
81% yield) of colorless crystals: H NMR (400 MHz, CDCl
3
) δ 4.90
1
9
1
3
(br, 2H), 3.9 br, 8 H), 1.30 (br, 8H); F{ H} NMR (376 MHz, CDCl )
δ -75.66; IR (thin film) 2985 (m), 2909 (m), 1458 (w), 1279 (s), 1150
(
vs), 1039 (w), 1010 (m), 952 (m), 920 (w), 856 (m), 774 (sh), 716
-
1
(
4
m), 628 (w) cm . Anal. Calcd for C10
.56; N, 0.00. Found: C, 31.50; H, 4.47; N, 0.02.
CpTiCl . The following synthesis is a modification of the procedure
3 3 3
H18Cl F O Ti: C, 30.22; H,
3
3
8
of Gorsich. In a drybox, a 250 mL three-necked flask with a large
egg-shaped stir bar was charged with 10.0 g (40.3 mmol) of Cp TiCl
1
2
2
.
mp ) steadily decomposed upon heating; H NMR (400 MHz, toluene-
4
1
The flask was closed, taken out of the drybox, and connected to a
Schlenk line. A reflux condenser with a joint-to-hose adapter was
attached, and via cannula 100 mL of toluene was added to the flask.
The mixture was stirred, while the flask was cooled to 0 °C using an
d
8
3
) δ (ppm) 5.90 (s, 5H), 1.71 (s, 3H) (lit. (benzene-d
6
) 5.87, 1.74);
1
1
C{ H} NMR (100 MHz, toluene-d
8
) δ (ppm) 118.89, 79.65; IR (thin
film) 3104 (s), 2974 (m), 2868 (m), 1863 (w), 1778 (w), 1682 (w),
1437 (s), 1389 (w), 1438 (w), 1389 (w), 1094 (sh), 1066 (m), 1019
-
1
ice-water bath. Then 13.5 mL (23.4 g, 123 mmol) of TiCl
4
was added
(m), 910 (w), 835 (s), 792 (m), 708 (w) cm . Anal. Calcd for C
Cl Ti: C, 36.23; H, 4.05; N, 0.00. Found: C, 35.75; H, 3.87; N, 0.00.
Cp TiClOCH CF , 10. To a mixture of 0.500 g (2.01 mmol) of
Cp TiCl in 50 mL of benzene was added 0.600 mL (4.30 mmol) of
Et N. The mixture was heated at 60 °C until all the solid dissolved.
6 8
H -
to the slurry via syringe. The ice bath was removed, and the flask
2
was immersed in an oil bath. The oil was heated to 135-140 °C so
the toluene reached reflux, and over time the Cp TiCl dissolved to
2 2
form a deep red solution. The reaction was stirred and heated at reflux
for 6 h.
After the reflux time was complete, the heat source was removed,
and stirring was stopped. The reaction mixture was left to cool
2
2
3
2
2
3
The resultant bright red solution was then treated with 0.140 mL (1.92
mmol) of CF CH OH. After several minutes the solution became bright
3
2
orange, and a white precipitate formed. The mixture was heated at 60
°C for 1 h and then heating was discontinued. The mixture was allowed
to stir for an additional 10 h at room temperature, after which it was
filtered. Benzene was removed under vacuum to give 0.487 g (78%
3
overnight, and as the solution cooled orange crystals of CpTiCl formed.
The supernatant was removed via cannula, and volatile materials were
removed from the product under vacuum, leaving 13.4 g (76% yield)
of crude product. A portion of the crude product (4.25 g) was
transferred to a vacuum sublimation apparatus, and sublimation at 90
1
yield) of an orange solid: mp ) 81-82.5 °C;
H NMR (400 MHz,
H-F
) 9.0 Hz); 13C-
benzene-d ) δ (ppm) 5.76 (s, 10H), 4.26 (q, 2H, J
H} NMR (100 MHz, CDCl ) δ (ppm) 124.33 (q, J
6
°
C and 0.01 Torr afforded 3.90 g (92% recovery) of orange crystals:
{
1
C-F
) 279 Hz),
3
3
8 1
19
1
mp ) 212-214 °C (lit. 208-211 °C); H NMR (400 MHz, CDCl
3
)
117.47, 76.62 (q, JC-C-F ) 34 Hz); F{ H} NMR (376 MHz, benzene-
d6) δ (ppm) -76.1; IR (thin film) 3100 (m), 2897 (m), 2852 (m), 1873
(vw), 1782 (vw), 1681 (vw), 1630 (vw), 1444 (m), 1365 (w), 1280 (s),
3
9
13
1
δ (ppm) 7.03 (s) (lit. (CH
CDCl ) δ (ppm) 123.40; IR (thin film) 3112 (m), 2971 (w), 2856 (w),
880 (w), 1801 (w), 1703 (w), 1436 (s), 1066 (w), 1019 (m), 843 (m),
3
CN) 7.25);
C{ H} NMR (100 MHz,
3
1
7
1145 (s), 1015 (m), 952 (m), 888 (m), 857 (m), 817 (s), 592 (w) cm-1
.
-
1
90 (sh) cm
CpTiCl OCH
g, 1.58 mmol) was dissolved in 10 mL of benzene. Then a solution of
.115 mL (1.58 mmol) of CF CH OH and 0.220 mL (1.58 mmol) of
Et N in 10 mL of benzene was added dropwise via a pressure-equalizing
addition funnel. A white precipitate of Et
.
Cp*TiCl
g, 1.75 mmol) was dissolved in 20 mL of THF; via syringe 0.241 mL
1.73 mmol) of Et N was added. Then a solution of 0.126 mL (1.73
mmol) of CF CH OH in 10 mL of THF was added dropwise via a
pressure-equalizing addition funnel. A white precipitate of Et
2 2 3 3
OCH CF , 9. In a 50 mL Schlenk flask, Cp*TiCl (0.505
2
2 3 3
CF , 2. In a 50 mL Schlenk flask, CpTiCl (0.346
(
3
0
3
2
3
2
3
NH Cl
+
-
3
+
-
3
NH Cl formed. The
formed, and the red solution turned bright orange. The solution was
solution was stirred for 2 h and then filtered. Benzene was removed
under vacuum, and the product (0.384 g, 86%) was transferred to a
stirred for 3 h and then filtered. Removal of solvent under vacuum
yielded 0.457 g (75% yield) of an orange solid: 1H NMR (400 MHz,
sublimation apparatus. Sublimation at 80 °C and 0.01 Torr afforded
) δ (ppm) 4.37 (q, 2H, JH-F ) 8.6 Hz), 1.90 (s, 15H); 13C-
toluene-d
8
1
0
.317 g (71% yield) of yellow crystals: mp ) 87.5-90 °C; H NMR
1
{
8
H} NMR (100 MHz, toluene-d ) δ (ppm) 133.38, 74.18 (q, JC-C-F
(400 MHz, CDCl
3
) δ (ppm) 6.81 (s, 5H), 4.79 (q, 2H, JH-F ) 8.3 Hz)
)
36 Hz), 12.71 (CF
3
not found).
4
0
13
1
(lit. (CCl
4
) 6.85 (s), 4.75 (q));
C{ H} NMR (100 MHz, CDCl
3
) δ
5
Bischloro-η -cyclopentadienyl-(R)-r(+)-1-(2-napthyl)ethoxy Ti-
(
ppm) 122.88 (q, JC-F ) 279 Hz), 121.51, 77.46 (q, JC-C-F ) 36 Hz);
tanium(IV), 5. This compound was prepared using the same procedure
for the synthesis of 2. The quantities of reagents used were 0.501 g
19
1
F{ H} NMR (376 MHz, CDCl
s), 2934 (w), 2869 (w), 1872 (w), 1793 (w), 1696 (w), 1459 (sh),
444 (s), 1398 (w), 1365 (m), 1282 (vs), 1154 (vs), 1065 (w), 1026
m), 1016 (m), 956 (m), 937 (sh), 833 (s), 784 (w), 724 (m), 668 (w),
3
) δ (ppm) -72.5; IR (thin film) 3116
(
1
(
(
2
2
2.28 mmol) of CpTiCl
-naphthalene methanol, 0.320 mL (0.232 g, 2.30 mmol) of Et N, and
× 25 mL of benzene. Yield: 0.682 g (84%) of a yellow solid. The
3
, 0.394 g (2.29 mmol) of (R)-(+)-R-methyl-
3
-
1
6
0
27 (m) cm . Anal. Calcd for C
.00. Found: C, 30.11; H, 2.49; N, 0.01.
CpTiCl N(CH , 3. In a 10 mL Schlenk flask, CpTiCl
.48 mmol) was dissolved in 5 mL of THF. Then 0.550 mL (3.48
SiN(CH was added via syringe. The solution turned
7 7 2 3
H Cl F OTi: C, 29.72; H, 2.49; N,
crude product was recrystallized from a solution of THF/hexanes to
yield 0.359 g (53% recovery) of yellow needle-like crystals: mp )
2
3
)
2
3
(0.723 g,
1
1
(
3
1
2
45-147 °C; H NMR (400 MHz, CDCl
3
) δ (ppm) 7.88 (m, 4H), 7.52
3
m, 3H), 6.54 (s, 5H), 5.95 (q, J ) 6.4 Hz, 1H), 1.72 (d, J ) 6.5 Hz),
H); C{ H} NMR (100 MHz, CDCl ) δ (ppm) 140.72, 133.10, 128.75,
3
28.15, 127.77, 126.51, 126.36, 124.46, 123.57, 122.32, 119.30, 91.90,
5.13; IR (thin film) 3098 (m), 3052 (w), 2981 (m), 2926 (w), 1868
mmol) of (CH
)
3 3
)
3 2
13 1
(
38) (a) Gorsich, R. D. J. Am. Chem. Soc. 1958, 80, 4744. (b) Gorsich,
R. D. J. Am. Chem. Soc. 1960, 82, 4211.
39) Cardoso, A. M.; Clark, R. J. H.; Moorhouse, S. J. Chem. Soc., Dalton
Trans. 1980, 1156.
40) Choukroun, R.; Gervais, D. J. Chem. Soc., Dalton Trans. 1980, 1800.
(
(41) Erskine, G. J.; Graham, J. B. H.; Weinberg, E. L.; Hunter, B. K.;
McCowam, J. D. J. Organomet. Chem. 1984, 267, 265.
(

Organotitanium(IV)-Catalyzed Polymerizations of Isocyanates
w), 1792 (w), 1700 (w), 1601 (w), 1506 (w), 1437 (m), 1369 (w),
J. Am. Chem. Soc., Vol. 118, No. 8, 1996 1915
(
and connected to a Schlenk line. n-Hexyl isocyanate (1.00 mL, 0.865
g, 6.80 mmol) was added via syringe. The Schlenk tube was fitted
with a ground-glass stopper, and the solution was stirred for 24 h.
Bulk Polymerization. In a drybox, an appropriate amount of
catalyst, depending upon the monomer-to-initiator ratio, was weighed
into a 25 mL Schlenk tube with a magnetic stir bar. A drop or two of
solvent was added to ensure rapid dissolution of the catalyst into the
monomer. The tube was capped with a rubber septum, taken out of
the drybox, and connected to a Schlenk line. n-Hexyl isocyanate (1.00
mL, 0.865 g, 6.80 mmol) was added via syringe. The Schlenk tube
was fitted with a ground-glass stopper, and the solution was stirred for
24 h.
1
326 (w), 1273 (w), 1171 (m), 1126 (m), 1087 (s), 1034 (m), 1020
(
7
m), 974 (w), 939 (w), 901 (w), 862 (w), 851 (m), 830 (m), 822 (m),
53 (m), 698 (w) cm . Anal. Calcd for C17H16Cl OTi: C, 57.50; H,
2
-
1
4
.54; N, 0.00. Found: C, 57.22; H, 4.44; N, 0.07.
Bischloro-η -cyclopentadienyl-(1S,2R,5S)-(+)-menthoxy Titanium-
5
(
IV), 6. This compound was prepared using the same procedure for
the synthesis of 2. The quantities of reagents used were 0.504 g (2.30
mmol) of CpTiCl , 0.358 g (2.29 mmol) of (1S,2R,5S)-(+)-menthol,
.700 mL (0.508 g, 5.02 mmol) of Et
3
0
3
N, and 2 × 25 mL of THF.
Yield: 0.707 g (91%) of a yellow solid. Vacuum sublimation of the
product at 130 °C yielded 0.305 g (39% recovery) of yellow needle-
1
like crystals: mp ) 99-101 °C; H NMR (400 MHz, CDCl
3
) δ (ppm)
Polymer Workup. After the polymerization was complete, the solid
6
1
.71 (s, 5H), 4.49 (m, 1H), 2.27 (m, 1H), 2.12 (m, 1H), 1.64 (m, 2H),
orange mass was dissolved in 20 mL of a 5% solution of CH OH in
3
13
1
.43 (m, 2H), 1.24 (m, 1H), 0.93 (m, 6H), 0.81 (m, 3H); C{ H} NMR
) δ (ppm) 118.76, 95.75, 50.50, 44.44, 33.97, 31.78,
5.71, 22.66, 21.97, 20.83, 15.67; IR (thin film THF) 3108 (m), 2958
THF. This solution was poured into 100 mL of CH OH, and a white
3
(
100 MHz, CDCl
3
solid precipitated. The polymer was isolated by filtration through a
2
0.2 µm nylon filter and washed with CH OH. Volatile materials were
3
(
(
s), 2928 (s), 2867 (s), 1855 (w), 1772 (w), 1684 (w), 1457 (m), 1438
m), 1385 (w), 1368 (m), 1346 (w), 1334 (w), 1267 (w), 1176 (w),
removed under vacuum. The polymer was purified by reprecipitation
1
from THF with CH OH: total range of yields 74-95%; H NMR (250
3
1
9
156 (w), 1095 (m), 1079 (m), 1062 (s), 1048 (s), 1019 (m), 1000 (w),
78 (w), 929 (w), 878 (w), 859 (w), 846 (m), 839 (m), 814 (s), 771
MHz, CDCl ) δ (ppm) 0.82 (br, 3H), 1.22-1.50 (a br and a very br
3
signal, 8H total), 3.60 (very br, 2H); 13C{ H} NMR (100.0 MHz,
1
-
1
(w), 734 (m), 713 (w), 592 (w) cm . Anal. Calcd for C15
H24Cl
2
OTi:
CDCl ) δ (ppm) 13.92, 22.53, 26.22, 28.42, 31.48, 48.52, 156.77 (lit.
3
14, 23, 26, 29, 32, 48, 157);42 IR (thin film) 2953 (s), 2923 (s), 2857
C, 53.12; H, 7.13; N, 0.00. Found: C, 53.20; H, 7.18; N, 0.11.
5
(
2-Carbethoxyprop-2-enoxy)bischloro-η -cyclopentadienyl Tita-
(s), 1700 (s), 1459 (m), 1349 (s), 1265 (s), 1180 (s), 1092 (s), 997 (m),
-
1
nium(IV), 7. This compound was prepared using the same procedure
for the synthesis of 2. The quantities of reagents used were 3.02 g
887 (m), 755 (s) cm ; TGA (onset temperature) 180 °C (100% wt
loss) (lit. 180 °C).43 Anal. Calcd for C
13NO: C, 66.11; H, 10.30;
7
H
(
13.8 mmol) of CpTiCl
3
, 1.45 mL (1.60 g, 13.7 mmol) of methyl
N, 11.01. Found: C, 66.18; H, 10.71; N, 11.16.
R-(hydroxymethyl)acrylate [CAUTION: Methyl R-(hydroxymethyl)-
acrylate is a powerful skin irritant which has been shown to cause
blisters and contact dermatitis upon exposure. Precautions should be
taken to prevent exposure to the material and its vapor during handling
and storing of the material. Storage in a darkened Schlenk flask at
Endcapping of a n-Hexyl Isocyanate Polymerization Initiated by
2 Using Trifluoroacetic Anhydride. A 25 mL Schlenk flask was
-
4
charged with 95.8 mg (3.39 × 10 mol) of 2 and 0.10 mL of THF.
-
2
To the slurry of catalyst in THF, 2.00 mL (1.73 g, 1.36 × 10 mol)
of n-hexyl isocyanate was added (M/I ) 40), and the polymerization
was stirred for 48 h. Afterwards, the Schlenk tube was brought into
the drybox where 0.864 g of the active polymer was transferred into a
flask and taken out of the drybox. This polymer sample was dissolved
in 5 mL of a 5% (v/v) CH OH in CH Cl solution and poured into 100
-
40 °C is recommended.], 2.00 mL (1.45 g, 14.3 mmol) of Et
0 + 25 mL of benzene. Yield: 3.39 g (82%) of a yellow solid. The
3
N, and
5
crude product was recrystallized by layering hexanes onto a toluene
solution of the compound yielding 1.14 g (34% recovery) of large
yellow-orange crystals: mp ) 102-104 °C; 1H NMR (400 MHz,
3
2
2
mL of CH OH. The polymer was isolated by filtration, and solvents
3
CDCl
3
) δ (ppm) 6.75 (s, 5H), 6.34 (m, 1H), 6.00 (m, 1H), 5.34 (s,
were removed under vacuum: yield 0.690 g (80%) of a white solid;
1
3
1
19
1
2
H), 3.78 (s, 3 H); C{ H} NMR (100 MHz, CDCl
3
) δ (ppm) 165.13,
F{ H} NMR (376 MHz, CDCl ) δ (ppm) -74.8.
3
1
37.03, 126.96, 119.93, 81.60, 52.11; IR (thin film) 3102 (m), 3000
The remaining active polymer was broken apart into pieces and
(
(
(
w), 2953 (w), 2853 (w), 1786 (w), 1715 (s), 1676 (w), 1635 (m), 1438
removed from the drybox in the Schlenk tube. Benzene (2.5 mL) was
added, and the polymer began dissolving to form an orange solution.
s), 1393 (w), 1377 (m), 1304 (s), 1194 (w), 1162 (w), 1040 (vs), 1021
-
1
m), 947 (m), 832 (s), 739 (m) cm . Anal. Calcd for C10
H12Cl
O
2 3
Ti:
-3
Then 1.00 mL (1.49 g, 7.08 × 10 mol) of trifluoroacetic anhydride
C, 40.17; H, 4.05; N, 0.00. Found: C, 40.24; H, 4.02; N, 0.00.
was added via syringe, and the solution turned red-orange. Another
2.5 mL of benzene was added to complete the dissolution of the
polymer. The solution was stirred for 1 h and then poured into 100
5
Bischloro-η -cyclopentadienyl(bicyclo[2.2.1]hept-5-enyl)-2-meth-
oxy Titanium(IV), 8. This compound was prepared using the same
procedure for the synthesis of 2. The quantities of reagents used were
mL of stirring CH
was washed with 5 × 10 mL of CH
trifluoroacetic acid. Volatile materials were removed under vacuum
3
OH. The mixture was filtered, and the polymer
0
.842 g (3.84 mmol) of CpTiCl
endo/exo-5-norbornene-2-methanol, 0.540 mL (0.392 g, 3.87 mmol)
of Et
3
, 0.465 mL (0.478 g, 3.85 mmol) of
3
OH to ensure removal of any
19
1
3
N, and 2 × 25 mL of benzene. Yield: 1.08 g (92%) of a yellow
to yield 0.941 g (90%) of a white solid: F{ H} NMR (376 MHz,
solid. The crude product was recrystallized by layering hexanes onto
3
CDCl ) δ (ppm) -70.5, -74.8 [ratio ) 1-1.09, 91% endcapped].
a toluene solution of the compound yielding 0.430 g (40% recovery)
Repeat trial using an optimized 19F NMR pulse sequence: 69.0 mg
1
of yellow crystals: mp ) 90-92 °C; H NMR (400 MHz, toluene-d
8
)
(0.244 mmol) of 2, 0.500 mL (0.433 g, 3.40 mmol) of n-hexyl
isocyanate (M/I ) 14), and 2.00 mL of trifluoroacetic anhydride; yield
δ (ppm) 6.06 (s, 5H), 5.96 (m, 1H), 5.84 (m, 1H), 3.91 (dd, 1H), 3.77
(dd, 1H), 2.97 (s, 1H), 2.57 (s, 1H), 2.18 (m, 1H), 1.54 (m, 1H), 1.39
0.296 g (68%) of a white solid; 19F{
1
H} NMR (376 MHz, CDCl ) δ
3
(m, 1H), 1.01 (m, 1H), 0.22 (m, 1H) [Note: Mostly one isomer, endo
(ppm) -70.5, -74.8 [ratio ) 1.02-1, 100% endcapped].
or exo, recrystallized. Small signals due to the other isomer were
Polymerization Kinetic Measurements. In a drybox, a volumetric
solution of the catalyst of interest in toluene was prepared. Several 5
mL volumetric flasks (chosen for the small capacity and for the stable
base) were charged with a microstir bar and 0.150 mL of the catalyst
solution and then fitted with a rubber septum. The flasks were removed
from the drybox. A needle attached to a Schlenk line hose was inserted
through the septum, so the flask could be kept under a pressure of
argon. The flasks were placed in a water bath thermostatted at 25.0 (
1
3
1
observed in the baseline of the spectrum.]; C{ H} NMR (100 MHz,
toluene-d ) δ (ppm) 137.58 [next to the solvent signal], 132.53, 118.7,
8.46, 49.61, 44.10, 42.51, 42.03, 28.58; IR (thin film) 3108 (m), 3057
8
8
(
(
(
Cl
0
m), 2967 (s), 2865 (m), 1867 (w), 1773 (w), 1654 (w), 1570 (w), 1449
m), 1364 (w), 1343 (m), 1256 (w), 1148 (w), 1076 (s), 1017 (m), 930
-1
w), 824 (s), 723 (m), 692 (m), 646 (m) cm . Anal. Calcd for C15
OTi: C, 53.12; H, 7.13; N, 0.00. Found: C, 53.20; H, 7.18; N,
.11.
General Procedure for the Polymerization of n-Hexyl Isocyanate.
24
H -
2
0
.2 °C, and stirring was started. Via syringe, 0.150 mL of n-hexyl
Solution Polymerization. In a drybox, a volumetric solution of the
(42) Cook, R.; Johnson, R. D.; Wade, C. G.; O’Leary, D. J.; Munoz, B.;
Green, M. M. Macromolecules 1990, 23, 3454.
43) (a) Aharoni, S. M. J. Polym. Sci., Polym. Phys. Ed. 1980, 18, 1303.
b) Durairaj, B.; Dimcock, A. W.; Samulski, E. T.; Shaw, M. T. J. Polym.
Sci., Part A: Polym. Chem. 1989, 27, 3211.
44) Itou, T.; Chikiri, H.; Teramoto, A.; Aharoni, S. M. Polym. J. (Tokyo)
1988, 20, 143.
catalyst, depending upon the monomer-to-initiator ratio, was prepared
using the appropriate solvent (for 1 or TiCl OCH CF (THF) toluene
3 2 3 2
was necessarily used). To a 25 mL Schlenk tube with a magnetic stir
bar, a measured amount (typically 0.100 mL) of this solution was added,
and the tube was capped with a rubber septum, taken out of the drybox,
(
(
(

1
916 J. Am. Chem. Soc., Vol. 118, No. 8, 1996
Patten and NoVak
isocyanate was added to the flask, and the time of the polymerization
was recorded using a stopwatch. The Schlenk line needle was removed,
and the septum and top of the flask were covered with wax film. At
timed intervals one of the flasks was opened, and the contents were
quantitatively transferred into a 25 mL volumetric flask using HPLC
isocyanate. After the polymerization had reached equilibrium (3 h),
the flask was fitted with a rubber septum, taken out of the drybox, and
attached to a Schlenk line. The polymerization was kept under a flow
of argon and stirred constantly. The flask was placed in an oil bath
equipped with a feedback temperature controller (temperature variation
(0.2 °C) and allowed to reach thermal equilibrium with the oil bath.
An aliquot was removed using a syringe and diluted volumetrically
grade CHCl
HPLC grade CHCl
using a 0.1 mm path length NaCl solution cell. The percent transmit-
3
. The solution was quantitatively diluted to 25.00 mL using
3
, and an IR spectrum of the solution was recorded
using a 2% (v/v) THF/HPLC grade CHCl
3
solution. An IR spectrum
-
1
tance at 2271 cm was noted for each of the data points. For
polymerizations using catalysts 2-4, depolymerization resulting from
the active catalyst end groups was insignificant for up to 1 h after the
dilution was performed. For polymerizations using 1, depolymerization
was inhibited by adding 2% (v/v) of THF to the HPLC grade
chloroform. The concentrations of monomer in the diluted solutions,
and therefore in the polymerizations themselves, were calculated using
a Beer’s law plot. The concentration of the catalyst in the polymeri-
zation was calculated taking into account the dilution factor upon
addition of n-hexyl isocyanate.
of the solution was recorded using a 0.1 mm path length NaCl solution
-
1
cell, and the percent transmittance at 2271 cm was noted. The
concentration of monomer in the diluted solution, and therefore in the
polymerization itself, was calculated using a Beer’s law plot. The
temperature of the oil bath was increased by an appropriate amount,
and the flask was allowed to reach thermal equilibrium with the oil
bath. Another aliquot was removed, and the concentration of monomer
was determined as described above.
Observation of the Active Polymer Endgroup Using IR Spec-
troscopy. In a drybox, an oligomerization of n-hexyl isocyanate was
conducted using the following quantities of reagents: 57.6 mg (0.253
mmol) of 3, 0.050 mL of THF, and 0.600 mL (0.519 g, 4.08 mmol) of
n-hexyl isocyanate. After completion (24 h), a sample of the oligo-
merization mixture was placed between NaCl plates, and an IR spectrum
was recorded. The plates were then separated, and the oligomerization
mixture was exposed to ambient moisture. An IR spectrum was then
recorded. Next, 0.174 g of the oligomerization mixture was transferred
into a volumetric flask and diluted to 2.00 mL using THF. Some of
this solution was placed in a 0.1 mm path length NaCl solution cell,
and an IR spectrum was recorded. Another IR spectrum was recorded
from which the IR spectrum of a separately prepared background
solution of 0.113 g of PHIC in THF was subtracted. The region of
Depolymerization Kinetic Measurements. A polymerization of
n-hexyl isocyanate was conducted using the following quantities of
-
2
reagents: 18.4 mg (7.26 × 10 mmol) of 1, 0.500 mL of toluene, and
.440 mL (0.381 g, 2.99 mmol) of n-hexyl isocyanate. After completion
3 h), volatile materials were removed from the polymerization under
0
(
vacuum, yielding 0.336 g (88% yield) of an orange solid. This was
dissolved in 1.50 mL of toluene, and the time of depolymerization was
kept beginning with the first addition of solvent. The solution was
transferred into a 10 mL round-bottom flask, which was then fitted
with a rubber septum and taken out of the box. The flask was placed
in a water bath thermostatted at 25.0 ( 0.2 °C. Timed aliquots were
removed from the depolymerization solution using a syringe and diluted
volumetrically using HPLC grade CHCl
were recorded using using a 0.1 mm path length NaCl solution cell.
3
. IR spectra of the solutions
interest in all of the spectra was identical. Next, a drop of CH
3
OH
was added to the oligomerization solution, and an IR spectrum was
recorded. Relevant IR stretching frequencies are included within the
text.
-
1
The percent transmittance at 2271 cm was noted for each of the data
points. The concentrations of monomer in the diluted solutions, and
therefore in the depolymerization itself, were calculated using a Beer’s
law plot. The concentration of the catalyst in the polymerization was
calculated assuming that all of the catalyst centers from the polymer-
ization were transferred into the depolymerization.
Acknowledgment. We gratefully acknowledge the National
Science Foundation (Presidential Faculty Fellow Award), du
Pont, Exxon, 3M, BFGoodrich, and Proctor and Gamble for
their financial support of this research. T.E.P. acknowledges
the U.S. Department of Defense for an NDSEG Predoctoral
Fellowship. We also thank Dr. Thomas Seery for help with
the static light-scattering measurements and Susanne Hoff for
information regarding titanium phenoxides as catalysts.
Thermodynamic Measurements. In a drybox, a polymerization
of n-hexyl isocyanate was conducted in a 10 mL Schlenk flask using
-
2
the following quantities of reagents: 11.6 mg (4.58 × 10 mmol) of
1
, 1.00 mL of toluene, and 1.00 mL (0.865 g, 6.80 mmol) of n-hexyl
(45) Berger, M. N.; Tidswell, B. M. J. Polym. Sci., Polym. Symp. 1973,
4
2, 1063.
JA9534516