Investigation of aromatic polyester synthesis by the chain-growth polycondensation method

Article abstract of DOI:10.1021/ma021449a

Synthesis of aromatic polyesters by chain-growth polycondensation of activated 4-hydroxybenzoic acid derivatives (1) was investigated. Model reactions of ester formation between active compounds (6 and 7) and phenolic nucleophile (8) gave the best result when 3-arylcarbonyl-2-benzothiazolone derivatives were used with tertiary amines in CH₂Cl₂. Therefore, we designed an synthesized 3-(4-hydroxy-3-octylbenzoyl)-2-benzothiazolone (1h) as a monomer. The M_n- and M_w/M_n-conversion curves of the polymerization of 1h with initiator suggested that transesterification between polymer and monomer occurred in the late stage of polymerization. Investigation of the effects of bases and initiators revealed that the transesterification at the initiator end was minimized by using 3-(4-benzoyloxybenzoyl)-2-benzothiazolone (7h) and diisopropylethylamine as an initiator and a base, respectively. When the ratio of initiator 7h to monomer 1h was 20 mol %, the polycondensation proceeded in a chain-growth manner to give aromatic polyester having a controlled molecular weight and a low polydispersity. However, the polymerization of 1h with 10 mol % or less of 7h could not be controlled and gave step-growth polymer as well as chain-growth polymer. The model reaction demonstrated that transesterification would occur at the polymer main chain even by use of weak tertiary amines as a base.

Full text of DOI:10.1021/ma021449a

4328
Macromolecules 2003, 36, 4328-4336
Investigation of Aromatic Polyester Synthesis by the Chain-Growth
Polycondensation Method
Ak ih ir o Yok oya m a ,† Ken -ich i Iw a sh ita ,† Kyota Hir a ba ya sh i,†
Ka zu sh ige Aiya m a ,† a n d Tsu tom u Yok oza w a *,†,‡
Department of Applied Chemistry, Kanagawa University, Rokkakubashi, Kanagawa-ku,
Yokohama 221-8686, J apan, and “Synthesis and Control”, PRESTO, J apan Science and
Technology Corporation (J ST), J apan
Received September 9, 2002; Revised Manuscript Received March 31, 2003
ABSTRACT: Synthesis of aromatic polyesters by chain-growth polycondensation of activated 4-hydroxy-
benzoic acid derivatives (1) was investigated. Model reactions of ester formation between active compounds
(6 and 7) and phenolic nucleophile (8) gave the best result when 3-arylcarbonyl-2-benzothiazolone
derivatives were used with tertiary amines in CH2Cl2. Therefore, we designed and synthesized 3-(4-
hydroxy-3-octylbenzoyl)-2-benzothiazolone (1h ) as a monomer. The Mn- and Mw/Mn-conversion curves
of the polymerization of 1h with initiator suggested that transesterification between polymer and monomer
occurred in the late stage of polymerization. Investigation of the effects of bases and initiators revealed
that the transesterification at the initiator end was minimized by using 3-(4-benzoyloxybenzoyl)-2-
benzothiazolone (7h ) and diisopropylethylamine as an initiator and a base, respectively. When the ratio
of initiator 7h to monomer 1h was 20 mol %, the polycondensation proceeded in a chain-growth manner
to give aromatic polyester having a controlled molecular weight and a low polydispersity. However, the
polymerization of 1h with 10 mol % or less of 7h could not be controlled and gave step-growth polymer
as well as chain-growth polymer. The model reaction demonstrated that transesterification would occur
at the polymer main chain even by use of weak tertiary amines as a base.
In tr od u ction
tions. Recently, we have developed chain-growth poly-
condensation methodology and synthesized aromatic
polyamides7 and polyethers8 with controlled molecular
weights and low polydispersities like living polymeri-
zation. In their syntheses, we took advantage of differ-
ent electron-donating abilities between the nucleophilic
site of monomer and the amide or ether linkage of
polymer propagating end. In this paper, we tried to
apply our chain-growth polycondensation to the syn-
thesis of poly(4-hydroxy-3-octylbenzoate) with a strategy
shown in Scheme 1. The aryl hydroxyl group of mono-
mer 1 having the leaving group X is deprotonated by a
base to afford 2. A strong electron-donating nature of
the phenoxide group of 2 reduces electrophilicity of the
active acyl moiety at the para position, and the reaction
between 2’s would be suppressed. If there is initiator 3
having the reactive acyl group in the reaction mixture,
the phenoxide group of 2 would react with 3 to give ester
4. Because this reaction changes the strong electron-
donating phenoxide group into the weak electron-
donating ester linkage, the electrophilicity of the acyl
group of 4 is much higher than that of 2. This change
of substituent effect will allow the phenoxide of 2 to
react with the acyl moiety of 4. In this way, the
deactivated acyl group in monomer is converted to the
reactive one by the reaction of monomer with the
polymer end group, which leads to chain-growth poly-
condensation.
A variety of methodologies have been developed for
the synthesis of poly(4-hydroxybenzoate) and its deriva-
tives because their unique properties such as liquid
crystallinity and morphology depend on the chemical
structure of monomers and the manner of polymeriza-
tion. The polyesters have been frequently synthesized
by molten polymerization of 4-acetoxybenzoic acid de-
rivatives.1 “Single-crystal” type poly(4-hydroxybenzoate)
was prepared by Economy using a solvent of high boiling
point, and the mechanism of polymerization and the
structure of polyester were discussed.2 Kricheldorf
investigated the condensations of various monomers for
polyesters and proposed a different mechanism and
structure.3 He also investigated the synthesis, structure,
and application of whisker of poly(4-hydroxybenzoate).4
It was also prepared by direct polycondensation of
4-hydroxybenzoic acid using a variety of condensing
reagents.4,5 Several substituted aromatic polyesters
have been synthesized in order to increase their solubil-
ity and exploit new characters such as liquid crystal-
linity.6
The Mn values of the polyesters were determined by
the 1H NMR analysis of the hydrolyzed compounds
because of very poor solubility of the polyesters,2,3 and
the Mn value of poly(4-hydroxybenzoate) prepared by
Kricheldorf was estimated to be 200 000 by this 1H NMR
method.3 On the other hand, the polydispersity of poly-
(4-hydroxybenzoate) derivatives has rarely been re-
ported. Since all of the polycondensation mentioned
above were step-growth polymerization, the synthesized
polyesters would have broad molecular weight distribu-
In the previous paper, we investigated the reaction
conditions for the chain-growth polycondensation of
4-trimethylsilyloxybenzoyl chloride using model reac-
tions.9 We attempted the chain-growth polycondensation
of this monomer under the optimized conditions, but
polymerization could not proceed in a chain-growth
manner because of unstable and highly reactive nature
of the monomer and of poor solubility of the polymer
obtained. Therefore, we first examined the leaving group
† Kanagawa University.
‡ J apan Science and Technology Corporation (J ST).
* To whom correspondence should be addressed. e-mail:
yokozt01@kanagawa-u.ac.jp.
10.1021/ma021449a CCC: $25.00 © 2003 American Chemical Society
Published on Web 05/13/2003

Macromolecules, Vol. 36, No. 12, 2003
Aromatic Polyester Synthesis 4329
Sch em e 1. Mech a n ism of Ch a in -Gr ow th
P olycon d en sa tion of 1
the reaction of 6h , 7h , and 8 are shown in Table 2.
When THF was used as a solvent, the reaction resulted
in low yield and moderate chemoselectivity. The reaction
in DMF, NMP, and CH2Cl2 gave quantitative yields of
the products. Since the selectivity in CH2Cl2 (9/10 )
14/86) was the highest of the three solvents, we next
studied the effect of base in CH2Cl2 (Table 3). The
reaction with pyridine resulted in low conversion of 8.
When 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) or 4-(di-
methylamino)pyridine (DMAP) was used, yields of the
products were very high but chemoselectivities were
low. Et3N, i-Pr2NEt, and K2CO3 gave products in high
yields and high selectivities. Of the three bases, the
reaction with i-Pr2NEt was the most selective. According
to these results, we concluded tentatively that a suitable
solvent and base for the chain-growth polycondensation
of 1 was CH2Cl2 and i-Pr2NEt, respectively.
Syn th esis of 1h . It was reported that unsubstituted
poly(4-hydroxybenzoate) had a poor solubility for or-
ganic solvents, but introduction of the alkyl substituents
to the aromatic ring increased its solubility.6 As poly-
amide synthesis we reported previously,7 we designed
a monomer having the octyl chain in order to gain high
solubility of polymer. Ballauf has synthesized 3-alkyl-
4-hydroxybenzoic acids by Fries rearrangement of ethyl
4-alkanoyloxybenzoate and subsequent Clemmensen
reduction.6a However, yields of Fries rearrangement
were not high, and harmful mercury was used in the
Clemmensen reduction. Therefore, we synthesized 1h
in another way (Scheme 3). 4-Hydroxybenzoic acid (11)
was brominated at the 3-position selectively,11 and then
the carboxylic acid 12 was converted to methyl ester 13.
We first protected the aryl hydroxyl group of 13 as the
benzyl ether, but this ether was not able to be cleaved
under standard hydrogenolysis conditions after intro-
duction of the octyl and 2-benzothiazolyl groups, and
deprotection under more vigorous conditions accompa-
nied hydrolysis of the 2-benzothiazolyl active amide
moiety. Therefore, we chose the p-methoxybenzyl (PMB)
ether as a protective group because this ether can be
easily removed by treatment with acids. PMB ether 14
was synthesized in acetone with PMB-Cl, K2CO3, and
18-crown-6. The octynyl group was introduced by the
Sonogashira reaction12 to afford 15, and subsequent
hydrolysis of the methyl ester gave carboxylic acid 16.
We first attempted to convert 16 into the acid chloride
with SOCl2 in order to introduce the 2-benzothiazolyl
group,10c but treatment of 16 with SOCl2 resulted in
deprotection of the PMB ether. The 2-benzothiazolyl
group was then introduced by condensation of 16 and
2-hydroxybenzothiazole at room temperature in the
presence of 1,3-dicyclohexylcarbodiimide (DCC) and
4-(dimethylamino)pyridine (DMAP). This reaction gave
the N-acyl product 17 in high yield without formation
of the O-acyl product. The triple bond was reduced by
catalytic hydrogenation in ethyl acetate-ethanol using
Pd/C under a H2 atmosphere, but the PMB ether could
not be cleaved under this condition, and addition of
aqueous acetic acid to the reaction mixture was inef-
fective, too. Treatment of 18 with ammonium cerium-
(IV) nitrate (CAN) in CH3CN-H2O13 resulted in cleav-
age of not only the PMB ether but also the 2-benzo-
thiazoyl moiety, and treatment with 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (DDQ)14 gave a complex mix-
ture. At last, we found that trifluoroacetic acid (TFA)
(X) of monomer 1 by model reactions. Then we synthe-
sized a monomer having an appropriate leaving group
and a linear alkyl chain on the aryl moiety and
investigated its chain-growth polycondensation.
Resu lts a n d Discu ssion
Mod el Rea ction s. To carry out the polymerization
of 1 in a chain-growth manner, the activated acyl group
(-COX) in 1 must have an adequate reactivity not only
to prevent the reaction between 2’s (resulting in step-
growth polymerization) but also to undergo the reaction
between 2 and 4 or 5 (resulting in chain-growth polym-
erization). As mentioned above, acid chloride as an
active acyl group was too reactive to prevent step-
growth polymerization. Furthermore, electrophilicity of
the active acyl group of 1 should be higher than that of
the aryl ester linkage of polymer formed. Considering
these factors, we chose eight leaving groups and studied
their abilities for chain-growth polycondensation by
model reactions (Scheme 2). We used 6a -h as models
of the electrophilic site of monomer, 7a -h as those of
the electrophilic site of the propagating end of polymer,
and 8 as a model of the nucleophilic site of monomer.
This reaction will give two products. Product 9 is formed
by the reaction between monomer electrophilic model
6 and its nucleophilic model 8 and corresponds to a
“step-growth polymerization” product. On the other
hand, the reaction of 8 with propagating end model 7
gives 10, which corresponds to a “chain-growth polym-
erization” product. Therefore, the leaving group X which
gives 10 most selectively should be the best one for
chain-growth polycondensation. A mixture of equimolar
amounts of 6, 7, and 8 was treated with Et3N in
CH2Cl2 at room temperature for 48 h, and the chemose-
lectivity of 8 toward 6 or 7 was evaluated (Table 1).
Trichloroethyl (a ) and 2,4,6-trichlorophenyl esters (b)
did not react under this condition at all, but 2,4-
dinitrophenyl ester (c) gave 10 selectively (9/10 )
17/83) in high yield. The reaction of thiolester (d )
proceeded much slower, but introduction of the nitro
group at the para position (e) accelerated the reaction.
When X’s were benzothiazole derivatives and benzothi-
azolone (f, g, and h ),10 3-acyl-2-benzothiazolone (h ) gave
the ester products quantitatively and most selectively
of all the eight leaving groups we examined.
According to the above results, we decided 2-ben-
zothiazolone (h ) as a leaving group of monomer and
optimized reaction conditions. The effects of solvents on
15
in CH2Cl2 could cleave the PMB ether of 18 to give
1h in 66% yield.

4330 Yokoyama et al.
Macromolecules, Vol. 36, No. 12, 2003
Sch em e 2. Mod el Rea ction for Lea vin g Gr ou p
Ta ble 1. Effect of Lea vin g Gr ou p in th e Rea ction of 6, 7,
a n d 8a
indicate that these polymerizations proceeded in a
typical step-growth polycondensation manner; the in-
crease of Mn and Mw/Mn was slow up to the middle
stage of polymerization but accelerated in the last stage.
The two bases gave similar polyesters, whose Mn and
Mw/Mn were 1.7 × 104 and 2.5-2.7, respectively. The
obtained polyester was easily soluble in CHCl3 and
CH2Cl2, modestly in THF and CCl4, and insoluble in
hydrocarbons (hexane, cyclohexane, and toluene), apro-
tic polar solvents (DMF, DMSO, and NMP), alcohols
(MeOH and EtOH), and other solvents such as CH3CN,
acetone, and diethyl ether.
Ch a in -Gr ow th P olycon d en sa tion of 1h fr om 19.
We then investigated the chain-growth polycondensa-
tion of 1h with 5 mol % of 3-(4-nitrobenzoyl)-2-ben-
zothiazolone (19) as an initiator (Scheme 4). 19 has the
strong electron-withdrawing nitro group at the para
position to the active acyl group, whose electrophilicity
would be much higher than that of monomer 1h . The
effect of bases on the polymerization are shown in Table
4. CH2Cl2 was used as a solvent in the polymerization
with amines (entries 1-4) because of the results of the
model reactions and of dissolving freely the polyester
formed. The reaction was continued until monomer was
consumed completely. The polymerization with 1,5-
diazabicyclo[4.3.0]non-5-ene (DBN) and DBU gave poly-
esters with higher Mn than the calculated one (Mn(calcd)
) 4900) and with high polydispersity. However, Et3N
or i-Pr2NEt showed better results: the Mn was close to
the calculated one, and the Mw/Mn was narrower than
others. These results were consistent with those of
model reactions shown in Table 3. Considering the
basicity of tertiary amines such as Et3N and i-Pr2NEt,
the phenol moiety of 1h would not be deprotonated
completely by these bases. As shown in Scheme 1, it
seemed that the acyl group of nondeprotonated 1h was
not deactivated so strongly as that of deprotonated
1h (that is, 2 in Scheme 1) and that the reaction
between 1h ’s each other might occur, resulting in rather
broad molecular weight distributions. We then used
stronger bases such as Et3SiH/CsF/18-crown-6 and
t-BuOK/18-crown-6. THF was used as a solvent de-
spite a modest solubility of the polymer in THF, because
CH2Cl2 was not thought to be inert for such strong
bases. The polymerization with these stronger bases
proceeded faster than those with the amines but gave
poor results like the polymerization with DBN and
DBU; the obtained polymer had high polydispers-
ities and much higher Mn values than the calculated
one.
a
The reaction was carried out in CH2Cl2 ([6]0 ) [7]0 ) [8]0
)
b
0.02 M) with Et3N at room temperature for 48 h. Determined
by GC.
Ta ble 2. Solven t Effect in th e Rea ction of 6h , 7h , a n d 8a
conv of
yield of
entry
solvent
8 (%)b
9 + 10 (%)b
9/10b
1
2
3
4
THF
26
100
100
100
21
100
97
24/76
15/85
17/83
14/86
DMF
NMP
CH2Cl2
100
a
The reaction was carried out with Et3N ([6h ]0 ) [7h ]0 ) [8]0
b
) 0.02 M) at room temperature for 48 h. Determined by GC.
Ta ble 3. Effect of Ba se in th e Rea ction of 6h , 7h a n d 8a
conv of
yield of
entry
base
8 (%)b
9 + 10 (%)b
9/10b
1
2
3
4
5
6
pyridine
DBU
DMAP
Et3N
i-Pr2NEt
K2CO3
20
100
100
100
100
100
17
98
91
100
96
90
30/70
51/49
23/77
14/86
12/88
17/83
a
The detailed behavior in polymerization with Et3N
and i-Pr2NEt is shown in Figures 2 and 3. The time-
conversion curves (Figure 2) show that the polymeri-
zation with i-Pr2NEt is faster than that with Et3N. The
conversion-Mn and -Mw/Mn relationships (Figure 3)
indicate that, in contrast to polycondensation in a step-
growth manner shown in Figure 1, the Mn increased in
proportion to conversion despite a slight difference from
the calculated value and the Mw/Mn ratios kept less than
The reaction was carried out in CH2Cl2 ([6h ]0 ) [7h ]0 ) [8]0
b
) 0.02 M) at room temperature for 48 h. Determined by GC.
Step -Gr ow th P olycon d en sa tion of 1h . To examine
the ability of polymerization of 1h and the solubility of
the polyester formed, the polymerization of 1h was
carried out with Et3N or i-Pr2NEt as a base in CH2Cl2
for 72 h at room temperature (conversion ) 97-100%).
Conversion-Mn and -Mw/Mn curves shown in Figure 1

Macromolecules, Vol. 36, No. 12, 2003
Aromatic Polyester Synthesis 4331
Sch em e 3. Syn th esis of 1h a
a
Reagents and conditions: (a) Br2, 1,4-dioxane, rt, 1 day; (b) H2SO4, MeOH, reflux, 2.5 h; (c) PMB-Cl, 18-crown-6, K2CO3,
acetone, reflux, 12 h; (d) 1-octyne, Pd2(dba)3, CuI, PPh3, Et3N, THF, 80 °C in a sealed tube, 37 h; (e) KOH, EtOH-H2O, reflux, 0.5
h; (f) 2-hydroxybenzothiazole, DCC, DMAP, CH2Cl2, rt, 1 h; (g) H2, Pd-C, EtOH-AcOEt, rt, 3 days; (h) CF3CO2H, CH2Cl2, rt, 3
h.
the same conditions as polymerization. Results are
shown in Table 5. Transesterification proceeded quickly
in the case of DBN, DBU, Et3SiH/CsF/18-crown-6, and
t-BuOK/18-crown-6 but did slowly in the case of Et3N
and i-Pr2NEt. These results agree with the effect of
bases on polymerization shown in Table 4; that is,
polymerization with Et3N or i-Pr2NEt gave better
results than those with other bases.
In vestiga tion of In itia tor . The transesterification
at the ester linkage between initiator unit and monomer
F igu r e 1. Mn (circle) and Mw/Mn (triangle) values of poly1h ,
obtained with (a) Et3N and (b) i-Pr2NEt in CH2Cl2 at room
unit is thought to be reduced by decreasing the electro-
philicity of the ester carbonyl carbon at this position.
We therefore designed two other initiators 22 and 7h
which have hydrogen and the electron-donating ben-
zoyloxy substituent instead of the nitro group of 19,
respectively. Especially, 7h is a dimer model of mono-
mer. The effect of these initiators on polymerization of
1h was evaluated (Scheme 7). The Mn of polymer was
estimated by 1H NMR, and the results are shown in
Table 6. While the polymerization using 19 or 22 as an
initiator resulted in smaller Mn than the calculated one,
the initiation by 7h gave a better result; the Mn of
polymer was close to the calculated one, and its poly-
dispersity was the lowest. These results show that 7h ,
which has the weak electron-donating benzoyloxy group
at the para position to the active ester group, has
enough reactivity to initiate polymerization and suggest
that such decrease of electrophilicity by a weak electron-
donating substituent is necessary in order to prevent
the transesterification at the linkage between initiator
and monomer units.
temperature, as a function of monomer conversion: [1h ]0
[Et3N]0 ) [i-Pr2NEt]0 ) 0.10 M.
)
1.3 up to 50% conversion in polymerization with both
bases. These results suggest that chain-growth poly-
condensation is a dominant reaction in the low conver-
sion stage. However, in the higher conversion stage,
increase of the Mn was once retarded and then both the
Mn and Mw/Mn were accelerated, implying that some
side reactions occurred in this stage.
The efficiency of initiator was checked by the model
reaction using initiator 19 and nucleophilic monomer
model 8 in the presence of Et3N or i-Pr2NEt as a base
(Scheme 5). The reaction proceeded rapidly at the same
rate irrespective of the kind of bases (Figure 4). The
yield of 20 was about 80% only in 5 min and quantita-
tive in 2 h. Therefore, we concluded that acceleration
of the Mn and increase of the polydispersity in the late
stage were caused not by the initiation step but by other
factor concerning polymerization process.
Tr a n sester ifica tion a t In itia tor En d . We assumed
tentatively that the curious behavior observed in Figure
3 might be caused by transesterification among mono-
mer, oligomer, and polymer. Because the ester linkage
formed in polymerization is a kind of aryl ester, this
linkage has a moderate reactivity for nucleophiles.
Especially, because of the strong electron-withdrawing
nature of the nitro group of initiator 19, electrophilicity
of the ester carbonyl carbon at the initiator unit in
polymer would be enhanced, and transesterification
would take place at this position, which might cause
deviations from the chain-growth polycondensation
manner we designed. Therefore, we examined the
possibility of the transesterification at the initiator unit
by a model reaction (Scheme 6). We used 21 as a
polymer initiator unit model and 8 as a nucleophilic
monomer model, and the reaction was carried out under
Ch a in -Gr ow th P olycon d en sa tion of 1h fr om 7h .
Using 7h as an initiator, 1h was polymerized with
i-Pr2NEt in CH2Cl2 at room temperature. The ratio of
chain-growth polymers to total polymers was estimated
by the 1H NMR spectrum. Figure 5 shows 1H NMR
spectra of (a) step-growth polymer (SP) prepared by the
polymerization of 1h without initiator and (b) polymer
obtained by the polymerization of 1h with initiator 7h ,
which was a mixture of SP and chain-growth polymer
(CP) propagated from 7h . The resonance of the aromatic
proton Ha placed between the octyl chain and the
carbonyl group of the propagating end unit of SP is the
same position (7.86 ppm) as that of CP. The phenol end
of the step-growth polymer gave unique signals at 8.01
(Hb), 7.98 (Hc), and 6.86 ppm (Hd). The ratio of chain-
growth polymerization (designated as “CP/(CP + SP)”)

4332 Yokoyama et al.
Macromolecules, Vol. 36, No. 12, 2003
Sch em e 4. P olym er iza tion of 1h w ith 19
Ta ble 4. Effect of Ba se in P olym er iza tion of 1h w ith 19a
[1h ]0/[7h ]0 ) 5 (Figure 6a), the Mn increased linearly
in proportion to conversion with low polydispersity
(Mw/Mn e 1.28), and this polymerization would progress
from initiator 7h because CP/(CP + SP) was always 1.0.
Although the Mn values were a little higher than the
calculated ones, these results indicate that the polym-
erization at this feed ratio proceeded in a chain-growth
manner. On the other hand, when [1h ]0/[7h ]0 g 10
(Figure 6b,c), the Mn-conversion curves were shown
like the polymerization with 19 in Figure 3; increase of
the Mn was reduced in the middle stage of polymeriza-
tion, and acceleration of the Mn and increase of the
Mw/Mn occurred in the late stage of polymerization.
Furthermore, deviations of CP/(CP + SP) from 1.0,
which were ascribed to the increased extent of step-
growth polymerization, became larger as deviations of
the Mn from the calculated values were large. These
tendencies enhanced as [1h ]0/[7h ]0 increased from 10
(Figure 6b) to 20 (Figure 6c). The results mentioned
above suggest that transesterification between monomer
and polymer chain occurred by the attack of the
monomer phenoxy moiety to the ester carbonyl carbon
of polymer main chain in the middle stage of polymer-
ization. The N-acyl-2-benzothiazolone moiety of polymer
end group is indeed a much better leaving group than
the aryl ester of polymer chain, and nucleophilic attack
of monomer occurred selectively at the polymer propa-
gating end rather than at the polymer chain in the low
conversion stage. In the higher conversion stage where
concentration of the polymer propagating ends was
much lower than that of the ester linkages in polymer
chain, however, the dominant concentration of the
polymer ester linkage would overcome the higher reac-
tivity of the polymer propagating end. Consequently, the
nucleophilic phenoxide of monomer might attack the
phenyl ester moiety of polymer chain, and the Mn might
not increase regardless of monomer consumption. This
transesterification would not only give “step-growth
polymer” bearing no initiator end but also cause slow-
down of increase of the Mn since such a transesterifi-
cation would decrease the molecular weight of the
polymer. In the last stage of polymerization, the step-
growth polymers would react with each other to give
polymer with high molecular weight, which led steep
increase of Mn.
time
solvent (h)b
Mn
Mw/Mn
c,d
c
entry
base
1
2
3
4
5
6
DBN
DBU
Et3N
i-Pr2NEt
Et3SiH/CsF/18-crown-6 THF
t-BuOK/18-crown-6 THF
CH2Cl2
CH2Cl2 48 15 300
CH2Cl2 36
CH2Cl2
2
9 300
2.69
2.87
1.54
1.43
2.67
2.75
5 700
4 600
15 200
10 900
6
3
3
a
The reaction was carried out with base (1.0 equiv) and 19 (5
mol %) in CH2Cl2 (entries 1-4, [1h ]0 ) 0.10 M) or THF (entries 5
b
and 6, [1h ]0 ) 0.17 M) at room temperature. Conversion ) 100%.
d
c Determined by GPC based on polystyrene standards. Mn(calcd)
) 4900.
F igu r e 2. Time-conversion curves for the polymerization of
1h in the presence of 19 (5 mol %) with Et3N (circle) and
i-Pr2NEt (square) in CH2Cl2 at room temperature: [1h ]0
[Et3N]0 ) [i-Pr2NEt]0 ) 0.10 M.
)
F igu r e 3. Mn (circle) and Mw/Mn (triangle) values of poly1h ,
obtained in the presence of 19 (5 mol %) with (a) Et3N and (b)
i-Pr2NEt in CH2Cl2 at room temperature, as a function of
monomer conversion: [1h ]0 ) [Et3N]0 ) [i-Pr2NEt]0 ) 0.10 M.
was calculated by the integral values of Ha (designated
as “Int[Ha]”) and Hd (Int[Hd]) in the following equation:
CP/(CP + SP) ) 1 - Int[Hd]/Int[Ha]
Tr a n sest er ifica t ion a t P olym er Ma in Ch a in .
Transesterification at the polymer backbone was studied
by a model reaction with various bases (Scheme 8).
Triester 23 was used as a polymer chain model, and this
compound has two aryl ester moieties (“a ” and “b”). The
carbonyl carbon “a ” is connected with two groups. One
The polymerizations were carried out with varying
feed ratio of [1h ]0/[7h ]0 ) 5, 10, and 20. The results of
the polymers obtained are listed in Table 7, and the
Mn, Mw/Mn, and CP/(CP + SP) as a function of mono-
mer conversion are shown in Figure 6. In the case of

Macromolecules, Vol. 36, No. 12, 2003
Aromatic Polyester Synthesis 4333
Sch em e 5. Mod el Rea ction for In itia tion
Sch em e 6. Mod el Rea ction of Tr a n sester ifica tion a t In itia tion En d
Sch em e 7. P olym er iza tion of 1h w ith Va r iou s In itia tor s
Ta ble 6. Effect of In itia tor in P olym er iza tion of 1h a
Mn
d
entry
initiator
timeb (h)
calcd
obsdc
Mw/Mn
1
2
3
19
22
7h
16
30
24
4957
4754
5071
3930
3630
4840
1.49
1.46
1.40
a
The reaction was carried out with i-Pr2NEt (1.0 equiv) and
initiator (5 mol %) in CH2Cl2 ([1h ]0 ) 0.10 M) at room temperature.
b
d
Conversion ) 100%. c Determined by 1H NMR. Determined by
GPC based on polystyrene standards.
frequently at the “b” position to give 26 as a exchanged
product and 27 as a leaving phenol. The reactivity at
the “a ” position would reflect the electrophilicity of the
ester moiety of the polyester backbone we synthesized,
and transesterification at the “b” position would cor-
respond to that at the initiator end of polyester when
22 or 7h is used as an initiator.
F igu r e 4. Yields of 20 in the reaction of 19 and 8 with Et3N
(filled cube) or i-Pr2NEt (open circle) in CH2Cl2 at room
temperature: [19]0 ) [8]0 ) 83 mM.
Ta ble 5. Tr a n sester ifica tion of 21 by 8a
However, the results of reaction were not so simple;
not only 24-27 but also other products including
oligomers were formed. This is because 24 and 26
having the aryl ester group could be attacked by the
phenol moiety of 8, 25, or 27, which led further trans-
esterification. Indeed, the total yields of 24 and 26 were
always smaller than conversions of 23 or 8. Since the
yields of 24 and 26 would not reflect the reactivities of
the “a ” and “b” position, respectively, the effect of base
on the transesterification was estimated by the conver-
sion of 23 and 8.
conv of
yield of
entry
base
solvent
8 (%)b
20 (%)b
1
2
3
4
5
6
DBN
DBU
Et3N
i-Pr2NEt
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
16.4
32.4
5.9
1.6
23.1
10.3
16.4
28.2
3.0
1.2
19.6
7.1
Et3SiH/CsF/18-crown-6
t-BuOK/18-crown-6
THF
a
The reaction was carried out with base (1.0 equiv) in CH2Cl2
(entries 1-4, [21]0 ) [8]0 ) 0.10 M) or THF (entries 5 and 6, [21]0
b
) [8]0 ) 0.17 M) at room temperature for 1 h. Determined by
GC.
Table 8 shows the conversion of 23 and 8 in the model
transesterification reaction using equimolar amounts of
23 and 8 in the presence of a variety of bases. In a
similar manner of the transesterification of initiator end
model 21 by 8 as shown in Scheme 6 and Table 5,
consumption of 23 was faster in the reaction with DBN,
DBU, Et3SiH/CsF/18-crown-6, and t-BuOK/18-crown-6,
and a considerable amount of 23 reacted even in a short
time (2 h). The results suggest that polymerization with
these bases causes transesterification readily to yield
polymers having the phenoxide moiety and the active
acyl group at each ends, followed by step-growth po-
lymerization to give polymer with a high molecular
is the aryloxy group substituted by the electron-
withdrawing ester group at the para position, and the
other is the aromatic ring having the electron-donating
acyloxy group at the para position. Nucleophilic reaction
by 8 at the “a ” position affords ester 24 and phenol 25
as an exchanged product and a leaving component,
respectively. The circumstance of the carbonyl carbon
“b” is similar to that of “a ”, but the aryl group connect-
ing to the carbonyl carbon “b” does not have the
electron-donating acyloxy substituent. Therefore, elec-
trophilicity of the carbonyl carbon “b” would be higher
than that of “a ”, and transesterification may occur more

4334 Yokoyama et al.
Macromolecules, Vol. 36, No. 12, 2003
F igu r e 5. 1H NMR spectra of the polyesters obtained by the polymerization of 1h in the (a) absence and (b) presence of initiator
7h (500 MHz, CDCl3).
Ta ble 7. P olym er iza tion of 1h in th e P r esen ce of 7h a
better results than that with 19. When the polymeri-
zation of 1h was carried out in the presence of 20 mol
% of 7h , it proceeded in a chain-growth manner to give
polymer with a controlled molecular weight and a low
polydispersity. However, the polymerization with 10 mol
% or less of 7h could not be controlled to give a certain
amount of step-growth polymer as well as chain-growth
polymer. Analysis of the polymerization behavior and
the results of model reactions suggest that transesteri-
fication at the polymer main chain interfered with
chain-growth polycondensation. We are currently in-
vestigating the reaction conditions preventing the trans-
esterification so as to synthesize aryl polyester with a
higher molecular weight and a low polydispersity by
chain-growth polycondensation.
Mn
entry [1h ]0/[7h ]0 (h)b calcd obsdc Mw/Mn
feed ratio time
CP/(CP +
d
SP) (%)c,e
1
2
3
5
10
20
24
30
30
1546 1950
2699 3000
5071 5810
1.28
1.30
1.45
100
92
69
a
The polymerization was carried out in the presence of 7h with
i-Pr2NEt (1.0 equiv) in CH2Cl2 ([1h ] ) 0.10 M) at room temper-
b
d
ature. Conversion ) 100%. c Determined by 1H NMR. Deter-
mined by GPC based on polystyrene standards. e Ratio of the
number of chain-growth polymers to that of total polymers.
weight and a broad molecular weight distribution. This
assumption agrees with the results of polymerization
of 1h with 19 shown in Table 4 in which reactions with
these bases gave polymers with a high molecular weight
and a broad molecular weight distribution. On the
other hand, transesterifications with Et3N or i-Pr2NEt
were very slow. Trace amounts of 23 and 8 were con-
sumed for 2 h (entries 3 and 5). When the reaction was
prolonged to 24 h, the conversions of 23 and 8 increased
up to 3-4% (entries 4 and 6). These results suggest
that transesterification indeed occur even by use of
weak tertiary amines such as Et3N and i-Pr2NEt and
is one of the factors of difficulty in achieving the chain-
growth polycondensation of 1h from a small amount of
initiators.
Con clu sion s. Model reactions of esterification re-
vealed that 2-benzothiazolone was the most suitable
leaving group for the synthesis of aryl polyesters by
chain-growth polycondensation, and we synthesized 1h
and investigated its polymerization. When 19 was used
as an initiator, the most suitable base was i-Pr2NEt,
but the polymerization did not proceed in a chain-
growth manner completely. The results of model reac-
tion suggest that transesterification occurred at the
initiator end. We examined the effect of initiators to find
that the polymerization with 7h as an initiator gave
Exp er im en ta l Section
Gen er a l. 1H and 13C NMR spectra were obtained on J EOL
A-500, EX-270, and FX-200 spectrometers using tetrameth-
ylsilane as an internal standard. IR spectra were recorded on
a J ASCO FT/IR-410. The Mn and Mw/Mn of polymers were
measured with a TOSOH HLC-8120 gel-permeation chroma-
tography (GPC) unit (eluent: tetrahydrofuran (THF); calibra-
tion: polystyrene standards) using two TSK-gel columns (2 ×
Multipore HXL-M). Isolation of polyesters was carried out with
a J apan Analytical Industry LC-908 recycling preparative
HPLC (eluent: CHCl3) using two TSK-gel column (2 ×
G2000HHR). All melting points were measured with a Yanagim-
oto hot stage melting point apparatus and were uncorrected.
Column chromatography was performed on silica gel (Kieselgel
60, 230-400 mesh, Merck) with a specified solvent. Com-
mercially available (Kanto) dehydrated dichloromethane
(CH2Cl2) and tetrahydrofuran (THF, stabilizer free) were used
as dry solvents. N,N-Dimethylformamide (DMF), 1-methyl-2-
pyridone (NMP), pyridine, triethylamine (Et3N), N,N-diiso-
propylethylamine (i-Pr2NEt), 1,8-diazabicyclo[5.4.0]undec-7-
ene (DBU), 1,5-diazabicyclo[4.3.0]non-5-ene (DBN), and triethyl-
silane (Et3SiH) were distilled by standard procedure.16 4-(Di-
methylamino)pyridine (DMAP), cesium fluoride (CsF), and
potassium tert-butoxide (t-BuOK) were used as received. 18-

Macromolecules, Vol. 36, No. 12, 2003
Aromatic Polyester Synthesis 4335
F igu r e 6. Mn (filled circle), Mw/Mn (filled triangle), and CP/(CP + SP) (open circle) values of poly1h , obtained in the presence of
7h with i-Pr2NEt at room temperature, as a function of monomer conversion: [1h ]0 ) [i-Pr2NEt]0 ) 0.10 M, [1h ]0/[7h ]0 ) (a) 5,
(b) 10, and (c) 20.
Sch em e 8. Mod el Rea ction of Tr a n sester ifica tion a t
P olym er Ch a in
by a syringe via the three-way stopcock with a stream of N2
during a certain period and analyzed by HPLC to determine
the degree of consumption of 1h and by GPC to determine Mn
and Mw/Mn of polymer formed. As 1h disappeared, 1 M
hydrochloric acid was added to the reaction mixture, which
was extracted with CH2Cl2. The organic layer was dried over
anhydrous MgSO4 and concentrated under reduced pressure.
The residue was purified by preparative HPLC (eluent: CHCl3)
1
using polystyrene gel column to give step-growth poly1h . H
NMR (500 MHz, CDCl3) δ: 8.20-8.17 (m, 2 H), 7.37 (br s, 1
H), 2.70 (br s, 2 H), 1.70 (br s, 2 H), 1.37-1.26 (m, 10 H), 0.87
(br s, 3 H). IR (KBr): 3455, 2925, 1742, 1587, 1498, 1246, 1167,
1090 cm-1
.
Ch a in -Gr ow th P olycon d en sa tion of 1h . Meth od
A
(Ba se ) Et3N, i-P r 2NEt, DBU, or DBN). A solution of 1h
(0.192 g, 0.500 mmol), initiator (0.025 mmol), and naphthalene
(internal standard for HPLC analysis, 64 mg, 0.50 mmol) in
dry CH2Cl2 (4 mL) was placed in a round-bottomed flask
equipped with a three-way stopcock under an Ar atmosphere.
Into the flask was added at room temperature a solution of a
base (0.50 mmol) in dry CH2Cl2 (1 mL) via a syringe from the
three-way stopcock with a stream of N2, and the mixture was
stirred at room temperature. A small portion of the reaction
mixture was withdrawn by a syringe during a certain period
and analyzed by HPLC to determine the degree of consumption
of 1h and by GPC to determine Mn and Mw/Mn of polymer
formed. When initiator was 7h , Mn was determined by 1H
NMR. As 1h disappeared, 1 M hydrochloric acid was added to
the reaction mixture, which was extracted with CH2Cl2. The
organic layer was dried over anhydrous MgSO4 and concen-
trated under reduced pressure. The residue was purified by
preparative HPLC (eluent: CHCl3) using a polystyrene gel
column.
Ta ble 8. Tr a n sester ifica tion of 23 by 8a
conv (%)b
time
entry
base
solvent
(h)
23
82
42
0.2
8
1
2
3
4
5
6
7
8
DBN
DBU
Et3N
CH2Cl2
CH2Cl2
CH2Cl2
2
2
2
75
50
0.6
24
2
4
3
i-Pr2NEt
CH2Cl2
0.2
3
0.4
3
24
2
2
Et3SiH/CsF/18-crown-6 THF
t-BuOK/18-crown-6 THF
94
92
76
58
a
The reaction was carried out with base (1.0 equiv) in CH2Cl2
(entries 1-6, [1h ]0 ) 0.10 M) or THF (entries 7 and 8, [1h ]0
)
Meth od B (Ba se ) Et3SiH/CsF /18-Cr ow n -6). CsF (0.076
g, 0.50 mmol) was placed in a round-bottomed flask equipped
with a three-way stopcock and dried at 250 °C under reduced
pressure for 20 min. The flask was cooled to room temperature
under an Ar atmosphere. Into the flask was added at room
temperature a solution of 1h (0.192 g, 0.500 mmol), initiator
(0.025 mmol), Et3SiH (0.058 g, 0.50 mmol), and naphthalene
(internal standard for HPLC analysis, 64 mg, 0.50 mmol) in
dry THF (2.5 mL) and then a solution of 18-crown-6 (0.264 g,
1.00 mmol) in THF (0.5 mL) successively via syringes from
the three-way stopcock with a stream of N2, and the mixture
was stirred at room temperature. Analysis and workup of the
reaction were carried out in the same manner as method A.
b
0.17 M) at room temperature. Determined by GC.
Crown-6 was dried under reduced pressure at 40 °C for more
than 6 h prior to use.
Mod el Rea ction for Lea vin g Gr ou p . Typ ica l P r oce-
d u r e for th e Rea ction of 8 w ith 6 a n d 7. To a solution of 8
(0.076 g, 0.50 mmol), Et3N (0.08 mL, 0.6 mmol), and phenan-
threne (internal standard for GC analysis, 0.36 g, 0.20 mmol)
in dry CH2Cl2 (2 mL) was added a solution of 6 (0.5 mmol)
and 7 (0.5 mmol) in dry CH2Cl2 (19 mL). The mixture was
stirred at room temperature for 48 h, and the conversion of 8
and the yields of 9 and 10 were determined by GC analysis of
the reaction mixture.
Step -Gr ow th P olycon d en sa tion of 1h . A solution of 1h
(0.192 g, 0.500 mmol) and naphthalene (internal standard for
HPLC analysis, 64 mg, 0.50 mmol) in dry CH2Cl2 (5 mL) was
placed in a round-bottomed flask equipped with a three-way
stopcock under an Ar atmosphere. Into the flask was added
Et3N (0.070 mL, 0.50 mmol) or i-Pr2NEt (0.087 mL, 0.50 mmol)
via a syringe from the three-way stopcock with a stream of
N2. A small portion of the reaction mixture was withdrawn
Meth od C (Ba se ) t-Bu OK/18-Cr ow n -6). A solution of
t-BuOK (0.056 g, 0.50 mmol) and 18-crown-6 (0.132 g, 0.500
mmol) in THF (1.5 mL) was placed in a round-bottomed flask
equipped with a three-way stopcock under an Ar atmosphere.
Into the flask was added at room temperature a solution of
1h (0.192 g, 0.500 mmol), initiator (0.025 mmol), and naph-
thalene (internal standard for HPLC analysis, 64 mg, 0.50
mmol) in dry THF (1.5 mL) via a syringe from the three-way
stopcock with a stream of N2, and the mixture was stirred at

4336 Yokoyama et al.
Macromolecules, Vol. 36, No. 12, 2003
(2) Economy, J .; Storm, R. S.; Matkovich, V. I.; Cottis, S. G.;
Nowak, B. E. J . Polym. Sci., Polym. Chem. Ed. 1976, 14,
2207-2224.
(3) Kricheldorf, H. R.; Schwarz, G. Makromol. Chem. 1983, 184,
475-496.
(4) For example, see: Kricheldorf, H. R.; Schwarz, G.; Ruhser,
F. Macromolecules 1991, 24, 3485-3488.
(5) See ref 3 and references therein.
(6) (a) Stern, R.; Ballauff, M.; Lieser, G.; Wegner, G. Polymer
1991, 32, 2096-2105. (b) Kalyvas, V.; Li, Z. G.; Mclntyre, J .
E.; Milburn, A. H.; Tomka, J . G.; Voice, A. M. Polymer 1993,
34, 3076-3082.
(7) Yokozawa, T.; Asai, T.; Sugi, R.; Ishigooka, S.; Hiraoka, S.
J . Am. Chem. Soc. 2000, 122, 8313-8314. Recently, another
group reported the synthesis of the same polyamide using a
different leaving group and base: Shibasaki, Y.; Araki, T.;
Okazaki, M.; Ueda, M. Polym. J . 2002, 34, 261-266.
(8) Yokozawa, T.; Suzuki, Y.; Hiraoka, S. J . Am. Chem. Soc.
2001, 123, 9902-9903.
(9) Yokozawa, T.; Horio, S. Polym. J . 1996, 28, 633-636.
(10) (a) Ueda, M.; Sato, A.; Imai, Y. J . Polym. Sci., Polym. Chem.
Ed. 1977, 15, 2731-2737. (b) Ueda, M.; Sato, A.; Imai, Y. J .
Polym. Sci., Polym. Chem. Ed. 1978, 16, 475-482. (c) Ueda,
M.; Sato, A.; Imai, Y. J . Polym. Sci., Polym. Chem. Ed. 1979,
17, 2013-2018.
room temperature. Analysis and workup of the reaction were
carried out in the same manner as method A.
Mod el Rea ction for In itia tion . A solution of 8 (0.076 g,
0.50 mmol), 19 (0.150 g, 0.500 mmol), and naphthalene
(internal standard for GC analysis, 64 mg, 0.50 mmol) in dry
CH2Cl2 (4 mL) was placed in a round-bottomed flask equipped
with a three-way stopcock under an Ar atmosphere. Into the
flask was added at room temperature a solution of base (0.50
mmol) in dry CH2Cl2 (1 mL) via a syringe from the three-way
stopcock with a stream of N2, and the mixture was stirred at
room temperature. A small portion of the reaction mixture was
withdrawn by a syringe during a certain period and analyzed
by GC to determine the yield of 20.
Mod el Rea ction of Tr a n sester ifica tion a t In itia tor
En d . Tr a n sester ifica tion of 21 by 8. The reactions were
carried out in the same manner as chain-growth polyconden-
sation of 1h using 21 (0.50 mmol) and 8 (0.50 mmol) as starting
materials instead of 1h and initiator. The yield of 20 was
determined by GC analysis of the reaction mixture.
Mod el Rea ction of Tr a n sester ifica tion a t P olym er
Ch a in Tr a n sester ifica tion of 23 by 8. The reactions were
carried out in the same manner as chain-growth polyconden-
sation of 1h using 23 (0.50 mmol) and 8 (0.50 mmol) as starting
materials instead of 1h and initiator. The conversions of 23
and 8 were determined by GC analysis of the reaction mixture.
(11) Tee, O. S.; Iyengar, N. R.; Kraus, B. J . Org. Chem. 1985, 50,
973-976.
(12) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett.
1975, 4467-4470.
(13) J ohansson, R.; Samuelsson, B. J . Chem. Soc., Perkin Trans.
1 1984, 2371-2374.
(14) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu,
O. Tetrahedron 1986, 42, 3021-3028.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectral data of the compounds in this paper. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(15) Yan, L.; Kahne, D. Synlett 1995, 523-524.
(16) Perrin, D. D.; Armarego, W. L. F. In Purification of Labora-
tory Chemicals, 3rd ed.; Pergamon: Oxford, 1988.
Refer en ces a n d Notes
(1) Gilkey, R.; Caldwell, J . R. J . Appl. Polym. Sci. 1959, 2, 198-
202.
MA021449A