Cationic organomagnesium complexes as highly active initiators for the ring-opening polymerization of ε-caprolactone

Article abstract of DOI:10.1021/om900669y

Species of the form [LH]⁺[BR₄]^- (R = pentafluorophenyl, phenyl, 1) were synthesized by reaction of Broensted acids with a novel bis-phosphinimine ligand (L = 4,6-(MesN=PPh₂) ₂dibenzofuran). Corresponding cationic complexes [LMg ⁿBu]⁺[BR₄]^- (2) were produced by reaction of [MgⁿBu₂] with [LH]⁺[BR ₄]^-. Organomagnesium species 2a and 2b exhibit extremely high activity as initiators for the polymerization of ε-caprolactone, yielding near-quantitative conversion of monomer to high molecular weight (> 2.0 × 10⁵ g/mol) polymer in 4 min at ambient temperature.

Full text of DOI:10.1021/om900669y

Organometallics 2010, 29, 1079–1084 1079
DOI: 10.1021/om900669y
Cationic Organomagnesium Complexes as Highly Active Initiators for the
Ring-Opening Polymerization of ε-Caprolactone
Benjamin J. Ireland, Craig A. Wheaton, and Paul G. Hayes*
Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive, Lethbridge,
Alberta T1K 3M4, Canada
Received July 28, 2009
Species of the form [LH]þ[BR4]- (R=pentafluorophenyl, phenyl, 1) were synthesized by reaction
€
of Bronsted acids with a novel bis-phosphinimine ligand (L = 4,6-(MesNdPPh2)2dibenzofuran).
Corresponding cationic complexes [LMgnBu]þ[BR4]- (2) were produced by reaction of [MgnBu2]
with [LH]þ[BR4]-. Organomagnesium species 2a and 2b exhibit extremely high activity as initiators
for the polymerization of ε-caprolactone, yielding near-quantitative conversion of monomer to high
molecular weight (>2.0 ꢀ 105 g/mol) polymer in 4 min at ambient temperature.
Introduction
for olefin polymerization.5 Only preliminary efforts have
been reported for extending similar strategies to lactones.6
Herein we describe the synthesis and characterization of a
family of cationic organomagnesium complexes that are
extremely active initiators for ε-caprolactone polymeriza-
tion. While some well-characterized examples of formally
cationic magnesium complexes have been reported,7 and a
variety of neutral organomagnesium species have been ap-
plied to lactone polymerization,8 to the best of our know-
ledge the results described herein constitute the first success-
ful use of a cationic organomagnesium species in ε-capro-
lactone polymerization.
Polylactones are an important class of biodegradable
materials that hold the potential to reduce waste associated
with the disposal of conventional plastics.1 The advancement
of polylactone technology, including the development of new
homogeneous catalysts for lactone polymerization, is thus an
important goal for many researchers, and many excellent
reviews are available.2 Of the commercially viable polylac-
tones (poly(ε-caprolactone), polylactide, and polyglycolide),
poly(ε-caprolactone) is particularly attractive, as it is pro-
duced from an inexpensive precursor, may be processed
under mild conditions (mp ∼60 °C3), and is readily degraded
by naturally occurring microorganisms.4
Magnesium- and zinc-based homogeneous catalyst
systems are particularly attractive for lactone polymeriza-
tion because of their low toxicity and cost paired with high
activity. Magnesium catalysts, however, generally exhibit
higher activity than more electron-rich zinc analogues.2a
Therefore, development of sterically and electronically
unsaturated magnesium complexes presents an opportunity
to achieve even higher levels of polymerization catalytic
activity. Catalyst activation by alkide abstraction from
organometallic species (which results in a sterically and
electronically unsaturated metal center) is well established
The polymerization of ε-caprolactone using a cationic
organomagnesium initiator may conceptually follow either
of two mechanisms: coordination-insertion or activated
chain-end (Scheme 1).8b-g,9 Typical organomagnesium cata-
lysts have been shown to polymerize lactones via the former
(5) Chen, E. Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391–1434.
(6) (a) Haddad, M.; Laghzaoui, M.; Welter, R.; Dagorne, S. Orga-
nometallics 2009, 28, 4584–4592. (b) Dagorne, S.; Le Bideau, F.; Welter, R.;
Bellemin-Laponnaz, S.; Maisse-Franc-ois, A. Chem.;Eur. J. 2007, 13,
3202–3217. (c) Milione, S.; Grisi, F.; Centore, R.; Tuzi, A. Organometallics
ꢀ
2006, 25, 266–274. (d) Lewinski, J.; Horeglad, P.; Dranka, M.; Justyniak, I.
Inorg. Chem. 2004, 43, 5789–5791. (e) Hannant, M. D.; Schormann, M.;
Bochmann, M. J. Chem. Soc., Dalton Trans. 2002, 4071–4073. (f) Emig,
ꢀ
N.; Nguyen, H.; Krautscheid, H.; Reau, R.; Cazaux, J. -B.; Bertrand, G.
Organometallics 1998, 17, 3599–3608.
*Corresponding author. E-mail: p.hayes@uleth.ca.
(7) (a) Pajerski, A. D.; Squiller, E. P.; Parvez, M.; Whittle, R. R.;
Richey, H. G. Organometallics 2005, 24, 809–814. (b) Sarazin, Y.;
Schormann, M.; Bochmann, M. Organometallics 2004, 23, 3296–3302.
(8) Selected examples: (a) Schofield, A. D.; Barros, M. L.; Cushion,
M. G.; Schwarz, A. D.; Mountford, P. Dalton Trans. 2009, 85–96. (b)
(1) (a) Arshady, R. Biodegradable polymers: concepts, criteria, defi-
nitions. In Biodegradable Polymers; Arshady, R., Eds.; PBM Series 2; Citus
Books: London, UK, 2003; pp 1-34. (b) Liebmann-Vinson, A.; Timmins, M.
Biodegradable polymers: degradation mechanisms, part 2. In Biodegradable
Polymers; Arshady, R., Eds.; PBM Series 2; Citus Books: London, UK,
2003; pp 329-372.
(2) (a) Wheaton, C. A.; Hayes, P. G.; Ireland, B. J. Dalton Trans.
2009, 4832–4846. (b) Platel, R. H.; Hodgson, L. M.; Williams, C. K. Polym.
Rev. 2008, 48, 11–63. (c) Wu, J.; Yu, T.-L.; Chen, C.-T., Lin, C.-C. Coord.
Chem. Rev. 2006, 250, 602-626. (d) Dechy-Cabaret, O.; Martin-Vaca, B.;
Bourissou, D. Chem. Rev. 2004, 104, 6147–6176. (e) O'Keefe, B. J.;
Hillmyer, M. A.; Tolman, W. B. J. Chem. Soc., Dalton Trans. 2001,
2215–2224.
ꢀ
Wang, Z.-X.; Qi, C.-Y. Organometallics 2007, 26, 2243–2251. (c) Sanchez-
ꢀ
ꢀ
Barba, L. F.; Garces, A.; Fajardo, M.; Alonso-Moreno, C.; Fernandez-Baeza,
~ ꢀ ꢀ
J.; Otero, A.; Antinolo, A.; Tejeda, J.; Lara-Sanchez, A.; Lopez-Solera, M. I.
Organometallics 2007, 26, 6403–6411. (d) Sanchez-Barba, L. F.; Hughes,
ꢀ
D. L.; Humphrey, S. M.; Bochmann, M. (e) Yu, T.-L.; Wu, C.-C.; Chen, C.-C.;
Huang, B.-H.; Wu, J.; Lin, C.-C. Polymer 2005, 46, 5909–5917. (f) Shueh,
M.-L.; Wang, Y.-S.; Huang, B.-H.; Kuo, C.-Y.; Lin, C.-C. Macromolecules
2004, 37, 5155–5162. (g) Chisholm, M. H.; Eilerts, N. W.; Huffman, J. C.;
Iyer, S. S.; Pacold, M.; Phomphrai, K. J. Am. Chem. Soc. 2000, 122, 11845–
11854.
(3) Di Maio, E.; Innace, S.; Marrazzo, C.; Narkis, M.; Nicolais, L.
Macromol. Symp. 2005, 228, 219–227.
(4) Shimao, M. Curr. Opin. Biotechnol. 2001, 12, 242–247.
(9) Odian, G. Principles of Polymerization, 4th ed.; John Wiley and
Sons: New York, 2004; pp 19-586.
r
2010 American Chemical Society
Published on Web 02/09/2010
pubs.acs.org/Organometallics

1080 Organometallics, Vol. 29, No. 5, 2010
Ireland et al.
Scheme 1. Possible Mechanisms for the Polymerization of ε-Caprolactone via a Cationic Organomagnesium Initiator:
Coordination-Insertion (above) or Activated Chain-End (below)a
a R = anionic functionality, L = neutral ancillary.
process;8b-g however, due to the cationic nature of the
complexes in the present study, both mechanisms have been
considered.
Since coordination-insertion lactone polymerization pre-
sents a scenario in which the catalytic species is intimately
connected to the active site of polymerization, it is considered
preferable, as it presents greater opportunity for subsequent
modulation of the polymerization process by way of ancil-
lary ligand modification. This mechanism requires an anio-
nic initiating group bound to the metal center; thus, an ideal
ancillary ligand for an activated magnesium catalyst of the
form [LMgR]þ[A]- (R=alkyl group, A=noncoordinating
anion) must be formally neutral and strongly electron donat-
ing. Phosphinimines possess such properties and in addition
are chemically robust and easily synthesized in a modular
manner, while providing a useful 31P NMR handle.10 In
addition, several phosphinimine-containing magnesium
complexes have been previously reported.8b,11
Synthesis and Characterization. The chelating bis(phos-
phinimine) architecture employed in the present work was
easily produced in high yield (93.9%) by reaction of
4,6-bis(diphenylphosphino)dibenzofuran with 2 equiv of
2,4,6-trimethylphenylazide (MesN3) under standard Stau-
dinger conditions.12 Single crystals of L suitable for X-ray
diffraction (Figure 1) were grown from a solution of
the compound in a toluene/pentane mixture at ambient
temperature.
Figure 1. Molecular structure of L (50% probability ellipsoids,
˚
H atoms omitted for clarity). Selected bond lengths (A) and
angles (deg): P(1)-N(1) 1.549(1), P(2)-N(2) 1.565(1); P(1)-N-
(1)-C(25) 129.5(1), P(2)-N(2)-C(46) 122.9(1).
Scheme 2. Synthesis of 1a, 1b, and 1c by Protonation of L Using
(i) [HNMe2Ph][B(C6F5)4], (ii) Na[BPh4] and H2O, and (iii) 2
Na[BPh4] and 2 HCl, and Synthesis of Complexes 2a and 2b by
Addition of (iv) [MgnBu2]
It has been previously demonstrated that direct reaction of
neutral organozinc precursors with protonated salts of imi-
nes and phosphinimines affords well-defined cationic zinc
complexes.6e,13,14 In an effort to utilize this methodology for
the preparation of analogous organomagnesium cations,
isolable salts of L, [(L)H]þ[BR4]- (R=C6F5 (1a), Ph (1b)),
(10) (a) Zhu, D.; Budzelaar, P. H. M. Organometallics 2008, 27, 2699–
2705. (b) Courtenay, S.; Walsh, D.; Hawkeswood, S.; Wei, P.; Das, A. K.;
Stephan, D. W. Inorg. Chem. 2007, 46, 3623–3631. (c) Welch, G. C.; Piers,
W. E.; Parvez, M.; McDonald, R. Organometallics 2004, 23, 1811–1818.
(11) (a) Ahmed, S. A.; Hill, M. S.; Hitchcock, P. B. Organometallics
2006, 25, 394–402. (b) Wei, P.; Stephan, D. W. Organometallics 2003, 22,
601–604.
(12) (a) Meyer, J.; Staudinger, H. Helv. Chim. Acta 1919, 2, 635–646.
(b) Alajarin, M.; Lopez-Leonardo, C. L.; Llamas-Lorente, P. L.; Bautista, D.
Synthesis 2000, 14, 2085–2091.
(13) Hannant, M. D.; Schormann, M.; Hughes, D. L.; Bochmann, M.
Inorg. Chim. Acta 2005, 358, 1683–1691.
(14) Wheaton, C. A.; Ireland, B. J.; Hayes, P. G. Organometallics
2009, 28, 1282–1285.
were generated in high yield (1a: 80.1%, 1b: 86.3%) by
€
reaction of L with an appropriate Bronsted acid (Scheme 2).
The doubly protonated species 1c was prepared in a parallel
fashion in 91.2% yield. Though 1a and 1b were not sufficiently
crystalline in our hands for single-crystal X-ray diffraction,
crystals of 1c were isolated and the molecular structure was
determined (Figure 2). By comparison to the structure of L
˚
(P1-N1 = 1.549(1) A), the P-N bonds of 1c (P1-N1 =
1.639(2) A) are elongated as a result of protonation of the
phosphinimine nitrogen atoms.
˚

Article
Organometallics, Vol. 29, No. 5, 2010 1081
Figure 2. Molecular structure of 1c (50% probability ellipsoids;
BPh4-, solvent (acetone), and all H atoms except N-H
(calculated) omitted for clarity). Selected bond lengths (A)
˚
and angles (deg): P(1)-N(1) 1.639(2), N(1)-H(1N) 0.894(2);
P(1)-N(1)-C(25) 125.6(1).
Figure 3. Molecular structure of 2b (30% probability ellipsoids;
BPh4- and H atoms omitted for clarity). Selected bond lengths
(A) and angles (deg): P(1)-N(1) 1.602(5), P(2)-N(2) 1.601(5),
Mg-N(1) 2.086(6), Mg-N(2) 2.077(5), Mg-C(55) 2.13(1);
P(1)-N(1)-Mg 129.6(3), P(2)-N(2)-Mg 130.7(3), N(1)-
Mg-N(2) 132.6(2), N(1)-Mg-C(55) 115.0(3), N(2)-Mg-C-
(55) 112.3(3).
˚
The species L, 1a, 1b, and 1c all display C2v symmetry in
solution, as indicated by a single 31P NMR resonance each at
δ -17.6 (L), 10.1 (1a, 1b), and 28.1 (1c). Both 1a and 1b
1
exhibit a broad H NMR resonance at δ 5.7 attributed to
the N-H proton, whereas the analogous resonance of 1c
appears as a doublet at δ 8.1 (2JPH=9.7 Hz). The molecular
symmetry established spectroscopically suggests that the
N-H protons of 1a and 1b are rapidly exchanging between
the two nitrogen atoms on the NMR time scale. At tempera-
tures as low as -80 °C, NMR experiments were unable to
show the Cs symmetric species; however, a rapid proton
exchange process is corroborated by the chemical shift of the
31P NMR resonances for 1a and 1b, which are intermediate
between those observed for 1c and L.
Table 1. Polymerization of ε-Caprolactone by Initiators 2a and 2b
temp
(°C)
initiator
conc (mM)
initiator
loading
conversion
after 4 min
initiator
2a
2b
2b
2b
23
23
23
0
4.2
4.4
0.37
2.1
0.77%
0.77%
0.19%
0.56%
95%
90%
75%
73%
2b were observed upfield of 0 ppm (δ -15.8 for 2a, δ -5.6 for
2b) and the 19F Δδmp of 2a was 3.8 ppm.16 Thus, it is
reasonable to suggest that both 2a and 2b are best described
as extremely weakly interacting ion pairs, and although
anions can have a significant effect on polymerization,17 that
does not appear to be the case here.
Treatment of 1a or 1b with 1 equiv of di-n-butylmagne-
sium resulted in loss of n-butane and concomitant formation
of 2a or 2b, respectively (Scheme 2). The remaining n-butyl
1
group was well resolved by H NMR spectroscopy, giving
rise to diagnostic resonances at δ 1.38-1.32 (CH2CH2), 0.99
(CH3), and -0.13 (MgCH2) in each case. C2v symmetry was
preserved, and a large downfield shift in the 31P NMR
resonance (δ -17.6 to 23.0 for 2a, δ -17.6 to 23.2 for 2b)
was observed upon coordination of L to the magnesium
center. This observation is consistent with the behavior of a
closely related mono-phosphinimine ligand upon coordina-
tion to zinc14 and with other phosphinimines upon coordina-
tion to magnesium.8b,11 Complexes 2a and 2b were isolated
as thermally stable, air-sensitive solids in good yield (2a:
73%, 2b: 91%). Notably, complexes 2a and 2b were prepared
in 55% and 74% overall yields, respectively, from 4,6-bis-
(diphenylphosphino)dibenzofuran.
Polymerization of ε-Caprolactone. Preliminary examina-
tion of 2a and 2b in the ring-opening polymerization of ε-
caprolactone revealed remarkably high activity. Investiga-
tions were conducted at initiator concentrations of approxi-
mately 4 mM (0.77 mol % relative to the lactone monomer)
in benzene-d6. In both cases, 1H NMR spectroscopy
indicated >90% conversion of ε-caprolactone to poly-
(ε-caprolactone) in 4 min at ambient temperature (Table 1).
-
As the BR4 anions were established to be extremely
weakly coordinating, they were not expected to substantially
influence the polymerization process. Indeed, similar activity
was observed for 2a and 2b at 23 °C (Table 1). Hence,
subsequent reactions focused on the less expensive initiator
2b. Notably, 2b maintains impressive activity at reduced
concentrations (0.37 mM) and loading (0.19%). Likewise,
high activity and a linear relationship between ln[monomer]
and time (pseudo-first-order kinetics) were observed at 0 °C.
In fact, 2b was active at temperatures as low as -40 °C;
however, at temperatures less than 0 °C the exceedingly high
viscosity of the reaction mixture rendered it difficult to
achieve complete conversion.
The polymerization of ε-caprolactone by 2a and 2b was
studied by NMR spectroscopy in an attempt to elucidate the
operative polymerization mechanism (Scheme 1). In situ
examination of the polymerization process indicated
cleavage of the Mg-C bond in both cases (evidenced by
The structure of 2b was confirmed by a single-crystal
X-ray diffraction study (Figure 3). The ligand is κ2-bound
˚
˚
to Mg (Mg-N 2.078(5), 2.086(5) A; Mg-C 2.13(1) A), result-
ing in trigonal-planar geometry about magnesium (sum of
angles=360.0°). No noteworthy cation-anion interactions
˚
were observed (shortest Mg-Canion contact=6.3 A).
Though 2a eluded exhaustive crystallization efforts, the
-
weakly coordinating nature of B(C6F5)4 has been well
established.15 Furthermore, the 1H, 31P, and 13C NMR
spectral properties of [MgnBu]þ in 2a and 2b are virtually
-
identical, suggesting minimal interaction with the BR4
counterions. Likewise, the 11B NMR resonances of 2a and
(15) Krossing, I.; Raabe, I. Angew. Chem., Int. Ed. 2004, 43, 2066–
2090.
(16) (a) Horton, A. D.; de With, J.; van der Linden, A. J.; van de Weg,
H. Organometallics 1996, 15, 2672–2674. (b) Drewitt, M. J.; Niedermann,
M.; Kumar, R.; Baird, M. C. Inorg. Chim. Acta 2002, 335, 43–51.
(17) Roberts, J. A. S.; Chen, M.-C.; Seyam, A. M.; Li, L.; Zuccaccia,
C.; Stahl, N. G.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 12713–12733.

1082 Organometallics, Vol. 29, No. 5, 2010
Ireland et al.
Table 2. GPC Results for Poly(ε-caprolactone) Produced Using
Initiator 2b: Weight Average (Mw), Number Average (Mn), and
Corresponding PDI (Mw/Mn) Values
solvents (Cambridge Isotopes) were dried with appropriate
drying agents, vacuum transferred, and stored under an inert
atmosphere prior to use. ε-Caprolactone was dried over CaH2,
distilled, and stored under an inert atmosphere prior to use. All
other materials were obtained in high purity (Sigma-Aldrich)
and used without additional purification. NMR spectra (1H
(300.13 MHz), 13C{1H} (75.47 MHz), 31P{1H} (121.48 MHz),
19F (282.42 MHz), and 11B (96.29 MHz)) were collected using a
Bruker Avance II NMR spectrometer equipped with a variable-
temperature unit. Spectra were collected at ambient temperature
unless otherwise noted and referenced to residual proteo solvent
resonances (1H), solvent 13C resonances (13C{1H}), or an exter-
nal standard (triphenylphosphine (31P{1H}), trifluorotoluene
(19F), or boron trifluoride diethyl etherate (11B)) depending on
the nucleus of interest. 1H and 13C NMR peak assignments were
facilitated by DEPT-45, DEPT-90, DEPT-135, COSY, and
HSQC experiments. X-ray crystal structures were collected
using a Bruker AXS SMART APEX II single-crystal X-ray
batch
Mw
Mn
PDI (Mw/Mn)
1
2
1.9 ꢀ 105
2.1 ꢀ 105
1.2 ꢀ 105
1.3 ꢀ 105
1.6
1.5
disappearance of a characteristic Mg-CH2 resonance at
δ -0.13 in the 1H NMR spectrum), and while the methylene
resonances could not be detected due to overlapping polymer
signals, the n-butyl-CH3 resonance is clearly visible at δ 0.86
in the 1H NMR spectrum. However, this same n-butyl
resonance could not be detected in the spectra of isolated
oligomer samples (generated using 10 mol % 2b), clearly
demonstrating that the polymer does not bear an n-butyl end
group. Corroboration of these results via MALDI-ToF mass
spectrometry was not successful despite an exhaustive exa-
mination of a variety of polymer samples.7b,18 Thus, while
the ultimate fate of the initiator is unknown, the absence of
an n-butyl end group on isolated oligo(ε-caprolactone)
samples suggests that lactone polymerizaton occurs via an
activated chain-end process (Scheme 1).
Poly(ε-caprolactone) samples produced using 2b (0.24%,
23 °C) possessed average molecular weights (Mw) as high as
2 ꢀ 105 g/mol and PDIs 1.6 or less, as determined by GPC
analysis (Table 2).2a The average molecular weights of the
polymers were higher than expected (i.e., a controlled poly-
merization utilizing 0.24 mol % initiator would be expected
to afford polymer with a molecular weight of 4.8 ꢀ 104 g/
mol), which may be attributed to slow initiation relative to
propagation in the polymerization process, and therefore
only moderate initiator efficiency. A relatively slow initia-
tion step is also consistent with the observed polydispersity
values. This could occur in the context of an activated chain-
end mechanism, whereby the cationic chain end may be less
stable and more reactive than the initiating cationic magne-
sium species.
˚
diffractometer (Mo KR (λ = 0.71073 A)). Elemental analyses
were performed using an Elementar Vario Microcube.
Synthesis of 4,6-(PPh2)2dibenzofuran. The ligand precursor
4,6-(PPh2)2dibenzofuran was prepared as described by Kranen-
burg et al.20 with several modifications. A 250 mL round-
bottom flask was charged with 2.6869 g (15.974 mmol) of
dibenzofuran, to which 100 mL of diethyl ether was added by
vacuum transfer at -78 °C. Tetramethylethylenediamine
(TMEDA) was injected slowly (7.2 mL, 5.6 g, 48 mmol), and
the suspension was allowed to warm to ambient temperature
over approximately 20 min. The dibenzofuran fully dissolved to
afford a light yellow solution. This solution was cooled back to
-78 °C, and a solution of sec-butyllithium (35 mL at 1.4 mol/L
in heptane, 49 mmol) was added dropwise. The reaction mixture
was stirred for 2 h, producing a light green suspension, which
became dark green upon slow warming to ambient temperature.
The reaction mixture was stirred for an additional 6 h and then
cooled to -78 °C. Beginning 9 h after the initial injection of sec-
butyllithium, 9.0 mL (11 g, 50 mmol) of neat chlorodiphenyl-
phosphine was injected rapidly. An immediate color change
from green to white was noted. The reaction mixture was
gradually warmed back to ambient temperature and stirred
for an additional 14 h, during which a light brown suspension
formed. The solvent was removed in vacuo. All subsequent
manipulations were performed under aerobic conditions. The
resulting light brown oil was dissolved in 80 mL of dichloro-
methane and quenched with 50 mL of distilled water. The
aqueous phase was removed, and the organic phase was washed
with three subsequent 50 mL fractions of distilled water. The
organic phase was dried thoroughly in vacuo, and the resultant
light brown oily solid was washed five times with 50 mL
fractions of pentane. During each washing procedure, the
mixture was sonicated and vigorously stirred for approximately
5 min prior to filtration. The resultant white solid was dried
thoroughly in vacuo, affording 6.06 g (11.3 mmol, 70.9%) of the
Conclusion
Cationic organomagnesium species (2a and 2b), prepared
by treating di-n-butylmagnesium with novel species 1a and
1b, respectively, have demonstrated exceptionally high ac-
tivity for the polymerization of ε-caprolactone. The observed
results are extremely promising, and future generations of
this initiator will be designed in hopes of maintaining high
activity while giving enhanced molecular weight control.
1
desired product. H and 31P{1H} NMR spectra matched pub-
Experimental Section
lished results.19 31P{1H} NMR (chloroform-d): δ -16.8 (s). 1H
NMR (chloroform-d): δ 7.95 (d, 3JHH=7.5 Hz, 2H, dbf-C1/9H),
General Procedures. All manipulations of air-sensitive mate-
rials and reagents were conducted using high-vacuum techni-
ques19 under a purified argon atmosphere or in a glovebox
(MBraun Labmaster 130). Proteo solvents were purified using
an MBraun solvent purification system (MB-SPS), stored in
Teflon-sealed glass vessels over appropriate drying agents, and
vacuum transferred directly to reaction vessels. Deuteurated
7.35-7.20 (ov m, 22H, dbf-C3/7H þ o-PPh2 þ m-PPh2
p-PPh2), 7.10 (m, 2H dbf-C2/8H).
þ
Synthesis of 4,6-(MesNdPPh2)2dibenzofuran (L). A 500 mL
Teflon-sealed glass reaction vessel was charged with 5.7732 g
(10.760 mmol) of 4,6-(PPh2)2dibenzofuran. The precursor dis-
solved fully in 110 mL of toluene, and then excess neat 2,4,6-
trimethylphenylazide (mesityl azide, MesN3)21 (4.255 g, 26.39
mmol) was added. Evolution of a colorless gas was noted within
(18) Fontaine, F.-G. Personal communication. High molecular
weight poly(ε-caprolactone) has proven impossible for the Fontaine
group to observe by MALDI-ToF mass spectrometry.
(19) Burger, B. J.; Bercaw J. E. Vacuum line techniques for handling
air-sensitive organometallic compounds. In Experimental Organometal-
lic Chemistry. A Practicum in Synthesis and Characterization; Wayda,
A. L., Darensbourg, M. Y., Eds.; ACS Symposium Series 357; American
Chemical Society: Washington DC, 1987; pp 79-98.
(20) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J.;
van Leeuwen, P. W. N. M.; Goubitz, K.; Fraanje, J. Organometallics
1995, 14, 3081–3089.
(21) Murata, S.; Abe, S.; Tomioka, H. J. Org. Chem. 1997, 62, 3055–
3061.

Article
Organometallics, Vol. 29, No. 5, 2010 1083
5 min, and the solution was stirred at ambient temperature, with
occasional venting, for 60 min. The temperature was then
gradually raised to 65 °C, and the solution was stirred for 16
h, over which time the color changed from yellow to light brown.
An additional 0.459 g (2.85 mmol) of neat MesN3 was added,
and the reaction mixture was allowed to stir at 65 °C until the 31P
NMR spectrum of crude reaction mixture aliquots indicated
that the reaction had reached completion (approximately 2
additional h). The solution was cooled to ambient temperature
and transferred to a 100 mL round-bottom flask in two fractions
of approximately 60 mL each. The solvent was removed in vacuo
between fractions and after the full volume had been trans-
ferred, yielding an oily yellow solid. All subsequent manipula-
tions were conducted under aerobic conditions. The product
was washed five times with 50 mL fractions of hexane. During
each washing procedure, the mixture was sonicated and vigor-
ously stirred for approximately 5 min prior to filtration. The
product was collected as a white powder and dried in vacuo.
Total yield was 93.9% (8.10 g, 10.1 mmol). 31P{1H} NMR
portions of distilled water. The organic layer was then tho-
roughly dried in vacuo for 14 h, yielding the desired product as an
analytically pure, light yellow solid in high yield (1.2508 g,
1.1136 mmol, 86.26%). 31P{1H} NMR (benzene-d6): δ 10.1
(s). 31P{1H} NMR (chloroform-d): δ 9.5 (s). 1H NMR
3
(chloroform-d): δ 8.14 (d, JHH = 6.3 Hz, 2H, dbf-C1/9H),
7.46-7.31 (ov m, 24H, dbf-C2/8H þ dbf-C3/7H þ o-PPh2 þ p-
PPh2 þ o-BPh4-), 7.30-7.19 (m, 8H, m-PPh2), 6.95 (dd, 3JHH
=
7.4 Hz, 3JHH=6.1 Hz, 8H, m-BPh4-), 6.82 (t, 3JHH=7.4 Hz, 4H,
p-BPh4-), 6.58 (s, 4H, m-Mes), 5.69 (br s, 1H, NH), 2.18 (s, 6H,
p-Mes), 1.56 (s, 12H, o-Mes). 13C{1H} NMR (chloroform-d): δ
1
164.4 (1:1:1:1 q, JCB = 49.1 Hz, ipso-BPh4-), 157.0 (s, dbf-
quaternary), 136.5 (s, m-BPh4-), 134.5 (s), 133.9 (s, dbf-C2/8),
133.6 (s), 132.5 (d, JCP=9.8 Hz, o-PPh2), 131.2 (s), 129.6 (s,
2
p-PPh2), 129.3 (s, m-PPh2), 128.5 (s, m-Mes), 127.4 (s, dbf-C1/9),
125.5 (s, o-BPh4-), 125.1 (s) 124.2 (d, 2JCP=6.0 Hz, dbf-C3/7),
123.4 (s), 121.6 (s, p-BPh4-), 20.8 (s, o-Mes (CCH3)), 20.2 (s, p-
Mes (CCH3)). ipso-PPh2 not observed. 11B NMR (chloroform-
d): δ -6.5 (br s). Anal. Calcd (%) for C78H69BN2OP2: C, 83.39;
H, 6.20; N, 2.49. Found: C, 83.24; H, 6.11; N, 2.51.
1
(benzene-d6): δ -17.6 (s). H NMR (benzene-d6): δ 7.82 (dd,
Synthesis of [H2-4,6-(MesNdPPh2)2dibenzofuran][BPh4]2
(1c). An excess of 1 M aqueous HCl (0.50 mL, 0.50 mmol) was
added to a suspension of L (0.110 g, 0.137 mmol) in methanol (5
mL). With stirring, a slight excess of NaBPh4 (0.112 g, 0.327
mmol) in MeOH (5 mL) was added, immediately yielding a
fluffy white precipitate. Stirring was continued for another
5 min. The precipitate was collected by filtration, washed three
times with methanol, and dried in vacuo to give the product in
91.2% yield (0.170 g, 0.125 mmol). 31P{1H} NMR (acetone-d6):
δ 28.1 (s). 1H NMR (acetone-d6): δ 8.56 (d, 2H, 3JHH=7.8 Hz,
J = 12 Hz, J = 9.2 Hz, 2H, dbf-C1/9H), 7.71-7.60 (m, 8H,
o-PPh2), 7.57 (m, 2H, dbf-C3/7H), 7.02-6.82 (br ov m, 18H, dbf-
C2/8H þ m-PPh2 þ p-PPh2 þ m-Mes), 2.27 (s, 6H, p-Mes), 1.93
(s, 12H, o-Mes). 13C{1H} NMR (benzene-d6): δ 157.0 (s, dbf-
quaternary), 145.2 (s, dbf-quaternary), 133.0 (s, p-Mes (CCH3)),
3
2
132.7 (d, JCP = 7.5 Hz, m-PPh2), 132.0 (d, JCP = 10.6 Hz,
o-PPh2), 131.9 (d, 1JCP=50.6 Hz, dbf-C4/6), 131.3 (s, dbf-C1/9),
129.0 (s, dbf-C2/8), 128.6 (s, m-Mes), 127.0 (d, JCP =3.0 Hz,
3
2
o-Mes (CCH3)), 124.5 (d, JCP = 6.8 Hz, ipso-Mes), 124.0 (s,
p-PPh2), 123.1 (d, 2JCP=9.8 Hz, dbf-C3/7), 121.5 (d, 1JCP=93.6
Hz, ipso-PPh2), 21.1 (s, o-Mes (CCH3)), 21.0 (s, p-Mes (CCH3)).
Anal. Calcd (%) for C54H48N2OP2: C, 80.76; H, 6.04; N, 3.48.
Found: C, 80.46; H, 6.03; N, 3.49.
dbf-C1/9H), 8.14 (d, 2H, 3JHP=9.7 Hz, NH), 8.08 (dd, 2H, 3JHP
=
14.3 Hz, 3JHH=7.8 Hz, dbf-C3/7H), 7.61 (td, 2H, 3JHH=7.8 Hz,
4JHP=1.4 Hz, dbf-C2/8H), 7.06 (t, 4H, 3JHH=7.4 Hz, p-PPh2),
6.84 (td, 8H, 3JHH-=7.8 Hz, 4JHH=3.8 Hz, m-PPh2), 6.78-6.65
Synthesis of [H-4,6-(MesNdPPh2)2dibenzofuran][B(C6F5)4]
(1a). A 50 mL round-bottom flask was charged with 0.2710 g
(0.3375 mmol) of L, 0.2659 g (0.3319 mmol) of [HNMe2Ph]-
[B(C6F5)4], and 10 mL of benzene. The solution was stirred for
10 min, and the benzene was removed in vacuo, affording an oily,
light yellow solid containing the desired product and Me2NPh.
The flask was attached to a swivel frit apparatus, and the solid
was washed three times with 10 mL portions of pentane. During
each washing procedure, the mixture was sonicated and stirred
for several minutes before filtration. The resultant light yellow
solid was dried in vacuo for 20 h. A total of 0.3945 g (0.2660
mmol) of [H-4,6-(MesNdPPh2)2dbf][B(C6F5)4] was recovered
as an analytically pure, light yellow solid (80.1% yield). 31P{1H}
NMR (benzene-d6): δ 10.1 (s). 31P{1H} NMR (chloroform-d): δ
9.4 (s). 1H NMR (chloroform-d): δ 8.30 (d, 3JHH=6.0 Hz, 2H,
dbf-C1/9H), 7.57-7.21 (br ov m, 24H, dbf-C2/8H þ dbf-C3/7H þ
o-PPh2 þ m-PPh2 þ p-PPh2), 6.58 (s, 4H, m-Mes), 5.72 (br s, 1H,
NH), 2.17 (s, 6H, p-Mes), 1.55 (s, 12H, o-Mes). 13C{1H} NMR
(chloroform-d): δ 157.2 (s, dbf-quaternary), 134.6 (d, 3JCP=5.3
3
(m, 24H, o-BPh4 þ o-PPh2), 6.47 (t, 16H, JHH=7.2 Hz, m-
3
BPh4-), 6.33 (t, 8H, JHH=7.2 Hz, p-BPh4-), 6.22 (s, 4H, m-
Mes), 1.63 (s, 6H, p-Mes), 1.26 (s, 12H, o-Mes). 13C NMR
(acetone-d6): 165.0 (1:1:1:1 q, 1JCB=49.4 Hz, ipso-BPh4-), 157.6
(d, JCP=3.4 Hz), 138.2 (s), 137.1 (1:1:1:1 q, 3JCB=1.4 Hz, m-
BPh4-), 137.0 (br s), 135.0 (s), 134.3 (d, JCP=11.4 Hz), 132.6 (s),
131.1 (s), 130.9 (s), 130.7 (d, JCP=1.9 Hz), 130.3 (s), 129.8 (s),
127.8 (s), 127.0 (d, JCP=11.7 Hz), 126.1 (1:1:1:1 q, 2JCB=2.8 Hz,
4
o-BPh4-), 122.3 (1:1:1:1 q, JCB=0.5 Hz, p-BPh4-), 116.1 (s),
20.8 (s, p-Mes (CCH3)), 19.8 (s, o-Mes (CCH3)). 11B NMR
(acetone-d6): δ -6.5 (br s). Anal. Calcd (%) for C102H90B2-
N2OP2 C3H6O: C, 83.99; H, 6.44; N, 1.87. Found: C, 83.89; H,
6.19; N, 1.94.
3
Synthesis of [4,6-(MesNdPPh2)2dibenzofuranMgBu][B(C6-
F5)4] (2a). Under argon, a 50 mL round-bottom flask was
charged with 0.1791 g (0.1207 mmol) of 1a to which 12 mL of
benzene was added. Di(n-butyl)magnesium (0.112 mL of 1.0 M
solution in heptane, 0.11 mmol) was slowly injected, and evolu-
tion of a colorless gas was noted. The solution was stirred for
50 min; then benzene was removed in vacuo. This afforded the
desired product as a pale yellow solid in 73% yield (0.1286 g,
0.08226 mmol). 31P{1H} NMR (benzene-d6): δ 23.0 (s). 1H
2
Hz, dbf-C2/8), 134.0 (s), 133.4 (s), 132.6 (d, JCP = 10.6 Hz,
o-PPh2), 131.2 (s), 130.0 (s, p-PPh2), 129.8 (s, m-PPh2), 129.5 (s,
m-Mes), 127.1 (s, dbf-C1/9), 125.4 (s), 124.0 (s), 123.3 (s), 20.8 (s,
p-Mes (CCH3)), 20.1 (s, o-Mes (CCH3)). B(C6F5)4- resonances
not reported. ipso-PPh2 not observed. 19F NMR (benzene-d6):
δ -130.8 (br d, 3JFF=11 Hz, 8F, o-C6F5), -161.7 (t, 3JFF=22
Hz, 4F, p-C6F5), -165.5 (m, 8F, m-C6F5). 11B NMR
(chloroform-d): δ -16.7 (br s). Anal. Calcd (%) for C78H49BF20-
N2OP2: C, 63.17; H, 3.34; N, 1.89. Found: C, 63.34; H, 3.37;
N, 1.95.
Synthesis of [H-4,6-(MesNdPPh2)2dibenzofuran][BPh4] (1b).
Under aerobic conditions, two solutions;one containing
1.0531 g (1.3115 mmol) of previously prepared L in 125 mL of
benzene, the other containing 0.4418 g (1.291 mmol) of NaBPh4
in 75 mL of distilled water;were prepared. The aqueous
solution was added to the organic solution in a 500 mL round-
bottom flask, and the mixture was stirred vigorously for 25 min.
The organic layer was decanted and washed with three 50 mL
3
NMR (benzene-d6): δ 7.80 (d, JHH=6.0 Hz, 2H, dbf-C1/9H),
7.28 (dd, 3JHP=12 Hz, 3JHH=9.2 Hz, 8H, o-PPh2), 7.09-6.97
(br ov m, 6H, dbf-C2/8H þ p-PPh2), 6.97-6.80 (br ov m, 10H,
dbf-C3/7H þ m-PPh2), 6.36 (s, 4H, m-Mes), 2.04 (s, 6H, p-Mes),
1.50 (s, 12H, o-Mes), 1.38-1.32 (ov m, 4H, MgCH2CH2CH2-
CH3), 0.99 (t, 3JHH=7.3 Hz, 3H, MgCH2CH2CH2CH3), -0.13
3
(t, JHH =9.2 Hz, 2H, MgCH2CH2CH2CH3). 13C{1H} NMR
(benzene-d6): δ 156.8 (s, dbf-quaternary), 137.3 (s, p-Mes
2
(CCH3)), 135.6 (d, JCP = 6.8 Hz, dbf-quaternary), 134.1 (d,
2JCP=9.1 Hz, o-PPh2), 133.9 (s), 133.7 (d, 1JCP=45.5 Hz, dbf-
C4/6), 133.1 (d, 2JCP=9.8 Hz, dbf-C3/7), 130.0 (s, m-Mes), 129.6
(s), 129.4 (s), 128.1 (s), 126.4 (s), 125.3 (d, JCP=8.3 Hz), 112.7 (d,
1JCP=106 Hz, ipso-PPh2), 32.0, 30.2 (s, MgCH2CH2CH2CH3),
20.6 (s, p-Mes (CCH3)), 20.1 (s, o-Mes (CCH3)), 14.1 (s,

1084 Organometallics, Vol. 29, No. 5, 2010
Ireland et al.
MgCH2CH2CH2CH3), 12.0 (s, MgCH2CH2CH2CH3). B-
An NMR tube was charged with 0.0010 g (0.00083 mmol) of
2b to which 2.2 mL of benzene-d6 was added. The tube was
capped with a rubber NMR tube septum, which was then
wrapped in parafilm and shaken vigorously. Dry, distilled ε-
caprolactone (48 μL, 0.43 mmol, 5.2 ꢀ 102 equiv) was measured
under an inert atmosphere into a 100.0 μL gastight micro-
syringe, which was sealed by inserting the needle into a rubber
septum until immediately before addition to the reaction mix-
ture. Prior to monomer injection, all appropriate instrumental
parameters were set and NMR spectra of the initiator were
collected. The sample was then removed from the instrument,
injected with the monomer, shaken, and reinserted into the
NMR spectrometer. Collection of NMR data began within
60 s of injection of the monomer. Conversion percentages were
determined by integration of the most downfield methylene
resonance (-COOCH2-) of the polymer (1H NMR (benzene-
d6): δ 3.98 (t, 3JHH=6.1 Hz, 2H)) relative to those of the residual
monomer (1H NMR (benzene-d6): δ 3.59 (t, 3JHH=6.1 Hz, 2H)),
as these resonances were most clearly resolved from all other
monomer, polymer, initiator, and residual solvent resonances.
Preparation of Isolated Poly(ε-caprolactone) Samples. A re-
presentative procedure for the polymerization of ε-caprolactone
by 2b is described herein. All preparations of polymer and
oligomer samples for GPC, NMR, or MALDI-ToF analysis
used a similar methodology.
Under an inert atmosphere, a 50 mL round-bottom flask
was charged with 0.0310 g (0.0258 mmol) of 2b, to which 10 mL
of benzene was added. The solution was stirred rapidly, and
1.20 mL (1.24 g, 10.8 mmol, 418 equiv) of ε-caprolactone was
injected, resulting in the immediate formation of a thick gel.
After 30 min, the reaction mixture was transferred to a 60 mL
syringe and added to 100 mL of rapidly stirring methanol under
aerobic conditions. This resulted in the immediate precipitation
of polymer, which formed a single solid mass that could easily be
mechanically separated. The polymer was dried in vacuo, yield-
ing 0.700 g of material (∼57%).
-
(C6F5)4 resonances not reported. 19F NMR (benzene-d6):
3
3
δ -130.7 (d, JFF = 11 Hz, 8F, o-C6F5), -161.7 (t, JFF
=
22 Hz, 4F, p-C6F5), -165.5 (m, 8F, m-C6F5). 11B NMR
(benzene-d6): δ -15.8 (br s). Anal. Calcd (%) for C82H57BF20-
MgN2OP2: C, 62.99; H, 3.68; N, 1.79. Found: C, 62.17; H, 3.86;
N, 1.84.
Synthesis of [4,6-(MesNdPPh2)2dibenzofuranMgBu][BPh4]
(2b). Under argon, a 100 mL round-bottom flask was charged
with 0.7422 g (0.6608 mmol) of 1b to which 40 mL of benzene
was added. A solution of Di(n-butyl)magnesium (0.67 mL of
1.0 M solution in heptane, 0.67 mmol) in 4 mL of benzene was
slowly injected. Evolution of a gas was noted, followed by a
color change from yellow to pale pink as the reaction mixture
was stirred for 30 min at ambient temperature. The solvent was
removed in vacuo, yielding the desired material as a white solid
(0.7242 g, 0.6017 mmol, 91.07%). 31P{1H} NMR (benzene-d6): δ
23.2 (s). 1H NMR (benzene-d6): δ 8.09-8.01 (br m, 8H,
3
o-BPh4-) 7.64 (d, JHH = 6.0 Hz, 2H, dbf-C1/9H), 7.26 (dd,
3JHP=12 Hz, 3JHH=9.1 Hz, 8H, o-PPh2), 7.19 (ov t, 3JHH=7.4
Hz, 4H, p-BPh4-), 7.09-6.97 (br ov m, 14H, dbf-C2/8H þ
p-PPh2 þ m-BPh4-), 6.97-6.84 (br ov m, 10H, dbf-C3/7H þ
m-PPh2), 6.35 (s, 4H, m-Mes), 2.03 (s, 6H, p-Mes), 1.52 (s, 12H,
o-Mes), 1.38-1.32 (ov m, 4H, MgCH2CH2CH2CH3), 0.99 (t,
3JHH=7.3 Hz, 3H, MgCH2CH2CH2CH3), -0.13 (t, 3JHH=9.2
Hz, 2H, MgCH2CH2CH2CH3). 13C{1H} NMR (benzene-d6): δ
1
165.4 (1:1:1:1 q, JCB = 48.3 Hz, ipso-BPh4-), 156.6 (s, dbf-
quaternary), 137.5 (s, m-BPh4-), 137.2 (s), 135.7 (d, 2JCP=6.8
Hz, dbf-quaternary), 133.9 (ov, o-PPh2), 133.8 (s), 133.1 (d, 2JCP
=9.1 Hz, dbf-C3/7), 129.9 (s, m-Mes), 129.7 (s), 129.5 (s), 128.1
(s), 126.5 (s), 126.2 (s, o-BPh4-), 125.4 (d, JCP=8.3 Hz), 122.2
(s, p-BPh4-), 112.2 (d, JCP = 107 Hz, ipso-Ph), 32.0, 30.2
1
(s, MgCH2CH2CH2CH3), 20.6 (s, p-Mes (CCH3)), 20.3 (s,
o-Mes (CCH3)), 14.0 (s, MgCH2CH2CH2CH3), 11.9 (s,
MgCH2CH2CH2CH3). Dbf-C4/6 not observed. 11B NMR
(benzene-d6): δ -5.6 (br s). Anal. Calcd (%) for C82H77BMgN2-
OP2: C, 81.81; H, 6.46; N, 2.33. Found: C, 80.85; H, 6.33;
N, 2.72.
General Procedures for the Polymerization of ε-Caprolactone.
Both 2a and 2b were found to be active in the polymerization of
ε-caprolactone. Representative procedures for examination of
polymerization activity by NMR spectroscopy and preparation
of polymer samples for GPC analysis are described herein.
GPC analyses were performed using a Viscotek Triple Detector
GPC system outfitted with a model 270 Dual Detector Plat-
form (four-capillary viscometer and light-scattering detector).
Polymer samples were run in THF at a concentration of
1 mg/mL.
Acknowledgment. P.G.H. acknowledges NSERC for a
Discovery Grant, the Canadian Foundation for Innova-
tion for a Leaders Opportunity Grant, and the University
of Lethbridge for a start-up fund. B.J.I. and C.A.W.
acknowledge NSERC (CGS-M and CGS-D, re-
spectively), Alberta Ingenuity Fund, and the University
of Lethbridge for student awards. The authors thank Dr.
Andrew R. McWilliams of Ryerson University for GPC
measurements and helpful discussions as well as Dr.
^ ꢀ
Adrien Cote (Xerox Research Centre of Canada) for
In Situ NMR Analysis of ε-Caprolactone Polymerization. A
representative procedure for the polymerization of ε-caprolac-
tone by 2b is described herein. All NMR-scale polymerization
procedures using 2a and 2b made use of similar methods.
Polymerization reactions at low temperature (-40 to 0 °C) were
performed by allowing the solution containing initiator to
equilibrate within the precooled instrument for 20 min prior
to monomer injection. Low-temperature reactions were run in
toluene-d8 rather than benzene-d6.
expert assistance with X-ray crystallography. Dr. Randy
Whittal (University of Alberta) and Ms. Qiao Wu
(University of Calgary) are acknowledged for mass spec-
trometry measurements.
Supporting Information Available: Supporting experimental
and X-ray crystallographic data in PDF format and CIF files are
available free of charge via the Internet at http://pubs.acs.org.