Full Paper
[
17]
lecting appropriate chiral catalysts and thus generate optically
active polymers. Although some examples have been pro-
catalyzed alcoholysis with diols III (Figure 1). A series of new
oligomers was thus obtained, the structures of which were de-
[
13]
1
13
posed, in particular by Itsuno and co-workers, asymmetric
polymerizations utilizing a step-growth approach have not
been investigated in great detail. In addition, most asymmetric
polymerizations based on either chain or step-growth mecha-
nisms use metallic catalysts that are moisture- and/or air-sensi-
tive and that are generally difficult to remove from the poly-
mer. Moreover, the use of metal-based (Pd, Ti, Zr, etc.) catalysts
also raise environmental concerns. An alternative route that
we envisage here relies on environmentally benign strategies,
that is, following organocatalyzed approaches. Whereas in-
termined by H and C NMR, MALDI-TOF, and SEC analyses.
The high level of enantioselectivity was evaluated after con-
trolled depolymerization of II, involving chemoselective reduc-
[
14]
tense research is underway in the field of organocatalysis,
[
15]
with ongoing applications to polymer synthesis,
to our
knowledge the synthesis of chiral polymers following organo-
catalyzed pathways has not yet been explored. To fill this gap,
we envisioned the development of such a polymerization by
using an organocatalyzed reaction that has proven to be suc-
cessful on monofunctional substrates. Organocatalyzed aldol
and Mannich reactions were first designed because a wealth of
[
14]
data is available on these reactions. However, our prelimi-
nary experiments in polyaldol reactions met with limited suc-
cess, producing mixtures of polymers in which crotonization
Figure 1. Desymmetrizing polymerization of bis-anhydrides and depolymeri-
zation of the resulting polyester through a chemoselective reduction/lacto-
nization sequence.
[
16]
by-products were present to a large extent. Moreover, to
avoid diastereocontrol issues, the methodology was restricted
to difunctional substrates generating a single stereocenter,
thus limiting the scope of the polymerization. This led us to
consider a more potent reaction that would circumvent the
above limitations. Organocatalyzed ring-opening of anhydrides
with alcohols appeared to be a reasonable choice because this
process has been studied in detail on monofunctional precur-
sors and is known to provide hemiesters in high yield (more
than 90%) and excellent enantioselectivities (more than 90–
tion of the carboxylic acid function followed by lactonization,
eventually forming the chiral bis-lactones IV. The influence of
the flexibility of the carbon backbone between reacting func-
tional groups (anhydrides and alcohols) was also investigated
by varying the nature of the spacers in both monomers.
[
17]
9
5%) within a few hours at room temperature. The reaction
Results and Discussion
is catalyzed by readily available small molecules obtained from
the chiral pool, including cinchona alkaloids. Unlike polycon-
densation, this reaction does not generate any by-product,
and constitutes an excellent illustration of an atom-economical
process. The ring-opening of anhydrides also increases func-
tional complexity, leading to ester and carboxylic acid func-
tions that allow further elaboration of the polymeric backbone.
The symmetrical nature of the precursors implies that a single
stereogenic center is created upon enantioselective ring open-
ing, avoiding the diastereoselectivity issues mentioned before.
A wide range of precursors can then be designed, offering
a larger reactivity pattern. Finally, the presence in the final
polymer of functional groups having different reactivities
allows a chemoselective degradation of the polymer, which
will be useful to evaluate the enantioselectivity of each ele-
mentary step. Controlled degradation of polymers is also of
concern for further recycling of polymer waste. The anhydride
ring-opening process thus combines features that might be
difficult to find in other efficient organocatalyzed reactions and
make this reaction suitable for application in polymer synthe-
sis. We report here the first preparation of optically active poly-
esters II, having pendant carboxylic acid functions, through de-
symmetrization of meso-bis-anhydrides I by cinchona-alkaloid
Bis-anhydride precursors 4a–c were prepared in a three-step
[18]
sequence,
starting from bis-benzaldehydes 1a–c, with 1a
and 1b being available through O-alkylation of p-hydroxy-
[19]
benzaldehyde (see the Supporting Information). The Knoeve-
nagel condensation between 1a–c and ethyl cyanoacetate led
to the desired a-cyanoesters 2a–c, which, after purification by
crystallization, were treated with Meldrum’s acid to give tetra-
cids 3a–c. The latter compounds were purified by column
chromatography over silica gel (CH Cl /MeOH/HCOOH,
2
2
95:5:0.5) (Scheme 1). The modest yield obtained during this
step might be attributed to the loss of material during chro-
matography. Dehydration of 3a–c using acetic anhydride final-
ly afforded the target bis-anhydrides 4a–c in satisfactory over-
all yield.
Prior to the “desymmetrizing polymerization” of bis-anhy-
drides 4a–c, model studies were carried out with mono-anhy-
dride 5 to optimize the reaction conditions, including the
nature of the alcohol (Scheme 2). It is worth mentioning that,
in contrast to standard conditions used with monofunctional
analogues, polymerization of 4a–c requires a strict stoichiome-
try between the bis-anhydride and the diol to achieve the
[20]
highest polymer molecular weight. Therefore, anhydride 5
Chem. Eur. J. 2014, 20, 11946 – 11953
11947 ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim