Construction of hyperbranched poly(alkenephenylene)s by diyne polycyclotrimerization: Single-component catalyst, glycogen-like macromolecular structure, facile thermal curing, and strong thermolysis resistance

Article abstract of DOI:10.1021/ma049871+

A simple catalyst of TaBr₅ NbBr₅, or NbBr₃CH₃O(CH₂₎₂OCH₃ is developed, which readily effects polycyclotrimerizations of α,ω- alkenediynes {HC≡C(CH_2)mC=CH, m = 4-6 [1(m)]} at room temperature, giving hyperbranched poly(alkenephenylene)s [hb-P1(m)] with high molecular weights (M_w up to ～270 × 10³) in high yields (normally >80%). The polymers prepared under optimal reaction conditions are completely soluble in common organic solvents such as toluene, THF, and chloroform. Spectroscopic characterizations prove that the polymers comprise of linear (L) and dendritic units (D) of 1,2,4/1,3,5-trialkylbenzenes and terminal unit (T) of 1,2,4-trialkylbenzene (or 6-alkyltetralin). The L and D units are generated by geostructurally different addition modes, similar to those used by nature to create L and D units of glycogen, a hyperbranched biopolymer. With the aids of model reactions and simulations, detailed structural analyses reveal that hb-P1(4) possesses a degree of branching of 64% and consists of 36%, 32%, and 32% of L, D, and T units, respectively, with 1,2,4-trialkylbenzene being the predominant isomeric structure (74%). The polymer shows outstanding thermal properties: it readily cures when baked at a moderate temperature of 100°C and loses little of its weight when heated to a high temperature of ～500°C.

Full text of DOI:10.1021/ma049871+

5
196
Macromolecules 2004, 37, 5196-5210
Construction of Hyperbranched Poly(alkenephenylene)s by Diyne
Polycyclotrimerization: Single-Component Catalyst, Glycogen-like
Macromolecular Structure, Facile Thermal Curing, and Strong
Thermolysis Resistance
†
†
†
†
Ron gh u a Zh en g, Hon gch en Don g, Ha n P en g, J a ck y W. Y. La m , a n d
,
†,‡
Ben Zh on g Ta n g*
Department of Chemistry, The Hong Kong University of Science & Technology, Clear Water Bay,
Kowloon, Hong Kong, China, and Department of Polymer Science and Engineering, Peking University,
Beijing 100871, China
Received J anuary 16, 2004; Revised Manuscript Received April 29, 2004
ABSTRACT: A “simple” catalyst of TaBr
effects polycyclotrimerizations of R,ω-alkenediynes {HCtC(CH
temperature, giving hyperbranched poly(alkenephenylene)s [hb-P1(m)] with high molecular weights (M
up to ∼270 × 10 ) in high yields (normally >80%). The polymers prepared under optimal reaction
conditions are completely soluble in common organic solvents such as toluene, THF, and chloroform.
Spectroscopic characterizations prove that the polymers comprise of linear (L) and dendritic units (D) of
5
, NbBr
5
, or NbBr
3
‚CH
3
O(CH
2 2 3
) OCH is developed, which readily
2 m
)
CtCH, m ) 4-6 [1(m)]} at room
w
3
1
,2,4/1,3,5-trialkylbenzenes and terminal unit (T) of 1,2,4-trialkylbenzene (or 6-alkyltetralin). The L and
D units are generated by geostructurally different addition modes, similar to those used by nature to
create L and D units of glycogen, a hyperbranched biopolymer. With the aids of model reactions and
simulations, detailed structural analyses reveal that hb-P1(4) possesses a degree of branching of 64%
and consists of 36%, 32%, and 32% of L, D, and T units, respectively, with 1,2,4-trialkylbenzene being
the predominant isomeric structure (74%). The polymer shows outstanding thermal properties: it readily
cures when baked at a moderate temperature of 100 °C and loses little of its weight when heated to a
high temperature of ∼500 °C.
In tr od u ction
Ch a r t 1
Naturally occurring macromolecules with highly
branched structures (or hyperbranched biopolymers)
serve important biological functions in the living world.
Glycogen, for example, is a polymer of glucose with a
treelike structure (Chart 1) that is used by animals as
solid fuels to store and transport energy.1,2 Nature has
admirably designed the molecular structure of glycogen
3
for fulfilling its biological missions. Its linear structure
(L) enables the carbohydrate chain to be readily syn-
thesized or degraded respectively by glycogenesis or
glycogenolysis catalyzed by the enzymes abundant in
the living cells, noting that most proteins are linear
polymers. The dendritic structure (D), on the other
hand, populates the terminal groups (T) in the three-
dimensional space, allowing many glucose monomers to
be added to, or released from, glycogen in parallel on
demand. This ensures fast polymer synthesis or degra-
dation and hence quick responses to physiological
requests or stimuli. The geometric distribution of the L
and D units in glycogen is random, which may be
aesthetically unattractive but practically useful. The
geostructural randomness, for example, endows the
hyperbranched biopolymers with great conformational
flexibilities to cope with variegated biological circum-
stances or situations in different species.4
Scrutiny of the glycogen structure reveals that the L
and D units are formed respectively via R-1,4 and R-1,6
glycosidic linkages (or ethereal bonds) of the glucose
3
residues. In other words, the L and D units of the
hyperbranched biopolymer are generated by the linking
modes that are same in chemical nature but different
in geometric structure. Creation of hyperbranched
synthetic polymers is a hot topic of current interest,6
and many such polymers have been prepared by poly-
merizations of ABm-type monomers, where A and B are
mutually reacting functional groups with m g 2. In such
systems, the D units are generated by polycoupling of
A and B groups, whereas the L units are formed when
some B groups are left unreacted inside the cores, as
,5
As discussed above, both L and D units are indispen-
sable for glycogen to perform its biological duties.
-8
†
The Hong Kong University of Science & Technology.
Peking University.
‡
*
To whom correspondence should be addressed (The Hong Kong
University of Science & Technology): phone +852-2358-7375; Fax
852-2358-1594; e-mail tangbenz@ust.hk.
+
1
0.1021/ma049871+ CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/11/2004

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5197
Sch em e 1
irregularity makes the polymers completely soluble in
common organic solvents and readily processable by
normal solution techniques. The high density of the
aromatic rings confers strong thermolysis resistance on
the polymers: hb-P1(4), for example, loses almost no
weight when heated to a temperature as high as ∼500
°
C.
Exp er im en ta l Section
Ma ter ia ls. Unless otherwise stated, all the solvents and
reagents used in this study were purchased from Aldrich and
used as received. Toluene, hexane, dioxane, and THF were
distilled from sodium benzophenone ketyl immediately prior
to use. Dichloromethane (DCM) was distilled over calcium
hydride. Metal halides MX
and trihalo(1,2-dimethoxyethane)niobium [NbX′
OCH CH OCH (DME)] were commercial products
Aldrich) of highest purities. 1,7-Octadiyne, 1,8-nonadiyne, and
5
(M ) Ta, Nb; X ) F, Cl, Br, I)
3
L: X ) Cl, Br;
L ) CH
(
3
2
2
3
diagrammatically illustrated in Scheme 1 for an AB2-
type polycondensation. The formation of the L units is
thus mechanistically associated with the incomplete
coupling or mismatching between A and B groups,
which may explain why the L units are often referred
to as structural defects in the research community of
dendritic polymers. Clearly, the biological and synthetic
systems are taking different mechanistic approaches to
the creation of the L and D units.
We here report a simple transition-metal-catalyzed
diyne polycyclotrimerization system that generates the
L and D units by a mechanism similar to that of the
biological system but different from that of the “con-
ventional” synthetic system. In the diyne polycyclotri-
merization, all the triple bonds are reacted, without any
unreacted ethynyl functional groups left inside the core,
which thus differs our system from the traditional ABm-
type polycondensation system. The polymerization propa-
gates in linking modes that are same in chemical nature
but different in geometric structure (Scheme 2), thus
sharing a mechanistic similarity to glycogenesis in the
living cells. The simultaneous propagations via the
different geometric growth patterns give rise to poly-
1
,9-decadiyne were purchased from Farchan, distilled from
calcium hydride, and stored in a dark, cold, dry place in an
exposure-proof refrigerator. Heteroatom-containing diyne
[(CHtC-p-C H ) Si(CH ) ] and triyne [(CHtC-p-C H ) N] were
6
4
2
3
2
6
4 3
synthesized in our laboratories according to our previously
published procedures.9
In str u m en ta tion . Weight (M
w
)- and number-average mo-
lecular weights (M ) and polydispersity indexes (PDI) of the
n
polymers were estimated by a Waters Associates gel perme-
ation chromatograph (GPC) system in THF using a set of
monodisperse polystyrenes as calibration standards. IR spectra
were recorded on a Perkin-Elmer 16 PC FTIR spectrophotom-
eter using solid films on, or liquid films between, sodium
1
13
chloride plates. H and C NMR spectra were measured on a
Bruker ARX 300 NMR spectrometer using chloroform-d as
solvent and tetramethylsilane (TMS) as internal reference (δ
) 0). Mass spectra were recorded on a Finnigan TSQ 7000
triple quadrupole spectrometer operating in a chemical ioniza-
tion (CI) mode using methane as carrier gas. UV-vis spectra
were measured on a Milton Ray Spectronic 3000 array
spectrophotometer, using DCM as solvent. Thermogravimetric
analyses (TGA) were carried out on a Perkin-Elmer TGA 7
analyzer at a heating rate of 20 °C/min under nitrogen or in
air.
(alkenephenylene)s with hyperbranched structures con-
P olym er iza tion . All polycyclotrimerization reactions were
carried out under nitrogen using standard Schlenk techniques.
A typical experimental procedure for the polymerization of 1(4)
sisting of both L and D repeat units of random dis-
tribution [hb-P1(m); Scheme 3]. This geostructural
Sch em e 2

5
198 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
Sch em e 3
is given below as an example. To a thoroughly baked and
carefully evacuated 15 mL Schlenk tube with a three-way
stopcock on the sidearm was placed 14.5 mg (0.025 mmol) of
P r ep a r a tion of Mod el Com p ou n d s: 1,2,4- a n d 1,3,5-
Tr ih exylben zen es (8). Trisubstituted benzenes were syn-
thesized as model compounds by monoyne(/diyne) (co)cycliza-
tions under conditions similar to those used in the diyne
polycyclotrimerization described above. For preparation of 8,
the reaction was carried out in toluene (5 mL), using 1-octyne
TaBr
2.3 mL) was injected into the tube using a hypodermic syringe,
5
under nitrogen in a glovebox. Freshly distilled toluene
(
and the mixture was stirred for 5 min. Monomer, 1(4) (0.20
mL, 1.5 mmol) was then syringed into the catalyst solution.
The resultant mixture was stirred at room temperature under
nitrogen for 6 h, after which the reaction was quenched by a
few drops of methanol. The mixture was added dropwise to
5
(0.74 mL, 5.0 mmol) as reactant and TaBr (29.0 mg, 0.05
mmol) as catalyst (Scheme 4). The reaction products were
purified by silica gel chromatography, using hexane as eluent.
A colorless liquid of 8 was isolated in ∼91% yield. Separation
of 1,2,4-8 from 1,3,5-8 was attempted but failed. The molar
ratio of 1,2,4-8 to 1,3,5-8 was estimated by NMR analysis to
∼
300 mL of methanol through a cotton filter under stirring.
The polymer precipitate was allowed to stand overnight and
was then collected by filtration. The isolated polymer was
washed with methanol and dried under vacuum at room
temperature to a constant weight. A white powdery product
-
1
be 3.0:1.0. IR (neat), ν (cm ): 3000 (Ar-H stretching), 2956,
2870 (CH stretching), 2927, 2857 (CH stretching), 1888, 1771
(overtone band, trisubstituted benzene ring), 1603, 1572, 1499
(-CdC- ring stretching), 1466 (CH scissoring), 1378 (CH
bending), 823 (Ar-H bending), 724 (CH
(300 MHz, CDCl ), δ (ppm): 7.04 [d, 1H, Ar-H (g)], 6.94 [s,
1H, Ar-H (h)], 6.92 [overlapping dd, 1H, Ar-H (i)], 6.80 [s,
3H, Ar-H (j)], 2.56 (m, 6H, Ar-CH -), 1.56 (m, 6H, Ar-CH
CH -), 1.30 (m, 18H, -CH -), 0.88 (m, 9H, -CH
(75 MHz, CDCl ), δ (ppm): 142.7, 140.3, 140.1, 137.7, 129.2,
128.9, 125.8, 125.7 (aromatic carbons), 36.0, 35.6 (Ar-CH -),
31.8, 31.6, 31.4 (Ar-CH -CH ), 32.8, 32.4, 29.5, 29.1, 22.7,
22.6 (-CH -), 14.1 (-CH ). MS (CI): m/e 331.3 [(M + 1) ].
3
2
of hb-P1(4) was obtained in 80%. M
w
: 30 800; PDI: 3.4 (GPC).
IR (thin film), ν (cm ): 3004 (Ar-H stretching), 2929 (CH
asymmetrical stretching), 2856 (CH symmetrical stretching),
893, 1770 (overtone band, trisubstituted benzene ring), 1602,
575, 1500, 1460 (-CdC- ring stretching), 1456, 1438 (CH
2
3
-
1
1
2
2
rocking). H NMR
2
3
1
1
2
2
2
-
1
13
scissoring), 825, 809, 755, 708 (Ar-H bending). H NMR (300
MHz, CDCl ), δ (ppm): 7.01, 6.94, 6.90, 6.86, 6.79 (Ar-H),
.71, 2.55 (Ar-CH
MHz, CDCl ), δ (ppm): 142.4, 139.6, 136.7, 134.2, 128.9, 125.5
aromatic (C ) carbons), 35.5, 31.3, 29.1, 23.4 [aliphatic (CH
carbons]. UV (DCM, 7.5 × 10 M), λmax (nm)/ꢀmax (mol
2
2
3
). C NMR
3
3
1
3
2
2
), 1.76, 1.63 (Ar-CH
2
-CH
2
). C NMR (75
2
3
2
2
+
[
6
H
3
2
)
L
2
3
-
4
-1
-4
-1
-1
UV (DCM, 7.5 × 10 mol/L), λmax (nm)/ꢀmax (mol L cm ):
-
1
cm ): 232/2275, 269/707, 277/616.
231/1863, 269/539, 277/447.

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5199
Sch em e 4
Ta ble 1. P olym er iza tion of 1,7-Octa d iyn e 1(4)a
Sch em e 5
run
catalyst
TaF5
TaCl5
TaBr5
TaI5
NbF5
NbCl5
NbBr5
yield (%)
Sb
Mwc
PDIc
4.7
1
2
3
4
5
6
7
8
9
0
3
76
81
d
4
x
55 300
0
58
49
0
79
0
4
x
20 900
95 500
3.4
6.5
NbI5
NbCl3‚DMEe
NbBr3‚DMEe
x
1
6
-H exyl-1,2,3,4-t et r a h yd r on a p h t h a len e or 6-H exyl-
a
Carried out in toluene under nitrogen at room temperature
tetr a lin (9). This model compound was prepared by the
cocyclization reaction of 1,7-octadiyne (0.13 mL, 0.98 mmol)
with 1-octyne (0.44 mL, 2.98 mmol) catalyzed by TaBr (29.0
5
mg, 0.05 mmol) in 4.4 mL of toluene by a procedure similar to
that used in the polymerization reaction (Scheme 5). The
product was purified by silica gel column chromatography,
using hexane as eluent. A colorless liquid of 9 was obtained
in an isolation yield of 6.3%. IR (neat), ν (cm ): 3002 (Ar-H
stretching), 2955, 2871 (CH stretching), 2927, 2857 (CH
stretching), 1602, 1578, 1503 (-CdC- ring stretching), 1460
b
for 7 h; [M]0 ) 0.6 M, [cat.] ) 20 mM (40 mM for run 9). Solubility
S) tested in common organic solvents such as THF, toluene, DCM,
(
and chloroform: 4 ) partially soluble; x ) completely soluble.
c
Measured by GPC in THF on the basis of a polystyrene calibra-
tion. d Trace amount. e DME ) 1,2-dimethoxyethane.
marized in Table 1. TaF5 catalyzes the polycyclotri-
merization of 1(4) in a sluggish way, producing little
amount of polymer (3%) after 7 h reaction. As mentioned
above, the TaCl5-Ph4Sn mixture is a good catalyst for
the diyne polycyclotrimerization, but TaCl5 alone affords
a polymeric product that is only partially soluble in
common solvents (Table 1, run 2). Delightfully, TaBr5
effectively catalyzes the polycyclotrimerization of 1(4),
giving a completely soluble polymer in a high yield
-
1
3
2
(
acyclic CH
bending), 826, 807 (Ar-H bending), 727 (CH
NMR (300 MHz, CDCl
2 2 3
scissoring), 1456 (cyclic CH scissoring), 1377 (CH
rocking). 1
2
H
3
), δ (ppm): 6.97 [d, 1H, Ar-H (k)], 6.90
overlapping dd, 1H, Ar-H (m)], 6.88 [s 1H, Ar-H (l)], 2.73
br, 4H, cyclic -CH -) 2.52 (t, 2H, acyclic -CH -), 1.78 (m,
H, cyclic Ar-CH -CH -), 1.56 (m, 2H, acyclic Ar-CH
CH -), 1.30 (m, 6H, -CH
75 MHz, CDCl ), δ (ppm): 140.0, 136.8, 134.2, 129.0, 128.9,
25.6 (aromatic carbons), 35.6 (acyclic Ar-CH -), 31.7 (acyclic
-), 23.3, 23.2 (cyclic
-), 14.1 (-CH ). MS
1
2
[
(
4
2
2
2
2
2
-
1
3
2
2
3
-), 0.88 (m, 3H, -CH ). C NMR
(
81%). The excellent solubility of the polymer allows its
(
1
3
molecular weights to be estimated by GPC, whose Mw
is found to be as high as 55 300. (The absolute Mw of
the polymer is probably much higher than this relative
value calibrated by the linear polystyrene standards
because it is well-known that the GPC analyses often
underestimate the molecular weights of hyperbranched
2
Ar-CH
Ar-CH
2
-CH
-CH
2
-), 29.4, 29.0 (cyclic Ar-CH
-), 31.8, 29.1, 22.6 (-CH
2
2
2
2
3
+
(
CI): m/e 216.2 (M ).
Resu lts a n d Discu ssion
1
0,13
polymers.
) TaI5 is, however, practically incapable of
P olym er iza tion Beh a vior s. We have previously
initiating the diyne polycyclotrimerization. Clearly, the
catalytic activity of the tantalum halide is dramatically
affected by its halogen anion, with TaBr5 being the best
catalyst for the diyne polycyclotrimerization. Too tight
or loose binding of the too small (F ) or big halogen
anion (I ) with the transition-metal center is not goods
the halogen anion with the “right” bonding force or
atomic size (Br ) showing the best performance as the
polycyclotrimerization catalyst.
found that alkyne polycyclotrimerizations can be ef-
fected by TaCl5-Ph4Sn.1
0,11
To prepare the binary-
mixture catalyst, stoichiometric amounts of TaCl5 and
Ph4Sn need to be mixed in a proper solvent and aged
for an adequate period at a desired temperature with
care. Compared to the binary mixture, a single-
component catalyst is preferable and is particularly
welcomed by industrial technologists because such a
catalyst will enjoy many advantages including low cost,
easy handling, and simple operation. In this work, we
explored the possibility of using the metal halides alone
to initiate the alkyne polycyclotrimerizations in the
absence of secondary components of “cocatalysts”, with
the aim of developing simple, monocomponent catalyst
systems.
-
-
-
Similar but more pronounced halogen effects are
observed in the case of niobium halides: the bromide
gives a soluble polymer of high molecular weight (Table
1
, run 7) and the chloride yields only partially soluble
polymer, whereas the fluoride and the iodide completely
fail to initiate the diyne polymerization. Overall, the
catalytic activities of the niobium halides are lower than
those of their tantalum counterparts, suggesting that
the metal center also plays an important role in
We first attempted to polycyclotrimerize 1,7-octadiyne
1(4)] by tantalum halides, and the results are sum-
[

5
200 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
Ta ble 2. P olym er iza tion s of Di- a n d Tr iyn es in th e
a
P r esen ce of Ta Br 5
[
M]0 yield
run
monomer
(M) (%) Sb
Mwc
PDIc
1
2
3
4
5
1,7-octadiyne [1(4)]
1,8-nonadiyne [1(5)]
1,9-decadiyne [1(6)]
(CHtC-p-C6H4)2Si(CH3)2 0.10
(CHtC-p-C6H4)3N 0.10
0.60
0.27
0.24
80
80
83
61
90
x
x
x
x
×
30 800 3.4
206 400 6.2
268 500 6.1
48 600 3.2
a
Carried out under nitrogen in toluene at room temperature
for 6 h. For runs 1-3, [TaBr5] ) 10 mM; for runs 4 and 5, [TaBr5]
b
)
5 mM. Solubility (S) tested in common organic solvent such
as THF, toluene, DCM, and chloroform: x ) completely soluble;
c
×
) insoluble. Measured by GPC in THF on the basis of a
polystyrene calibration.
determining the catalytic activity of the metal halide.
When a niobium chloride complex with a 1,2-dimethoxy-
ethane ligand (NbCl3‚DME) is used to initiate the
polycyclotrimerization, excellent catalytic activity is
observed: a high molecular weight polymer is obtained
in a high yield (Table 1, run 9). In sharp contrast, its
bromide congener (NbBr3‚DME) cannot catalyze the
diyne polymerization at all.14 This set of experimental
data, coupled with the halogen effects discussed above,
clearly indicates that the catalytic activity of the transi-
tion-metal halide can be greatly tuned by molecular
engineering of halogen anion or organic ligand. We are
currently working on the development of new catalysts
with superior catalytic performance through ligand
coordination chemistry,1
5,16
and the results will be
published in a separate paper in due course.
After identifying that TaBr5 was an excellent catalyst
for the polycyclotrimerization of 1(4), we tried to use it
to polycyclotrimerize other diyne and triyne monomers.
As can be seen from Table 2, TaBr5 can effectively
catalyze the polycyclotrimerizations of all the three
alkenediynes tested in this study, that is, 1(4), 1(5), and
F igu r e 1. Effects of concentrations of (A) monomer and (B)
catalyst on the polycyclotrimerizations of 1,7-octadiyne [1(4)]
1
(6), giving hyperbranched poly(alkenephenylene)s with
3
catalyzed by TaBr
5
in toluene at room temperature for 6 h;
) 0.6 M.
high molecular weights (Mw up to ∼270 × 10 ) in high
for (A), [cat.] ) 10 mM; for (B), [M]
0
yields (g80%). TaBr5 can also initiate polycyclotrimer-
1
0,11
izations of heteroatom-containing arylenediynes.
further increased to beyond 0.90 M, the polymerization
becomes too vigorous to control; as a result, the result-
ant polymers become only partially soluble. There exists
a monomer concentration window, with the optimal [M]0
being ∼0.60 M. This concentration effect is understand-
able because at too low [M]0, the polymerization reaction
will be actively terminated by the intramolecular cy-
clization of the propagating species with the diyne
molecules (vide post), whereas at too high [M]0, inter-
molecular reactions between the propagating species
will lead to the formation of cross-linking networks.
The polycyclotrimerization of a silicon-containing diyne,
namely, bis(4-ethynylphenyl)dimethylsilane, proceeds
smoothly at low catalyst and monomer concentrations,
resulting in the formation of a completely soluble
polymer. These results clearly prove that TaBr5 is an
excellent catalyst for diyne polycyclotrimerizations.
Triynes can also be polycyclotrimerized by TaBr5, but
the resultant polymers are often insoluble: for example,
the polymerization of tris(4-ethynylphenyl)amine cata-
lyzed by TaBr5 proceeds rapidly, and the polymerization
solution quickly gels (Table 2, run 5). The triynes can
be polymerized into soluble polymers through their
copolycyclotrimerizations with di- and/or monoynes,
Compared to [M]0, the catalyst concentration [cat.]
has less effect on the polycyclotrimerization of 1(4)
(Figure 1B). Whereas doubling [M]0 (from 0.3 to 0.6 M)
can greatly accelerate the polymerization (cf. Figure 1A),
increasing [cat.] by 6-fold (from 5 to 30 mM) does not
cause as much change. The isolation yields and molec-
ular weights of all the polymers are high in the whole
range of catalyst concentrations, with [cat.] of 10 mM
being the best when the polymer yield is concerned. All
the isolated polymeric products are completely soluble
in common organic solvents such as toluene, benzene,
THF, dioxane, DCM, and chloroform.
1
7
which will be the subject of our next publication.
Using 1(4) as a model monomer, we systematically
studied the effects of reaction conditions on the alkyne
polycyclotrimerization in an effort to understand the
polymerization behaviors and to optimize the polymer-
ization process. Figure 1A shows the effect of initial
monomer concentration ([M]0) on the polycyclotrimer-
ization of 1(4). When [M]0 is lower than 0.30 M, the
polymerization is sluggish: the isolation yields of the
polymers and their molecular weights are both low,
3
being ∼3% and e10 × 10 , respectively. Increasing [M]0
Table 3 shows the effect of solvent on the polycyclo-
trimerization of 1(4). The solvent strongly affects the
diyne polymerization. In the electron-donating solvents
to 0.6 M dramatically increases the polymer yield (80%)
3
and molecular weight (Mw ∼30 × 10 ). But if [M]0 is

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5201
Ta ble 3. Solven t Effect on P olym er iza tion of
a
1
,7-Octa d iyn e 1(4)
run
solvent
yield (%)
Sb
Mwc
PDIc
1
2
3
4
5
THF
dioxane
DCM
toluene
hexane
0
0
31
80
10
x
x
4
27 800
30 800
3.0
3.4
a
Carried out at room temperature under nitrogen for 6 h using
b
TaBr5 as catalyst; [cat.] ) 10 mM, [M]0 ) 0.60 M. Solubility (S)
tested in common organic solvents such as THF, toluene, DCM,
and chloroform: x ) completely soluble; 4 ) partially soluble.
c
Measured by GPC in THF on the basis of a polystyrene calibra-
tion.
of THF and dioxane, the polymerization of 1(4) is
completely inhibited. The electron-rich solvent mol-
ecules may have coordinately complexed with the elec-
tron-deficient transition-metal center and have thus
deactivated the metal halide catalyst. The polymeriza-
tion occurs in a less polar solvent of DCM, but the
polymer yield is low (31%). In a nonpolar solvent of
toluene, the diyne polycyclotrimerization propagates
smoothly, resulting in the formation of a polymer with
a high molecular weight in a high yield (Table 3, run
4
). Hexane is also a nonpolar solvent, but it does not
perform well as a solvent for the polycyclotrimerization
of 1(4): the yield is low (10%), and the polymer is only
partially soluble. This is probably due to the precipita-
tion of high molecular weight polymers at the early
stage of the polycyclotrimerization reaction as well as
the partial cross-linking of the closely located propagat-
ing species in the polymer precipitates. Among all the
solvents tested, toluene gives the best polycyclotrimer-
ization results and has thus been the choice of poly-
merization solvent in our further investigations.
In the discussion above, we have assumed that high
molecular weight polymers are formed at the early stage
of the polycyclotrimerization reaction of 1(4). To check
whether this is indeed the case, we followed the time
course of the polymerization reaction. As can be seen
from Figure 2A, the reaction is really fast: a high
molecular weight polymer is formed in a high yield in
F igu r e 2. (A) Time course of, and (B) temperature effect on,
the polymerizations of 1,7-octadiyne 1(4) ([M]
0
) 0.6 M)
catalyzed by TaBr (10 mM) in toluene; for (A), at room
5
temperature; for (B), for 6 h.
Str u ctu r a l Ch a r a cter iza tion . In the polymeriza-
tion reactions described above, we used TaBr5 as a
single-component replacement for the binary mixture
of TaCl5-Ph4Sn, with the belief that the TaBr5-
catalyzed polymerizations propagate in the same reac-
tion mechanism, namely, polycyclotrimerization. To
make sure that TaBr5 has indeed catalyzed alkyne
cyclization, we carried out a model reaction of 1-octyne,
a monoyne (cf. Scheme 4). Under the reaction conditions
similar to those used in the diyne polymerization, the
monoyne is cyclotrimerized quantitatively (or in 100%
conversion), although the isolation yield of the trihexyl-
benzenes (8) is lower (∼91%) due to the operational loses
incurred during the product purification processes. This
model reaction clearly rules out the possibility that a
terminal alkyne may be polymerized by TaBr5 into a
linear polyene with an alternative double-bond back-
0
.5 h, the shortest time at which we stopped a poly-
merization reaction and collected a sample for analysis.
This obviously differentiates the alkyne polycyclotri-
merization from the ABm-type polycondensation: the
former is a chain polymerization while the latter is a
step polymerization. Prolonging the polymerization time
of 1(4) does not help increase the polymer yield. The
Mw and PDI of the polymer change little in the wide
time span of 0.5-8 h.
We also investigated the influence of temperature on
the polymerization of 1(4) (Figure 2B). When the reac-
tion temperature is increased from 0 to 90 °C, the yield
of the polymer is decreased from 100% (data not shown)
to 49%. The polymer formed at 0 °C is insoluble because
the very fast polymerization at the low temperature
makes the reaction difficult to control. All of the
polymers obtained from the polymerizations carried out
at other temperatures are, however, completely soluble.
High temperature is not good to the diyne polycyclo-
trimerization because both molecular weight and poly-
dispersity index of the polymer decrease with an in-
crease in the temperature. Room temperature (∼23 °C)
is used in most of the polymerization reactions con-
ducted in this work because the room temperature
reactions can afford high molecular weight polymers
with excellent solubility in high yields.
1
8,19
bone via a metathesis mechanism.
We tried to
separate the 1,3,5- and 1,2,4-isomers of the cyclotri-
merization product 8 by column chromatography and
distillation under controlled pressures but failed to
achieve our goal. Spectroscopic methods are thus used
to evaluate the molar ratio of the isomers, and it is found
that 1,2,4-8 is the major product of the cyclotrimeriza-
tion reaction (vide infra).
The model reaction evidently proves that the TaBr5-
catalyzed cyclotrimerization has converted 1(4) into a

5
202 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
hyperbranched poly(alkenephenylene) or hb-P1(4). To
further verify the polycyclotrimerization mechanism and
to learn more about the molecular structure of hb-P1(4),
we measured its spectroscopic data. Figure S1 shows
IR spectra of hb-P1(4) and its monomer (Supporting
Information). Monomer 1(4) absorbs strongly at 3298
-
1
and 2117 cm
due to tC-H and CtC stretching
vibrations, respectively. These absorption bands com-
pletely disappear in the spectrum of hb-P1(4) (Figure
S1B). Alternatively, new absorption bands appear at
-
1
1
602 and 1500 cm due to the skeletal vibrations of
2
0
benzene rings. These results suggest that the ethynyl
groups of 1(4) have been converted to the benzene rings
of hb-P1(4) by the TaBr5-catalyzed alkyne polycyclotri-
merization.
The molecular structure of hb-P1(4) is further inves-
tigated by UV analysis. Figure S2 shows the UV
absorption spectrum of hb-P1(4), along with those of its
monomer [1(4)] and model compound (8) (Supporting
Information). The spectrum of 1(4) is a flat line parallel
to the abscissa with no absorption signals recorded in
the whole spectral region of 200-900 nm because its
isolated acetylenic triple bonds are known to absorb at
F igu r e 3. 1H NMR spectra of chloroform-d solutions of (A)
2
0
<
200 nm. Polymer hb-P1(4), however, absorbs in the
1,7-octadiyne 1(4), (B) hyperbranched polymer hb-P1(4) (sample
taken from Table 3, no. 4), (C) 1,2,4/1,3,5-trihexylbenzenes 8,
and (D) 6-hexyltetralin 9. The solvent peaks are marked with
asterisks (/).
UV region, giving three peaks at 232, 269, and 277 nm,
with respective molar absorptivities of 2275, 707, and
-
1
-1
6
16 mol L cm . All these three absorption peaks are
associated with π f π* transitions of the isolated
trisubstituted benzene rings, with the peak at 232 nm
assignable to the E2 band and those at 269 and 277 nm
to the B bands. The absorption spectrum of model
compound 8, a mixture of 1,2,4- and 1,3,5-trihexylben-
zenes, is similar to that of hb-P1(4), further supporting
that the repeat units of the polymer is comprised of
trialkylbenzenes, with the benzene rings separated or
isolated by the nonconjugating alkyl spacers.
intramolecular cyclization of the diyne molecules is
favored. An excess amount of monoyne 7 (0.6 M) is
added to the reaction mixture as a cautionary measure
to suppress the formation of oligomeric species. Al-
though homo- and cocyclizations of diyne 1(4) and
monoyne 7 can give many different kinds of cyclic
products, it is anticipated that 9 will be formed as one
of the products. Indeed, pure form of 9 can be isolated,
albeit in a low yield (6%). The protons of the cyclic
methylene groups show, as expected, resonance peaks
at δ 2.73 (e) and 1.78 (d; Figure 3D), validating our
assumption that there exist terminal units of tetralin
rings in the molecular structure of the polymer.
2
0
NMR spectroscopy is the most powerful technique for
structural characterization, and indeed the NMR spec-
1
tra of hb-P1(4) are very informative. Comparing the H
NMR spectrum of 1(4) with that of hb-P1(4) shown
respectively in panels A and B of Figure 3, it is clear
that the resonance peaks of the ethynyl (a) and pro-
pargyl protons (b) of 1(4) at δ 1.97 and 2.23 have
disappeared after polymerization, with its resonance
peak at δ 1.63 (c) being the only one left. Several new
peaks appear at δ ∼ 6.50-7.20, 2.71, 2.55, and 1.76 in
the NMR spectrum of the polymer. Model compound 8
gives resonance peaks of phenyl and benzyl protons at
δ ∼ 6.8-7.0 (g-j) and 2.56 (f; Figure 3C), which helps
assign the peaks of the polymer at δ ∼ 6.50-7.20 and
13
The C NMR spectrum of hb-P1(4) exhibits resonance
peaks well corresponding to its expected molecular
structure, with no any peaks unassignable. As shown
in Figure 4, the resonance peaks of the ethynyl carbon
atoms at δ 68.5 and 83.8 seen in the spectrum of the
monomer (panel A) are absent in spectrum of the
polymer (panel B). The new resonance peaks of the
polymer in the aromatic (δ ∼ 142-126) and aliphatic
regions (δ ∼ 36-23) can be readily identified by com-
parison with those of model compounds 8 and 9 (panels
C and D, respectively) in the corresponding spectral
regions. Summarizing the spectral data of the IR, UV,
and NMR analyses, it can be concluded beyond doubt
that the diyne molecules have been polymerized into
hyperbranched poly(alkenephenylene)s via a polycyclo-
trimerization mechanism.
2
.55 respectively to the proton absorptions of the phenyl
and benzyl groups formed by the alkyne polycyclotri-
merization.
What are the peaks at δ 2.71 (e) and 1.76 (d) in the
NMR spectrum of hb-P1(4) responsible for? They are
probably arising from the absorptions of the methylene
protons of the 1,2,3,4-tetrahydronaphthalene (or tetra-
lin) rings formed by an end-capping dimerization reac-
tion (vide infra). To substantiate these peak assign-
ments, we designed another model reaction (cf. Scheme
TGA analyses reveal that hb-P1(4) is resistant to
thermolysis: a 10% weight loss is recorded when it is
heated to 493 °C under nitrogen or to 465 °C in air
(Figure 5). Poly(1-hexyne) {-[HCdC(C4H9)]n-}, a linear
polyalkyne with an alternative double-bond backbone,
loses a similar amount of weight at ∼220 °C under nitro-
5
) and synthesized another model compound, 6-hexyl-
tetralin (9). From Figure 1, we have learned that the
alkyne polycyclotrimerization is sensitively affected by
monomer concentration: at low [M]0 (e0.3 M), little
polymer has been formed. It is therefore possible to
prepare the model compound 9 using a low concentra-
tion of 1(4) (e.g., at [M]0 ) 0.2 M), at which the
2
1
gen, whereas linear poly(p-phenylene) [-(C6H4)n-] is
stable at ∼450 °C in air or at ∼550 °C in inert
2
2
atmosphere. The thermal stability of hb-P1(4) is thus
similar to that of the polyphenylene but different from
that of the polyalkyne. This stability data serve as a

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5203
F igu r e 4. 13C NMR spectra of chloroform-d solutions of (A)
1
,7-octadiyne 1(4), (B) hyperbranched polymer hb-P1(4) (sample
taken from Table 3, no. 4), (C) 1,2,4/1,3,5-trihexylbenzenes 8,
and (D) 6-hexyltetralin 9. The solvent peaks are marked with
asterisks (/).
1
F igu r e 6. Expanded H NMR spectra of (A) hyperbranched
polymer hb-P1(4) (sample taken from Table 3, no. 4), (B) 1,2,4/
1
,3,5-trihexylbenzenes 8, and (C) 6-hexyltetralin 9 in the
chemical shift region of δ 6.54-7.18. The overlapping peaks
are deconvoluted by a multipeak Lorentzian fit.
tion, the next issue we need to tackle is the geoisomeric
structures of the repeat units of the trialkylbenzenes
1
in the resultant polymer. Integration of H NMR peak
area has been commonly used for quantitative analy-
2
0
sis, but as can be seen from Figure 3B, the “peaks” of
the polymer in the aromatic spectral region (δ ∼ 6.50-
7
.20) seem to be hopelessly complex due to severe
overlapping of the resonance signals of the isomeric
protons of the 1,2,4/1,3,5-trisubstituted benzenes. With
the aid of model compounds 8 and 9, we can, however,
deconvolute the broad peaks of the polymer. The ex-
1
panded H NMR spectrum of 8 in the aromatic region
is shown in Figure 6B, from which the peaks associated
with the absorptions of the phenyl protons of 1,2,4- and
F igu r e 5. TGA thermograms of hyperbranched polymer hb-
P1(4) (sample taken from Table 3, no. 4) recorded under
nitrogen (solid line) and in air (dashed line) at a heating rate
of 20 °C/min.
1,3,5-trihexylbenzenes can be unambiguously identi-
20,25
fied.
Using eq 1, it is calculated that molar ratio of
1
,3,5-8 to 1,2,4-8 is 1.0:3.0 (cf. Scheme 4), in agreement
2
3,24
with the general trends
experimental results;
and also with our previous
i.e., 1,2,4-geoisomers are the
1
0,11
collateral evidence to support the conclusion derived
from the spectroscopic analysis data that hb-P1(4) has
a polyarylene structure. Clearly, the high thermal
stability of hb-P1(4) stems from its phenylene units.
Interestingly and usefully, hb-P1(4) readily cures when
it is heated to a moderate temperature. The polymer
remains completely soluble when warmed to 80 °C but
becomes insoluble when heated to 100 °C. Baking the
polymer at 100 °C for 5 h in air or under nitrogen causes
it to cross-link in 95 or 96% gelation yields, and heating
the polymer to 200 °C quantitatively converts it to a
thermoset network. The thermal curing mechanisms of
the polymer are unclear at present but may be associ-
ated with its benzyl units. Thermally activated hydro-
gen abstractions from the benzyl sites of hb-P1(4) may
have generated numerous radical species in the macro-
molecular branches, the coupling of which may have
facially cross-linked the polymer.
major products of transition-metal-catalyzed alkyne
poly)cyclotrimerizations:
(
N1,2,4-8
A /3
Aj
j
)
)
(1)
N1,3,5-8 (A + A + A )/3 A + A + A
g
h
i
g
h
i
where Aj, Ag, Ah, and Ai are the integrated areas of the
resonance peaks of protons j, g, h, and i of 1,3,5- and
1
1
,2,4-8, respectively. The expanded H NMR spectrum
of 9 in the aromatic region can also be elucidated with
ease, as clearly marked in Figure 6C. The spectral data
of 9 further confirm that the aromatic protons of 1,2,4-
isomers always show resonance peaks in the downfield
chemical shift region (δ > 6.82), in comparison to their
20,25
counterparts of 1,3,5-isomers.
The spectrum of hb-P1(4) in the aromatic region is
evidently more complicated than those of model com-
pounds 8 and 9. The overlapping peaks of the polymer
can, however, be deconvoluted or fitted by a Lorentzian
function into six peaks of different intensities (given in
the parentheses): δI 7.01 (0.17), δII 6.94 (0.54), δIII 6.90
(0.42), δIV 6.86 (0.42), δV 6.79 (0.38), and δVI 6.73 (0.17),
It is well-known that transition-metal-catalyzed alkyne
cyclotrimerizations give rise to the formation of 1,2,4-
and 1,3,5-trisubstituted benzenes, with the 1,2,4-
2
3,24
isomers being normally the major products.
After
confirming that 1(4) has undergone polycyclotrimeriza-

5
204 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
Ch a r t 2
F igu r e 7. Expanded 1H NMR spectra of hyperbranched
polymer hb-P1(4) (sample taken from Table 3, no. 4) in the
chemical shift region of δ 2.2-3.9.
1
,2,4- and 1,3,5-geoisomeric structures are indicated in
the parentheses. Diad 10(1,2,4; 1,2,4) has the highest
probability to form from the viewpoint of reaction
2
3,24
mechanism,
but diad 10(1,3,5; 1,3,5) has the lowest
steric energy, as revealed by the calculations using the
built-in programs in the CS ChemBats3D Ultra of
Cambridge Soft. It is recognized that the individual
peaks may not be assigned to the particular molecular
structures with confidence, unless all the diad models
are synthesized and isolated in pure forms. It is,
however, also noted that the preparation and purifica-
tion of these model compounds, especially 10(1,3,5;
1,3,5), are real synthetic challenges. We are now taking
the challenges in our laboratories, with the aim of
unambiguously assigning all the individual peaks.
In hb-P1(4), there exist two types of benzyl struc-
tures: one is a cyclic (C) 6-substituted tetralin type, and
another is an acyclic (A) trisubstituted benzene type,
as shown in Figure 7. What is the ratio of the two (RC/A)?
Knowing the ratio will help understand the polycyclo-
trimerization mechanisms because the formation of the
tetralin-type cyclic structure terminates the growth of
the propagating species. The somewhat overlapped
peaks of hb-P1(4) at δ 2.71 (e) and 2.55 (f) are readily
deconvoluted by a multipeak fitting function, the inte-
gration of which finds that the areas of peaks e (Ae) and
f (Af) are 0.77 and 2.81, respectively. Peak e at δ 2.71 is
clearly associated with absorptions of four (4) protons
of tetralin-type structure C, whereas peak f at δ 2.55 is
related to the absorptions of two (2) protons of structure
C and six (6) protons of structure A. The RC/A value can
thus be calculated by eq 3:
as shown in Figure 6A. From the above discussions on
the peak identifications of 8 and 9 (cf. panels B and C
of Figure 6), we can assign with confidence that the
peaks of hb-P1(4) located in the downfield with chemical
shifts of δ > 6.825 on the left side of the dotted vertical
line, i.e., peaks I-IV, are associated with the absorp-
tions of the phenyl protons of the 1,2,4-isomers of the
polymer, whereas those in the upfield with δ < 6.825,
i.e., peaks V and VI, are related to the absorptions of
the phenyl protons of its 1,3,5-isomeric units. This peak
differentiation enables estimation of molar ratio of the
1
,2,4- and 1,3,5-isomers, N1,2,4-Ph/N1,3,5-Ph, by peak area
integration using eq 2:
N1,2,4-Ph A - (A + A )
t
V
VI
)
(2)
N1,3,5-Ph
A + A
V
VI
where N1,2,4-Ph is the number of 1,2,4-substituted ben-
zene rings, N1,3,5-Ph the number of 1,3,5-substituted
benzene rings, At the total integrated area of resonance
peaks of all aromatic protons, AV the integrated area of
peak V at δ 6.79, and AVI the integrated area of peak
VI at δ 6.73. The calculated value of N1,2,4-Ph/N1,3,5-Ph
is 2.8, close to the value (3.0) for the model reaction of
monoyne cyclization (cf. Scheme 4). This is reasonable
because one may not expect that the cyclotrimerization
of 1(4), a diyne with two triple bonds separated by a
flexible alkyl spacer, would differ much from that of the
monoyne.23 This agreement in the molar ratios also
supports the validity of the peak deconvolution and
elucidation processes.
A /4
3Ae
e
RC/A
)
)
(3)
[
A - (A /4) × 2]/6
2A - Ae
f
e
f
Putting the integrated areas of peaks e (Ae ) 0.77) and
f (Af ) 2.81) into above equation gives an experimental
RC/A value of 0.48 (or 48%).
Further detailed assignments of the individual peaks
(
I-VI) generated by the Lorentzian multipeak fit may
Mech a n istic Discu ssion . It has become clear that
the diynes have undergone polycyclotrimerizations in
the presence of the transition-metal catalysts. Let us
discuss how the polymerization propagates. If the
propagating species grows exclusively in a linear mode,
the diyne molecules will be knitted into a linear chain
(4), but if the propagation is exclusively in a dendritic
mode, a dendritic sphere (6) will be formed (cf. Scheme
2). In either case, the resultant polymer should bear
triple bonds at the chain ends or on the sphere periph-
eries. None of the spectroscopic analyses have, however,
be done but are not attempted because of our cautions
in trying not to mislabel the peaks. Different from the
isolated small molecules of the model compounds, the
hyperbranched polymer is in a much more complex
macromolecular environment, where not only the geo-
metric structures of 1,2,4- and 1,3,5-isomers but also
the diad sequences of different geoisomeric combinations
will affect chemical shifts of the aromatic protons.
Chart 2 shows a few examples of simplified diad
sequences, in which the different combinations of the
2
6

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5205
Sch em e 6
Ch a r t 3
detected any absorption peaks associated with the
resonance of the ethynyl groups. Such triple bonds thus
must have been consumed by their dimerization reac-
tions with the diyne molecules, as exemplified in
Scheme 6 for the case of polycyclotrimerization of 1(4).
When a pure linear chain [4(4)] is terminated by the
dimerization end-capping reactions, a linear polymer
experimental value for the molar ratio of cyclic to acyclic
units (RC/A) is 48%. Let us examine what kinds of cyclic
and acyclic units exist in the polymer. Theoretically
there are five such structures: one dendritic unit (D),
two liner units (L′ and L), and two terminal units (T′
and T), as illustrated in Chart 3. The spectroscopic
analyses have, however, found no triple bonds in the
polymersthe existence of units L′ and T′ can thus be
ruled out. Among the three units left, T is a cyclic
structure and D and L are acyclic ones. The molar
ratio of cyclic T unit to acyclic D and L units thus should
be equal to RC/A, viz., 0.48. This value is between 0
(exclusive linear growth) and 1 (exclusive dendritic
growth), indicating that the alkyne polycyclotrimeriza-
tion has proceeded in both linear and dendritic propaga-
tion modes. The simultaneous linear and dendritic
propagations lead to the formation of a hyperbranched
polymer with a glycogen-like molecular structure, con-
sisting of the repeat units of acyclic linear chains (L)
[l-P1(4)] will be resulted. In this extreme case, the molar
ratio of the terminal groups to the repeat units is
virtually 0 when the degree of polymerization or the
2
9
2
7
molecular weight of the polymer becomes high enough.
On the other hand, when a pure dendritic sphere [6(4)]
is end-capped by the dimerization reactions, a dendritic
polymer [d-P1(4)] will be formed. In this extreme case,
the molar ratio of the terminal groups to the repeat
units rapidly approaches to 1 with an increase in the
number of generations.10,28
The real situation is in between these two extreme
cases. As discussed above (cf. Figure 7 and eq 3), the

5
206 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
Ta ble 4. Id ea l Gr ow th P a tter n s of Den d r itic a n d Ter m in a l Un its in Diyn e P olycyclotr im er iza tion a
(ND′, NT′)
no.
L
G1
G2
G3
G4
G5
‚‚‚
Gx
x-1
- 2, 3 × 2x-1)
- 2 - 3, 3 × 2
- 2 - 6, 3 × 2
- 2 - 9, 3 × 2
1
2
3
4
5
6
7
8
L0
L1
L2
L3
L4
L5
‚‚‚
Ly
(1, 3)
(4, 6)
(1, 3)
(10, 12)
(7, 9)
(4, 6)
(22, 24)
(19, 21)
(16, 18)
(13, 15)
(10, 12)
(7, 9)
(46, 48)
(43, 45)
(40, 42)
(37, 39)
(34, 36)
(31, 33)
‚‚‚
‚‚‚
‚‚‚
‚‚‚
‚‚‚
‚‚‚
‚‚‚
‚‚‚
(3 × 2
(3 × 2
(3 × 2
(3 × 2
(3 × 2
(3 × 2
‚‚‚
x-1
x-1
x-1
x-1
x-1
x-1
x-1
x-1
-3)
-6)
- 9)
(1, 3)
x-1
- 2 - 12, 3 × 2
- 2 - 15, 3 × 2
- 12)
- 15)
x-1
‚‚‚
x-1
- 2 - 3y, 3 × 2x-1 - 3y)
(3 × 2
a
Abbreviations: L ) number of linear growths, G ) number of generations, ND′ ) number of dendritic units in the ideal model, and
NT′ ) number of terminal units in the ideal model.
Sch em e 7
If the linear propagation mode participates, the
geostructures of the resultant polymers will vary (Chart
4
) and their ND′ and NT′ values will decrease with an
increase in the number of linear growths (y), when
compared to those of their counterparts of ideal den-
drimers with the same generation numbers (Table 4).
Whereas both NT′ and ND′ change with y, their ratio
RT/D′ remains almost constant. Indeed, when x is high
enough, RT/D′ approaches unity, viz., NT′ ) ND′. This can
be readily appreciated by comparing the general for-
mulas given in the last column of Table 4: when x is
x-1
large enough, the difference between NT′ (3 × 2
-
x-1
3
y) and ND′ (3 × 2
- 2 - 3y) is marginally small.
This can also be easily understood by looking over the
example of propagation reaction given in Scheme 8.
When a propagating species (11) with a high x value
and a unity RT/D ratio undergoes a series of liner
growths, a new propagating species (14) is formed.
During the linear growths, the ND and NT values
experience little change. The following relationship thus
holds in the diyne polycyclotrimerization, regardless
whether there are any linear units in the polymer
structure:
N ) N (x f ∞)
(5)
T
D
and dendritic branches (D), with all their ends capped
by cyclic terminal units (T; Scheme 7).
As discussed above, in the real case of the diyne
polycyclotrimerization, there are one type of cyclic
terminal unit (T) and two types of acyclic repeat units
The degree of branching (DB) is an important struc-
tural parameter for a hyperbranched polymer. It is,
however, difficult to directly measure the DB of hb-P1(4)
because the spectrometers cannot differentiate its L and
(D and T; cf. Chart 3). Equation 3 can thus be changed
to eq 6, where NL is the number of linear units:
2
9
NT
3Ae
D units. This technical difficulty forces us to take a
different approach, and ideal models are thus built in
an effort to deduce the necessary equations for calculat-
ing the DB value. Assuming that the propagating
species grow in an ideal dendritic mode with their
peripheral triple bonds all end-capped by the tetralin-
type cyclic terminal units, perfect dendrimers with
globular geostructures will be generated. A few ex-
amples of such ideal dendrimers with generation num-
bers 1-3 (or G1-G3) are given in the first row of Chart
RC/A
)
)
) 0.48
(6)
N + N
2A - Ae
D
L
f
When N is used to represent the number of all struc-
tural units in the polymer, the following equation holds
by definition:
N + N + N ) N
(7)
D
L
T
Dividing eqs 5-7 by N gives the following equations:
4
[noting that no linear growth is involved in this
specific case; thus, all the dendrimers possesses 0 linear
unit (L0)]. The sequential growth patterns of the D and
T units in the ideal dendrimers are given in Table 4
fT ) fD
(8)
fT
fD + f
(no. 1), which shows clearly that with an increase in
) 0.48
(9)
L
the number of generation (x) the numbers of D (ND′) and
T units (NT′) increase, and their ratio (RT/D′) rapidly
approaches 1 (eq 4).
f + f + f ) 1
(10)
D
L
T
x-1
where fT, fD, and fL are the molar fractions of terminal
NT/N), dendritic (ND/N), and linear units (NL/N), re-
spectively. From eqs 8-10, fT, fD, and fL are determined
N ′
T
3 × 2
(
RT/D′ ) lim
) lim
) 1
(4)
x-1
xf∞ N ′
xf∞
D
3 × 2
- 2

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5207
Ch a r t 4

5
208 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
Sch em e 8
by simple calculations to be 32%, 32%, and 36%,
respectively.
For a hyperbranched polymer prepared from a poly-
condensation reaction of an ABm-type monomer, its DB
is defined by eq 11.30 This equation can be rewritten as
eq 12 by dividing its denominator and numerator by N:
(3) Structural determination. The spectra of the
polymers seemed simple at first glance but troubled us
when we scrutinized the details. The most confusing
result was the complete absence of triple-bond reso-
nance in the spectra because a hyperbranched polymer
prepared from alkyne polycyclotrimerization was ex-
pected to carry many triple bonds on its peripheral
surface. With the aid of the model reactions, the spectra
are fully elucidated, from which the structures and
ratios of the L, D, and T units are determined and the
hyperbranched nature of the polymer structure is
confirmed.
N + N
D
T
DB )
(11)
(12)
N + N + N
D
L
T
fD + f
T
DB )
f + f + f
T
D
L
(4) Mechanistic understanding. The structural analy-
ses and the model simulations enable mechanistic
deciphering. The propagations in the diyne polycyclo-
trimerization share some similarity to glycogenesis: the
active centers simultaneously grow in L and D modes,
giving the L and D units, respectively. [It is the L
growth that has considerably decreased the number of
T units (by ∼50 mol % from that obtainable from the
pure D growth).] The triple bonds at the ends of the L
chains and on the peripheries of the D spheres are all
consumed by their dimerizations with the alkyne mono-
mers, yielding the tetralin-type T units. The end-
capping reaction is so beautifully done that no triple-
bond moieties in the polymer structures can be detected
by all the spectrometers we employed for the structural
analyses.
Putting the fD, fL, and fT values into eq 12 yields a DB
value of 64%. This DB value, in addition to the firm
spectroscopic analysis data, further confirms the hyper-
branched nature of the molecular structure of our
polymer.3
0,31
Con clu d in g Rem a r k s
In summary, in this work, we have synthesized a
group of new hyperbranched poly(alkenephenylene)s by
transition-metal-catalyzed alkyne polycyclotrimeriza-
tions. What we have achieved in this work can be
summarized as follows:
(1) Catalysts exploration. Effective, single-component
catalysts (TaBr5, NbBr5, and NbBr3‚DME) for alkyne
polycyclotrimerization are developed. These catalysts
enjoy such advantages as high activity and operational
simplicity as well as commercial availability and low
costs. The excellent performance of NbBr3‚DME in the
diyne polycyclotrimerization suggests the possibility of
exploring new catalysts with even higher activity from
the organometallic complexes by tuning molecular
structures of their ligands, which is currently pursued
in our laboratories.
(5) Property discovery. The new polymers are expected
to exhibit many interesting properties, and in this study,
their unique, useful thermal properties are uncovered.
Polymer hb-P1(4) can be readily cured at moderate
temperatures (e.g., ∼100 °C) and lose almost no weight
at high temperatures (up to ∼500 °C). Similarly, no
weight losses are observed when its structural con-
geners, i.e., hb-P1(5) and hb-P1(6), are heated to com-
parably high temperatures. The polymers containing
silicon and nitrogen heteroatoms exhibit even more im-
pressive thermal behaviors: they graphitize or ceramize
in high yields (up to 80%) when pyrolyzed at 900 °C in
(2) Polymerization control. Alkyne polycyclotrimer-
ization is in some sense a difficult reaction, which often
runs out of control to give cross-linked polymers of
1
0-12
intractability.
The effects of the reaction parameters
1
7
an inert atmosphere. These outstanding thermal
properties may enable the polymers to find high-tech
applications as, e.g., solvent- and moisture-resistant
coatings, thermally and dimensionally stable films, and
processable precursors to nanostructured materials such
including [M]0, [cat.], solvent, time, and temperature are
systematically investigated in this work, and under
optimal reaction conditions, high molecular weight
3
polymers (with Mw up to ∼270 × 10 ) are reproducibly
prepared in high yields (normally >80%), which are
3
2,33
34,35
completely soluble and readily processable.
as carbon nanotubes
and SiC ceramics.

Macromolecules, Vol. 37, No. 14, 2004
Hyperbranched Poly(alkenephenylene)s 5209
Ack n ow led gm en t. The work described in this paper
was partially supported by the Research Grants Council
Projects N_HKUSR606_03, 604903, HKUST6085/02P,
and 6121/01P) and the University Grants Committee
of Hong Kong through an Area of Excellence Scheme
(10) (a) Chen, J .; Peng, H.; Law, C. C. W.; Dong, Y.; Lam, J . W.
Y.; Williams, I. D.; Tang, B. Z. Macromolecules 2003, 36,
4
319-4327. (b) Xu, K.; Peng, H.; Sun, Q.; Dong, Y.; Salhi,
(
F.; Luo, J .; Chen, J .; Huang, Y.; Zhang, D.; Xu, Z.; Tang, B.
Z. Macromolecules 2002, 35, 5821-5834. (c) Peng, H.; Cheng,
L.; Luo, J .; Xu, K.; Sun, Q.; Dong, Y.; Salhi, F.; Lee, P. P. S.;
Chen, J .; Tang, B. Z. Macromolecules 2002, 35, 5349-5351.
(
Project AoE/P-10/01-1A).
(d) Peng, H.; Luo, J .; Cheng, L.; Lam, J . W. Y.; Xu, K.; Dong,
Y.; Zhang, D.; Huang, Y.; Xu, Z.; Tang, B. Z. Opt. Mater. 2002,
Su p p or tin g In for m a tion Ava ila ble: IR spectra of 1(4)
2
1, 315-320. (e) Xu, K.; Tang, B. Z. Chin. J . Polym. Sci. 1999,
and hb-P1(4) and UV absorption spectra of DCM solutions of
hb-P1(4), 8, and 1(4). This material is available free of charge
via the Internet at http://pubs.acs.org.
17, 397-402.
(11) (a) Mi, Y.; Tang, B. Z. Polym. News 2001, 26, 170-176. (b)
Lam, J . W. Y.; Luo, J .; Peng, H.; Xie, Z.; Xu, K.; Dong, Y.;
Cheng, L.; Qiu, C.; Kwok, H. S.; Tang, B. Z. Chin. J . Polym.
Sci. 2001, 19, 585-590. (c) Lam, J . W. Y.; Chen, J .; Law, C.
C. W.; Peng, H.; Xie, Z.; Cheuk, K. K. L.; Kwok, H. S.; Tang,
B. Z. Macromol. Symp. 2003, 196, 289-300. (d) Xie, Z.; Peng,
H.; Lam, J . W. Y.; Chen, J .; Zheng, Y.; Qiu, C.; Kwok, H. S.;
Tang, B. Z. Macromol. Symp. 2003, 195, 179-184. (e)
H a¨ ussler, M.; Lam, J . W. Y.; Zheng, R.; Peng, H.; Luo, J .;
Chen, J .; Law, C. C. W.; Tang, B. Z. C. R. Chim. 2003, 6,
Refer en ces a n d Notes
(
1) Biopolymers from Polysaccharides and Agroproteins; Gross,
R. A., Scholz, C., Eds.; American Chemical Society: Wash-
ington, DC, 2001.
2) Small saccharides other than glucose can also form hyper-
branched biopolymers. Galactose, for example, can be poly-
merized into galactogen, a polysaccharide with a branched
molecular structure similar to that of glycogen.
3) Teusink, B.; Passarge, J .; Reijenga, C. A.; Esgalhado, E.; van
der Weijden, C. C.; Schepper, M.; Walsh, M. C.; Bakker, B.
M.; van Dam, K.; Westerhoff, H. V.; Snoep, J . L. Eur. J .
Biochem. 2000, 267, 5313-5329.
(
8
33-842.
12) Xu, K. Ph.D. Dissertation, Hong Kong University of Science
Technology, Dec 2000.
(
(
&
(
13) (a) Muchtar, Z.; Schappacher, M.; Deffieux, A. Macromol-
ecules 2001, 34, 7595. (b) Grayson, S. M.; Frechet, J . M. J .
Macromolecules 2001, 34, 6542. (c) Uhrich, K. E.; Hawker,
C. J .; Frechet, J . M. J .; Turner, S. R. Macromolecules 1992,
25, 4583.
(
4) (a) Saier, M. H. Mechanisms and Regulation of Carbohydrate
Transport in Bacteria; Academic Press: Orlando, FL, 1985.
(
14) Zheng, R.; Peng, H.; Lam, J . W. Y.; Chen, J .; H a¨ ussler, M.;
(
b) Shimeno, S. Studies on Carbohydrate Metabolism in Fish;
Tang, B. Z. Polym. Prepr. 2003, 44 (1), 956-957.
A.A. Balkema: Rotterdam, 1982. (c) Carbohydrate Reserves
in Plants: Synthesis and Regulation; Gupta, A. K., Kaur, N.,
Eds.; Elsevier: New York, 2000.
(
15) (a) Xu, K.; Peng, H.; Lam, J . W. Y.; Poon, T. W. H.; Dong, Y.;
Xu, H.; Sun, Q.; Cheuk, K. K. L.; Salhi, F.; Lee, P. P. S.; Tang,
B. Z. Macromolecules 2000, 33, 6918-6924. (b) Tang, B. Z.;
Xu, K.; Sun, Q.; Lee, P. P. S.; Peng, H.; Salhi, F.; Dong, Y.
ACS Symp. Ser. 2000, 760, 146-164. (c) Tang, B. Z.; Poon,
W. H.; Leung, S. M.; Leung, W. H.; Peng, H. Macromolecules
1997, 30, 2209-2212. (d) Tang, B. Z.; Kotera, N. Macromol-
ecules 1989, 22, 4388-4390.
(
(
(
5) Amylopectin, a hyperbranched biopolymer of glucose in
plants, is less branched than glycogen, probably because the
energy storage and transport in plants are less demanding
in speed than those in animals, taking into consideration that
the living machineries in plants experience less vigorous
1
,4
changes than those in animals.
(16) Lam, J . W. Y.; Tang, B. Z. J . Polym. Sci., Part A: Polym.
Chem. 2003, 41, 2607-2629.
(17) (a) Preliminary results have been reported in ref 9b. (b)
Zheng, R.; Dong, H.; Lam, J . W. Y.; Tang, B. Z. Manuscript
in preparation.
6) (a) Dendrimers V: Functional and Hyperbranched Building
Blocks, Photophysical Properties, Applications in Materials
and Life Sciences; Schalley, C. A., Vogtle, F., Eds.; Springer-
Verlag: Berlin, 2003. (b) Dendrimers and Other Dendritic
Polymers; Frechet, J . M. J ., Tomalia, D. A., Eds.; Wiley: New
York, 2002.
(18) (a) Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH:
Weinheim, 2003. (b) Choi, S. K.; Gal, Y. S.; J in, S. H.; Kim,
H. K. Chem. Rev. 2000, 100, 1645-1681. (c) Metathesis
Polymerization of Olefins and Polymerization of Alkynes;
Imamoglu, Y., Ed.; Kluwer: Boston, 1998. (d) Ivin, K. J .; Mol,
J . C. Olefin Metathesis and Metathesis Polymerization;
Academic Press: San Diego, 1997. (e) Ginsburg, E. J .;
Gorman, C. B.; Grubbs, R. H. In Modern Acetylene Chemistry;
Stang, P. J ., Diederich, F., Eds.; VCH: Weinheim, 1995;
Chapter 10. (f) Shirakawa, H.; Masuda, T.; Takeda, K. In The
Chemistry of Triple-Bonded Functional Groups, Supplement
C2; Patai, S., Ed.; Wiley: New York, 1994; Chapter 17.
(19) (a) Lam, J . W. Y.; Dong, Y.; Cheuk, K. K. L.; Luo, J .; Xie, Z.;
Kwok, H. S.; Mo, Z.; Tang, B. Z. Macromolecules 2002, 35,
1229-1240. (b) Tang, B. Z.; Xu, H.; Lam, J . W. Y.; Lee, P. P.
S.; Xu, K.; Sun, Q.; Cheuk, K. K. L. Chem. Mater. 2000, 12,
1446-1455. (c) Huang, Y. M.; Lam, J . W. Y.; Cheuk, K. K.
L.; Ge, W.; Tang, B. Z. Macromolecules 1999, 32, 5976-5978.
(d) Tang, B. Z.; Kong, X.; Wan, X.; Peng, H.; Lam, W. Y.; Feng,
X.; Kwok, H. S. Macromolecules 1998, 31, 2419-2432.
(20) (a) Silverstein, R. M.; Webster, F. X. Spectrometric Identifica-
tion of Organic Compounds, 6th ed.; Wiley: New York, 1998.
(b) Pretsch, E.; Buhlmann, P.; Affolter, C. Structure Deter-
mination of Organic Compounds: Tables of Spectral Data,
3rd ed.; Springer: Berlin, 2000. (c) Field, L. D.; Sternhell,
S.; Kalman, J . R. Organic Structures from Spectra, 3rd ed.;
Wiley: New York, 2002.
7) (a) Zhao, D. H.; Moore, J . S. Chem. Commun. 2003, 807-
8
18. (b) Tomalia, D. A.; Frechet, J . M. J . J . Polym. Sci., Part
A: Polym. Chem. 2002, 40, 2719-2728. (c) Crooks, R. M.
Chem. Phys. Chem. 2001, 2, 644-654. (d) Watson, M. D.;
Fechtenkotter, A.; Mullen, K. Chem. Rev. 2001, 101, 1267-
1
1
2
300. (e) J ikei, M.; Kakimoto, M. Prog. Polym. Sci. 2001, 26,
233-1285. (f) Voit, B. J . Polym. Sci., Part A: Polym. Chem.
000, 38, 2505-2525. (g) Frey, H.; Schlenk, C. Top. Curr.
Chem. 2000, 210, 69-129. (h) Siemsen, P.; Livingston, R. C.;
Diederich, F. Angew. Chem., Int. Ed. 2000, 39, 2633-2657.
(
i) Bunz, U. H. F. Chem. Rev. 2000, 100, 1605-1644. (j) Hult,
A.; J ohansson, M.; Malmstrom, E. Adv. Polym. Sci. 1999, 143,
-34. (k) Scherf, U. Top. Curr. Chem. 1999, 201, 163-222.
l) Hawker, C. J . Curr. Opin. Colloid Interface Sci. 1999, 4,
1
(
1
17-121. (m) Newkome, G. R.; He, E. F.; Moorefield, C. N.
Chem. Rev. 1999, 99, 1689-1746. (n) Bosman, A. W.; J ans-
sen, H. M.; Meijer, E. W. Chem. Rev. 1999, 99, 1655-1688.
(o) Kim, Y. H. J . Polym. Sci., Part A: Polym. Chem. 1998,
3
6, 1685-1698.
(
8) (a) Miller, T. M.; Neenan, T. X.; Zayas, R.; Bair, H. E. J . Am.
Chem. Soc. 1992, 114, 1018-1025. (b) Tour, J . M. Adv. Mater.
1
994, 6, 190-198. (c) J ohnen, N. A.; Kim, H. K.; Ober, C. K.
ACS Symp. Ser. 1994, 579, 298-315. (d) Hecht, S.; Frechet,
J . M. J . J . Am. Chem. Soc. 1999, 121, 4084-4085. (e)
Mitsutoshi, J .; Ken, F.; Yang, G.; Kakimoto, M. Macromol-
ecules 2000, 33, 6228-6234. (f) Cameron, C.; Fawcett, A. H.;
Hetherington, C. R.; Mee, R. A. W.; McBride, F. V. Macro-
molecules 2000, 33, 6551-6568. (g) Weil, T.; Wiesler, U.-M.;
Herrmann, A.; Bauer, R.; Hofkens, J .; De Schryver, F. C.;
Mullen, K. J . Am. Chem. Soc. 2001, 123, 8101-8108. (h)
Komber, H.; Ziemer, A.; Voit, B. Macromolecules 2002, 35,
(21) Masuda, T.; Tang, B. Z.; Higashimura, T.; Yamaoka, H.
Macromolecules 1985, 18, 2369-2373.
(22) Goosy, M. In Specialty Polymers; Dyson, R. W., Ed.; Blackie:
London, 1987; Chapter 5, pp 83-109.
(23) (a) Saito, S.; Yamamoto, Y. Chem. Rev. 2000, 100, 2901-
2915. (b) Lautens, M.; Klute, W.; Tam, W. Chem. Rev. 1996,
96, 49-92. (c) Schore, N. E. Chem. Rev. 1988, 88, 1081-1119.
(d) Vollhardt, K. P. C. Angew. Chem., Int. Ed. Engl. 1984,
23, 539-644. (e) Maitlis, P. M. Acc. Chem. Res. 1976, 9, 93-
99.
3
514-3519. (i) Markoski, L. J .; Thompson, J . L.; Moore, J .
S. Macromolecules 2002, 35, 1599-1603.
(
9) (a) Zheng, R.; Lam, J . W. Y.; Peng, H.; H a¨ ussler, M.; Tang,
B. Z. Polym. Mater. Sci. Eng. 2003, 89, 442-443. (b) Zheng,
R.; Lam, J . W. Y.; Peng, H.; H a¨ ussler, M.; Law, C. C. W.;
Tang, B. Z. Polym. Prepr. 2003, 44 (2), 770-771.
(24) (a) Cotton, F. A.; Hall, W. T. J . Am. Chem. Soc. 1979, 101,
5094-5095. (b) Bruck, M. A.; Copenhaver, A. S.; Wigley, D.

5
210 Zheng et al.
Macromolecules, Vol. 37, No. 14, 2004
E. J . Am. Chem. Soc. 1987, 109, 6525-6527. (c) Strickler, J .
R.; Wexler, P. A.; Wigley, D. E. Organometallics 1988, 7,
helpful because it would make the estimation of the L/D
molar ratio a straightforward, easy job.)
2
067-2069. (d) Bianchini, C.; Caulton, K. G.; Chardon, C.;
(
30) (a) Hawker, C. J .; Lee, R.; Fr e´ chet, J . M. J . J . Am. Chem.
Soc. 1991, 113, 4583-4588. (b) Frey, H.; H o¨ lter, D. Acta
Polym. 1999, 50, 67-76.
Eisenstein, O.; Folting, K.; J ohnson, T. J .; Meli, A.; Peruzzini,
M.; Rauscher, D. J .; Streib, W. E.; Vizza, F. J . Am. Chem.
Soc. 1991, 113, 5127-5129. (e) Strickler, J . R.; Wexler, P.
A.; Wigley, D. E. Organometallics 1991, 10, 118-127. (f)
Kwon, D.; Real, J .; Curtis, M. D.; Rheingold, A.; Haggerty,
B. S. Organometallics 1991, 10, 143-148. (g) Pope, R. M.;
Vanorden, S. L.; Cooper, B. T.; Buckner, S. W. Organo-
metallics 1992, 11, 2001-2003. (h) Pelissier, H.; Rodriguez,
J .; Vollhardt, K. P. C. Chem.sEur. J . 1999, 5, 3549-3561.
(31) (a) Wang, M.; Gan, D.; Wooley, K. L. Macromolecules 2001,
3
2
4, 3215-3223. (b) Parker, D.; Feast, W. J . Macromolecules
001, 34, 2048-2059. (c) Thompson, D. S.; Markoski, L. J .;
Moore, J . S.; Sendijarevic, I.; Lee, A.; McHugh, A. J . Macro-
molecules 2000, 33, 6412-6415. (d) Radke, W.; Litvinenko,
G.; Muller, A. H. E. Macromolecules 1998, 31, 239-248. (e)
Yan, D. Y.; Muller, A. H. E.; Matyjaszewski, K. Macromol-
ecules 1997, 30, 7024-7033. (f) Kim, Y. H.; Webster, O. W.
Macromolecules 1992, 25, 5561-5572.
1
3
(
25) (a) Pouchert, C. J .; Behnke, J . The Aldrich Library of C and
1
H FT NMR Spectra; Aldrich Chemical Co., Ltd.: Milwaukee,
WI, 1993. (b) Feinstein, K. Guide to Spectroscopic Identifica-
tion of Organic Compounds; CRC Press: Boca Raton, FL,
1
(
32) (a) Gherghel, L.; Kubel, C.; Lieser, G.; Rader, H. J .; Mullen,
K. J . Am. Chem. Soc. 2002, 124, 13130-13138. (b) Dai, L.
M.; Patil, A.; Gong, X. Y.; Guo, Z. X.; Liu, L. Q.; Liu, Y.; Zhu,
D. B. Chem. Phys. Chem. 2003, 4, 1150-1169. (c) Park, C.;
Yoon, J .; Thomas, E. L. Polymer 2003, 44, 6725-6760.
995.
(
26) (a) Matsuzaki, K.; Uryu, T.; Asakura, T. NMR Spectroscopy
and Stereoregularity of Polymers; J apan Scientific Societies
Press: Tokyo, 1996. (b) Bovey, F. A.; Mirau, P. A. NMR of
Polymers; Academic Press: San Diego, CA, 1996. (c) NMR
Spectroscopy of Polymers in Solution and in the Solid State;
Cheng, H. N., English, A. D., Eds.; American Chemical
Society: Washington, DC, 2003.
27) (a) Carraher, C. E., J r. Seymour/ Carraher’s Polymer Chem-
istry, 6th ed.; M. Dekker: New York, 2003. (b) Stevens, M.
P. Polymer Chemistry: an Introduction, 3rd ed.; Oxford
University Press: New York, 1999.
28) (a) Markoski, L. J .; Thompson, J . L.; Moore, J . S. Macromol-
ecules 2002, 35, 1599-1603. (b) H o¨ lter, D.; Burgath, A.; Frey,
H. Acta Polym. 1997, 48, 30-35.
29) Although L may also be viewed as a cyclic structure, the NMR
spectrometry does not treat it as a cyclic unit but an acyclic
one because the big ring (12-membered ring when m ) 4)
with little ring strain gives resonance signals identical to
(33) Tang, B. Z.; Xu, H. Macromolecules 1999, 32, 2569-2576.
(34) (a) Simon, P. F. W.; Ulrich, R.; Spiess, H. W.; Wiesner, U.
Chem. Mater. 2001, 13, 3464-3486. (b) Corriu, R. J . P.
Angew. Chem., Int. Ed. 2000, 39, 1376-1398. (c) Interrante,
L. V.; Liu, Q.; Rushkin, I.; Shen, Q. Appl. Organomet. Chem.
1996, 10, 1-10. (d) Birot, M.; Pillot, J . P.; Dunogues, J . Chem.
Rev. 1995, 95, 1443-1477. (e) Seyferth, D. Adv. Chem. Ser.
1995, 245, 131-160.
(
(
(
35) (a) Sun, Q. H.; Xu, K. T.; Peng, H.; Zheng, R. H.; Haussler,
M.; Tang, B. Z. Macromolecules 2003, 36, 2309-2320. (b) Sun,
Q.; Xu, K.; Lam, J . W. Y.; Cha, J . A. P.; Zhang, X.; Tang, B.
Z. Mater. Sci. Eng., C 2001, 16, 107-112. (c) Sun, Q.; Lam,
J . W. Y.; Xu, K.; Xu, H.; Cha, J . A. P.; Wong, P. C. L.; Wen,
G.; Zhang, X.; J ing, X.; Wang, F.; Tang, B. Z. Chem. Mater.
(
2
000, 12, 2617-2624.
1
0,12,20,25
those of the open branch (D).
(Indeed, if L exhibited
resonance peaks different from those of D, that would be very
MA049871+