Eun Sil Jang1, Jeremy M John1, Richard R Schrock1. 1. Department of Chemistry 6-331, Massachusetts Institute of Technology , Cambridge, Massachusetts 02139, United States.
Abstract
Cis,syndiotacticA-alt-B copolymers, where A and B are two enantiomerically pure trans-2,3-disubstituted-5,6-norbornenes with "opposite" chiralities, can be prepared with stereogenic-at-metal initiators of the type M(NR)(CHR')(OR")(pyrrolide). Formation of a high percentage of alternating AB copolymer linkages relies on an inversion of chirality at the metal with each propagating step and a relatively fast formation of an AB sequence as a consequence of a preferred diastereomeric relationship between the chirality at the metal and the chirality of the monomer. This approach to formation of an alternating AB copolymer contrasts dramatically with the principle of forming AB copolymers from achiral monomers and catalysts.
Cis,syndiotacticA-alt-B copolymers, where A and B are two enantiomerically pure trans-2,3-disubstituted-5,6-norbornenes with "opposite" chiralities, can be prepared with stereogenic-at-metal initiators of the type M(NR)(CHR')(OR")(pyrrolide). Formation of a high percentage of alternating AB copolymer linkages relies on an inversion of chirality at the metal with each propagating step and a relatively fast formation of an AB sequence as a consequence of a preferred diastereomeric relationship between the chirality at the metal and the chirality of the monomer. This approach to formation of an alternating AB copolymer contrasts dramatically with the principle of forming AB copolymers from achiral monomers and catalysts.
Copolymers
in which monomers A and B are
incorporated in an alternating manner, poly(A-alt-B), are rare.[1−5] Examples are alternating AB copolymers formed from
CO and olefins or CO2 and epoxides. In these cases alternation
is greatly assisted by the fact that one partner (CO or CO2) does not itself polymerize. A few alternating AB copolymers
have been formed through ring-opening metathesis polymerization (ROMP)
of cyclic olefin monomers,[6−28] but in these circumstances both A and B usually can be homopolymerized and the stereochemistry of the C=C
bond in the polymer is not fixed. One exception is the alternating AB copolymer which has all trans C=C
bonds and <5% AA errors formed from a norbornene-like
monomer (B) that is slow to homopolymerize and cyclooctene
or cycloheptene (A). The most successful initiators are
of the type Mo(NR)(CHCMe2Ph)[OCMe(CF3)2]2 (R = 2,6-Me2C6H3 or
2,6-i-Pr2C6H3),
two well-defined alkylidene initiators that contain Mo or W out of
many that have proven useful for preparing stereoregular polymers
from norbornenes and norbornadienes.[29−35]Among the well-defined Mo or W initiators
are those that contain
a stereogenic metal, e.g., 1a–1c.
These initiators can produce a special category of stereoregular A-alt-B copolymers made from
a racemic chiral monomer where A and B are
enantiomers. These “A-alt-A*” copolymers have a basic cis, syndiotactic structure (eq ), which is readily proven through 1H and 13C NMR studies.[36,37] Only one example of
a stereoregular A-alt-A* copolymer has been reported in the older literature.[38,39] The cis structure is formed when 1a or 1b reacts with monomer to yield all cis metallacycles in trigonal bipyramidal (TBP) intermediates in which
the terphenoxide and the imido ligands are in apical positions, while
syndiotacticity results from an inversion of chirality at
the metal center with each step in the polymerization. Inversion
of chirality at the metal forces the olefin to approach first one
side of the M=C bond and then the other. Incorporation of enantiomers
in an alternating fashion is a consequence of one enantiomer of the
racemic monomer reacting more rapidly with one enantiomer (at the
metal center) of each propagating species. We have called this mode
of control of polymer structure “stereogenic metal control”;
although the chirality of the chain end nearest the metal that results
from last inserted monomer is not necessarily irrelevant, the determining
feature is the lowest energy diastereomeric combination of chirality
at the metal and chirality of the monomer. The “errors”
in the cis,syndiotactic-poly(A-alt-A*) structure arise through
formation of AA and A*A*cis,syndiotactic and trans,isotactic dyads. Trans,isotactic dyads arise through formation of a trans metallacyclobutane
intermediate (instead of a cis metallacyclobutane),
which “flips over” before opening, a rearrangement that
preserves the configuration at the metal and leads to a trans C=C linkage.[37] This mechanistic
proposal is based on the fact that polymerization of (+)-DCMNBE (DCMNBE
= 2,3-dicarbomethoxynorbornene) by 1a yields a polymer
that contains ∼75% trans,isotactic dyads and 25% cis,syndiotactic dyads, while 1c yields a polymer that contains ∼92% trans,isotactic dyads and ∼8% cis,syndiotactic dyads (Figure ).[37] The olefinic protons in trans,isotactic polymer are inequivalent and on the same C=C bond (and therefore
coupled to each other with JHH ∼
16 Hz[37]) while the olefinic protons in cis,syndiotactic polymer are inequivalent
and on different C=C bonds (and therefore not coupled to each
other). It was also shown that W(O)(CH-t-Bu)(OHMT)(Pyr)(PMe2Ph) polymerizes (+)-DCMNBE to give only cis,syndiotactic-poly[(+)-DCMNBE].
Figure 1
Olefinic region of the 1H NMR spectrum of the poly[(R,R)-DCMNBE] prepared from initiator 1c. Reprinted with
permission from ref (37). Copyright 2012 American
Chemical Society.
Olefinic region of the 1H NMR spectrum of the poly[(R,R)-DCMNBE] prepared from initiator 1c. Reprinted with
permission from ref (37). Copyright 2012 American
Chemical Society.An interesting question is whether stereogenic metal control will direct formation of a copolymer where A and B are not strictly enantiomers, but have similar structures and reactivities toward homopolymerization, are enantiomerically
pure, and have “opposite” chirality. If A and B are significantly different chemically, the resulting
polymer could be further manipulated through selective reactions that
involve one of the two components within the polymer. We show here
that several such cis,syndiotacticA-alt-B copolymers can
be prepared with Mo (primarily) and W alkylidene initiators.
Results
and Discussion
Initial screening experiments employed the
four monomers shown
in Figure , where
the A monomers have the (2R,3R) configuration and the B monomers have the
(2S,3S) configuration. The 13C NMR spectra of cis,syndiotactic-poly(A-alt-B) should
reveal four different olefinic carbon resonances, and 1H NMR spectra could reveal up to four first order
resonances for four different olefinic protons that are coupled pairwise
(eq ). (Overlap of proton
resonances could result in non first order 1H NMR spectra.)
Racemic A is known to be polymerized
by 1a to give cis,syndiotactic-poly[A(,)-A(,)];[37] we find that cis,syndiotactic,alt polymers are also formed from racemic A, B, and B (see Supporting Information).
Figure 2
First
four monomers employed in this study.
First
four monomers employed in this study.Copolymerization of a mixture of 25 equiv of A and 25 equiv of B with 1a (0.1 M in toluene-d8) as the initiator was complete within seconds to give cis,syndiotactic-poly(A-alt-B). Its partial 13C NMR spectrum in CDCl3 showed primarily four different olefinic resonances (Figure , right), while its 1H NMR spectrum showed four overlapping first order (pseudo
triplet) olefinic proton resonances (Figure , left). (See Supporting Information for details.) The broad resonance shown between
5.45 and 5.50 ppm in Figure we propose is half of the pattern that arises from trans,isotacticAA and BB “errors”
(see Figure ). The
other half of the pattern, along with any (minor) pattern that is
characteristic of cis,syndiotacticAA and BB errors (see Figure ), is buried under the main pattern of four
triplets for cis,syndiotactic-poly(A-alt-B) around 5.30 ppm. If we assume that only trans,isotacticAA and BB dyad resonances are present under the main four
triplet resonance, we can estimate that ∼94% of the polymer
contains cis,syndiotactic-poly(A-alt-B) dyads. The olefinic carbon resonances
for any errors cannot be identified reliably in the partial carbon
NMR spectrum shown in Figure .
Figure 3
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).Copolymerization of 25 equiv of (R,R)-2,3-(CO2Me)2-norbornene (A) and 25 equiv (S,S)-2,3-(CO2CH2CF3)2-norbornene
(B) with 1a (0.1
M in toluene-d8) as the initiator was
also complete within seconds. The 13C NMR spectrum of the
resulting polymer again showed primarily four different olefinic carbon
resonances (Figure , right), while its 1H NMR spectrum showed four overlapping
first order (pseudo triplet) olefinic proton resonances (Figure , left), consistent
with the formation of cis,syndiotactic-poly(A-alt-B). A virtually identical cis,syndiotactic-poly(A-alt-B) polymer was prepared employing 1b as
the initiator (see Supporting Information). It is clear from the spectra in Figure that this cis,syndiotacticA-alt-B copolymer contains more trans,isotactic errors than the cis,syndiotacticA-alt-B copolymer described above, most likely as a consequence of
the more significant differences in reactivity between A and B than between A and B. The 5.50 ppm resonance was integrated,
and the % cis,syndiotactic-poly(A-alt-B) dyads were calculated to be ∼90%.
Two olefinic carbon resonances for the AA and BB “errors”
in this case can be seen at ∼133.0 and 128.3 ppm (Figure ). The 19F NMR spectrum of cis,syndiotactic-poly(A-alt-B) also reveals two types
of overlapping fluorine resonances for AB and BB errors
(see Supporting Information), integration
of which suggests that the % cis,syndiotactic-poly(A-alt-B) dyads is ∼80%.
Figure 4
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).
Figure 5
1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).A third cis,syndiotactic polymer
was prepared through copolymerization of a mixture of 25 equiv of
(R,R)-2,3-(CH2OAc)2-norbornene (A) and
25 equiv of (S,S)-2,3-(CO2Et)2-norbornene (B) with 1b as the initiator (Figure ). The 13C NMR spectrum showed
primarily four different olefinic carbon resonances, while the 1H NMR spectrum showed two essentially first order triplet
resonances for protons coupled to one another, along with a second
order resonance at ∼5.23 ppm for two coupled olefinic protons.
On the basis of the carbon NMR spectrum we can estimate the number
of AA and BB errors to be on the order of 5%.A fourth example is cis,syndiotactic-poly(A-alt-B). The 1H NMR
spectrum of cis,syndiotactic-poly(A-alt-B) (Figure ) provides little evidence that the polymer is relatively
regular. However, inspection of the 13C NMR spectrum shows
primarily four olefinic resonances, consistent with a relatively high
percentage (estimated ∼90%) of the proposed structure. The
complexity seen in the 1H NMR spectrum can be traced to
the overlap and second order nature of the proton resonances. At least
four carbon resonances for AA and BB errors can
be seen in the 13C NMR spectrum.
Figure 6
1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).Twelve other Mo and W initiators
were explored for making cis,syndiotactic-poly(A-alt-B), but none was as efficient
as 1a or 1b, at least according to proton
NMR spectroscopy (see Supporting Information for details). The failure
of more than two initiators (so far) to produce cis,syndiotacticA-alt-B copolymers of the type described here is not surprising
if one considers the complexity of the stereoregular ROMP reaction[29] and the need to control formation of AA and BB errors. The requirements that the metal has a stereogenic
center, that its configuration must switch with each insertion of A or B, and that the polymerization be controlled
primarily by the chirality of the stereogenic metal are demanding.Four additional enantiomerically pure monomers (Figure ) were prepared, and five AB combinations were found to
give copolymers with >90% alternating AB dyads using 1b as the initiator, according to their 1H and 13C NMR spectra (see Supporting Information for a complete list of reactions employing A and B). The percentage of trans,isotactic and/or cis,syndiotactic errors was estimated to be in the range of 5–10%.
Figure 7
Four additional
monomers.
Four additional
monomers.Copolymerization of A and B ((S,S)-(CO2-t-Bu)2-norbornene)
using 1b yielded a CDCl3-soluble polymer whose 1H NMR spectrum showed primarily two resonances, a triplet
at 5.34 ppm and a second order resonance at 5.23 ppm that integrated
to three times its relative intensity (Figure ). Weak and broad resonances near 5.50 and
5.45 can be attributed to trans,isotactic errors. The presence of primarily four olefinic resonances in the 13C NMR spectrum at 134.0, 133.3, 130.8, and 130.6 ppm suggests
that the cis,syndiotactic-poly(A-alt-B) structure is of the order of 90%.
Figure 8
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).Copolymerization of A and
(S,S)-(CH2OAc)2-norbornene (B) proceeded smoothly
to give another CDCl3 soluble polymer. The 1H NMR spectrum of the isolated polymer showed two pairs of overlapping
olefinic proton resonances and weak resonances at 5.45–5.50
for trans,isotactic errors (Figure ). However, the 13C NMR spectrum showed primarily four olefinic resonances,
which confirm that the polymer has largely the cis,syndiotactic,alt structure. Given
the successful copolymerization of (R,R)-(CH2OAc)2-norbornene (A) with (S,S)-(CO2Et)2-norbornene (B; Figure ), the formation of cis,syndiotactic-poly(A-alt-B) is not surprising. The
resonance attributed to trans,isotactic errors is much more pronounced when 1a is used as the
initiator (see Supporting Information).
Figure 9
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125
MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic
resonances only).The copolymer derived
from A ((R,R)-(CO2-t-Bu)2-norbornene)
and B ((S,S)-(CH2OMe)2-norbornene) showed two
higher ordered olefinic proton
resonances of equal intensity in CDCl3, (Figure ) along with resonances for AA and BB errors. The two overlapping proton resonances
were not well-resolved, making assessment of the errors in the polymer
structure difficult. However, the 13C NMR spectrum showed
primarily four olefinic resonances, consistent with formation of largely cis,syndiotactic-poly(A-alt-B). Copolymerization of A and B gives cis,syndiotactic-poly(A-alt-B) (Figure ), the proton NMR spectrum of which resembles that of cis,syndiotactic-poly(A-alt-B) (Figure ). The second order olefinic proton resonance at 5.33 ppm
was shifted to higher frequency with respect to two coupled triplet
proton resonances, revealing the resonances for trans,isotactic errors. Copolymerization of A with B gave cis,syndiotactic-poly(A-alt-B) (Figure ), which contains <10% errors.
Figure 10
1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
Figure 11
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3,
right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
Figure 12
1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3,
right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).
1H NMR (500
MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3, right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3,
right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).1H NMR (500 MHz, CDCl3, left) and 13C NMR (125 MHz, CDCl3,
right) spectra of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only).It is important to establish whether
the tungsten analogue of 1b (1b) is an equally
efficient catalyst. Addition of 50 equiv of rac-DCBNBE
(DCBNBE = 2,3-dicarbo-t-butoxynorbornene) to a toluene
solution of 1b led to full consumption
of the monomer within 10 min. Only two pseudo triplet olefinic proton
resonances (3JHH = 10 Hz) are
present in the 1H NMR spectrum of the resulting polymer
(Figure a). The 13C NMR spectrum is also sharp and free of any significant
fine structure associated with structural irregularities (see Supporting Information). These results are consistent
with a cis,syndiotactic,alt structure for the polymer. The two small broad resonances
assigned to trans,isotactic dyads,
in the olefin region of poly(rac-DCBNBE) prepared
from 1a, are absent from the spectrum of poly(rac-DCBNBE) prepared from 1b. The reason is that polymerization of (S,S)-DCBNBE with 1b gives cis,syndiotactic-poly[(S,S)-DCBNBE] with the 1H NMR spectrum
shown in Figure b; there is no evidence for a trans,isotactic structure. The olefinic proton resonances of cis,syndiotactic-poly[(S,S)-DCBNBE] are located in the middle of the olefinic proton resonances
for poly(rac-DCBNBE) (Figure a), which makes it difficult to assess the
percentage of microstructural errors formed in this copolymer using 1H NMR spectroscopy. However, when racemic DCENBE (DCENBE =
2,3-dicarboethoxynorbornene) and rac-DCMNBE are polymerized
by 1b under similar conditions,
the olefinic proton triplet resonances are broadened, Figures c and 13e. The positions of the minor component within the olefinic proton
resonances of poly(rac-DCENBE) and poly(rac-DCMNBE) are visible and can be unambiguously ascribed to cis,syndiotactic dyads in the largely cis,syndiotactic,alt structure
(compare Figures c–13e).
Figure 13
1H NMR (500
MHz, CDCl3) spectra for cis,syndiotactic polymers synthesized using 1b (olefinic resonances only).
1H NMR (500
MHz, CDCl3) spectra for cis,syndiotactic polymers synthesized using 1b (olefinic resonances only).As reported previously,[37] (+)-DCMNBE
and rac-DCMNBE are polymerized at approximately the
same rate using W(O)(CHCMe3)(OHMT)(Pyr)(PMe2Ph). In contrast, (−)-DCBNBE and rac-DCBNBE
are polymerized at different rates. The sharp olefinic resonances
in the 1H and 13C spectra of poly(rac-DCBNBE) prepared from 1a as the initiator are consistent
with a lower percentage of cis,syndiotactic errors. A similar trend is seen when one inspects the 1H and 13C NMR spectra of poly(rac-DCBNBE)
using 1b (Figure a).The proton and carbon
NMR spectra of the A-alt-B copolymers derived
from 1a, 1b,
and 1b are compared in Figures and 15. Initiators 1b and 1b appear to yield the highest percentages
of cis,syndiotacticA-alt-B structures with trans,isotactic errors being formed when 1b is employed
and cis,syndiotactic errors being
formed when 1b is employed.
Figure 14
13C NMR spectra (125 MHz, CDCl3) of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only) formed with initiators 1a, 1b, and 1b.
Figure 15
1H NMR spectra (500 MHz, CDCl3) of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only) formed with
initiators 1a, 1b, and 1b.
13C NMR spectra (125 MHz, CDCl3) of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only) formed with initiators 1a, 1b, and 1b.1H NMR spectra (500 MHz, CDCl3) of cis,syndiotactic-poly(A-alt-B) (olefinic resonances only) formed with
initiators 1a, 1b, and 1b.
Conclusions
Cis,syndiotacticA-alt-B copolymers, where A and B are two enantiomerically pure trans-2,3-disubstituted-5,6-norbornenes with “opposite”
chiralities, can be prepared with stereogenic-at-metal initiators
of the type M(NR)(CHR′)(OHMT)(pyrrolide) (R = 1-adamantyl or
2,6-Me2C6H3; R′ = CMe2Ph; M = Mo or W). The errors when Mo initiators are employed
are primarily trans,isotacticAA and BB dyads, while the errors when a W initiator
is employed are cis,syndiotacticAA and BB dyads. Formation of a high percentage
of alternating AB copolymer linkages relies on an inversion
of chirality at the metal center with each propagating step and faster
formation of an AB sequence than an AA or BB sequence as a consequence of a preferred diastereomeric
relationship between the chirality at the metal and the chirality
of the monomer.
Experimental Section
Representative Polymerization
A mixture of 21.0 mg
(0.1 mmol, 25 equiv) of A and
34.6 mg (0.1 mmol, 25 equiv) of B in 0.5 mL of toluene-d8 was added to
a solution of 3.0 mg (0.004 mmol) of 1b in 0.5 mL of
toluene-d8. The reaction mixture thickened
within seconds. 1H NMR spectroscopy was used to monitor
the course of the reaction. Once complete, the reaction mixture was
exposed to air and poured into 35 mL of MeOH. The precipitated cis,syndiotactic-poly(AB) was
allowed to settle. The solvent was decanted, and the polymer was dried in vacuo.
Authors: Hiroshi Nakade; M Firat Ilker; Brian J Jordan; Oktay Uzun; Nicholas L Lapointe; E Bryan Coughlin; Vincent M Rotello Journal: Chem Commun (Camb) Date: 2005-05-27 Impact factor: 6.222
Authors: William P Forrest; Jonathan G Weis; Jeremy M John; Jonathan C Axtell; Jeffrey H Simpson; Timothy M Swager; Richard R Schrock Journal: J Am Chem Soc Date: 2014-07-22 Impact factor: 15.419
Authors: Konstantin V Bukhryakov; Richard R Schrock; Amir H Hoveyda; Charlene Tsay; Peter Müller Journal: J Am Chem Soc Date: 2018-02-16 Impact factor: 15.419
Authors: Konstantin V Bukhryakov; Richard R Schrock; Amir H Hoveyda; Peter Müller; Jonathan Becker Journal: Org Lett Date: 2017-05-01 Impact factor: 6.005
Authors: Konstantin V Bukhryakov; Sudarsan VenkatRamani; Charlene Tsay; Amir Hoveyda; Richard R Schrock Journal: Organometallics Date: 2017-10-23 Impact factor: 3.876