Folding-driven reversible polymerization of oligo(m-phenylene ethynylene) imines: Solvent and starter sequence studies

Article abstract of DOI:10.1021/ma034159i

Bis(imino) end-functionalized oligo(m-phenylene ethynylene)s were equilibrated in a closed system under conditions that promote reversible imine metathesis. The metathesis reaction joins two oligomers and produces a small molecule byproduct. In polar solvents, equilibration gave high molecular weight polymers while equilibration in chloroform produced only low molecular weight oligomers. This polymerization is hypothesized to be driven by the free energy gained from the folding of the long polymer chains directed by the noncovalent, intramolecular aromatic stacking and solvophobic interactions. This polymerization was also conducted in a series of solvents in which m-phenylene ethynylene oligomers have previously shown varied, intermediate folding stabilities. These experiments revealed a good correlation of the product molecular weight with the stability of the m-phenylene ethynylene helix. The equilibrium state of the metathesis reaction was also demonstrated to depend on the chain length of the starter sequences. With a pair of trimeric precursors, macrocyclization instead of polymerization takes place. Consistent with the notion that the polymerization is a consequence of the intramolecular solvophobic chain association, higher degrees of polymerization followed from enhanced solvophobicity of the m-phenylene ethynylene backbone, achieved by appending a methyl substituent to half of the repeat units. The considerably longer equilibration time required by these more stabilized sequences suggests that the elongation process may involve unfolding or partial unfolding of the chain; alternatively, intermolecular association may be responsible for the slow chain growth.

Full text of DOI:10.1021/ma034159i

2712
Macromolecules 2003, 36, 2712-2720
Folding-Driven Reversible Polymerization of Oligo(m-phenylene
ethynylene) Imines: Solvent and Starter Sequence Studies
Da h u i Zh a o a n d J effr ey S. Moor e*
Departments of Chemistry and Materials Science & Engineering, University of Illinois at Urbanas
Champaign, Urbana, Illinois 61801
Received February 6, 2003; Revised Manuscript Received February 17, 2003
ABSTRACT: Bis(imino) end-functionalized oligo(m-phenylene ethynylene)s were equilibrated in a closed
system under conditions that promote reversible imine metathesis. The metathesis reaction joins two
oligomers and produces a small molecule byproduct. In polar solvents, equilibration gave high molecular
weight polymers while equilibration in chloroform produced only low molecular weight oligomers. This
polymerization is hypothesized to be driven by the free energy gained from the folding of the long polymer
chains directed by the noncovalent, intramolecular aromatic stacking and solvophobic interactions. This
polymerization was also conducted in a series of solvents in which m-phenylene ethynylene oligomers
have previously shown varied, intermediate folding stabilities. These experiments revealed a good
correlation of the product molecular weight with the stability of the m-phenylene ethynylene helix. The
equilibrium state of the metathesis reaction was also demonstrated to depend on the chain length of the
starter sequences. With a pair of trimeric precursors, macrocyclization instead of polymerization takes
place. Consistent with the notion that the polymerization is a consequence of the intramolecular
solvophobic chain association, higher degrees of polymerization followed from enhanced solvophobicity
of the m-phenylene ethynylene backbone, achieved by appending a methyl substituent to half of the repeat
units. The considerably longer equilibration time required by these more stabilized sequences suggests
that the elongation process may involve unfolding or partial unfolding of the chain; alternatively,
intermolecular association may be responsible for the slow chain growth.
In tr od u ction
The goal of the current work, as the first step toward
identifying tyligomer sequences through reversible po-
lymerizations, is to demonstrate that the free energy
gained from folding provides the driving force for high
polymer formation. We show that high molecule weight
can be achieved from a reversible polymerization by
appropriately choosing starter sequences that are only
stabilized in long polymer chains that adopt compact
form. Toward this end, a system that can undergo a
reversible unfolded-to-folded conformational transition
must first be identified. The oligo(m-phenylene eth-
ynylene)s that have extensively been studied in our
laboratory exhibit a reversible conformational transition
in solution.5 These oligomers are comprised of a hydro-
carbon-rich, aromatic backbone appended with tris-
(ethylene glycol) monomethyl ether side chains via an
ester linkage. They adopt an ordered helical conforma-
tion in polar solvents such as acetonitrile but exist in a
random conformational state in chlorohydrocarbon sol-
vents such as chloroform. Such a coil-to-helix transition
was explained to result from solvophobically favored
aromatic stacking interactions among the aromatic
backbone units in polar media.6 For folding to take
place, a critical chain length (nucleation size) is required
for the oligomers to achieve sufficient enthalpy gain
from stacking interactions to overcome the entropy loss
of restricted conformation.5b Therefore, the polymeri-
zation of the m-phenylene ethynylene oligomers may
exhibit a unique nucleation process (i.e., formation of
short sequences incapable of folding followed by favor-
able elongation of the chain into stable, folded poly-
mers).7
The folding of proteins into well-defined globular
structures requires very specific sequences and compo-
sition of the constituent amino acids.1,2 Both theoretical
and experimental studies have suggested that only a
fraction of all possible sequences give rise to globular,
soluble structures.2,3 For synthetic polymers, the design
of tyligomers4a that possess a sequence that is able to
adopt a desired secondary or tertiary structure (i.e.,
the “masterpiece” sequence)4b presents a significant
challenge for chemists. One of the major obstacles comes
as a consequence of the enormous magnitude of all
possible sequences of a heteropolymerization. Even for
a binary monomer system, the number of available
sequences is large enough to make the identification of
a specific sequence almost impossible. Conventional
irreversible polycondensations inevitably generate rep-
resentatives of all possible sequences in a copolymeri-
zation, regardless of the structures resulting from the
sequences.3e On the other hand, in a reversible polym-
erization system wherein all sequences are inter-
convertible by dissociation and reassociation of the
comprising subunits, the relative population of each
different sequence is determined by their relative
thermodynamic stabilities. Therefore, it is reasonable
to hypothesize that a reversible polymerization would
give rise to a smaller subset of the entire sequence
library, selectively producing structures that are ther-
modynamically most stable. To induce the production
of certain sequences, conditions can be selected to
stabilize the desirable structures. If a polymerization
is set up under conditions that favor folding, uniquely
folded structures may be generated.
To clearly elucidate the role of folding in the polym-
erization of m-phenylene ethynylenes, imine metathe-
sis8 (Scheme 1) was specifically chosen to perform the
reversible, covalent ligation of starter sequences in the
* Corresponding author. Fax: (217) 244-0181. E-mail: moore@
scs.uiuc.edu.
10.1021/ma034159i CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/19/2003

Macromolecules, Vol. 36, No. 8, 2003
Folding-Driven Reversible Polymerization 2713
Sch em e 1
polymerization reaction. Many features of the meta-
thesis reaction9 qualify it as a suitable choice for
coupling monomers or sequences to test the concepts put
forth in. Most notably, the sum of the bond energies on
one side of the equilibrium will be nearly identical to
those on the other. Thus, for a metathesis reaction, the
equilibrium distribution will not be biased to a particu-
lar product by bond energy changes. Therefore, the
formation of high molecular weight products, should
they be observed, can reasonably be attributed to the
energy gained by folding or collapsing of the polymer
chains. For several reasons that are noted below,
metathesis reaction between imine units appeared
favorable for m-phenylene ethynylene system. First, the
imine unit is nondisruptive to aromatic interactions10
and imine metathesis can be conducted in a variety of
solvents. Additionally, while the metathesis is reversible
and quickly equilibrates in the presence of catalytic acid,
the interconversion can conveniently be “switched off”
(e.g., via addition of excess triethylamine) to “quench”
the equilibrated product distribution.11 As expected, the
equilibrium constant of this reaction is nearly unity,11a
which ensures that little, if any, favorable energy is
gained from new covalent bonds generated during
polymerization.
In a previous communication, we demonstrated that
the folding of the m-phenylene ethynylene chains can
drive the imine metathesis polymerization to generate
high polymers in acetonitrile.12 As a control, only low
molecular weight oligomers were observed when the
metathesis took place under nonfolding conditions, i.e.,
in chloroform. The correlation of the molecular weight
of the resulting polymers with the solvent composition
of a series of acetonitrile/chloroform binary mixtures,
which dictates the stability of the folded conformation
of m-phenylene ethynylene molecules,5b supported the
proposition that folding shifts the metathesis equilib-
rium toward high polymers. Furthermore, the insensi-
tivity of the equilibrated molecular weight of the
products to the concentration indicated that intra-
molecular association (i.e., folding) rather than inter-
molecular aggregation was the dominant driving force
for the high polymer formation.12 Herein we report
detailed studies elucidating the effect of solvent, starter
sequence, and backbone chemical features on the po-
lymerization reaction.
F igu r e 1. Histogram of the calculated chain length distribu-
tions (slashed bar, mole fraction; solid bar, weight fraction)
at equilibrium for a closed-system metathesis polymerization
in which the equilibrium constant of the coupling reaction is
unity. The average degree of polymerization is calculated to
be 2.0.
distribution can be calculated by statistical methods
(Figure 1).13 An average degree of polymerization of 2.0
is predicted with the products distributed in the low
molecular weight region. In other words, only oligomers
with low degrees of polymerization are expected for a
reversible polymerization conducted in a closed system.
As the equilibrium constant for imine metathesis be-
tween monofunctional m-phenylene ethynylene oligo-
mers has been shown to be close to unity, we expect to
observe low molecular weight products under nonfolding
conditions. Significant deviation from this calculated
distribution suggests the presence of an alternate driv-
ing force that shifts the equilibrium.
Solven t Stu d ies: P olym er iza tion of 1 a n d 2. As
previously reported,12 the imine metathesis polymeri-
zation of a pair of R,ω-imine functionalized, tetrameric
m-phenylene ethynylenes 1 (N-terminal bis(imine)) and
2 (C-terminal bis(imine)) in chloroform resulted in a set
of low molecular weight oligomers, the distribution of
which was in qualitative agreement with the calculation
mentioned above. In contrast, high polymers were
obtained when the same metathesis reaction was car-
ried out in acetonitrile under otherwise identical condi-
tions (Scheme 2). Since the m-phenylene ethynylene
chains are believed to be stabilized by compact, helical
conformations in acetonitrile, as opposed to a random
conformational state in chloroform, it can be concluded
that the free energy gained from folding shifted the
equilibrium toward high polymer formation (assuming
there is no aggregation, vide supra).
If folding is responsible for the equilibrium shifting
and consequently the elongation of the polymer chains,
the degree of polymerization should correlate to the
folding energy of the resulting products. In previous
studies, we showed that the degree of polymerization
systematically depended on the amount of chloroform
in acetonitrile/chloroform binary solvent mixtures.12 The
observed dependence of the molecular weight on the
solvent composition supported the role of folding as the
driving force for shifting the equilibrium toward high
polymers.
Resu lts a n d Discu ssion
Ca lcu la tion s of MW Distr ibu tion u n d er Non -
fold in g Con d ition s. To explicitly elucidate the effect
of folding in driving high polymer formation, calcula-
tions were first conducted to determine the product
molecular weight distribution of a metathesis polym-
erization in the absence of folding. If the equilibrium
constant for reversible coupling of monomers or oligo-
mers is assumed to be unity,11a the molecular weight
An obvious alternative approach to confirming the
relationship between the folding energy and the degree
of polymerization was to perform the equilibration in
solvents that provide various folding stabilities. On the

2714 Zhao and Moore
Macromolecules, Vol. 36, No. 8, 2003
Sch em e 2
Ta ble 1. Molecu la r Weigh t Da ta of 3 fr om Sta r ter
Sequ en ces 1 a n d 2a
that for these sequences equilibrium could be reached
within this time. The size exclusion chromatography
(SEC) traces of the equilibrium products obtained from
selected solvents are shown in Figure 2, and the
corresponding molecular weight data are provided in
Table 1. A consistent increase in the number- and
weight-average molecular weight of the products was
observed as the A313/A295 absorbance ratio decreased.
This further demonstrated that folding was responsible
for shifting the equilibrium, driving the chain to elon-
integration
Mn
Mw polydispersity
b
solvent
A
313/A295 limits (min) (kDa) (kDa)
(Mw/Mn)
a
b
c
d
e
f
26.0-27.8
22.7-27.6
22.6-27.6
22.3-27.6
19.7-27.4
18.0-27.5
1.8
3.0
3.0
3.4
5.8
7.5
1.8
3.9
3.9
4.5
10.3
34.0
1.0
1.3
1.3
1.3
1.8
CHCl3
THF
1,4-dioxane
MeOAc
EtOAc
0.938
0.860
0.811
0.739
0.719
0.672
4.5
26.3
g
CH3CN
15.3-27.5 13.5 350.2
a
Metathesis conditions: 1:1 equiv of reactants at
5 mM
equilibrated in the presence of 0.1 equiv of oxalic acid at room
temperature for 6 days. The first column is keyed to Figure 2.b UV
absorbance data are as previously reported.14
basis of the results from independent solvent studies,14
metathesis polymerization was conducted in a series of
solvents in which the folding propensity of a m-phen-
ylene ethynylene oligomer has been shown to vary
between that in chloroform and acetonitrile (Table 1).
Previous studies have found that the ratio of UV
absorbances at 313 and 295 nm (A313/A295) serves as a
qualitative index of the helical stability and thus reflects
the solvent quality for the promoting the folding of
m-phenylene ethynylene oligomers.5 The cisoid (helical)
and transoid (nonhelical) states of the m-phenylene
ethynylene unit give rise to different absorbance ratios
at these wavelengths, and a decrease in this ratio
corresponds to a larger fraction of the helical conforma-
tion.
Equimolar quantities of tetrameric starter sequences
1 and 2 were dissolved in the various solvents listed in
Table 1 at a concentration of 5 mM in the presence of
0.1 equiv of oxalic acid as a catalyst. These sequences
were equilibrated at room temperature for 6 days at
which time the reaction was quenched with an excess
amount of triethylamine. Exposing the reactants to the
catalyst solution for a longer period of time led to no
significant change in the molecular weight, indicating
F igu r e 2. SEC traces of starter sequences 1 and 2 (a) and
products 3 equilibrated for 6 days at 5 mM and room temper-
ature in chloroform (b), tetrahydrofuran (c), dioxane (d), methyl
acetate (e), ethyl acetate (f) and acetonitrile (g) solutions (the
traces eluted at ca. 30 min are from the byproduct 4). Samples
were eluted from SEC columns with THF and detected by a
refractive index (RI) detector. The molecular weight calibration
is based on narrow molecular weight polystyrene standards.
In trace a, starter sequences 1 and 2 coelute.

Macromolecules, Vol. 36, No. 8, 2003
Folding-Driven Reversible Polymerization 2715
F igu r e 3. SEC (A) and HPLC (B) traces of starter sequences 5 and 6 (a) and their products following equilibration for 40 h at
5 mM and room temperature in chloroform (b and d) and acetonitrile (c and e). Starter sequence 6 was in excess in both reactions.
SEC analyses were conducted under the same conditions as in Figure 2. Samples were eluted from silica gel HPLC columns with
CHCl3/iPrOH (3.0 vol %) and detected by a UV detector operating at 290 nm. In the SEC trace a, starter sequences 5 and 6
coelute.
Sch em e 3
gate into high polymers. These results also indicate that
folding driven polymerization is a general feature of
m-phenylene ethynylene imine sequences under solvo-
phobic conditions, rather than a unique observation that
only takes place in acetonitrile.
Ma cr ocycliza tion of Oligom er s 5 a n d 6. In con-
trast to the polymerization of tetrameric starter se-
quences 1 and 2, no high molecular weight product was
obtained when trimeric m-phenylene ethynylene oligo-
mers 5 and 6 were allowed to equilibrate under the
same metathesis conditions. This pair of starter se-
quences contains the same imine end groups as 1 and
2, but each is shorter in length by one phenylene
ethynylene unit. On the basis of SEC studies, a mono-
dispersed product with a molecular weight approxi-
mately twice that of either starting material was
obtained from a catalyzed mixture of 5 and 6 equili-
brated in either chloroform or acetonitrile (Figure 3A).
Subsequent HPLC analyses on the same reaction mix-
tures (Figure 3B) confirmed the generation of a single,
low molecular weight product. This product was found
to coexist with a considerable amount of the starting
materials when the equilibration was conducted in
chloroform, while the reaction was driven to nearly

2716 Zhao and Moore
Macromolecules, Vol. 36, No. 8, 2003
F igu r e 4. Schematic illustration of the metathesis polymerization and macrocyclization.
quantitative conversion when it takes place in acetoni-
trile. This monodisperse product was then identified by
1H NMR and MS (FAB) as macrocycle 7. On the basis
of the integrations of the end group resonance in the
1H NMR spectrum, the equilibrium constant of the
reaction shown in Scheme 3 was determined to be ca.
13 mM in chloroform. However, this macrocycle exhib-
ited extraordinary stability in acetonitrile. Even when
one of the reactants (i.e., 6) was present in excess, 7
was obtained in nearly quantitative yield (based on the
amount of the quantity limiting compound 5).
Macrocycle 7 is known to aggregate into columnar
assemblies in polar solvents, including acetonitrile,10
driven by the solvophobically favored aromatic stacking.
On the basis of this evidence, we propose that the
macrocycle is stabilized by the free energy gained from
the intermolecular, noncovalent interactions upon ag-
gregation in acetonitrile and that the aggregates there-
fore became the thermodynamically most stable species.
Accordingly, the metathesis equilibrium was shifted in
favor of cyclization and the macrocycle was afforded as
the predominant product. Since unimolecular cyclization
is intrinsically disfavored by entropy, low reaction yield
has long presented a challenge for macrocycle syntheses.
Significantly, the aggregation-driven synthesis of mac-
rocycle 7 illustrates a possible approach to synthesize
macrocyclic structures in high yield. From the stand-
point of polymerization, however, these cyclic products
are recognized as another type of “deep trap” that can
interfere with high polymer formation.
The different product distributions obtained from the
two sets of starter sequences (i.e., 1 + 2 vs 5 + 6) under
the same metathesis conditions may be explained by the
structural differences of the corresponding macrocycles
(Figure 4). On the basis of our previous studies,15
macrocycle 7 is expected to possess a planar or nearly
planar conformation in solution, which is optimal for
efficient intermolecular π-stacking interactions and
aggregation. Unlike 7, an analogous octameric macro-
cycle (resulting from 1 and 2) should be strained, having
a nonplanar conformation, and would not be able to
efficiently form stacked aggregates. Instead, a crescent
broken ring would be formed, which upon further
reaction would develop into long polymeric chains that
maximize aromatic stacking intramolecularly. Conse-
quently, polymerization dominates for 1 and 2.
F igu r e 5. SEC traces of starter sequences 2 and 5 (a) and
products following equilibration for 6 days at 5 mM and room
temperature in chloroform (b) and acetonitrile (c).
tetrameric oligomers gives rise to higher polymers while
macrocyclization occurs when the reaction starts with
a pair of trimeric species, a question naturally arises
as to what would happen if a trimer is metathesized
with a tetramer. To answer this question, a metathesis
reaction was performed under the same conditions as
described earlier, with a mixture of oligomers 2 and 5.
The SEC traces of the equilibrated products showed that
polymerization dominated in acetonitrile and low mo-
lecular weight oligomers were generated in chloroform.
In the MALDI MS of the reaction mixture equilibrated
in chloroform, both the open chain and the cyclic dimers
of 2 and 5 were detected. The presence of two different
forms of dimers in chloroform explained the dispropor-
tionally large peak that was eluted right before the
starting material mixture in the SEC trace (Figure 5b).
Although MALDI MS was not able to provide informa-
tion on the sequence identity of the products obtained
from acetonitrile solution due to the broad molecular
weight distribution, the presence of high polymers was
evidenced by SEC analysis (Figure 5c). This suggests
that the cyclic dimer of 2 and 5 is at most of modest
stability and is insufficient to produce an energy trap
that prevents polymerization.
Meta th esis P olym er iza tion of 2 a n d 5. Since it has
been illustrated that metathesis between a pair of

Macromolecules, Vol. 36, No. 8, 2003
Folding-Driven Reversible Polymerization 2717
Sch em e 4
Ta ble 2. Molecu la r Weigh t Da ta of Meta th esis P r od u cts
3 a n d 10a
Sta r ter Sequ en ces w ith a La r ger P olym er iza -
tion Dr ivin g F or ce. Since the extent of equilibrium
shifting is dependent on the stability of the polymer’s
solution conformation, higher molecular weights should
be attainable from sequences giving rise to polymers
that self-associate more strongly or fold with greater
stability. Previous studies in our laboratory revealed
that replacing the cavity-directed hydrogen atoms of the
m-phenylene ethynylene oligomers with methyl groups
generates a more stable folded structure.16 The en-
hanced folding stability is attributed to the higher
solvophobicity of the hydrocarbon backbone, which
further strengthens the π-stacking interactions.6,17 There-
fore, a higher degree of polymerization should arise
when corresponding methyl-substituted starter se-
quences are employed. To test this hypothesis, a new
pair of starter sequences 8 and 9 (the methyl-substi-
tuted analogues of 1 and 2) were synthesized and
subjected to polymerization investigations (Scheme 4).
For the purpose of simplifying the synthesis, only half
of the phenylene units contained in 8 and 9 were
substituted with methyl groups. Such a structural
alteration was envisioned to increase solvophobicity to
a degree that would cause a distinct molecular weight
change in the products.
Oligomers 8 and 9 were prepared in a similar fashion
as previously reported for 1 and 2.12 They were then
reacted under the same metathesis conditions as applied
to 1 and 2 (i.e., equilibration in catalytic solution at room
temperature for 6 days before quenching with base).
After analysis of the reaction products by SEC, it was
apparent that the methyl-containing polymers afforded
a higher number-average molecular weight (Mn) than
the H-substituted analogues in the entire series of
solvents investigated. However, when the weight-aver-
age molecular weights (Mw) were compared, the methyl
modified sequences gave lower values than their H
analogues in ethyl acetate and acetonitrile (Table 2).18
These unexpected observations were explained upon
examining the product molecular weight at prolonged
reaction times (Figure 6). In solvents such as chloroform,
THF and dioxane that furnish low to medium molecular
3b
10
b
b
c
c
solvent
CHCl3
Mn
Mw
3.9
3.9
4.5
Mn
Mw
Mn
Mw
Mn Mw
3.0
3.0
3.4
5.8
7.5
3.1
3.1
4.4
3.6
3.7
6.0
THF
1,4-dioxane
MeOAc
EtOAc
10.3 12.0 27.0 18.1d
701d 21.6 268
34.0 13.3 27.8 19.9d 1470d 23.1 355
18.7 41.7 33.5e 382e 40.0 234
CH3CN
13.5 350
a
All molecular weight data are in kDa and metatheses were
performed at 5 mM reactant concentration in the presence of 0.1
equiv oxalic acid. b Data were obtained after equilibrating reaction
mixtures at room temperature for 6 days. c After equilibration at
d
40 °C for 6 days followed by 10 days at room temperature. After
equilibration at room temperature for 19 weeks. e After equilibra-
tion at room temperature for over 25 weeks (the equilibrium state
may have yet to be reached).
weight products, no discernible change in molecular
weight distribution occurred after 6 days indicating that
the equilibrium state had been reached in each case.
However, much higher molecular weight products were
observed in acetonitrile solution after the reaction was
allowed to equilibrate for 44 days at room temperature.
Further extending the reaction time revealed that in
either methyl acetate or ethyl acetate, equilibrium could
be reached within approximately 19 weeks,19 while in
acetonitrile the product molecular weight continued
growing slowly even after 25 weeks at which point the
monitoring was stopped (Figure 6). On the basis of the
molecular weight data from these experiments (Figure
7 and Table 2), it can be concluded that given sufficient
reaction time to reach the equilibrium much higher
molecular weights are achieved by the methyl-substi-
tuted sequences in methyl acetate, ethyl acetate, and
acetonitrile.20 These results agreed with our original
premise that the product molecular weight can becon-
trolled through structural modifications that vary the
folding energy.
The significantly longer equilibration time required
by the methyl-substituted sequences indicates much
slower kinetics as these more solvophobic chains elon-

2718 Zhao and Moore
Macromolecules, Vol. 36, No. 8, 2003
disfavored dissociation and the imine bonds buried
within stacked helices.
Additional evidence for slow kinetics came from
experiments with metathesis between 8 and 9 at varied
temperatures. At a given equilibration time (e.g., 6 d),
the molecular weight of polymer 10 was found to
increase with the reaction temperature up to 30 °C;
however, it decreased at even higher equilibration
temperatures (Figure 8). This result suggests that the
observed lower molecular weight of the more stabilized
polymer is a kinetic rather than thermodynamic limita-
tion. At elevated temperatures the unfolded state (or
the dissociated state of the aggregated species) becomes
more accessible and chain elongation via metathesis is
facilitated. The rate at which equilibrium is approached
increased and higher molecular weight products were
achieved within a shorter period of time. On the other
hand, although increasing the reaction temperature
increases the approach to equilibrium, the thermody-
namic stability of the folded conformation is probably
reduced at higher temperature, leading to a diminished
molecular weight at equilibrium. Therefore, the maxi-
mum observed in the molecular weight as a function of
temperature can be rationalized as slow kinetics at low
temperature and decreased stability at high tempera-
ture. Further investigations at elucidating the kinetics
of this polymerization are underway.
F igu r e 6. SEC traces of starter sequences 8 and 9 (a) and
products equilibrated at 5 mM and room temperature in
chloroform (b, 6 days), tetrahydrofuran (c, 6 days), dioxane (d,
6 days), methyl acetate (e, - - - 6 days; - 19 weeks), ethyl
acetate (f, - - -
6 days; - 19 weeks) and acetonitrile
(g, - - - 6 days; s‚s 44 days; - 25 weeks) solutions. In trace a,
starter sequences 8 and 9 coelute.
Since raising the temperature was effective in ac-
celerating the approach to equilibrium, efforts were
made to modify the polymerization procedure for 8 and
9 in order to achieve high molecular weight within a
reasonable period of equilibration time. As the experi-
mental results showed that heating the catalyzed
reactant mixture at 40 °C for 6 days followed by
equilibration at room temperature for ca. 10 days
generates products with a molecular weight close to that
obtained at the most prolonged equilibration time (Table
2), the subsequent polymerizations of 8 and 9 in a series
of chloroform/acetonitrile binary solvent mixtures were
all conducted under such modified conditions (Figure
9). Although the equilibrium state is most likely not
reached by the end of 16 days in the solvents with a
high acetonitrile composition, these experiments allowed
us to obtain information on the product molecular
weight close to equilibrium at room temperature. The
molecular weight variation of 10 with solvent composi-
tion was compared to that of 3. As the plot in Figure 10
shows, an escalation of the average molecular weight
occurred at higher chloroform composition for 10 than
for 3, and an overall higher molecular weight was
obtained from the methyl-substituted polymers than
from the parent analogues at corresponding solvent
composition (except for Mw in pure acetonitrile).18 Thus,
even though the equilibrium may not have been reached
for the methyl-substituted polymers in high acetonitrile
composition solutions, these results are consistent with
our expectation of higher molecular weight polymers
arising from better stabilized, folded structures. Again,
the drop in molecular weight at high acetonitrile
composition was likely a consequence of slow kinetics
with an insufficient equilibration time.
F igu r e 7. Molecular weight (number-average Mn (main plot)
and weight-average Mw (inset)) of the metathesis products 3
(s0s) and 10 (s9s) vs the UV absorbance ratio A313/A295 of
a m-phenylene ethynylene oligomer in given solvents.
gate. On the basis of this observation, we hypothesize
that the metathesis of the imine bond may require at
least partial unfolding the m-phenylene ethynylene
chain. Under conditions that strongly stabilize the
helical conformation, the unfolded state is considerably
disfavored. Thus, a larger energy barrier must be
overcome before additional stabilizing energy can be
attained from incorporation of more monomer units.
Consequently, chain elongation is kinetically impeded.
An alternative explanation is that intermolecular as-
sociation, which is absent or less severe in the case of
H-substituted sequences,11a becomes significant for the
more solvophobic backbone in polar media. In either
case, slow kinetics of metathesis likely resulted from
Con clu sion
The solvent and sequence effects on the closed-system
reversible, imine metathesis polymerization of m-phe-
nylene ethynylene oligomers have been investigated. In
these polymerizations, the folding or collapsing of the

Macromolecules, Vol. 36, No. 8, 2003
Folding-Driven Reversible Polymerization 2719
F igu r e 8. SEC traces (A) and corresponding molecular weight of metathesis products 10 equilibrated at 5 mM in CH3CN for 6
days at 23 (a), 33 (b), 40 (c), and 48 °C (d).
F igu r e 10. Molecular weight (number-average Mn (main plot)
and weight-average Mw (inset)) of metatheses products 3
(s0s, equilibrated at room temperature for 6 d) and 10
(s9s, equilibrated for 6 days at 40 °C followed by 10 days at
room temperature) at 5 mM in CHCl3/CH3CN mixtures.
F igu r e 9. SEC traces of metathesis products 10 equilibrated
at 5 mM for 6 days at 40 °C followed by 10 days at room
temperature in CHCl3 (a), CHCl3/CH3CN (vol % of CHCl3: b,
75%; c, 50%; d, 25%; e, 22%; f, 12.5%; g, 5%), and CH3CN (h)
solutions.
weight polymers. This agrees with the folding-driven
nature of the polymerization.
polymer chain was shown to provide a sufficient driving
force to push the equilibrium toward high polymers. By
means of tuning the solvent quality and/or temperature,
the folding propensity of m-phenylene ethynylene imine
chains can be modulated, and as a result, the control
over the molecular weight of the resulting polymers can
be achieved. By varying the chain length of the starter
sequences, macrocyclization can be either favored or
circumvented. When the chain growth proceeds via an
oligomeric intermediate containing six m-phenylene
units, macrocycles form exclusively. This was attributed
to the stabilization of the intermolecular aggregates by
solvophobic and π-stacking interactions. When the
starter sequence lengths do not allow the formation of
the hexameric macrocycle, chain elongation into poly-
mers dominates. Besides the chain length, the struc-
tural characteristics of the starter sequence strongly
influence the kinetics and equilibrium state of the final
product. Improving the folding capability of m-phen-
ylene ethynylene chain resulted in higher molecular
Ack n ow led gm en t. This material is based upon
work supported by the National Science Foundation
under Grant No. 0117792 and the U.S. Department of
Energy, Division of Materials Sciences, under Award
No. DEFG02-91ER45439, through the Frederick Seitz
Materials Research Laboratory at the University of
Illinois at UrbanasChampaign.
Su p p or tin g In for m a tion Ava ila ble: Text giving calcula-
tions on molecular weight distribution of a reversible polym-
erization based on statistical methods, syntheses, and char-
acterization data of 5, 6, 8, and 9 and a scheme showing the
syntheses of these compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
Refer en ces a n d Notes
(1) Rose, G. D.; Wolfenden, R. Annu. Rev. Biophys. Biomol.
Struct. 1993, 22, 381-415.
(2) (a) Dill, K. A. Biochemistry 1985, 24, 1501-1509. (b) Gates,
R. E.; Fisher, H. F. Proc. Natl. Acad. Sci. U.S.A. 1971, 68,
2928-2931.

2720 Zhao and Moore
Macromolecules, Vol. 36, No. 8, 2003
(3) (a) Yamauchi, A.; Yomo, T.; Tanaka, F.; Prijambada, I. D.;
Ohhashi, S.; Yamamoto, K.; Shima, Y.; Ogasahara, K.;
Yutani, K.; Kataoka, M.; Urabe, I. FEBS Lett. 1998, 421,
147-151. (b) Prijambada, I. D.; Yomo, T.; Tanaka, F.;
Kawama, T.; Yamamoto, K.; Hasegawa, A.; Shima, Y.;
Negoro, S.; Urabe, I. FEBS Lett. 1996, 382, 21-25. (c)
Camacho, C. J .; Thirumalai, D. Phys. Rev. Lett. 1993, 71,
2505-2508. (d) Lau, K. F.; Dill, K. A. Macromolecules 1989,
22, 3986-3997. (e) Rao, S. P.; Carlstrom, D. E.; Miller, W.
G. Biochemistry 1974, 13, 943-952. (f) Fisher, H. F. Proc.
Natl. Acad. Sci. U.S.A. 1964, 51, 1285-1291.
(15) Shetty, A. S.; Zhang, J .; Moore, J . S. J . Am. Chem. Soc. 1996,
118, 1019-1027.
(16) (a) Prince, R. B. Ph.D. thesis, University of Illinois, 1999. (b)
Gin, M. S.; Yokozawa, T.; Prince, R. B.; Moore, J . S. J . Am.
Chem. Soc. 1999, 121, 2643-2644. (c) Gin, M. S.; Moore, J .
S. Org. Lett. 2000, 2, 135-138. (d) Prince, R. B.; Barnes, S.
A.; Moore, J . S. J . Am. Chem. Soc. 2000, 122, 2758-2762.
(e) Mio, M. J .; Prince, R. B.; Moore, J . S.; Martin, C.; Kuebel,
D. C. J . Am. Chem. Soc. 2000, 122, 6134-6135.
(17) For examples of solvophobicity promoted aromatic stacking
interactions, see: (a) Cubberly, M. S.; Iverson, B. L. J . Am.
Chem. Soc. 2001, 123, 7560-7563. (b) Smithrud, D. B.;
Diederich, F. J . Am. Chem. Soc. 1990, 112, 339-343.
(18) Small variations in the values of Mn and Mw observed in a
number of cases shown here result from different polydis-
persities. Number average molecular weight value is much
more sensitive than weight average molecular weight to the
presence of the nonfoldable, low molecular weight oligomers.
The stability and correspondingly the relative amount of
these short oligomers vary with solvent and temperature
conditions in a different manner than the long, foldable
polymers. This inconsistency complicated the distribution of
the products, which leads to the variation in Mn and Mw. For
example, the narrow distribution of 10, in which nearly all
the low molecular weight species that are present in 3 have
disappeared, gives rise to the higher Mn. Polymer 3 obviously
contains some high molecular weight components that are
absent from 10, which accounted for the higher Mw of 3.
(19) Equilibrium may have been reached in methyl acetate and
ethyl acetate in less than 19 weeks, but the product was not
monitored day-to-day. The equilibration time and the molec-
ular weight were also found to be sensitive to the amount of
water introduced into the system; longer reaction times
corresponding to higher molecular weights were necessary
for reactions in solvents that had been rigorously dried.
(4) (a) Hill, D. J .; Mio, M. J .; Prince, R. B.; Hughes, T. S.; Moore,
J . S. Chem. Rev. 2001, 101, 3893-4012. (b) Moore, J . S.;
Zimmerman, N. W. Org. Lett. 2000, 2, 915-918.
(5) (a) Nelson, J . C.; Saven, J . G.; Moore, J . S.; Wolynes, P. G.
Science 1997, 277, 1793-1796. (b) Prince, R. B.; Saven, J .
G.; Wolynes, P. G.; Moore, J . S. J . Am. Chem. Soc. 1999, 121,
3114-3121.
(6) Lahiri, S.; Thompson, J . L.; Moore, J . S. J . Am. Chem. Soc.
2000, 122, 11315-11319.
(7) Precedents of synthetic polymerizations with a nucleation
process are limited. See: (a) Kern, W.; J aacks, V. J . Polym.
Sci. 1960, 48, 399-404. (b) Leese, L.; Baumber, M. V. Polymer
1965, 6, 269-286. (c) Kojmoto, T.; Akaishi, T.; Oh ya, M.; Kawai,
T. Makromol. Chem. 1972, 154, 151-159. (d) Fujie, A.;
Kojmoto, T.; Oh ya, M.; Kawai, T. Makromol. Chem. 1973, 169,
301-321.
(8) To´th, G.; Pinte´r, I.; Messmer, A. Tetrahedron Lett. 1974, 735-
738.
(9) For other examples of acyclic metathesis polymerization,
see: Schwendeman, J . E.; Church, A. C.; Wagener, K. B. Adv.
Synth. Catal. 2002, 344, 597-613.
(10) Zhao, D.; Moore, J . S. J . Org. Chem. 2002, 67, 3548-3554.
(11) (a) Oh, K.; J eong, K.-S.; Moore, J . S. Nature (London) 2001,
414, 889-893. (b) Nishinaga, T.; Tanatani, A.; Oh, K.; Moore,
J . S. J . Am. Chem. Soc. 2002, 124, 5934-5935.
(12) Zhao, D.; Moore, J . S. J . Am. Chem. Soc. 2002, 124, 9996-
9997.
(13) Odian, G. Principles of Polymerization, 3rd ed.; J ohn Wiley
& Sons: New York, 1991; pp 68-91.
(14) Hill, D. J .; Moore, J . S. Proc. Natl. Acad. Sci. U.S.A. 2002,
99, 5053-5057.
(20) The lower Mw observed in acetonitrile compared to methyl
acetate and ethyl acetate is mostly likely the result of a very
slow approach to equilibrium in this solvent.
MA034159I