Cyclic Polymers As Building Block For Cyclic Brush Polymers and Gels

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Reactive & Functional Polymers 80 (2014) 40–47

Contents lists available at ScienceDirect

Reactive & Functional Polymers


journal homepage: www.elsevier.com/locate/react

Cyclic polymers as a building block for cyclic brush polymers and gels
Ke Zhang a,b,⇑, Gregory N. Tew a,⇑
a
Department of Polymer Science and Engineering, University of Massachusetts, Amherst, MA 01003, USA
b
State Key Laboratory of Polymer Physics and Chemistry, Institute of Chemistry, The Chinese Academy of Sciences, Beijing 100190, China

a r t i c l e i n f o a b s t r a c t

Article history: Cyclic polymers, as one of the oldest topological polymers, are undergoing resurgence. This is largely
Available online 12 February 2014 ascribed to the significant achievements in modern polymer chemistry. The novel ring-expansion tech-
niques have conveniently produced varied cyclic polymers with highly topological purity and on large
Keywords: scales, which should facilitate their use in the near future. Beyond the monocyclic molecular conforma-
Cyclic polymer tions, the combination of controlled polymerization techniques and click chemistry have established a
Brush polymer robust strategy for preparing cyclic polymers with more complex architectures, such as theta, eight,
Polymer gel
and tadpole shapes. This diversification in cyclic polymer composition and conformation significantly
Ring-Expansion metathesis polymerization
Click chemistry
broadens interest in the cyclic polymers. However, compared to the synthesis achievements, the explo-
ration of cyclic polymer property and application are lagging behind. Recently, we explored the ring-
expansion metathesis polymerization on various functional ring-strained olefin monomers to produce
cyclic functional polymers, which were then used as the building blocks to fabricate cyclic brush poly-
mers and cyclic gel materials and will be discussed here.
Published by Elsevier B.V.

1. Introduction Alternatively, ring-expansion is based on the insertion of monomer


units into an activated cyclic chain. Because the cyclic polymers re-
Cyclic polymers have received significant attention recently with main intact during the whole ring-expansion process, this method
several important reviews [1–7] and a specific cyclic polymer sym- can produce high molecular weight cyclic polymers with high pur-
posium celebrated at the 245th ACS national meeting. Due to their ity, even from concentrated solutions or the bulk. One disadvan-
unique endless molecular topology, cyclic polymers have markedly tage is that it can be difficult to control the molecular weight
different characteristics from their linear counterparts, including a and polydispersity of the resultant cyclic polymers. The ring-clo-
smaller hydrodynamic volume and radius of gyration, lower melt sure and ring-expansion methods have enabled the preparation
viscosity, higher thermostability, and increased rate of crystalliza- of many types of cyclic polymers, including cyclic polystyrene
tion [1–7]. [8], polymethacrylate [9], polyacrylate [10] polyacrylamide [11]
To date, known synthetic strategies for cyclic polymers can be polyester [12] polycyclooctene [13] and polynorbornene [14]. In
generalized into two categories: ring-closure and ring-expansion addition, cyclic polymers with complex architectures have been
methods both have advantages and disadvantages. In the ring-clo- developed, including those with theta [15], eight [15], tadpole
sure method, cyclic polymers are prepared by applying highly effi- [16], block [17], and brush [18–20] shapes. The various synthetic
cient coupling chemistry to end-functionalized linear telechelic methods, polymer types, and various cyclic architectures have
polymers. Progress in combining click chemistry with controlled been thoroughly described in recent reviews [1–7].
polymerization has increased the viability of this method for syn- As a complement in this tutorial review, we mainly focus on the
thesizing cyclic polymers with controlled molecular weight and recent progress on cyclic polymers from our research group fo-
low polydispersity, but there are still limitations. To assure high cused on ring-expansion metathesis polymerization (REMP) to
topological purity, ring-closure reactions typically require dilute produce the functional cyclic polyolefins, which were then used
conditions leading to reduced yields. In addition, it is difficult to as the universal building blocks to prepare cyclic brush polymers
obtain cyclic polymers with high molecular weight by this method. and cyclic gels.

⇑ Corresponding authors at: Department of Polymer Science and Engineering,


2. Ring-expansion metathesis polymerization
University of Massachusetts, Amherst, MA 01003, USA (K. Zhang). Fax: +1 413 545
0082.
E-mail addresses: [email protected] (K. Zhang), [email protected] (G.N. REMP was firstly reported by Grubbs and coworkers in 2002
Tew). [13]. Following significant contributions from Bielawski, Boydston,

http://dx.doi.org/10.1016/j.reactfunctpolym.2014.01.012
1381-5148/Published by Elsevier B.V.
K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47 41

and Xia in the Grubbs’s research group, REMP has become a pow-
erful method for preparing cyclic polyolefins [13,14,19,21–23].
Fig. 1 shows the key mechanistic steps involved in REMP. Initiation
occurs via the cyclic olefin monomer inserting into the Ru-alkyl-
idene bond to expand the cyclic catalyst ring. Propagation then
proceeds by a repeated insertion of monomer into the growing cyc-
lic polymer. As monomer is consumed, a competing intramolecular
chain transfer reaction decreases the molecular weight of the
growing cyclic polymer, broadens its polydispersity, and generates
cyclic polymers. An additional catalyst release process can produce
the original catalyst and more catalyst-free cyclic polymer.
To date, two series of cyclic Ru-alkylidene catalysts (UC series
and SC series) of five main types (UC-5, 6, 7 and SC-5, 6) (Fig. 2)
have been developed. It was found that both the heterocycle size
and the N-heterocyclic carbene (NHC) ligand electronics signifi-
cantly impact the catalyst efficiency and the polymerization
behavior as well. The catalyst with larger ring leads to faster poly-
merization rates and lower catalyst release from cyclic polymer.
The saturated NHC ligand endows the catalyst with much faster
polymerization rate than that of the unsaturated NHC ligand. Fur-
thermore, various cyclic olefin monomers have been successfully
polymerized by REMP. As shown in Fig. 2, they are composed of
not only the commercial available cyclic olefins and their simple
derivatives having a small functional group, but also those having Fig. 2. Cyclic Ru-alkylidene catalysts and appropriate cyclic olefin monomers.

bulky side groups such as dendrons and even polymers.


As is typical of ring-expansion approaches, REMP can produce
ing the backbone topology [27–30]. When cyclic polymers are cho-
large amounts of pure cyclic polymers at high polymerization con-
sen as backbones, the novel cyclic brush polymers are designed to
centrations. Furthermore, REMP is unique in that it can obtain cyc-
bear a donut-like molecular topology.
lic polymers with the highest reported molecular weight. This
Although a few reports have been published on synthesizing
makes it a good candidate to prepare cyclic polymer building
cyclic brush polymers [18,31–33], the preparation of brush poly-
blocks for the fabrication of cyclic materials, such as the nano-scale
mers demonstrating stiff toroidal macromolecular topologies has
cyclic brush polymers and the macro-scale cyclic gels.
remained elusive [13,23,34,35]. REMP is unique here because it
meets many of the necessary requirements for building the back-
3. Cyclic brush polymers bone of cyclic brush polymers. Inspired by this, REMP has been
successfully utilized to prepare cyclic brush polymers by Grubbs’
Brush polymers are composed of a long polymer backbone with and Tew’s research groups independently [14,19,20,36,37]. Apply-
ing the three traditional routes to linear brush polymers, [grafting
densely grafted polymer side chains. Due to the steric repulsion be-
tween densely grafted bulky side chains, these macromolecules through (polymerization of macromonomers), grafting onto (cou-
pling of side chains to backbone), and grafting from (polymeriza-
have an extended, rigid worm-like conformation when the back-
bone is much longer than the side chains [24,25]. It is this unique tion of side chains from a backbone macroinitiator)], cyclic brush
molecular conformation that renders brush polymers as a unique polymers have been prepared by REMP.
building block for generating advanced nanomaterials, such as cre-
ating well-defined single-molecular hybrid nano-objects and for 4. Grafting through
fabricating periodic nanomaterials with large domain spacings by
self-assembly [26]. In addition to the linear morphology, the The first report to synthesize and visualize cyclic dendronized
molecular shapes of brush polymers can be enriched by manipulat- polymers by the grafting through technique was published by
the Fréchet and Grubbs’ research groups in 2009 [14]. A norborn-
ene-based macromonomer was designed with a second generation
dendron side chain. When using the SC-5 or UC-6 as catalyst, the
cyclic dendronized polymers were obtained by REMP (Fig. 3A).
The donut-like molecular topology was visualized by AFM with a
diameter of 30–40 nm (Fig. 3B). Following the same concept, they
further investigated the direct REMP of norbornene-based
macromonomers with a bulky polymer side chain (Fig. 3C) [19].
A series of cyclic brush polymers were then produced with ultra-
high molecular weight and various polymer side chains, including
polystyrene, polylactide, and poly(n-butyl acrylate). In this work,
the macromonomers were efficiently prepared by controlled poly-
merization techniques and click chemistry. For a moderate to high
polymerization conversion of these bulky macromonomers, the
more active SC-series catalysts were required. The donut-like
molecular topology was visualized by AFM and most cyclic brush
polymers have diameters in the range of 100–180 nm (Fig. 3D),
which is much larger than that from the cyclic dendronized poly-
Fig. 1. Key mechanistic steps involved in REMP [22]. mers. The increased molecular size is presumably caused by the
42 K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47

Fig. 3. (A) Synthesis scheme of cyclic dendronized polymers from the combination of REMP and grafting from technique, (B) the corresponding AFM characterization
(reproduced and adapted from Ref. [14]), (C) synthesis scheme of cyclic brush polymers from the combination of REMP and grafting from technique, and (D) the
corresponding AFM characterization (reproduced and adapted from Ref. [19]. Copyright 2011 Wiley).

longer and significantly extended backbone conformation. In addi- demonstrate this concept, monofunctional poly(ethylene glycol)
tion to the cyclic topology, linear contaminants were occasionally (PEG) with an amine end group and with an average of 13 repeat
observed by AFM in this method (Fig. 3D), however, it was demon- units was chosen as the coupling polymer side chains (Fig. 4(ii)).
strated that they were caused by mechanical scission of the cyclic A high grafting density of >95% was confirmed in this case. The
backbone during AFM characterization. conformation of the resultant cyclic brush polymers was character-
ized by AFM and a cyclic morphology with an outer diameter of ca.
5. Grafting onto 30 nm was observed.
Considering the difficulty of preparing amine end-functional-
The combination of REMP and the grafting onto technique for ized polymers, we derivatized the cyclic Poly-2 template with a
the formation of cyclic brush polymers was firstly developed in clickable alkyne side group on each repeat unit. Given that azide
our group in 2011 [20]. In this approach (Fig. 4), a functional nor- end-functionalized polymers are conveniently obtained by atom-
bornene-based monomer (M1) was designed to have an activated transfer radical polymerization (ATRP), cyclic Poly-2 was easily
pentafluorophenol ester side group. When applying REMP on M1 functionalized with various polymer side chains when employing
(Fig. 4(i)), high molecular weight functional cyclic polymers (cyclic the Cu(I)-catalyzed azide-alkyne click chemistry. A cyclic polymer
Poly-1) were produced bearing activated ester side groups on each brush containing polystyrene (PS) side chains was prepared to
repeat unit. The highly efficient nucleophilic substitution of acti- illustrate the concept (Fig. 4(iii)). PS-N3 was obtained by substitut-
vated esters by amines enabled cyclic Poly-1 to be used as a tem- ing the terminal bromine of PS-Br from ATRP by an azide moiety.
plate for preparing cyclic brush polymers when coupling with The highly efficient Cu(I)-catalyzed azide-alkyne click chemistry
amine end-functionalized polymer side chains. As an example to resulted in quantitative grafting density.

Fig. 4. Synthesis of cyclic brush polymers from the combination of REMP and grafting onto technique and the corresponding AFM characterization [20].
K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47 43

Most recently, we explored the use of supramolecular chemis- polymers. As a result, this initial concept, based on the combina-
try techniques to prepare cyclic brush polymers by combining tion of REMP and the grafting from technique, is expected to be-
REMP and the grafting onto technique [37]. In order to include or- come a powerful toolbox for the preparation of functionalized
ganic/inorganic hybrid concept as well, the metal–ligand supramo- cyclic polymers and brushes with varied molecular architectures
lecular interaction was chosen as the building unit for the and functionalities.
preparation of cyclic brush polymers. Due to the outstanding che-
lating ability to various transition-metal ions, the terpyridine li- 7. Cyclic gels
gand was introduced as the side group on each repeat unit of the
cyclic Poly-3 (Fig. 5). The classic two-step terpyridine-Ru(III)/Ru(II) The properties of gels are determined by their inherent chemi-
supramolecular chemistry was specifically chosen to demonstrate cal structures. Two excellent examples have demonstrated the
this technique. In this approach, various terpyridine end-function- general rule that manipulating the cross-linking structure signifi-
alized polymer side chains were synthesized by reversible addi- cantly alters the final gel characteristics. Topological (TP) gels
tion-fragmentation chain-transfer polymerization (RAFT) with a [38], developed by Okumura used a-cyclodextrin (a-CD) and
terpyridine-functionalized RAFT agent. As the first step, the poly- poly-(ethylene glycol) (PEG). By chemically cross-linking the R-
merterpyridineRuCl3 monocomplex was obtained from the ter- CD component, ‘figure-of-eight’ cross-linkages were generated that
pyridine end-functionalized polymer precursor by refluxing with allowed ‘sliding’ of the PEG chains. This structural topology gave
RuCl3. Subsequently, the metallo-supramolecular cyclic brush good tensile strength and a large swelling capacity to the gel. An-
polymers were prepared by reacting the terpyridine-Ru(III) other case is the double network (DN) gels [39], developed by Gong
mono-complex onto the cyclic backbone under reducing condi- that contain a highly cross-linked, stiff poly(2-acrylamido-2-meth-
tions (ethanol, N-ethylmorpholine), to produce the stable bis(ter- ylpropanesulfonic acid) network and a loosely cross-linked
pyridine)–Ru(II) linkage connecting the polymer side chains and poly(acrylamide) network.
the cyclic polynorbornene main chains (Fig. 5). Because of the Structural defects in networks, such as loops and dangling chain
inherent metal-containing molecular structure, these novel supra- ends, deteriorate gel properties as they do not contribute to the
molecular hybrid cyclic brush polymers provide an ideal model for elasticity of the crosslinked networks. Eliminating these defects
direct imaging of single polymer chains by TEM. As shown in Fig. 5, is difficult, if not impossible using, many traditional employed
a representative TEM image clearly shows the cyclic molecular methods for network formation, including radical copolymeriza-
topology with an average diameter of ca. 40 nm. This is the first tion of monomers (with a small amount of cross-linker), or the di-
time that TEM has been used to directly image the molecular rect cross-linking of linear polymer chains. In contrast, if cyclic
topology of cyclic polymers. polymer chains are used rather than their linear counterparts,
then, in theory, networks without dangling chain ends could be
6. Grafting from formed when more than two cross-linkages per polymer chain
are constructed. This should result in gels with improved proper-
In 2012, we reported the third method for the preparation of ties. Inspired by this idea, we prepared cyclic gels from chemically
cyclic brush polymers based on REMP by combining it with the cross-linking cyclic polymers [40].
grafting from technique [36]. The formation of cyclic brush poly- REMP was used to synthesize cyclic polymer building block be-
mers with polyester side chains was selected as a model system cause it not only produces polymers with high molecular weight,
to demonstrate this approach (Fig. 6). The functional norbornene- but also generates large quantities of pure cyclic polymers. Cyclic
based monomer M2 was designed to have a hydroxyl side group and linear poly(5-acetoxy-1-cyclooctene) (PACOE) were then pre-
which was used as the initiating group for ring-opening polymer- pared by REMP and ring-opening metathesis polymerization
ization of cyclic ester monomers. Following cyclic polymer forma- (ROMP) respectively (Fig. 7). To eliminate the influence of molecu-
tion, a series of cyclic brush polymers were subsequently prepared lar conformation on the gel formation and properties, the cyclic
with various polyester side chains, such as poly(d-valerolactone) and linear PACOE were designed to have similar molecular weight
(PVL), poly(L-lactide) (PLLA), and poly(e-caprolactone) (PCL). (Mn,cyclic = 595,700; Mn,linear = 571,000) and polydispersity
Employing similar ideas, when the cyclic macroinitiators from (PDIcyclic = 1.77; PDIlinear = 1.70). The remaining internal double
REMP have various initiating sites, such as alkyl bromide or thio- bonds enabled facile cross-linking. As shown in Fig. 8, the classical
carbonylthio groups, controlled polymerization techniques like thiol-ene chemistry was utilized to initiate radical formation and
ATRP or RAFT coupled with grafting from generated cyclic brush the subsequent carbon-centered radical coupling cross-linked the

Fig. 5. Synthesis of metallo-supramolecular cyclic brush polymers by the combination of REMP and grafting onto technique [37].
44 K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47

Fig. 6. Synthesis of cyclic brush polymers by combination of REMP and grafting from technique [36].

Table 1
Gel fraction, swelling ratio in different organic solvents, and shear modulus of cyclic
and linear gels in Xylene [40].

Runa GF (%)b Q (mL/g)c G (Pa)d


DCM THF Benzene Anisole Xylene
Cyclic gels
1e 51.3 126.2 112.8 110.1 101.3 94.74 111
2f 47.5 124.1 104.5 103.8 100.6 87.39 72
g
3 35.7 199.6 160.5 157.9 150.8 125.4 ND
Linear gels
4e 74.3 40.95 38.56 37.14 34.62 31.16 3979
5f 59.8 71.92 61.29 62.96 58.25 51.46 1028
6g 42.9 174.3 146.4 137.9 132.9 116.8 ND
a
For runs 1–3, gels were obtained from cyclic PACOE; for runs 4–6, gels were
obtained from linear PACOE.
b
GF: Gel fractions.
c
Q: Swelling ratios.
d
G: Shear moduli, where ND means ‘not determined’.
e
The initial PACOE concentration was 84 mg/mL.
f
The initial PACOE concentration was 42 mg/mL.
g
The initial PACOE concentration was 21 mg/mL.

Fig. 7. Synthesis of cyclic PACOE from REMP and linear PACOE from ROMP [40].
larger statistical mesh size (n), and a higher swelling ratio (Q)
at the same initial PACOE concentration (C0). These differences
became more pronounced with increasing C0 (Table 1 and
internal double bonds in PACOE, leading to gel formation. By this Fig. 9A–D).
cross-linking condition, the gels were conveniently obtained in 2. For the linear gels, as expected, the increment of C0 increased in
air at room temperature in just a few minutes. GF, l, and G, and decreased in n and Q [41–45]. The cyclic gels
Table 1 and Fig. 9 compared the differences in gel formation and demonstrated similar behavior at low C0; however, at higher
properties between the cyclic and linear gels. Several general rules C0, the increase of C0 led to the unusual increase of GF, Q, G,
can be extracted from these data as follows: and n simultaneously (Fig. 9A–D).
3. At the same preparation state (same C0), the swelling ability and
1. Compared to conventional networks formed by cross-linking maximum strain at break of the cyclic gels were always greater
linear polymer chains, cyclic gels had a lower gel fraction than those of the linear gels, and these differences became more
(GF), statistic cross-link density (l), and shear modulus (G), a pronounced as C0 increased (Fig. 9E and F).

Fig. 8. Preparation of cyclic and linear gels from the thiol-ene chemistry.
K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47 45

240
A Cyclic gels B 3.0
Cyclic gels/Linear gels
210 Linear gels

180 2.5

150

Q ratio
2.0
Q
120

90 1.5

60
1.0
30
20 30 40 50 60 70 80 90 20 30 40 50 60 70 80 90
C 0 (mg/mL) C 0 (mg/mL)

C 1.2 Cyclic gels


Linear gels
D 70
Cyclic gels
Linear gels
1.0
60
0.8
µ (µmol/mL)

ξ (nm)
50
0.6

0.4 40

0.2 30
0.0
20 30 40 50 60 70 80 90 20 30 40 50 60 70 80 90
C 0 (mg/mL) C 0 (mg/mL)

E 0 F 0

-5
-10 -10
Stress (KPa)
Stress (KPa)

-15
-20
-20

-30 -25
Cyclic gels Cyclic gels
Linear gels -30 Linear gels
-40 Mooney-Rivlin Fit Mooney-Rivlin Fit
-35
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 -0.8 -0.6 -0.4 -0.2 0.0
Strain Strain

G 210 H 700
180
2 600
150 -0.31 (R =0.68)
120 500
ξ (nm)
Q'

90 -0.80 (R =0.99)
2
2
400 -0.58 (R =0.99)

60

300
Cyclic gels Cyclic gels
Linear gels Linear gels

10 20 30 40 50 60 0.005 0.01 0.015 0.02 0.025


Φ0 ν2,s

Fig. 9. Swelling ratio (Q) (A), the ratio of Q (B), statistical cross-link density (l) (C), and statistical mesh size (n) (D) between cyclic and linear gels as function of the initial
PACOE concentration (C0). Stress–strain curves for cyclic and linear gels prepared at C0 of 84 mg/mL (E) and 42 mg/mL (F). (G) Volumetric equilibrium swelling ratio (Q0 ) as a
function of C0 for cyclic and linear gels and (H) statistical network mesh size (n) as a function of equilibrium polymer volume fraction (t2,s) for cyclic and linear gels [40].

4. Although the linear gels follow the theoretical scaling predictions ture. Due to the large number of internal double bonds remaining
well [45,46], the behavior of cyclic gels strongly deviated from the in the PACOE backbone, there is a competition between
description of the classical gel scaling laws (Fig. 9G and H). intramolecular and intermolecular carbon centered radical cou-
pling. This competition is influenced by C0, but also by the confor-
These differences were explained based on the unique molecu- mation of cyclic and linear polymer chains. The mean square radius
lar topology and its influence on network formation and architec- of gyration ðhR2g iÞ of the linear polymer is two times larger than
46 K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47

the corresponding cyclic polymer [13], meaning that cyclic poly- weight. Thus, while the classical mesh size is reduced at higher
mers are more compact and are more likely to react by intramolec- C0 as expected, the larger ring size of the cyclic polymer chains cre-
ular versus intermolecular cross-linking at a given C0 compared to ates a new mesh spacing, so that statistically, more space is avail-
their linear counterparts. This results in linear gels having a higher able in the networks formed from cyclic polymers at this
GF, l, and G, and a smaller n and Q than the corresponding cyclic concentration. This leads to the cyclic gels with the non-traditional
gels at each C0. trends of GF, G, Q, l and n as a function of C0, and again with the
By varying concentration, it was shown that at lower C0, strong derivation from the classical scaling laws for describing lin-
intramolecular cross-linking reactions dominate over intermolecu- ear gels.
lar ones. Using linear chains (Fig. 10A), this produces loop defects An increasingly popular strategy for modifying the mechanical
that essentially waste the cross-linking points, resulting in lower and swelling properties of organic polymer gels is the incorpora-
l, since the loops do not contribute to the elasticity of the network. tion of inorganic components [47,48]. Taking advantage of the ver-
When cross-linked at lower C0, linear gels are formed with a large satility of the metallo-supramolecular chemistry to cross-link
n, which is reflected by a smaller GF and G, and a larger Q. In con- cyclic polymer chains, we recently created novel gels with the
trast, for cyclic polymer chains (Fig. 10B), as intramolecular cross- structural merits of both the cyclic gels and organic/inorganic hy-
links effectively reduce the cyclic ring size, the network structure brid materials [37]. Such gels are expected to have unique and
of cyclic gels is similar to that of linear gels obtained in this situa- interesting properties. The cyclic polymer/inorganic hybrid gels
tion. At lower C0, the differences in the behavior of the networks are produced with inorganic transition-metal ions(II) serving to
from linear and cyclic PACOE are not as obvious. Upon increasing cross-link the terpyridine-functionalized cyclic Poly-3 (Fig. 5).
C0, the overlap of polymer chains in solution increases, generating Two model transition metal ions, Fe2+ and Ni2+, were chosen to
more intermolecular cross-linking reactions. In addition, it is demonstrate this concept, and, as shown in Fig. 11A, free-standing
known that for linear chains this increase in intermolecular metallo-supramolecular gels were obtained in air at room temper-
cross-links generates a higher l, a tighter network structure, and ature in less than 5 min, using a molar ratio of 1:5 between metal
thus a smaller n [41–45], which can be observed by a larger GF ions and terpyridine groups. In addition to the ‘cyclic’ and ‘hybrid’
and G, and a smaller Q. The gels made from linear PACOE showed molecular structure, another appealing feature of these novel gels
this behavior (Fig. 10C). However, for cyclic polymer chains is their tunable properties under certain external stimuli, which is
(Fig. 10D), the decrease in intramolecular cross-links reduces the enabled by the characteristic reversibility of the metallo-supramo-
number of these ‘effectively smaller rings’ so that as more lecular bonds. To demonstrate this, in one example a large excess
intermolecular cross-links are formed, the effective cyclic ring size of free 2,20 ;60 ,200 -terpyridine was added to the Ni2+ cross-linked
‘‘expands’’ and begins to approach the cyclic polymer’s molecular metallo-supramolecular gel and allowed to stand at room temper-

Fig. 10. The ideal network structural units are shown to highlight the most important features in each system. The top two systems correspond to low C0 for linear PACOE (A)
and cyclic PACOE (B), while the bottom two panels correspond to high C0 for linear PACOE (C) and cyclic PACOE (D) [40].
K. Zhang, G.N. Tew / Reactive & Functional Polymers 80 (2014) 40–47 47

the classical hybrid or double network concepts founded in linear


gels, a series of novel gel materials could be expected with not only
hierarchy network structure but also the advanced gel properties.

Acknowledgments

Generous support was primarily provided by NSF (DMR-


0820506 and CMMI-0531171), NSFC (21090353 and 21374122).
Partial support was provided from ARO W911NF-09-1-0373 and
ONR N00014-10-1-0348. Shared facilities support also comes from
DMR-0820506. K.Z thanks the Bairen project from CAS for support.

References

[1] Y. Tezuka, Polym. J. 44 (2012) 1159.


[2] Z. Jia, M.J. Monteiro, J. Polym. Sci., Part A: Polym. Chem. 50 (2012) 2085.
[3] T. Yamamoto, Y. Tezuka, Polym. Chem. 2011 (1930) 2.
[4] J.N. Hoskins, S.M. Grayson, Polym. Chem. 2 (2011) 289.
[5] H.R. Kricheldorf, J. Polym. Sci., Part A: Polym. Chem. 48 (2010) 251.
[6] B.A. Laurent, S.M. Grayson, Chem. Soc. Rev. 38 (2009) 2202.
[7] K. Endo, Adv. Polym. Sci. 217 (2008) 121.
[8] B.A. Laurent, S.M. Grayson, J. Am. Chem. Soc. 128 (2006) 4238.
[9] M. Glassner, J.P. Blinco, C. Barner-Kowollik, Macrom. Rapid Commun. 32
(2011) 724.
[10] R. Nicolay, K. Matyjaszewski, Macromolecules 44 (2011) 240.
[11] J. Xu, J. Ye, S. Liu, Macromolecules 40 (2007) 9103.
[12] W. Jeong, J.L. Hedrick, R.M. Waymouth, J. Am. Chem. Soc. 129 (2007) 8414.
[13] C.W. Bielawski, D. Benitez, R.H. Grubbs, Science 297 (2002) 2041.
Fig. 11. Pictures of a metallo-supramolecular cyclic gel and their stimulus–
[14] A.J. Boydston, T.W. Holcombe, D.A. Unruh, J.M.J. Frechet, R.H. Grubbs, J. Am.
response [37]. Chem. Soc. 131 (2009) 5388.
[15] H. Oike, H. Imaizumi, T. Mouri, Y. Yoshioka, A. Uchibori, Y. Tezuka, J. Am. Chem.
Soc. 122 (2000) 9592.
ature. After 20 days, the free-standing gel became fluid (Fig. 11B). [16] D.E. Lonsdale, M.J. Monteiro, Chem. Commun. 46 (2010) 7945.
[17] S. Honda, T. Yamamoto, Y. Tezuka, J. Am. Chem. Soc. 132 (2010) 10251.
Detailed structure–property studies on these novel gels, specifi- [18] M. Schappacher, A. Deffieux, Science 319 (2008) 1512.
cally their dynamics, are currently underway. [19] Y. Xia, A.J. Boydston, R.H. Grubbs, Angew. Chem. Int. Ed. 50 (2011) 5882.
[20] K. Zhang, M.A. Lackey, Y. Wu, G.N. Tew, J. Am. Chem. Soc. 133 (2011) 6906.
[21] Y. Xia, A.J. Boydston, Y. Yao, J.A. Kornfield, I.A. Gorodetskaya, H.W. Spiess, R.H.
Grubbs, J. Am. Chem. Soc. 131 (2009) 2670.
8. Conclusion and outlook [22] A.J. Boydston, Y. Xia, J.A. Kornfield, I.A. Gorodetskaya, R.H. Grubbs, J. Am. Chem.
Soc. 130 (2008) 12775.
Three general methods have been developed for the formation [23] C.W. Bielawski, D. Benitez, R.H. Grubbs, J. Am. Chem. Soc. 125 (2003) 8424.
[24] M.F. Zhang, A.H.E. Muller, J. Polym. Sci., Part A: Polym. Chem. 43 (2005) 3461.
of cyclic brush polymers by combining REMP with grafting
[25] S.S. Sheiko, B.S. Sumerlin, K. Matyjaszewski, Prog. Polym. Sci. 33 (2008) 759.
through, grafting onto, and grafting from techniques, respectively. [26] J. Rzayev, Acs Macro Lett. 1 (2012) 1146.
These methods illustrate that various cyclic brush polymers can be [27] K.L. Beers, S.G. Gaynor, K. Matyjaszewski, S.S. Sheiko, M. Moller,
conveniently produced and that many other molecular combina- Macromolecules 31 (1998) 9413.
[28] Y. Xia, J.A. Kornfield, R.H. Grubbs, Macromolecules 42 (2009) 3761.
tions can be easily envisaged. For example, when using the cyclic [29] K. Matyjaszewski, S.H. Qin, J.R. Boyce, D. Shirvanyants, S.S. Sheiko,
brush polymers with appropriate functional polymer side chains Macromolecules 2003 (1843) 36.
as single molecule templates, various well-defined cyclic nano-ob- [30] M. Schappacher, A. Deffieux, Macromolecules 38 (2005) 4942.
[31] M. Schappacher, A. Deffieux, J. Am. Chem. Soc. 130 (2008) 14684.
jects such as cyclic hybrid nano-cylinders and cyclic nano-tubes [32] M. Schappacher, A. Deffieux, Angew. Chem. Int. Ed. 48 (2009) 5930.
could be expected. In addition, the self-assembly of amphiphilic [33] B.A. Laurent, S.M. Grayson, J. Am. Chem. Soc. 133 (2011) 13421.
cyclic brush polymers is likely to produce nanomaterials with [34] D.E. Lonsdale, C.A. Bell, M.J. Monteiro, Macromolecules 43 (2010) 3331.
[35] D.E. Lonsdale, M.J. Monteiro, J. Polym. Sci., Part A: Polym. Chem. 48 (2010)
attractive morphologies and properties. Unimolecular micelles 4496.
with various size and functionality are of considerable interest [36] K. Zhang, G.N. Tew, ACS Macro Lett. 1 (2012) 574.
and should be available through these synthetic methods. [37] K. Zhang, Y. Zha, B. Peng, Y. Chen, G.N. Tew, J. Am. Chem. Soc. (2013).
[38] Y. Okumura, K. Ito, Adv. Mater. 13 (2001) 485.
Although we have produced a fundamentally new gel system by [39] J.P. Gong, Y. Katsuyama, T. Kurokawa, Y. Osada, Adv. Mater. 15 (2003) 1155.
cross-linking the cyclic polymers and discovered its unique gel for- [40] K. Zhang, M.A. Lackey, J. Cui, G.N. Tew, J. Am. Chem. Soc. 133 (2011) 4140.
mation and properties, research on cyclic gels is just beginning and [41] A.S. Hoffman, Adv. Drug Deliver. Rev. 54 (2002) 3.
[42] K.S. Anseth, C.N. Bowman, L. BrannonPeppas, Biomaterials 17 (1996) 1647.
many further studies are needed. For example, elaborate cross-
[43] O. Okay, Makromol. Chem. 189 (1988) 2201.
linking methods are urgently needed to efficiently manipulate [44] G.M. Cruise, D.S. Scharp, J.A. Hubbell, Biomaterials 19 (1998) 1287.
the intramolecuar and intermolecular cross-link in the formation [45] T. Canal, N.A. Peppas, J. Biomed. Mater. Res. 23 (1989) 1183.
of cyclic gels, since the intramolecular cross-link effectively re- [46] S.P. Obukhov, M. Rubinstein, R.H. Colby, Macromolecules 27 (1994) 3191.
[47] K. Haraguchi, T. Takehisa, Adv. Mater. 14 (2002) 1120.
duces the cyclic ring size and seriously degrades the cyclic gel’s [48] Q. Wang, J.L. Mynar, M. Yoshida, E. Lee, M. Lee, K. Okuro, K. Kinbara, T. Aida,
properties. In addition, by merging the cyclic topology effects with Nature 463 (2010) 339.

You might also like