Advances in Ni(0)-catalyzed coupling for the synthesis of polythiophenes and polyphenylenes

Article abstract of DOI:10.1021/ma0355453

The effects of solvent and monomer structure on the Ni(0)-catalyzed polymerization of 2,5-dichloro-3-(2′-thiophenecarbonyl)thiophene, 3-benzenesulfonyl-2,5-dichlorothiophene, 2,5-dichlorobenzophenone, and 2-benzenesulfonyl-1,4-dichlorobenzene were explored. Poly[3-(2′- thiophenecarbonyl)-2,5-thiophene] with an intrinsic viscosity of 0.38 dL/g and a 10% weight loss temperature in nitrogen of 480 °C (T_g = 149 °C) was synthesized. Oligomeric material was prepared from 3-benzenesulfonyl-2,5-dichlorothiophene. Poly(2,5-benzophenone) of 58 × 10³ g/mol ([η] = 1.15 dL/g) with a T_g of 180 °C and a 10% weight loss temperature in nitrogen of 576 °C was synthesized. Oligomeric materials were also obtained from the polymerization of 2-benzenesulfonyl-1,4-dichlorobenzene. It was shown that the solvent choice has drastic effects upon the catalyst system and the final polymer. The differences in the catalyst resulting from solvent choice were explored by nuclear magnetic resonance. Effects arising from the structure of the monomer are discussed in relationship to the results obtained and the proposed mechanism.

Full text of DOI:10.1021/ma0355453

4748
Macromolecules 2004, 37, 4748-4754
Articles
Advances in Ni(0)-Catalyzed Coupling for the Synthesis of
Polythiophenes and Polyphenylenes
Er ik C. Ha gber g, Da vid A. Olson , a n d Va ler ie V. Sh ea r es*,†
Department of Chemistry, Iowa State University, Ames, Iowa 50011
Received October 14, 2003; Revised Manuscript Received April 13, 2004
ABSTRACT: The effects of solvent and monomer structure on the Ni(0)-catalyzed polymerization of 2,5-
dichloro-3-(2′-thiophenecarbonyl)thiophene, 3-benzenesulfonyl-2,5-dichlorothiophene, 2,5-dichloroben-
zophenone, and 2-benzenesulfonyl-1,4-dichlorobenzene were explored. Poly[3-(2′-thiophenecarbonyl)-2,5-
thiophene] with an intrinsic viscosity of 0.38 dL/g and a 10% weight loss temperature in nitrogen of 480
°C (Tg ) 149 °C) was synthesized. Oligomeric material was prepared from 3-benzenesulfonyl-2,5-
dichlorothiophene. Poly(2,5-benzophenone) of 58 × 103 g/mol ([η] ) 1.15 dL/g) with a Tg of 180 °C and a
10% weight loss temperature in nitrogen of 576 °C was synthesized. Oligomeric materials were also
obtained from the polymerization of 2-benzenesulfonyl-1,4-dichlorobenzene. It was shown that the solvent
choice has drastic effects upon the catalyst system and the final polymer. The differences in the catalyst
resulting from solvent choice were explored by nuclear magnetic resonance. Effects arising from the
structure of the monomer are discussed in relationship to the results obtained and the proposed
mechanism.
In tr od u ction
Despite the numerous materials that have been
synthesized and the current level of understanding of
these Ni(0)-catalyzed polymerizations, many issues
remain to be addressed, specifically with polythiophenes
and polyphenylenes containing electron-withdrawing
side groups. Previously, our group explored the polym-
erization of 3-benzoyl-2,5-dichlorothiophene (M1) as a
route to high-performance materials for applications
such as field-effect transistors (FET), optical sensors,
and light-emitting diodes.11 Polymers with intrinsic
viscosities of 0.37 dL/g were synthesized, but several
problems were encountered. Purification of the mono-
mer, a viscous, yellow oil, was difficult. The removal of
all protic impurities, which is essential for Ni(0) cou-
pling, presented particular difficulty. During the po-
lymerization partial degradation of this material, lead-
ing to a mixture of polymer and low molecular weight
side product, was observed. Degradation of poly-
thiophene derivatives at long reaction times in Ni(0)
polymerization was also seen by Ueda.12 In addition, the
final product only exhibited partial solubility in most
common solvents, including tetrahydrofuran and N,N-
dimethylacetamide, which led to difficult processing and
characterization.
Since its development by Colon and Kelsey, nickel-
(0)-catalyzed coupling has proven to be a powerful
synthetic method for the formation of carbon-carbon
aryl bonds. The reaction conditions tolerate many
functionalities with the only known exceptions being
protic, nitro, and amine-containing substituents. A wide
range of polymeric materials have been synthesized
from inexpensive arylene chlorides1 and mesylates2,3 as
well as the more costly bromide, iodide, and triflate
derivatives.4-6 The mechanism of the reaction has been
of great interest, and many advances have been made
in understanding the role of ligands, temperature, and
reducing metal on the polymerization.7
Our interests are in the synthesis of polythiophenes
and polyphenylenes substituted with electron-withdraw-
ing substituents. The presence of electron-withdrawing
groups ortho or para to the reactive site accelerates the
reaction rate by activating that site to oxidative addition
by the Ni(0) complex.8 Substituents such as the benzoyl
and benzenesulfonyl pendant group have been shown
to exhibit many additional beneficial properties. The
benzoyl pendant is known to increase the solubility of
the wholly aromatic polyphenylene material, while
maintaining the outstanding thermal and mechanical
properties of the conjugated backbone.9 The benzene-
sulfonyl pendant imparts very high thermal stability
with 10% weight loss temperatures for oligomeric
materials near 500 °C.10
In our work with polyphenylenes we noted the excel-
lent reactivity of benzophenone derivatives. Materials
with number-average molecular weights of up to 25 ×
103 g/mol with outstanding mechanical and thermal
properties were synthesized, but these polymers formed
brittle films.13 The formation of flexible films is a
requirement for many of the potential applications of
this material, such as separation membranes and
organic electronics. Also, reduction of nearly 15% of the
carbonyl group was observed.14 When the carbonyl
* Corresponding author. E-mail: ashby@email.unc.edu.
† Current address: Department of Chemistry, University of
North Carolina at Chapel Hill, 15-26 Venable Hall, CB#3290,
Chapel Hill, NC 27599-3290.
10.1021/ma0355453 CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/26/2004

Macromolecules, Vol. 37, No. 13, 2004
Polythiophenes and Polyphenylenes 4749
J ) 7.8 Hz), 8.50 (d, 1H, J ) 2.4 Hz). 13C NMR: δ (ppm) )
128.95 (CH), 129.23 (CH), 131.05 (CH), 131.42 (q-C), 132.45
(q-C), 132.51 (q-C), 133.38 (CH), 133.07 (CH), 134.76 (CH),
139.63 (q-C). Anal. Calcd for C12H8O2SCl2: C, 50.19; H, 2.81;
S, 11.17. Found: C, 49.06; H, 3.38; S, 11.09.
P olym er Syn th esis. In DMAc: Nickel(II) chloride (0.1
equiv), zinc, (3.1 equiv), triphenylphosphine (0.4 equiv), and
bipy (0.1 equiv) under an inert atmosphere were added to a
flask equipped with an overhead stirrer and a nitrogen inlet.
Solvent (10 equiv) was added via a syringe, and the mixture
was stirred at 80 °C until a deep red color was observed. The
monomer (1 equiv) was added and allowed to react for a
specified amount of time or until the reaction mixture became
too viscous to stir. The polymer was then precipitated in
hydrochloric acid and methanol (1:4), filtered, washed on a
glass frit, and reprecipitated from DMAc.
group was replaced with a sulfone functionality, only
oligomeric materials were synthesized.10
In this paper, the synthesis of a new polymer from
2,5-dichloro-3-(2′-thiophenecarbonyl)thiophene (M2) and
a study of the behavior of 3-benzenesulfonyl-2,5-dichlo-
rothiophene (M3), 2,5-dichlorobenzophenone (M4), and
2-benzenesulfonyl-1,4-dichlorobenzene (M5) are pre-
sented. Reaction conditions were varied and analyzed
to address the limitations presented above. The factors
considered were the monomer structure, type of solvent,
and type of ligand present. The catalytic systems were
explored via the use of proton (1H) and phosphorus (31P)
nuclear magnetic resonance (NMR). The polymers and
oligomers were analyzed by gel permeation chromatog-
raphy (GPC), viscometry, thermal gravimetric analysis
(TGA), differential scanning calorimetry (DSC), UV-
vis spectroscopy, and fluorescence spectroscopy.
In THF or anisole: Dichlorobis(triphenylphosphine)nickel-
(II) (0.1 equiv), zinc (3.1 equiv), triphenylphosphine (0.2 equiv),
and bipy (0.1 equiv) were added to a flask equipped with an
overhead stirrer in a glovebox. The same procedure was
followed as above, except in THF the reaction was carried out
at 60 °C.
Exp er im en ta l Section
Ma ter ia ls. All reagents were purchased from Aldrich,
unless otherwise noted, and used without further purification.
N,N-Dimethylacetamide (DMAc) and anisole were dried over
CaH2 and vacuum-distilled before use. Tetrahydrofuran (THF)
was distilled from sodium and benzophenone. 2,2′-Bipyridine
(bipy) and triphenylphosphine were purified by recrystalliza-
tion from ethanol and cyclohexane, respectively.
Mon om er Syn th esis. The monomers were synthesized by
adding aluminum chloride (1 equiv) to a stirring mixture of
the aromatic substrate (1 equiv) and acid chloride (1.1 equiv)
in 5 equiv of nitromethane at 0 °C. The reaction was allowed
to warm to room temperature and stir for 24 h. The reaction
was stopped by pouring the solution into acidic ice water. The
resulting oil was separated by extraction with ethyl acetate.
For solid products, the crude monomer was isolated by
filtration. The oil was concentrated by removal of the ethyl
acetate under vacuum. The oil (or crude solid) was treated with
activated carbon in boiling ethyl acetate/heptane (1:8) and
filtered. Upon cooling to room temperature, the resulting
crystals were isolated by filtration on a glass frit and dried
under vacuum.
2,5-Dich lor o-3-(2′-th ioph en ecar bon yl)th ioph en e). From
2,5-dichlorothiophene and thiophene-2-carbonyl chloride (76%
yield). Yellow-orange crystalline solid, melting point 64 °C. 1H
NMR: δ (ppm) ) 7.06 (s, 1H), 7.17 (dd, 1H, J ) 3.9, 2.7 Hz),
7.64 (d, 1H, J ) 2.7 Hz), 7.76 (d, 1H, J ) 3.9 Hz). 13C NMR:
δ (ppm) ) 126.76 (CH), 127.27 (q-C), 128.28 (CH), 130.05
(q-C), 135.18 (CH), 135.45 (CH), 136.23 (q-C), 143.30 (q-C),
180.23 (CO). Anal. Calcd for C9H4OS2Cl2: C, 41.08; H, 1.54;
S, 24.37. Found: C, 41.09; H, 2.03; S, 24.49.
P oly[3-(2′-th iop h en eca r bon yl)-2,5-th iop h en e]. 1H NMR:
δ ) (ppm) 7.15 (broad), 7.53 (broad), 7.64 (broad), 7.79 (broad).
13C NMR: δ (ppm) ) 131.94 (CH), 133.57 (CH), 135.31 (CH),
136.80 (CH), 139.06 (q-C), 140.30 (q-C), 141.93 (q-C), 148.39
(q-C), 187.37 (CO). Anal. Calcd for C9H4OS2: C, 56.22; H, 2.10;
S, 33.35. Found: C, 55.57; H, 2.39; S, 31.05.
P oly(3-ben zen esu lfon yl-2,5-th ioph en e). 1H NMR: δ (ppm)
) 7.19 (broad), 7.37 (broad), 7.46 (broad), 7.57 (broad). 13C
NMR: δ (ppm) ) 126.79 (CH), 127.10 (q-C), 129.31 (CH),
131.75 (q-C), 133.82 (q-C), 133.98 (CH), 138.54 (q-C), 139.54
(CH). Anal. Calcd for C10H6O2S2: C, 54.03; H, 2.72; S, 28.85.
Found: C, 53.69; H, 3.35; S, 26.88.
P oly(2,5-ben zop h en on e). 1H NMR: δ (ppm) ) broad
signal 6.5-8.2 (peaks at 6.5, 6.7, 7.2, 7.5, 7.75). 13C NMR: δ
(ppm) ) broad signal 125-135 (peaks at 128.14, 129.22,
129.81, 131.03, 131.85), broad signal at 137.23, broad signal
at 138.94, signal at 197.6 (CO). Anal. Calcd for C13H8O: C,
86.14; H, 4.48. Found: C, 87.23; H, 4.76.
P oly(3-ben zen esu lfon yl-1,4-ben zen e). 1H NMR: δ (ppm)
) broad signal 6.9-8.9 (peaks at 6.95, 7.4, 7.75, 8.0, 8.4, 8.8).
13C NMR: δ (ppm) ) 126.94, 127.34, 128.84, 129.56, 131.75,
133.615, 134.55, 138.15, 139.85. Anal. Calcd for C12H8O2S: C,
66.65; H, 3.73; S, 14.83. Found: C, 62.46; H, 4.28; S, 14.10.
Ca ta lyst Stu d y. NiCl2 (9.4 × 10-2 mmol, 12.2 mg), Zn (2.91
mmol, 190.2 mg), PPh3 (0.38 mmol, 99.9 mg), and bipy (9.4 ×
10-2 mmol, 14.7 mg) were added to an NMR tube along with
1 mL of deuterated N,N-dimethylformamide. Following three
freeze-pump-thaw cycles, the tube was sealed and heated
at 80 °C for 24 h. NMR spectra were obtained at 0, 4, 8, 16,
and 24 h. With deuterated tetrahydrofuran as solvent (1 mL)
the same procedure was followed except NiCl2(PPh3)2 (9.4 ×
10-2 mmol, 61.5 mg) and PPh3 (0.19 mmol, 49.8 mg) were used,
and the reaction was heated at 60 °C.
3-Ben zen esu lfon yl-2,5-d ich lor oth iop h en e. From 2,5-
dichlorothiophene and benzenesulfonyl chloride (60% yield).
White crystalline solid, melting point 134 °C. 1H NMR:
δ
Ch a r a cter iza tion . 1H and 13C NMR spectra were acquired
in deuterated DMSO, chloroform, or methylene chloride on a
Varian VXR-300 or Bruker AC200. 31P NMR were obtained
on a Bruker AC200 in deuterated DMF or THF using triphe-
nylphosphine as an internal reference (-5.2 ppm relative to
85% phosphoric acid external reference). Molecular weights,
relative to narrow polystyrene standards, were measured using
a Waters GPC system consisting of a Waters 510 pump, a
Waters 717 autosampler, a Wyatt Optilab DSP interferometric
refractometer, and a Wyatt Dawn EOS light scattering detec-
tor. The measurements were taken at 40 °C with THF as the
mobile phase on four columns (Polymer Labs PLgel 100, 500,
1 × 104, 1 × 105 Å). Viscosity measurements were performed
with an Ubbelohde capillary viscometer with DMAc as solvent
at 30 °C. UV-vis spectra were measured on a HP 8542A diode
array spectrophotometer, and fluorescence spectra were ob-
tained on a J obin Yvon Spex FluoroMax-2 with an excitation
wavelength equal to that of the λmax of absorbance. Thermo-
gravimetric analysis was performed on a Perkin-Elmer TGA
7 with a heating rate of 10 °C/min. Glass transitions were
(ppm) ) 7.25 (s, 1H), 7.56 (dd, 2H, J ) 7.2, 7.2 Hz), 7.65 (t,
1H, J ) 7.2 Hz), 8.00 (d, 2H, J ) 7.2 Hz). 13C NMR: δ (ppm)
) 126.18 (CH), 127.68 (q-C), 127.82 (CH), 129.31 (CH), 131.89
(q-C), 133.97 (CH), 137.37 (q-C), 140.33 (CH). Anal. Calcd for
C
10H6O2S2Cl2: C, 40.97; H, 2.06; S, 21.87. Found: C, 40.75;
H, 2.36; S, 21.82.
2,5-Dich lor ob en zop h en on e. From benzene and 2,5-
dichlorobenzoyl chloride (72% yield). White crystalline solid,
melting point 93 °C. 1H NMR: δ (ppm) ) 7.37 (s, 1H), 7.41
(m, 2H), 7.49 (dd, 2H, J ) 7.2, 7.2 Hz), 7.63 (t, 1H, J ) 7.2
Hz), 7.80 (d, 2H, J ) 7.2 Hz). 13C NMR: δ (ppm) ) 128.75
(CH), 128.86 (CH), 129.54 (q-C), 130.03 (CH), 131.07 (CH),
131.24 (CH), 132.89 (q-C), 134.07 (CH), 135.79 (q-C), 139.83
(q-C), 193.67 (CO). Anal. Calcd for C13H8OCl2: C, 62.18; H,
3.22. Found: C, 62.47; H, 3.28.
2-Ben zen esu lfon yl-1,4-d ich or oben zen e. From benzene
and 2,5-dichlorobenzenesulfonyl chloride (54% yield). White
crystalline solid, melting point 134 °C. 1H NMR: δ (ppm) )
7.36 (d, 1H, J ) 8.4 Hz), 7.50 (dd, 1H, J ) 8.4, 2.4 Hz), 7.54
(dd, 2H, J ) 7.8, 7.8 Hz), 7.64 (t, 1H, J ) 7.8 Hz), 7.96 (d, 2H,

4750 Hagberg et al.
Macromolecules, Vol. 37, No. 13, 2004
Ta ble 1. Con d ition s a n d Resu lts for th e P olym er iza tion
of 2,5-Dich lor o-3-(2′-th iop h en eca r bon yl)th iop h en e (M2)
sample
solventa
time (h)
% yieldb
[η]c dL/g
PTC1
PTC2
PTC3
PTC4
PTC5
DMAc
anisole
anisole
THF
24
4
24
4
42
78
88
98
97
0.20
d
d
0.28
0.35
THF
24
a
DMAc ) N,N-dimethylacetamide, THF ) tetrahydrofuran.
b Determined by mass recovered. c Determined using an Ubbelohde
capillary viscometer with DMAc as solvent at 30 °C. Intrinsic
d
viscosities were too low to be measured.
Ta ble 2. P h ysica l P r op er ties of
P oly[3-(2′-th iop h en eca r bon yl)-2,5-th iop h en e]
λ
max(abs)
(nm)a
λ
max(em)
(nm)b
10% weight
loss tempc
d
F igu r e 1. Structures of 3-benzoyl-2,5-dichlorothiophene (M1),
2,5-dichloro-3-(2′-thiophenecarbonyl)thiophene (M2), 3-ben-
zenesulfonyl-2,5-dichlorothiophene (M3), 2,5-dichlorobenzophe-
none (M4), and 2-benzenesulfonyl-1,4-dichlorobenzene (M5).
sample
Tg
PTC1
PTC2
PTC3
PTC4
PTC5
412
418
414
452
450
560
605
610
610
610
391
351
387
496
480
134
99
92
135
149
determined at the inflection point of the endotherm with a
Perkin-Elmer Pyris 1 differential scanning calorimeter with
a heating rate of 10 °C/min.
a
Determined by UV-vis spectroscopy; abs ) absorbance.
b
Determined by fluorescence spectroscopy with an excitation
wavelength equal the λmax(abs); em ) emission. c Determined by
Resu lts a n d Discu ssion
d
TGA. Determined by DSC.
In previous work with the 2,5-dichloro-3-benzoylthio-
phene monomer (Figure 1, M1), limitations arising from
monomer purity, polymer degradation during the reac-
tion, and partial polymer solubility were encountered.11
To overcome these difficulties, structurally similar
derivatives of monomer M1 were synthesized. The
monomers considered here are shown in Figure 1 and
include the two thiophene derivatives, 2,5-dichloro-3-
(2′-thiophenecarbonyl)thiophene (M2) and 3-benzene-
sulfonyl-2,5-dichlorothiophene (M3). Monomer M1 was
a viscous oil and subsequently very difficult to purify.
It was shown that monomer purity is highly important
in Ni(0) polymerization due to reduction of the aryl
chloride bond caused by protic impurities.5 Side reac-
tions, such as this reduction, result in low molecular
weights. The purification issues associated with M1
were easily overcome when the benzoyl pendant was
replaced with a thiophenecarbonyl or a benzenesulfonyl
electron-withdrawing group. The resulting monomers,
M2 and M3, were synthesized in high yield and greater
than 99% purity by GC. As previously mentioned, the
effects of the benzenesulfonyl group are known, but the
structure-property relationships of the thiophenecar-
bonyl pendant are unknown. It is expected to have
properties similar to the benzoyl pendant, but the effects
on the glass transition temperature (Tg), solubility, and
optoelectronic properties are undetermined.
The next step was to examine the polymerization
behavior of these two monomers in addition to the final
polymer properties. The question was whether the same
problems of degradation and limited solubility observed
with M1 would occur with M2 and M3. Polymerization
of M2 was carried out in three solvents N,N-dimethyl-
acetamide (DMAc), anisole, and tetrahydrofuran (THF)
(Table 1). Polymerization in DMAc gave yields of only
40-50%. Increasing the reaction time led to similar
degradation as observed with M1 and lowered the yield.
Conducting the polymerization in anisole or THF in-
creased the yields and, more importantly, did not lead
to degradation. Using THF as a solvent led to near
quantitative yields with short reaction times.
improvement allowed for the molecular weight charac-
terization by GPC and intrinsic viscosity. All polymers
from M2 were completely soluble in DMAc. The materi-
als synthesized in anisole were completely soluble in
THF. All other polymers of M2 exhibited partial solubil-
ity in THF. Molecular weights were measured by GPC
for the fractions of material soluble in THF, and
intrinsic viscosities were measured in DMAc. The
polymers synthesized in near quantitative yields had
the highest molecular weight THF soluble fractions of
approximately 2.0 × 103 g/mol. The oligomers from the
reactions in anisole were completely soluble in THF and
had number-average molecular weights of 700-900
g/mol. The material from reaction in DMAc had a
viscosity of 0.20 dL/g but a low yield due to side
reactions that led to a large fraction of methanol-soluble
material. The intrinsic viscosities of the material pro-
duced in anisole were too low to measure due to low
molecular weights. However, 4 h in THF gave a polymer
with an intrinsic viscosity of 0.28 dL/g. Increasing the
reaction time to 24 h led to a polymer with an intrinsic
viscosity of 0.35 dL/g.
The UV-vis and fluorescence spectra of the polymers
and oligomers prepared from M2 were measured (Table
2). All spectra were measured with DMAc as the solvent.
The polymers synthesized in THF had the longest
measured λmax of absorbance values, 450-452 nm (PTC4
and 5). The oligomers produced in anisole had shorter
λ
max of absorbance values, corresponding with very low
molecular weights. Comparison of the spectra shows a
marked difference in the resulting material from the
polymerizations conducted in DMAc and those con-
ducted in anisole or THF. In Figure 2, the spectrum of
the product of the reaction in DMAc (PTC1) does not
show a large valley at 360 nm like the products from
reaction in anisole or THF. The absorbance of the same
material also extends further toward the red than the
higher molecular weight polymers from reaction in THF.
It is important to note that the spectra of the materials
from the ether solvents are qualitatively the same and
only differ in the values of the maxima and minima.
This difference arises from the varying molecular
In addition to overcoming the degradation, the solu-
bility of the final polymer from M2 was improved. This

Macromolecules, Vol. 37, No. 13, 2004
Polythiophenes and Polyphenylenes 4751
Ta ble 3. Con d ition s a n d Resu lts for th e P olym er iza tion
of 3-Ben zen esu lfon yl-2,5-d ich lor oth iop h en e (M3)
time
(h)
Mn
% yieldb (g/mol)c PDIc (dL/g)
[η]d
sample solventa
PTS1
PTS2
PTS3
DMAc
DMAc
THF
24
72
24
77
62
99
1700e
1300e
2200
1.89
1.59
2.05
0.10
f
f
a
DMAc ) N,N-dimethylacetamide, THF ) tetrahydrofuran.
Determined by mass recovered. c Determined by GPC. Deter-
b
d
mined using an Ubbelohde capillary viscometer with DMAc as
solvent at 30 °C. e Not completely soluble in THF. f Intrinsic
viscosities were too low to be measured.
was used as solvent. The material resulting from the
reaction in DMAc had poor general solubility. It was
soluble only in DMAc and partly soluble in THF. The
oligomer from reaction in THF was completely soluble
in polar solvents such as THF, chloroform, and DMAc.
Viscosity measurements were attempted, but only PTS1
exhibited a measurable intrinsic viscosity of 0.1 dL/g.
The physical properties of the materials resulting from
these two solvents again were quite different. PTS3
exhibited longer λmax of absorbance and emission than
the material produced in DMAc (PTS1) (λmax(abs) ) 378
nm, λmax(em) ) 555 nm from THF, λmax(abs) ) 340 nm
and λmax(em) ) 470 nm from DMAc). In addition, the
oligomer from THF had higher 10% weight loss tem-
perature of 434 °C compared to 335 °C for the polymer
from DMAc. A glass transition temperature of 146 °C
was also observed for the material synthesized in THF.
The material from DMAc (PTS1) had a lower Tg of 101
°C.
F igu r e 2. UV-vis spectra of poly[3-(2′-thiophenecarbonyl)-
2,5-thiophene] synthesized in DMAc (]), anisole (0), and THF
(4).
F igu r e 3. Fluorescence spectra of poly[3-(2′-thiophenecarbo-
nyl)-2,5-thiophene] synthesized in DMAc (]), anisole (0), and
THF (4).
With the knowledge of the differing results in these
solvents, the polymerization chemistry of two materials
that we earlier reported, 2,5-dichlorobenzophenone (M4)
and 2-benzenesulfonyl-1,4-dichlorobenzene (M5), was
revisited in order to address the limitations encoun-
tered. In previous experiments, polymers of M4 with
number-average molecular weights of up to approxi-
mately 25 × 103 g/mol were synthesized in quantitative
yields. Unfortunately, polymerization in amide solvents
led to reduction of nearly 15% of the carbonyl function-
alities.14 This limitation was overcome by simply oxidiz-
ing the material postpolymerization or by the addition
of small quantities of triphenylphosphine oxide (TPO)
to the reaction mixture.15 In addition to this difficulty,
these polymers formed brittle films. Many of the target
applications for polymers such as this require the
formation of flexible, creasable films. In the case of the
polymerization of M5 only oligomeric materials were
produced. However, this material showed great promise
with 10% weight loss temperatures near 450 °C.10
Utilizing THF as solvent and the conditions described
for M2 and M3, polymerization of M4 resulted in a
number-average molecular weight of 55.6 × 103 g/mol
and an intrinsic viscosity of 1.15 dL/g. In addition to
higher molecular weights than previously reported, no
reduction of the carbonyl was observed, even without
the addition of TPO. This result eliminated the need
for oxidation reactions postpolymerization. The polymer
was soluble in solvents such as chloroform, THF, and
DMAc. Thin films of the polymer were cast from
chloroform. Although not highly creasable, the films
were flexible and showed much better overall film-
forming properties than the polymers synthesized pre-
viously in amide solvents. In addition to overcoming the
two previous limitations, the physical properties of this
polymer were greatly enhanced as well. The polymer
weights. A similar trend in the fluorescence spectra of
the polymers was observed (Figure 3). The higher
molecular weight material produced in THF had a
longer λmax of emission than did the lower molecular
weight materials synthesized in anisole. Once again,
these two spectra were very similar with their differ-
ences arising from the molecular weights of the materi-
als. The material from reaction in DMAc exhibited a
much lower λmax of emission (560 nm) and broader
emission spectra, different from the spectra of the
materials produced in the ether solvents. Thermal
analysis of these polymers again showed a difference
between the materials produced in DMAc and those
synthesized in ether solvents. PTC1 exhibited a 10%
weight loss temperature of 391 °C and a glass transition
temperature of 134 °C. The oligomers from anisole
(PTC2 and -3) showed a similar weight loss to PTC1
but lower glass transition temperatures. The polymers
synthesized in THF had respectable 10% weight loss
temperatures near 500 °C and glass transitions of 135
and 149 °C. It was evident from the optical and thermal
data that the polymers produced in anisole and THF
were structurally the same and only differed in molec-
ular weight, while the product from reaction in DMAc
was a wholly different material. Possible reasons for
these differences were determined and will be addressed
later in the text.
Given the results for monomer M2, it was chosen to
study the polymerization of monomer M3 in DMAc and
THF (Table 3). A reaction time of 24 h in DMAc gave a
yield of 77%. Upon increasing the reaction time, deg-
radative side reactions again were evident, and the yield
decreased to 62% in 72 h. When THF was employed as
solvent, a yield of 99% was obtained. However, no large
increase in molecular weight was observed when THF

4752 Hagberg et al.
Macromolecules, Vol. 37, No. 13, 2004
F igu r e 4. 1H NMR spectra of the reaction of NiCl2 (0.1 equiv),
Zn (3.1 equiv), TPP (0.4 equiv), and bipy (0.1 equiv) over 24 h
at 80 °C in DMF-d7.
had a 10% weight loss temperature of 576 °C and a glass
transition of 180 °C. Polymerization of M5 in THF only
yielded soluble oligomeric material with an Mn of 1.6
× 103 g/mol and a PDI of 1.3. As in the previous study
of this material, excellent potential thermal properties
were observed with a 10% weight loss of 578 °C in
nitrogen and a Tg of 136 °C. From these results, it was
apparent that THF was the best solvent for coupling
these arylene chlorides. Also, it was clear that the
substitution of the carbonyl with a sulfone drastically
changed the polymerization chemistry. No high molec-
ular weight sulfone-containing materials were synthe-
sized. It should be noted here that, on the basis of these
data, high molecular weight polymer from M1 would be
expected, if the monomer purity was sufficient. These
results led to two questions. The first was regarding the
role of solvent choice in the polymerization. The second
was concerning the role of the monomer structure.
Solvent choice was shown previously to play an
important role in Ni(0) coupling, but the mechanism
that gives rise to degradation and reduction was un-
clear.3,16 Therefore, a study of the catalyst in the
different solvents was undertaken. The catalyst com-
ponents (NiCl2 in DMF, NiCl2(PPh3)2 in THF, Zn, PPh3,
and bipy) were sealed in NMR tubes and heated, and
spectra were taken over a 24 h period (Figure 4). N,N-
Dimethylformamide-d7 (DMF-d7) was substituted for
DMAc due to availability. Several important differences
in these series of spectra were apparent. Free bipy (8.4
and 8.7 ppm) was observed throughout the reaction in
DMF-d7 by 1H NMR. No signal from free bipy was
observed in THF-d8 (Figure 5). The 31P NMR in DMF-
d7 showed an increase in intensity and narrowing of the
triphenylphosphine (TPP) signal (-5.2 ppm) from 8 to
24 h (Figure 6). In THF-d8, the TPP signal remained
broad and constant in intensity over the same time
period (Figure 7).
F igu r e 5. 1H NMR spectra of the reaction of NiCl2(PPh3)2 (0.1
equiv), Zn (3.1 equiv), TPP (0.2 equiv), and bipy (0.1 equiv)
over 24 h at 60 °C in THF-d8.
F igu r e 6. 31P NMR spectra of the reaction of NiCl2 (0.1 equiv),
Zn (3.1 equiv), TPP (0.4 equiv), and bipy (0.1 equiv) over 24 h
at 80 °C in DMF-d7.
the bipy, free bipy was not expected to be observed in
either solvent. The observation of free bipy in DMF
suggested that less bipy is complexed to Ni in amide
solvents than in THF. The ligands in Ni(0) coupling
were shown to play an important role in preventing side
reactions. In one example, the addition of bipy prevented
phenyl migration from TPP to the Ni atom.5 Bipy, not
complexed with the Ni center, would be expected to have
little utility in this role. Also, Ni(0)-phosphine com-
plexes have been observed to insert into the carbon-
sulfur bonds of certain thiophene derivatives.17 A hy-
drodesulfurization mechanism, such as this insertion,
could play a role in the degradation of polythiophene
chains observed at long reaction times. Bipy, complexed
to the Ni(0), may play a role in preventing such a side
reaction.
The observation of free TPP in solution was expected
due to the large stoichiometric excess compared to the
nickel salt. However, because of a greater binding
affinity and a one-to-one ratio between the nickel and

Macromolecules, Vol. 37, No. 13, 2004
Polythiophenes and Polyphenylenes 4753
that early in the reaction (conversions less than 80%)
the reduction of the ArNi(II)ClL2 (step 2) was the rate-
limiting step. As the concentration of ArCl approached
that of the nickel species, the oxidative addition of ArNi-
(I)L3 (step 3) became rate limiting. The degree of
polymerization is related to the conversion through eq
1 ([Xn] ) average degree of polymerization, F ) conver-
sion).18 Oligomers of 10 repeat units (such as the sulfone
materials presented here) equated to a 90% conversion
of the aryl chloride. At this concentration, the oxidative
addition of ArCl to the ArNi(I)L3 would be rate limiting.
Sulfonyl substituents are stronger withdrawing groups
than ketones and thus deactivate the ArNi(I)L3 complex
for oxidative addition to a greater extent.
F igu r e 7. 31P NMR spectra of the reaction of NiCl2(PPh3)2
(0.1 equiv), Zn (3.1 equiv), TPP (0.2 equiv), and bipy (0.1 equiv)
over 24 h at 60 °C in THF-d8.
[Xn] ) 1/(1 - F)
(1)
Sch em e 1. Rea ction Mech a n ism P r op osed by Colon
for th e Ni(0)-Ca ta lyzed P olym er iza tion of Ar yl
Ch lor id es
From this mechanism and our experimental results,
we propose that in Ni(0)-catalyzed coupling there exists
a window of reactivity in which the activating character
of the electron-withdrawing ability of the subsistent in
the first two steps is balanced with the deactivating
character of the same group in the third step. Ketone
pendants such as the benzoyl group or the thiophene-
carbonyl group meet this requirement. Pendants, such
as sulfones, deactivate the third step to a large enough
degree to inhibit the reaction.
Con clu sion s
The choice of solvent and the monomer structure were
shown to be key considerations in the Ni(0) polymeri-
zation of arylene dichloride monomers containing elec-
tron-withdrawing substituents. Polymerization in DMAc
resulted in side reactions, such as degradation of the
thiophene polymers in the reaction mixture and reduc-
tion of the carbonyl groups, which limited the materials
utility. Through the use of THF as solvent, high molec-
ular weight poly[3-(2′-thiophenecarbonyl)-2,5-thiophene]
and poly(2,5-benzophenone) were synthesized with no
side reactions being observed. Although the exact cause
of the differing results was not found, drastic differences
in the catalytic environments in the two solvents were
observed by NMR. Also, high molecular weight materi-
als from monomers containing carbonyl functionalities
were obtained, while substitution with a sulfone re-
sulted in oligomeric material. Analysis of the results and
consideration of the mechanism showed that in Ni(0)-
catalyzed polymerization there is a window of electron-
withdrawing ability for which functionalities such as
carbonyl-containing pendants meet the criteria and
greatly accelerate the reaction. Increasing the electron-
withdrawing ability by substitution with a sulfone slows
the reaction due to the stabilization of a reactive
intermediate. Further work is planned to quantify the
electron-withdrawing effect through the study of ben-
zophenone monomers with varying electron-donating
and -withdrawing functionalities in the 4′-position.
The difference in the catalyst stability observed in the
two solvents was also of interest. In DMF, the increase
in intensity and narrowing of the TPP signal showed
that over time less ligand was complexed with nickel.
It then follows that there was less active catalyst
present over time in DMF. Observation of the color of
the two systems supports this conclusion. In DMF, the
solution turned the deep red color of the active Ni(0)
species within minutes, but concurrently with the rise
in the TPP signal a black precipitate formed. The
reaction in THF remained the deep red color throughout
the 24 h period. Although the active catalyst species
could not be identified, the greater number and concen-
tration of phosphorus species in THF pointed to higher
active catalyst concentration.
To understand the role of the monomer structure in
the reaction, the mechanism was considered. The cata-
lytic cycle proposed by Colon and Kelsey contains three
steps in which the structure of the monomer could
possibly play a role (Scheme 1).5 These steps were
oxidative addition of Ni(0) across the aryl chloride bond
(step 1), reduction resulting in an arylnickel(I) species
(step 2), and oxidative addition of the arylnickel(I)
species to a second aryl chloride (step 3). The first two
steps were shown to be accelerated by the presence of
an electron-withdrawing substituent on the aryl chlo-
ride.8 In the third step, the presence of an electron-
withdrawing group ortho to the Ni-C bond would be
expected to increase the stability of the electron-rich
ArNi(I)L3 (where L ) PPh3 or L2 ) bipy) complex and
thus slow the oxidative addition reaction of this complex
with a second aryl chloride. Colon and Kelsey proposed
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from Dupont and 3M awards and the
U.S. Department of Energy, Office of Basic Energy
Sciences, Division of Chemical Science, Contract W-7405-
Eng-82.

4754 Hagberg et al.
Macromolecules, Vol. 37, No. 13, 2004
(11) Wang, J .; Sheares, V. V. Macromolecules 1998, 31, 6769.
(12) Ueda, M.; Miyaji, Y.; Ito, T.; Obu, Y.; Sone, T. Macromolecules
1991, 24, 2694.
Refer en ces a n d Notes
(1) Phillips, R. W.; Sheares, V. V.; Samulski, E. T.; DeSimone,
J . M. Macromolecules 1994, 27, 2354.
(2) Percec, V.; Bae, J . Y.; Zhao, M.; Hill, D. H. J . Org. Chem.
1995, 60, 176.
(3) Grob, M. C.; Feiring, A. E.; Auman, B. C.; Percec, V.; Zhao,
M.; Hill, D. H. Macromolecules 1996, 29, 7284.
(4) Percec, V.; Okita, S.; Weiss, R. Macromolecules 1992, 25,
1816.
(5) Colon, I.; Kelsey, D. R. J . Org. Chem. 1986, 51, 2627.
(6) Pei, J .; Yu, W.-L.; Ni, J .; Lai, Y.-H.; Huang, W.; Heeger, A.
J . Macromolecules 2001, 34, 7241.
(13) Pasquale, A. J .; Sheares, V. V. J . Polym. Sci., Part A: Polym.
Chem. 1998, 36, 2611.
(14) Bloom, P. D.; Sheares, V. V. Macromolecules, in press.
(15) Salazar, L. A.; Sheares, V. V. J . Polym. Chem., Part A: Polym.
Chem. 2003, 41, 2277.
(16) Kende, A. S.; Liebeskind. L. S.; Braitsch, D. M. Tetrahedron
Lett. 1975, 39, 3375.
(17) Herna´ndez, M.; Miralrio, G.; Are´valo, A.; Berne`s, S.; Garc´ıa,
J . J .; Lo´pez, C.; Maitlis, P. M.; del Rio, F. Organometallics
2001, 20, 4061.
(7) Wang, Y.; Quirk, R. P. Macromolecules 1995, 28, 3495.
(8) Stille, J . K.; Lau, K. S. Y. Acc. Chem. Res. 1977, 10, 434.
(9) Marrocco, M. L.; Gange, R. R. U.S. Patent 5,227,457, 1993.
(10) Bloom, P. D.; Sheares, V. V. Macromolecules 2001, 34, 1627.
(18) Odian, G. Principles of Polymerization, 3rd ed.; J ohn Wiley
and Sons: New York, 1991; p 79.
MA0355453