LiCl-promoted chain growth kumada catalyst-transfer polycondensation of the "reversed" thiophene monomer

Article abstract of DOI:10.1021/ma201192p

The effect of LiCl on the chain growth Kumada catalyst-transfer polycondensation (KCTP) of the "reversed" thiophene monomer, 5-bromo-2-chloromagnesio-3-hexylthiophene (3a) (that has bulky substituent adjacent to the chloromagnesium group), was investigate

Full text of DOI:10.1021/ma201192p

ARTICLE
pubs.acs.org/Macromolecules
LiCl-Promoted Chain Growth Kumada Catalyst-Transfer
Polycondensation of the “Reversed” Thiophene Monomer
†
,‡
†,‡
†
,†
†
Shupeng Wu, Li Huang, Hongkun Tian, Yanhou Geng,* and Fosong Wang
†
State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences,
Changchun 130022, P. R. China
‡
Graduate School of Chinese Academy of Sciences, Beijing 100049, P. R. China
S Supporting Information
b
ABSTRACT: The effect of LiCl on the chain growth Kumada
catalyst-transfer polycondensation (KCTP) of the “reversed” thio-
phene monomer, 5-bromo-2-chloromagnesio-3-hexylthiophene (3a)
(
that has bulky substituent adjacent to the chloromagnesium group),
was investigated. LiCl promotes the polymerization of 3a, and the
polymerization also exhibits living characteristics similar to those
of the “normal” monomer, 2-bromo-5-chloromagnesio-3-hexylthio-
phene (3b). However, initiation is much slower than chain propaga-
tion due to the steric hindrance between hexyl groups in the
transmetalation (TM) step that leads to the formation of the initiator
via head-to-head (HH) coupling. Consequently, regioregular poly(3-hexylthiophene) (P3HT) with large polydispersity (PDI) and
higher number-average molecular weight (M ) than the theoretical value was obtained. Because of the slow initiation with 3a, the
n
polymerization of 3a occurs primarily after the consumption of 3b in the copolymerization of 3a and 3b, and P3HT with high
regioregularity was still obtained.
’
INTRODUCTION
Scheme 1. Homopolymerizations of (A) 3a and (B) 3b in the
Presence of 1 equiv of LiCl
Chain growth Kumada catalyst-transfer polycondensation
(
KCTP) has become an attractive method for the controlled
1ꢀ4
synthesis of conjugated polymers.
Particularly, KCTP with
living characteristics has been demonstrated for several polymers,
5
ꢀ17
17,18
such as polythiophenes (PThs),
polyphenylenes (PPs),
and poly(bithienylmethylene)s
PBTMs). This has enabled the synthesis of fully conjugated
1
7,19,20
polypyrroles (PPys),
2
1
(
8
,19,22ꢀ26
27ꢀ31
block copolymers,
rodꢀcoil block copolymers,
3
2ꢀ36
and polymer brushes
in a well-defined way. It has been
demonstrated that KCTP is strongly affected by both catalyst
8
,13ꢀ15,23,24,37ꢀ39
and the monomer structures.
One of those
or 1,3-bis(diphenylphosphino)ethane nickel dichloride (Ni(dppe)-
13,15
factors, the effect of alkyl substituents on monomer reactivity,
Cl ) as catalyst,
since the HH coupling of two 3a molecules,
2
8,13,15,39
15
plays an important role in KCTP of certain monomers.
For example, KCTP of a 85:15 mixture of the “normal” monomer,
-bromo-5-chloromagnesio-3-hexylthiophene (3b), and the “re-
versed” monomer, 5-bromo-2-chloromagnesio-3-hexylthiophene
3a), prepared by the metalation of 2,5-dibromo-3-hexylthio-
phene (4) affords P3HT with regioregularity values of 95%ꢀ
which is required for the formation of initiator, was prohibited.
In the surface-initiated KCTP to prepare poly(p-phenylene)
PPP) brushes, it was found that monomers with shorter side
chains could be more easily polymerized and result in smoother
2
(
35
(
and thicker films. In the synthesis of all conjugated diblock
copoly(3-alkylthiophene)s, Yang et al. found that the monomer
with the shorter side chain had to be polymerized first in order to
obtain the copolymer with the targeted molar ratio of the two
segments, implying the influence of the different bulkiness of
alkyl substituents in the monomer on block copolymerization
8
9
7%. Previous studies attribute this regioselectivity to the steric
hindrance caused by the hexyl group that is adjacent to the
chloromagnesio group in the “reversed” monomer 3a. Therefore,
the polymerization takes place predominantly with the “normal”
8
monomer 3b. Recently, Kiriy and Luscombe independently
studied the polymerization of the “reversed” monomer 3a. They
Received: May 26, 2011
found that no polymer was produced with either 1,3-bis-
Revised:
August 14, 2011
(diphenylphosphino)propane nickel dichloride (Ni(dppp)Cl2)
Published: September 06, 2011
r 2011 American Chemical Society
7558
dx.doi.org/10.1021/ma201192p
|Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
Figure 1. (A) M and PDI versus conversion of 3a. (B) M and PDI versus [converted 3a]/[Ni(dppp)Cl ] in monomer addition experiments.
n
n
2
(
C) GPC elution curves of P3HT in monomer addition experiments. (D) MALDI-TOF mass spectrum of polymer from 3a. All polymerizations were
carried out at room temperature in the presence of LiCl with [1] = 0.1 mol/L. For (A), the polymerization was conducted with 4 mol % Ni(dppp)Cl
For (B) and (C), the polymerization was conducted with 1.8 mol % Ni(dppp)Cl to total 1. For (D), the polymerization was carried out with 10 mol %
Ni(dppp)Cl (P3HT: M = 5.1 ꢁ 10 , PDI = 1.48 measured by GPC with polystyrene as the standard).
0
2
.
2
3
2
n
2
5
results. Accordingly, it is clear that understanding the steric
effects on KCTP and developing approaches to promote the
polymerization are important for the development of this new
polymerization method.
’ RESULTS
Homopolymerizations of 3a and 3b. A factor in the poly-
1
5
merization of 3a that cannot be ignored is the purity of 1. Kiriy
et al. found that a small amount of 2 can significantly affect the
polymerization. Therefore, in our experiments, 1 was exhaus-
tively purified and the measured purity of 1 was 99.5% (see
Figure S4 in the Supporting Information). As shown in
Scheme 1A, 3a was synthesized by Grignard metathesis of 1
Grignard reagents form complicate aggregates in solution,
and dispersing the aggregates can significantly improve reac-
tivity toward electrophiles; thus, the steric hindrance issues
in the TM step might be alleviated. The literature has indica-
tions that lithium salts such as LiCl exhibit this function.
It was found that the addition of 1 equiv of LiCl in the reaction
4
0
4
1ꢀ43
i
with PrMgCl in the presence of 1 equiv of LiCl. The magnesiumꢀ
4
1,42
halogen exchange took place selectively with iodine to afford
medium could promote magnesiumꢀhalogen exchange.
Grignard reagent 3a in a yield of ∼95% in 1 h. The resulting
This phenomenon was also observed in KCTP for conjugated
polymers. For example, Yokozawa et al. found that the polymer-
ization of 1-bromo-4-chloromagnesio-2,5-dihexyloxybenzene ex-
1
1
solution was characterized by H NMR. Compared with the H
NMR spectrum of 3a without the addition of LiCl, an obvious
1
18
peak shift was observed in the H NMR spectra of 3a with the
hibited living characteristics in the presence of 1 equiv of LiCl.
Recent studies by McNeil indicated that LiCl can form aggre-
gates with the monomer and that the effect of LiCl on the
polymerization depends on the catalyst used.
dppp)Cl was used as the catalyst, a faster polymerization was
observed in the presence of LiCl at monomer concentrations
lower than 0.4 M. For polymerization with Ni(dppe)Cl , LiCl
1
addition of LiCl (Figure S5, H NMR spectra of the other
monomers mentioned in this paper with and without LiCl are
also shown in Figure S5), suggesting that the addition of LiCl
37,38
When Ni-
may inhibit the aggregation of 3a and that a new aggregate
(
37,38
2
between LiCl and 3a was formed.
Polymerization of 3a with
4
1
mol % Ni(dppp)Cl as the catalyst was almost complete after
3
8
2
2
0 min to render P3HT in a yield of 85% with M , PDI, and
37
n
may affect only the initiation, but not chain propagation. In the
synthesis of polyfluorenes, Geng et al. found that the addition of
LiCl could enhance the conversion of the magnesiumꢀhalogen
4
regioregularity of 1.3 ꢁ 10 , 1.50, and 93%, respectively. As
shown in Figure 1A, M s of the polymers increased with
n
conversions, but not linearly. In contrast, no polymer was
afforded even after 5 h in the absence of LiCl, consistent with
44
exchange and M of the resulting polymers. However, the effect
n
13,15
of LiCl on the polymerization of thiophene monomers has not
been studied in detail yet, although we and others have inde-
pendently reported the synthesis of P3HT in the presence of
the results of Luscombe and Kiriy.
This indicates that the
addition of LiCl can overcome the steric hindrance issue in the
polymerization of 3a. Polymerizations of 3b were also conducted
under the same conditions for comparison (Figure S6). In
the presence of LiCl, the polymerization was also complete
1
6,24
LiCl.
In the current contribution, we systematically studied
the KCTP homopolymerization and copolymerization of thio-
phene monomers in the presence of 1 equiv of LiCl and found
that LiCl can promote the polymerization of the “reversed”
monomer 3a. However, the polymerization is characterized by
slow initiation and fast chain propagation.
in ∼10 min (Figure S6A), M increased linearly versus conver-
n
sion with PDI <1.23 (Figure S6B), and P3HT with M , PDI, and
n
3
regioregularity of 6.5 ꢁ 10 , 1.23, and 92%, respectively, was
afforded in a yield of 86%. Clearly, the polymerization of 3a is
7
559
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
Scheme 2. Copolymerization of 3a and 3b in the Presence of 1 equiv of LiCl
Figure 2. (A) Conversions of 3a and 3b and regioregularity of P3HT versus time. (B) M
n
versus conversion of 3b. (C) M
n
and PDI versus the sum of the
conversions of 3a and 3b. (D) GPC elution curves of P3HT at different polymerization times. The polymerization was carried out at room temperature
in the presence of 1 equiv of LiCl with 2 mol % Ni(dppp)Cl and [4] = 0.10 mol/L.
2
0
remarkably different from that of 3b. To determine whether the
polymer chain is still living after the polymerization of 3a,
monomer addition experiments were carried out. Gel permea-
tion chromatography (GPC) profiles as shown in Figure 1C
shifted to higher molecular weight regions and remained unim-
odal after the addition of each portion of 3a, indicating that
the polymer chain remains active during the whole process.
Figure 1D shows the matrix-assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectrum of P3HT from 3a;
the main signals correspond to polymers with H/Br ends. These
data clearly confirm the living character of the polymerization of
three stages. In the first 5 min (the first stage), 3a essentially did
not polymerize while over 80% 3b was consumed. The second
stage involved a very short period of time during which both 3a
and 3b were involved in the polymerization accompanied by a
slight decrease of regioregularity. In the third stage, the conver-
sion of 3b terminated at ∼96% while the polymerization of
the remaining 3a continued until reaching a conversion of 91%.
M
increased during the whole process. As shown in Figure 2B,
after the consumption of 3b stopped, M still increased with the
polymerization of 3a. Meanwhile, M versus conversion of [3a +3b]
n
n
n
was almost linear. The GPC profiles as shown in Figure 2D are
also consistent with the above observation and show that the peak
gradually moves to the higher molecular weight region over the
course of polymerization. This indicates that 3a propagates the living
polymer chains instead of new chains. Considering that polymeriza-
tion of 3b is living and the resting state of the Ni catalyst after the
3
a. We attribute the nonlinear increase of M versus conversion
n
of 3a and [converted 3a]/[Ni(dppp)Cl ] to reasons other than the
2
presence of chain termination or transfer (see Discussion section).
Copolymerization of 3a and 3b. Copolymerization of 3a and
b was initially carried out using a 3a/3b ratio of 16/84
3
38
(Scheme 2A). The 3a/3b mixture was prepared from magne-
polymerization is P3HTꢀNi(dppp)Br as reported in the literature,
siumꢀhalogen exchange of 2,5-dibromo-3-hexylthiophene (4)
the block copolymerization of 3a and 3b is reasonable.
8
following McCullough’s method, with the exception that 1
All the above results confirm that LiCl can promote the
polymerization of the “reversed” monomer 3a. Polymerizations
with different amounts of catalyst were also conducted for the
3a/3b (16:84) mixture. As shown in Table 1, all three of the
polymerizations in the presence of LiCl afforded P3HT with
equiv of LiCl was added. Figure 2A depicts the conversions of
3
a and 3b versus time. In contrast to the polymerization without
LiCl (only 3b polymerized), both 3a and 3b were consumed
when LiCl was present. The polymerization can be divided into
7
560
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
a
Table 1. Polymerization Results from 4 Catalyzed by Different Amounts of Ni(dppp)Cl without and with 1 equiv LiCl
2
regioregularity (%)
yield (%)
n
M /PDI
amount of catalyst (mol %)
without LiCl
with LiCl
without LiCl
with LiCl
without LiCl
with LiCl
4
2
1
81
92
94
85
91
92
62
59
62
85
89
80
3500/1.13
9200/1.13
16000/1.20
6400/1.27
12600/1.34
20000/1.26
a
0
All polymerizations were carried out at room temperature with [4] = 0.1 mol/L.
n n
Figure 3. (A) Conversions of 3a and 3b and regioregularity of P3HT versus time. (B) M versus the conversion of 3b. (C) M and PDI versus the sum of
the conversions of 3a and 3b. (D) GPC elution curves of P3HT at different polymerization times. The polymerization was carried out at room
temperature with 1 equiv of LiCl and 2 mol % Ni(dppp)Cl to the sum of 1 and 2. [1] = [2] = 0.05 mol/L.
2
0
0
yields g80%, which are significantly higher than those without
LiCl. This is consistent with the fact that 3a does not polymerize
Therefore, this difference in the polymerization rate must origi-
nate from the differenceofthe starting materials (4 versus 1 and 2),
although the actual reason is not currently clear.
13,15
in the absence of LiCl.
We emphasize that the regioregularity
of the resulting P3HT decreased only slightly. The above results
indicate that the addition of LiCl can significantly enhance the
yield of regioregular P3HT. This would be of obvious importance
for the mass production of regioregular P3HT for commercial
purposes.
’
DISCUSSION
According to the polymerization mechanism of 3b proposed
7
by Yokozawa (Scheme 3B), the initiator 7b is formed by the
To further verify the polymerization of 3a in the presence of
b, a 50/50 mixture of 3a and 3b was prepared for copolymer-
coupling of Ni(dppp)Cl with 2 equiv of 3b followed by an
2
3
immediate insertion of Ni into one CꢀBr bond of the bithio-
phene with two bromine atoms. Polymerization then proceeds
following an intramolecular catalyst transfer polycondensation
ization (Scheme 2B). As shown in Figure 3A, similar to the
situation of the 16/84 mixture of 3a and 3b, polymerization of 3a
also began after 80% 3b had been consumed (Figure 3A). How-
ever, GPC profiles become bimodal along with an increase of
PDI from ∼1.3 to ∼2.0. The peak in the low molecular weight
region only moved slightly to lower elution time, and a new peak
appeared in the higher molecular weight region. This interesting
phenomenon indicates that chain propagation for polymeriza-
tion of 3a occurs in only some P3HT chains, although all P3HT
chains should be living. The conversion of 3a was complete
in ∼30 min, noticeably faster than that of 3a prepared from 4
13,15
mechanism. As reported by Luscombe and Kiriy,
in the ab-
sence of LiCl, the coupling of Ni(dppp)Cl with 3a stopped at
2
the stage of 5a due to the large steric hindrance caused by the
hexyl group. As a result, the polymerization of 3a could not
proceed. The successful polymerization of 3a with 1 equiv of
LiCl in our case implies that the coupling of 5a with 3a and
intramolecular catalyst transfer to afford the initiator 7a can
proceed smoothly. To validate that the polymerization of 3a
follows a mechanism similar to that of 3b (Scheme 3A), the
structures of the chain ends of the polymers were characterized
(
about 100 min as shown in Figure 2A). We also conducted the
polymerization of the 16/84 mixture of 3a and 3b which was
1
by an H NMR spectrometer. An additional P3HT sample was
prepared from 1 and 2; the conversion of 3a was also very fast.
also prepared by polymerizing 3b with an external initiator, i.e.,
7
561
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
Scheme 3. Polymerization Mechanism of 3a and 3b (L = dppp)
2
3
1
Figure 5. P NMR spectra (162 MHz, THF) of the polymerization
mixture after consumption of 3a (A) and then further stored for 1 h in a
N2 atmosphere (B). The polymerization was carried out at room
temperature with 1 equiv of LiCl, 10 mol % Ni(dppp)Cl
.10 mol/L.
2 0
, and [1] =
0
45
1
Ni(dppp)Cl as the catalyst. Figure 4C is the H NMR spec-
2
trum of P3HT prepared by the polymerization of 3a with
Ni(dppp)Cl in the presence of LiCl. The two doublets at 7.16
2
and 6.93 ppm, identical to those in Figure 4B, clearly confirm the
presence of the TH-1 chain end, which was produced by
quenching the active chain ends. It was, however, difficult to
identify the characteristic signals corresponding to Br-terminated
chain ends formed by HH coupling of 3a (Figure 4A, TB-1 in
structure 2). The P3HT sample was treated with excessive
1
Figure 4. Chemical structures of P3HT (A) and H NMR spectra
(
600 MHz, CDCl ) of P3HT from polymerization of 3b with 4%
external initiator 10 in the absence of LiCl [(B) structure 1 in (A): M =
3
t
BuMgCl and then quenched with dilute HCl to convert the
n
3
3
1
5
.7 ꢁ 10 , PDI = 1.14 by GPC), P3HT from polymerization of 3a with
0% Ni(dppp)Cl in the presence of LiCl [(C) structure 2 in (A): M
.1 ꢁ 10 , PDI = 1.48 by GPC), and P3HT after Grignard metathesis of
Br-terminated chain ends to H-terminated chain ends following
8
1
2
n
=
the method proposed by Mccullough. Figure 4D displays the H
3
NMR spectrum of the resulting polymer. A new doublet ap-
t
the sample in (C) with BuMgCl [(D) structure 3 in (A). The
polymerizations were carried out with [1] or [2] = 0.1 mol/L. /
peared at 7.31 ppm which can be ascribed to H of the HH
d
1
0
0
coupling chain end TH-3 according to the H NMR spectrum
13
and # in the spectra indicate C satellites.
0
0
of 3,3 -dihexyl-2,2 -bithiophene (Figure S7). Although another
doublet due to the same thiophene unit, which may overlap with
other strong signals, could not be distinguished, the presence of
this new doublet clearly confirms the occurrence of HH coupling
in the initiation stage. Moreover, the integration value of the Hd
cis-bromo(3-hexyl-2-thienyl)-1,3-bis(diphenylphosphino)propaneꢀ
nickel(II) (10), to help in the assignment of the chain ends.
According to the polymerization mechanism, the structure of
the afforded P3HT should be structure 1 in Figure 4A, and two
types of H-terminated chain ends (TH-1 and TH-2) should be
observed in the H NMR spectrum of this polymer. As shown in
Figure 4B, the two doublets at 7.16 and 6.93 ppm are attributed
doublet is close to that of H doublet, implying that initiation is
a
dominated by HH coupling of 3a. The catalyst resting state after
1
31
the polymerization was also monitored by a P NMR spectro-
31
meter. Figure 5A shows the P NMR spectrum of the reac-
to H and H in the TH-1 chain end, according to the assignment
tion mixture with 10 mol % Ni(dppp)Cl . Two peaks at 18.84
a
b
2
1
of the H NMR signals of the model compounds and P3HT
and ꢀ3.25 ppm were observed, indicating that the resting state of
32,45
38
chain ends reported in the literature.
The singlet peak at
the catalyst after polymerization is P3HTꢀNi(dppp)Br. It is
6
.90 ppm is ascribed to H in the TH-2 chain end, which is
noticeable that no obvious changes were found after 1 h storage
c
apparent in P3HT prepared by the polymerization of 3b with
in a N atmosphere, as shown in Figure 5B, implying that this
2
7
562
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
catalyst resting state is relatively stable. This confirms the living
characteristic of the polymerization of 3a and indicates that
polymer chains are still active after the polymerization. All the
above discussion indicates that polymerization of 3a also follows
a mechanism similar to that of 3b, as shown in Scheme 3A.
According to Kiriy’s recent study, the presence of ∼3% 3b
along with a small peak corresponding to Ni(dppp) were ob-
2
served (Figure 6A). The complex 12 was stable after 30 min
storage (Figure 6B). However, once 1 equiv of LiCl was added, in
10 min, the peaks assigned to 12 almost disappeared, and the
signal attributed to Ni(dppp) became dominant (Figure 6C).
2
These results clearly confirm the promotion effect of LiCl on the
HH coupling.
1
5
could cause the polymerization of 3a in the absence of LiCl. To
exclude the effect of the trace amount of 3b on the polymeriza-
tion, polymerizations of 3a were carried out with the external
initiator 10 in the presence and absence of LiCl, respectively, as
shown in Scheme 4. In this situation, the effect of a trace amount
of 3b can be reduced substantially since the reaction of 3b with
As shown in Figure 1, the M of P3HT from the polymeriza-
n
tion of 3a is higher than the theoretical value, and the corre-
sponding PDI is relatively large. According to theoretical studies
on living polymerization, these phenomena could indicate a slow
46,47
initiation and fast propagation rate.
To explore the charac-
1
ꢀ
0 still afforded a Ni complex with an alkyl group neighboring
Ni(dppp)Cl. As expected, in the presence of LiCl, the polym-
teristics of the polymerization of 3a in the presence of LiCl, the
reaction of 2 equiv of Grignard reagent 3a (added 1 equiv by 1
erization proceeded smoothly (see Supporting Information). In
contrast, no polymer was formed after 6.5 h in the absence of
LiCl. Kiriy et al. have also shown that the reverse monomer 3a
can not polymerize with Ni(dppp)Cl or Ni(dppe)Cl as the
equiv) with 1 equiv of Ni(dppp)Cl was conducted. A similar
2
reaction of 3b was also carried out for comparison. The reaction
mixtures were quenched with 5 M HCl for characterization by
GCꢀmass spectroscopy (GC-MS). 1,4-Dipentyloxybenzene was
used as a standard, and its concentration was normalized as 100.
The relative concentrations of thiophene and bithiophene deri-
vatives are listed in Table 2. Polymers or oligomers with numbers
of repeat units g3 could not be detected by GC/MS in the
current experimental condition due to their high boiling points. A
small amount of bithiophene byproducts (HThThH and BrTh-
ThBr) was also observed. As shown in Table 2, lines A and B
show the concentrations of thiophene and bithiophene deriva-
tives in the quenched solutions of the fresh Grignard reagents
2
2
catalyst in the absence of LiCl, since the HH coupling of two 3a
molecules leading to the formation of initiator was prohibited.
The above results imply that LiCl can promote the HH coupling,
leading to the polymerization of 3a.
To further confirm the promotion effect of LiCl on the HH
coupling, reactions of 4 equiv of 2-chloromagnesio-3-hexylthio-
phene (9a) and 1 equiv of Ni(dppp)Cl in the absence and
presence of LiCl as shown in Scheme 5A were carried out, and
the products were analyzed using gas chromatography (GC).
Without LiCl, only 5% 9a was converted to 3,3 -dihexyl-2,2 -
bithiophene in 10 min. In contrast, in the presence of LiCl, the
conversion of 9a was 33% after the same time. The chemical
2
0
0
0
0
structure of 3,3 -dihexyl-2,2 -bithiophene was confirmed by the
1
H NMR spectrum. Reactions were also carried out as shown in
31
Scheme 5B for in situ 1 P NMR characterization. Consistent
5,39
with Kiriy’s observation,
without LiCl, only the signals of cis-
chloro-(3-hexyl-2-thienyl)-1,3-bis(diphenylphosphino)propaneꢀ
nickel(II) (12) (two doubletsat 17.68 and ꢀ3.97 ppm, J = 63 Hz)
Scheme 4. Polymerizations of 3a with External Initiator 10 in
the (A) Absence and (B) Presence of LiCl
3
1
Figure 6. P NMR spectra (162 MHz, THF) of reaction mixture of 4
equiv of 9a and 1 equiv of Ni(dppp)Cl without the addition of LiCl
2
after reaction for 10 min (A) and 30 min (B) and after the addition of
0
LiCl for 10 min (C). The reactions were carried out at 0 °C with [9a] =
3
9
0
.09 mol/L. The signals were assigned according to the literature and
31
the P NMR spectrum of the external initiator 10 (in the same
measurement condition, 10 showed two doublets at 19.00 and ꢀ3.22
ppm with J = 57 Hz).
31
Scheme 5. Reactions of 3a (4 equiv) with Ni(dppp)Cl (1 equiv) for Kinetics Study (A) and in-situ P NMR Characterization (B)
2
7
563
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
and reaction mixtures after the addition of 1 equiv of Grignard
reagent, respectively. Over 80% 3b was converted to the initiator
7a rapidly grows to an oligomer or polymer once it is formed, as
supported by the presence of high molecular weight products as
indicated by GPC measurements (Figure S8). Moreover, no
signals corresponding to 6a were observed in P NMR spectra
(Figure S9). This phenomenon together with the fact that most
5a remained and only a small amount of 8a was observed indi-
cates that the TM step leading to 6a is the rate-limiting step in the
formation of the initiator 7a.
7
b (the quenching product is 8b), indicating that the formation
31
of the initiator is a very fast process. However, the reaction
behavior of 3a and Ni(dppp)Cl was remarkably different. Only a
2
small amount of 8a, which is the quenched product of 7a as
shown in Scheme 3C, was observed. According to the kinetics
study as shown in Figure 1A, the polymerization of 3a is very fast
(
over in ∼10 min with 4 mol % catalyst). Kiriy et al. also showed
To further confirm the slow initiation and fast propagation
characteristics of the polymerization of 3a, the reaction of
the external initiator 10 with a mixture of 9a (5 equiv) and 2-
chloromagnesio-4-hexylthiophene (9b, 5 equiv) was carried out
in the presence of 1 equiv of LiCl, and the resulting product was
analyzed (Scheme 6, for experimental details please see Support-
39
that 3a reacts smoothly with Ni(dppp)Cl to yield 5a. There-
2
fore, we postulate that the conversion of 5a to 6a and then to the
initiator 7a is much slower than the formation of 5a and the chain
propagation process. The reaction results with 2 equiv of Grig-
nard reagent also support above conclusion. Before characteriza-
tions, the reaction mixture was stirred for 10 min to ensure 3a was
0
ing Information). Two bithiophene derivatives, i.e., 3,4 -dihexyl-
0
0
0
completely consumed. All Ni(dppp)Cl was converted as in-
2,2 -bithiophene (11b) and 3,3 -dihexyl-2,2 -bithiophene (11a),
with a ratio of 15:1 (11b:11a) were found. This implies that the
propagation rate is an order of magnitude faster than the initia-
tion rate in the polymerization of 3a. According to the theoretical
2
31
dicated by its complete dissolution. Meanwhile, in situ P NMR
spectra (Figure S9D) only exhibit signals corresponding to aryl-
Ni(dppp)-X (X = Br or Cl). Therefore, 5d was the quenched
product of 5a (Scheme 3C). From the concentration of 5d, we
can estimate that about 60% 3a was consumed by ∼20% cat-
alyst. These observations indicate that the bithiophene initiator
46,47
studies on living polymerizations,
initiator not involved in
the polymerization and PDI can be estimated from the ratio of
initiation and propagation rates, the ratio of monomer and ini-
tiator, and the conversion of the monomer. In the polymerization
as shown in Scheme 1A, it was estimated that about 20% initiator
remained after the polymerization. However, the theory assumes
that initiator is soluble in the solvent, and in the current system
the formation of the initiator involves a dissolution process of the
Table 2. Relative Concentrations of Thiophene and Bithio-
phene Derivatives in Quenched Solutions of 3a and 3b before
a
(Line A) and after (Lines B and C) Polymerization
3
a
3b
catalyst since the solubility of Ni(dppp)Cl in THF is poor.
2
Therefore, the catalyst which was not involved in the polymer-
5
d
HThThH 8a BrThThBr 5e HThThH 8b BrThThBr
ization should be more than 20%, leading to a M that is ∼2 times
n
46,47
A
570
440
230
0
0
0
0
11
4
0
2
3
592
102
48
0
2
3
0
240
94
0
2
2
of the theoretical value. According to the theory,
the PDI of
b
B
P3HT should be in the range of 1.3ꢀ1.4, slightly lower than the
c
observed value.
C
a
On the basis of above discussion, we can conclude that LiCl
can promote the coupling of 3a to form HH bithiophene initiator
7a for polymerization (Scheme 3A). However, this process is
much slower than chain propagation due to the large steric
hindrance between the two hexyl groups in HH coupling.
Therefore, the number of active chains in the polymerization
process was not constant, resulting in the large PDI and non-
Concentration of standard 1,4-dipentyloxybenzene was normalized as
b
1
00, and the reaction was carried out at room temperature. After
c
addition of 1 equiv of Grignard reagent in 1 min. 10 min after addition
of 2 equiv of Grignard reagent.
Scheme 6. Reaction of the External Initiator 10 with
Grignard Reagents 9a and 9b
linear increase of M versus conversion of 3a and [converted
n
3
a]/[Ni(dppp)Cl ] (Figure 1A,B). The low molecular weight
2
tail was also observed as shown in Figure 1C due to a slow
initiation relative to propagation. Moreover, molecular weight
was higher than the theoretical value since not all catalyst was
involved in the polymerization. The KCTP with slow initiation
Scheme 7. Proposed Mechanism of the Copolymerization of 3a and 3b in the Presence of 1 equiv of LiCl
7
564
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
was also observed recently by MeNeil et al. in the polymerization
of 1-bromo-4-chloromagnesio-2,5-dihexyloxybenzene with Ni-
NMR (Figure S10). This indicates HH coupling is unfavorable in
the coupling process. Therefore, in the copolymerization, in-
itiator 7a is formed in very limited amount while initiator 7b is
the most abundant. Head-to-tail (HT) and tail-to-head (TH)
coupling can afford both 7c and 7d due to the random ring-
(depe)Cl (depe =1,3-bis(diethylphosphino)ethane) as the cat-
2
17 31
alyst. On the basis of P NMR studies, they proposed that the
slow initiation was caused by the slow reductive elimination of a
symmetric biaryl Ni complex, which is the precursor to form the
initiator via intramolecular catalyst transfer. However, in the
current paper, the slow initiation in the polymerization of 3a is
attributed to the slow TM step for the formation of intermediate
14
walking process. 7c should exhibit similar polymerization acti-
vity to 7b. Both 7b and 7c prefer chain propagation with 3b due
to the large steric hindrance between two hexyl groups in the HH
coupling. 7a and 7d should be able to initiate the polymerization
of both 3a and 3b. However, the overall amount of 7a and 7d
should be much less than that of 7b and 7c. In addition, once 3b
is coupled with 7a and 7d, the resulting active centers are the
same as those of 7c and 7b. Therefore, growing chain ends be-
come identical in a short time, and the polymerization proceeds
predominantly with monomer 3b in the early stage of the co-
polymerization. With the decrease of the concentration of 3b, 3a
begins to take part in chain growth, leading to the formation of
HH coupling. This HH coupling is similar to the formation of the
initiator 7a and is a slow process. However, once the polymer
chain terminated with 3a unit is formed, this new active chain can
initiate the polymerization of 3a to afford block copolymer. Since
HH coupling is slow and chain propagation is fast, only part of
the P3HT chains from 3b can grow to block copolymers. There-
fore, as shown in Figures 2D and 3D, the shoulder peak for the
16/84 mixture of 3a and 3b and the new peak for the 50/50
mixture of 3a and 3b in the high molecular weight region of the
GPC profiles are attributed to the diblock copolymers compris-
ing poly(3a) and poly(3b) blocks, and the peaks in the low
molecular weight region should be ascribed to the 3b-dominated
polymers.
6
a due to the large steric hindrance (reductive elimination of 6a
and subsequent intramolecular catalyst transfer lead to bithio-
phene initiator 7a). Kiriy et al. also observed the slow initiation
phenomenon in the polymerization of 3a (containing ∼3% 3b)
with Ni(dppp)Cl or Ni(dppe)Cl as the catalyst in the absence
2
2
1
5
of LiCl. However, the mechanism is different. In their system,
the slow formation of the initiator 7d (as shown in Scheme 7) is
due to the low concentration of 3b.
The copolymerizations of 3a and 3b are more complicate
since four possible initiators (7aꢀd) may be involved, as shown
in Scheme 7. In order to qualitatively elucidate the probability of
the formation of these initiators, a model reaction as shown in
Scheme 8 was carried out. Ni(dppp)Cl (1 equiv) was reacted
2
with a mixture of 9a (5 equiv) and 9b (5 equiv), and the resulting
1
products were analyzed by GC-MS and H NMR. Three bith-
iophene derivatives (11aꢀc) were found in a ratio of 1:14:29
1
(
11a:11b:11c) by GC-MS and 1:13:26 (11a:11b:11c) by H
Scheme 8. Reaction of Ni(dppp)Cl with 9a and 9b
2
To provide more evidence for the aforementioned block
copolymerization mechanism, the copolymerization of 14 and
3
a was conducted (Scheme 9). As expected, the copolymeriza-
tion of monomer 14 (synthesized by Grignard metathesis of
Scheme 9. Copolymerization of 3a and 14 in the Presence of 1 equiv of LiCl
4
Figure 7. (A) GPC profiles of copolymerization mixture of 14 and 3a (M
n
= 1.1 ꢁ 10 , PDI = 2.38) and its two fractions (poly(14)-b-poly(3a): M
n
=
4
3
2
.1 ꢁ 10 , PDI = 1.55; poly(14): M = 4.8 ꢁ 10 , PDI = 1.21). (B) MALDI-TOF mass spectrum of poly(14); the inset is the expanded spectrum. The
n
2 0 0
polymerization was carried out at room temperature with 1 equiv of LiCl and 2 mol % Ni(dppp)Cl to the sum of 13 and 1. [1] = [13] = 0.05 mol/L.
7
565
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
i
PrMgCl and 13) and 3a gave a polymer with a bimodal GPC
of 3a occurs primarily after the consumption of 3b, yielding fully
conjugated block copolymers. This finding further deepens the
understanding of KCTP.
profile (Figure 7A), identical to that of the copolymerization
of 3a and 3b as shown in Figure 3D. Moreover, the polymeri-
zation of 3a began at ∼30 min, after 14 was almost consumed
(
Figure S11). The slow polymerization of 14 can be ascribed to
’
ASSOCIATED CONTENT
the larger steric hindrance of ethylhexyl than hexyl. According to
the mechanism shown in Scheme 7, the peaks at the shorter and
longer elution time in Figure 7A should correspond to the block
copolymer and monomer 14-dominated polymer, respectively,
which are named as poly(14)-b-poly(3a) and poly(14). The
branched alkyl group (ethylhexyl) endows poly(14) good solu-
bility in petroleum ether (PE). In contrast, poly(14)-b-poly(3a)
exhibits poor solubility in this solvent due to the strong inter-
molecular interaction of poly(3a) or P3HT block. Therefore,
poly(14)-b-poly(3a) and poly(14) can be easily separated by
washing the copolymerization mixture with PE, as shown in
Figure 7A.
S
Supporting Information. All experimental details, GPC
b
profiles of the polymerization mixture of 3a and 3b with
50 mol % catalyst, H NMR spectra of Grignard reagents
and 3,3 -dihexyl-2,2 -bithiophene, conversion versus time for
the copolymerization of 14 and 3a, and H NMR spectra of
the resulting polymers. This material is available free of charge
via the Internet at http://pubs.acs.org.
1
0
0
1
’ AUTHOR INFORMATION
Corresponding Author
*
E-mail: yhgeng@ciac.jl.cn.
The separated poly(14)-b-poly(3a) and poly(14) were char-
1
acterized by an H NMR spectrometer. As shown in Figure S12,
the ratio of poly(14) and poly(3a) blocks in poly(14)-b-poly-
’
ACKNOWLEDGMENT
(
1
3a) is 23:77, and poly(14) is indeed dominated by monomer
4. Poly(14) was further characterized by MALDI-TOF mass
This work is supported by National Basic Research Program of
China (973 Project, No. 2009CB623603) of Chinese Ministry of
Science and Technology and NSFC (Nos. 20921061, 20923003,
and 21074131).
spectrometer (Figure 7B). There are five series of peaks. For easy
analysis, the spectrum was expanded with a group of peaks
including one from each series. The peak with molecular weight
of 3967.6 is corresponding to H/Br end-capped poly(14)
(
194.35 ꢁ 20 + 1 + 79.9 = 3967.9), and the other four peaks
’
REFERENCES
with molecular weight of 3939.1, 3911.3, 3883.4, and 3855.6 can
be attributed to H/Br end-capped polymers consisting of poly-
(
3
1
4
(
(
(
1) Yokoyama, A.; Yokozawa,T. Macromolecules2007, 40, 4093–4101.
2) Yokozawa, T.; Yokoyama, A. Chem. Rev. 2009, 109, 5595–5619.
3) Osaka, I.; Mccullough, R. D. Acc. Chem. Res. 2008, 41,
14) and one to four 3a units (194.35 ꢁ 19 + 166.3 + 1 + 79.9 =
939.9, 194.35 ꢁ 18 + 166.3 ꢁ 2 + 1 + 79.9 = 3911.8, 194.35 ꢁ
7 + 166.3 ꢁ 3 + 1 + 79.9 = 3883.8 and 194.35 ꢁ 16 + 166.3 ꢁ
1202–1214.
(4) Geng, Y. H.; Huang, L.; Wu, S. P.; Wang, F. S. Sci. China, Ser. B:
Chem. 2010, 53, 1620–1633.
(
1
+ 1 + 79.9 = 3855.7), respectively. The H NMR spectrum of
5) Loewe, R. S.; Khersonsky, S. M.; McCullough, R. D. Adv. Mater.
poly(14)-b-poly(3a) indicates the presence of an H-end from 3a
Figure S13A). However, this type of chain-end was not ob-
1
999, 11, 250–253.
(
1
(6) Loewe, R. S.; Ewbank, P. C.; Liu, J. S.; Zhai, L.; McCullough,
served in the H NMR spectrum of poly(14) (Figure S13B),
implying that a small amount of 3a (∼5%) in poly(14) is incor-
porated in the initiation step or the polymerization process via
HT or TH coupling process. This is consistent with the proposed
mechanism shown in Scheme 7, and thus the bimodal GPC
profile resulted from the slow initiationꢀfast chain propagation
KCTP of the reversed monomer 3a instead of the termination
of poly(14) chain. In fact, when another portion of 14 was
added after the copolymerization of 14 and 3a (see Supporting
Information), the GPC peak with longer elution time (corres-
ponding to poly(14)) shifted to the high molecular weight region
R. D. Macromolecules 2001, 34, 4324–4333.
7) Miyakoshi, R.; Yokoyama, A.; Yokozawa, T. J. Am. Chem. Soc.
005, 127, 17542–17547.
8) Iovu, M. C.; Sheina, E. E.; Gil, R. R.; McCullough, R. D.
Macromolecules 2005, 38, 8649–8656.
(9) Jeffries-EL, M.; Sauv ꢀe , G.; McCullough, R. D. Adv. Mater. 2004,
16, 1017–1019.
(
2
(
(10) Doubina, N.; Ho, A.; Jen, A. K.-Y.; Luscombe, C. K. Macro-
molecules 2009, 42, 7670–7677.
(
11) Bronstein, H. A.; Luscombe, C. K. J. Am. Chem. Soc. 2009, 131,
2894–12895.
12) Achord, B. C.; Rawlins, J. W. Macromolecules 2009, 42,
634–8639.
13) Boyd, S. D.; Jen, A. K.-Y.; Luscombe, C. K. Macromolecules
009, 42, 9387–9389.
14) Tkachov, R.; Senkovskyy, V.; Komber, H.; Sommer, J.-U.; Kiriy,
1
8
2
(
(Figure S14), indicating the living nature of the polymer chain.
(
’
CONCLUSION
(
Literature studies of the KCTP of 5-bromo-2-chloromagne-
A. J. Am. Chem. Soc. 2010, 132, 7803–7810.
(15) Tkachov, R.; Senkovskyy, V.; Komber, H.; Kiriy, A. Macro-
molecules 2011, 44, 2006–2015.
sio-3-hexylthiophene (3a) showed that this “reversed” monomer
does not polymerize due to the difficulty of the formation of the
bithiophene initiator. Our current study indicates that LiCl can
promote the formation of the initiator permitting the polymer-
ization of 3a and that the polymerization exhibits living char-
acteristics. However, the formation of the initiator is much slower
due to the large steric hindrance between hexyl groups at the TM
step, leading to a slowinitiationꢀfast chain propagation polymeri-
zation process. As a result, the PDI of the obtained polymer is
rather large, and the molecular weight is higher than the theor-
etical value. In the copolymerization of 3a and 3b, the polymerization
(
16) Lohwasser, R. H.; Thelakkat, M. Macromolecules 2011, 44,
388–3397.
17) Lanni, E. L.; Locke, J. R.; Gleave, C. M.; McNeil, A. J. Macro-
molecules 2011, 44, 5136–5145.
18) Miyakoshi, R.; Shimono, K.; Yokoyama, A.; Yokozawa, T. J. Am.
Chem. Soc. 2006, 128, 16012–16013.
19) Yokoyama, A.; Kato, A.; Miyakoshi, R.; Yokozawa, T. Macro-
3
(
(
(
molecules 2008, 41, 7271–7273.
(20) Stefan, M. C.; Javier, A. E.; Osaka, I.; McCullough, R. D.
Macromolecules 2009, 42, 30–32.
7
566
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567

Macromolecules
ARTICLE
(
21) Wu, S. P.; Sun, Y. Q.; Huang, L.; Wang, J. W.; Zhou, Y. H.;
Geng, Y. H.; Wang, F. S. Macromolecules 2010, 43, 4438–4440.
22) Zhang, Y.; Tajima, K.; Hirota, K.; Hashimoto, K. J. Am. Chem.
Soc. 2008, 130, 7812–7813.
23) Miyakoshi, R.; Yokoyama, A.; Yokozawa, T. Chem. Lett. 2008,
7, 1022–1023.
24) Wu, S. P.; Bu, L. J.; Huang, L.; Yu, X. H.; Han, Y. C.; Geng,
Y. H.; Wang, F. S. Polymer 2009, 50, 6245–6251.
25) Ge, J.; He, M.; Qiu, F.; Yang, Y. L. Macromolecules 2010, 43,
422–6428.
26) Smeets, A.; Willot, P.; Winter, J. D.; Gerbaux, P.; Verbiest, T.;
Koeckelberghs, G. Macromolecules 2011, 44, 6017–6025.
27) Iovu, M. C.; Jeffries-EL, M.; Sheina, E. E.; Cooper, J. R.;
McCullough, R. D. Polymer 2005, 46, 8582–8586.
28) Radano, C. P.; Scherman, O. A.; Stingelin-Stutzmann, N.;
(
(
3
(
(
6
(
(
(
M €u ller, C.; Breiby, D. W.; Smith, P.; Janssen, R. A. J.; Meijer, E. W.
J. Am. Chem. Soc. 2005, 127, 12502–12503.
(
29) Iovu, M. C.; Jeffries-EL, M.; Zhang, R.; Kowalewski, T.;
McCullough, R. D. J. Macromol. Sci., Part A 2006, 43, 1991–2000.
30) Iovu, M. C.; Craley, C. R.; Jeffries-EL, M.; Krankowski, A. B.;
Zhang, R.; Kowalewski, T.; McCullough, R. D. Macromolecules 2007,
0, 4733–4735.
31) Boudouris, B. W.; Frisbie, C. D.; Hillmyer, M. A. Macromole-
cules 2008, 41, 67–75.
32) Senkovskyy, V.; Khanduyeva, N.; Komber, H.; Oertel, U.;
Stamm, M.; Kuckling, D.; Kiriy, A. J. Am. Chem. Soc. 2007, 129,
626–6632.
33) Khanduyeva, N.; Senkovskyy, V.; Beryozkina, T.; Bocharova,
(
4
(
(
6
(
V.; Simon, F.; Nitschke, M.; Stamm, M.; Gr €o tzschel, R.; Kiriy, A.
Macromolecules 2008, 41, 7383–7389.
(34) Khanduyeva, N.; Senkovskyy, V.; Beryozkina, T.; Horecha, M.;
Stamm, M.; Uhrich, C.; Riede, M.; Leo, K.; Kiriy, A. J. Am. Chem. Soc.
2
4
4
1
009, 131, 153–161.
35) Marshall, N.; Sontag, S. K.; Locklin, J. Macromolecules 2010,
3, 2137–2144.
36) Marshall, N.; Sontag, S. K.; Locklin Chem. Commun. 2011,
7, 5681–5689.
37) Lanni, E. L.; McNeil, A. J. J. Am. Chem. Soc. 2009, 131,
6573–16579.
(
(
(
(
(
38) Lanni, E. L.; McNeil, A. J. Macromolecules 2010, 43, 8039–8044.
39) Senkovskyy, V.; Sommer, M.; Tkachov, R.; Komber, H.; Huck,
W. T. S.; Kiriy, A. Macromolecules 2010, 43, 10157–10161.
(
(
40) Ashby, E. C. Pure Appl. Chem. 1980, 52, 545–569.
41) Krasovskiy, A.; Knochel, P. Angew. Chem., Int. Ed. 2004,
4
3, 3333–3336.
42) Ren, H.j; Krasovskiy, A.; Knochel, P. Org. Lett. 2004, 6,
215–4217.
43) Buttrus, N. H.; Eaborn, C.; El-Kheli, M. N. A.; Hitchcock, P. B.;
Smith, J. D.; Sullivan, A. C.; Tavakkoli, K. J. Chem. Soc., Dalton Trans.
988, 381–391.
44) Huang, L.; Wu, S. P.; Qu, Y.; Geng, Y. H.; Wang, F. S.
Macromolecules 2008, 41, 8944–8947.
45) Trznadel, M.; Pron, A.; Zagorska, M.; Chrzaszcz, R.;
Pielichowski, J. Macromolecules 1998, 31, 5051–5058.
(
4
(
1
(
(
(46) Gold, L. J. Chem. Phys. 1958, 28, 91–99.
(47) Johnson, A. F.; Mohsin, M. A.; Meszena, Z. G. J. Macromol. Sci.,
Rev. Macromol. Chem. Phys. 1999, C39, 527–560.
7
567
dx.doi.org/10.1021/ma201192p |Macromolecules 2011, 44, 7558–7567