Verification of radical and anionic polymerization mechanisms in the sulfinyl and the Gilch route

Article abstract of DOI:10.1021/ma020992o

Mechanistic studies were performed on the sulfinyl and the Gilch routes, both p-quinodimethane-based polymerizations, toward OC₁C₁₀-PPV. The influence of a combination of additives was investigated to verify the nature of their polym

Full text of DOI:10.1021/ma020992o

Macromolecules 2003, 36, 3035-3044
3035
Verification of Radical and Anionic Polymerization Mechanisms in the
Sulfinyl and the Gilch Route
Lieve Hon tis,† Veer le Vr in d ts,‡ Dir k Va n d er za n d e,*,†,‡, a n d La u r en ce Lu tsen ‡
Research Group of Polymer and Organic Chemistry, Instituut voor MateriaalOnderzoek (IMO),
Limburgs Universitair Centrum (LUC), Universitaire Campus, gebouw D, B-3590 Diepenbeek,
Belgium, and division IMOMEC, IMEC, Universitaire Campus, gebouw D, B-3590 Diepenbeek,
Belgium
Received J une 24, 2002; Revised Manuscript Received J anuary 15, 2003
ABSTRACT: Mechanistic studies were performed on the sulfinyl and the Gilch routes, both p-
quinodimethane-based polymerizations, toward OC1C10-PPV. The influence of a combination of additives
was investigated to verify the nature of their polymerization mechanisms. In contrast to the sulfinyl
route, the Gilch route was irreproducible when performed in THF. Therefore, a reproducible Gilch
procedure was developed in dioxane at room temperature. The results of the additives were evaluated by
size exclusion chromatography (SEC). All observed effects of both the sulfinyl and Gilch routes are
consistent with a radical polymerization mechanism.
In tr od u ction
active layer in P-LEDs and photovoltaic devices is the
dehydrohalogenation or Gilch route. Even though this
synthesis is currently being applied, the mechanism of
the polymerization of these p-quinodimethane-based
polymerizations is still the subject of an ongoing discord.
The two mechanistic possibilitiessanionic or radicals
have been hard to distinguish. For example, several
recent publications on the Gilch route state or favor an
anionic polymerization mechanism,6,7 while others ar-
gue that, e.g., the Wessling route proceeds via a radical
mechanism.8 Also a previous mechanistic study on the
sulfinyl route in N-methyl formamide (MMF) led to the
conclusion that the main pathway occurs via radicals.9
A common way to investigate the nature of a polym-
erization mechanism is to study the effects of additives.
In the current case of in situ monomer formation,
additives meant to cause an anionic polymerization
reaction (in step 2) could also intervene with the
equilibrating carbanion chemistry involved in the p-
quinodimethane formation (step 1). This definitely
complicates a careful interpretation of the observed
phenomena. However, clarity on this mechanistic aspect
is crucial from an application viewpoint, as control of
material properties relies on an understanding of the
mechanism.
Previous research on the polymerization mechanism
of the sulfinyl route toward PPV precursors concluded
that radical as well as anionic polymerization mecha-
nisms can occur simultaneously.10 Competition between
the anionic and radical mechanism strongly depends on
the solvent and monomer ring substituents. The anionic
mechanism-always yielding oligomers (Mw < 10 000
g/mol)-is not observed in protic solvents and is pro-
moted by electron withdrawing substituents. Electron-
donating substituents suppress the anionic polymeri-
zation in all types of solvent. Mostly, the main polymer-
ization mechanism of the sulfinyl route is radical in
nature, and leads to high(er) molecular weight polymer.
The aim of this report is to contribute fundamentally
to the ongoing discussion on the polymerization mech-
anism of p-quinodimethane-based routes toward PPV
derivatives. The described mechanistic study on
poly[2-methoxy-5-(3,7-dimethyloctyloxy)-p-phenylene-
vinylene] (OC1C10-PPV) aims to make a link between
Over the last 2 decades research in the field of (semi-)
conducting polymers has made impressive progress.
Thanks to the efforts of synthetic chemists, applied
physicists, theoreticians, and materials scientists, plas-
tic electronics such as field-effect transistors (FETs),1
photovoltaic (solar) cells,2 and polymer light-emitting
devices (P-LEDs)3 can now (or are almost able to) meet
the necessary specifications. The major advantages of
polymers are the ease of designing new materials and
the low-cost solution processing. Light-emitting displays
and integrated circuits, for example, can in principle be
manufactured using simple inkjet printer techniques.4
At first, the poor processability of conducting polymers
hampered their development, as solution processing is
in most cases the preferable method for device fabrica-
tion. However, several clever synthetic routes (involving
precursor and side-chain approaches) were developed
to process these polymers in various aqueous and
organic solvents, while maintaining their electrical
properties. Many of these routes involve p-quinodi-
methane systems as actual monomers, and can there-
fore all be described by a general scheme consisting of
three steps (Figure 1). The first step is a base-induced
1,6-elimination from a p-xylene derivative, 1, leading
to the in situ formation of the p-quinodimethane system
2. Second, this intermediate-which can be represented
by three structures contributing differently to the
character of the molecule-polymerizes spontaneously
to the precursor polymer 3. The conjugated structure 4
is obtained in a third step, directly or after thermal
treatment depending on the specific chemical structure
of the starting monomer and the polymerization condi-
tions. The synthetic routes differ in the leaving group
(L) and a polarizer (P), both functionalities at the
benzylic positions. The sulfinyl route uses different L
and P functionalities, implying better control over the
distinct processes.5
The synthetic route commonly used in industry to
obtain poly(p-phenylenevinylenes) (PPVs) for use as the
† Limburgs Universitair Centrum (LUC).
‡ IMEC.
10.1021/ma020992o CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/05/2003

3036 Hontis et al.
Macromolecules, Vol. 36, No. 9, 2003
F igu r e 1. General scheme for p-quinodimethane-based routes toward OC1C10-PPV.
F igu r e 2. Representation of the additives used in the mechanistic study.
precipitate was collected, washed with n-hexane, and dried
under reduced pressure, yielding the corresponding bis(tet-
rahydrothiophenium) salt as a white hygroscopic solid (96.5
the polymerization mechanisms of the sulfinyl and the
Gilch routes, which are both p-quinodimethane-based
polymerizations. To elucidate the mechanism of both
routes, multiple polymerizations were performed in
apolar, aprotic solvents in the presence of different types
of additives, and the effects were evaluated by NMR and
SEC analyses.
1
g, 85%). H NMR (D2O, 400 MHz, numbering based in Figure
5): δ 0.82 (d, J ) 3.0 Hz, 6H, H8′), 0.90 (d, J ) 6.3 Hz, 3H,
H9′), 1.12 (m, 2H, H6′), 1.12 + 1.26 (m, 2H, H4′), 1.28 (m, 2H,
H5′), 1.46 (m, 1H, H7′), 1.51 + 1.71 (m, 2H, H2′), 1.51 (m, 1H,
H3′), 2.21 (m, 8H, SCH2CH2), 3.40 (m, 8H, SCH2), 3.80 (s, 3H,
H13), 4.07 (m, 2H, H1′), 4.41 (s, 4H, ArCH2S), 7.12 (s, 2H, ArH)
ppm. A mixture of NatBuO (7.54 g, 78.4 mmol) and n-
butanethiol (7.07 g, 78.4 mmol) in 150 mL of methanol was
stirred for 30 min at ambient temperature. This clear solution
was added in one portion to a stirred solution of the bis-
(tetrahydrothiophenium) salt (46.5 g, 80 mmol) in 250 mL of
methanol. After 1 h, the reaction mixture was neutralized with
aqueous HCl (1.0 M), and concentrated under reduced pres-
sure. The crude product was diluted with chloroform (250 mL)
and filtrated to remove precipitate. The filtrate was concen-
trated in vacuo, thus yielding a yellow oil. To remove the
tetrahydrothiophene, an azeotropic distillation with n-octane
was repeated three times to obtain 1-[2-[(butylsulfanyl)-
methyl]-5-(chloromethyl)-4-methoxyphenoxy]-3,7-dimethyloc-
tane (and isomer) in a yield of 90%. To obtain the correspond-
ing sulfoxide an aqueous solution (35 wt %) of hydrogen
peroxide (19.4 g, 0.2 mol) was added dropwise to a solution of
crude thioether (32.8 g, 78.4 mmol) in a mixture of 1,4-dioxane
(300 mL), TeO2 (1.55 g, 9.8 mmol), and some drops of
concentrated HCl. After 3 h the reaction was quenched by
addition of 250 mL of a saturated NaCl solution. Extraction
with chloroform, drying of the combined organic layers over
MgSO4, filtration, and concentration under reduced pressure
Exp er im en ta l Section
Ma ter ia ls. The commercially available chemicals NatBuO,
tetrahydrothiophene, n-butanethiol, MgSO4, TeO2, KtBuO,
TEMPO, carbon tetrabromine, 4-tert-butylbenzyl chloride,
4-tert-butylbenzyl bromide, 4-methoxyphenol, 2-tert-butyl-4-
methoxyphenol, 2,6-di-tert-butyl-4-methoxyphenol, tert-butyl-
benzyl chloride, tert-butylbenzyl bromine, and p-methoxyphenol
were purchased from Aldrich. The solvents used were all of p.a.
quality, purchased from Acros, and used without further
purification unless stated otherwise. The solvents THF and
1,4-dioxane were dried over sodium and distilled prior to use.
All reactions were performed under a nitrogen atmosphere.
2,5-Bis(chloromethyl)-1-(3,7-dimethyl-octyloxy)-4-methoxyben-
zene (BCDM) was donated by Covion Organic Semiconductors.
All glassware were dried overnight in a drying oven prior to
use.
Mon om er Syn t h esis of 1-[2-[(Bu t ylsu lfin yl)m et h yl]-
5-(ch lor om et h yl)-4-m et h oxyp h en oxy]-3,7-d im et h yloct -
a n e (a n d Isom er ). A solution of BCDM (70.5 g, 0.195 mol)
and tetrahydrothiophene (70 mL, 0.79 mol) in methanol (150
mL) was stirred for 70 h at ambient temperature. The reaction
mixture was precipitated in diethyl ether (1700 mL). The

Macromolecules, Vol. 36, No. 9, 2003
Radical and Anionic Polymerization Mechanisms 3037
F igu r e 3. Proposed elementary reactions for a radical polymerization mechanism.
(SiO2, eluent: hexane/ethyl acetate 60/40) to give pure mono-
mer as a light yellow oil (82%), which solidified upon standing.
1H NMR (CDCl3, 300 MHz, numbered according to Figure 5):
δ 0.83 (d, J ) 6.6 Hz, 6H, H8′), 0.90 (d, J ) 6.3 Hz, 3H, H9′),
0.89 (t, J ) 7.2 Hz, 3H, H12), 1.12 (m, 2H, H6′), 1.12 + 1.26
(m, 2H, H4′), 1.28 (m, 2H, H5′), 1.41 (m, 2H, H11), 1.46 (m,
1H, H7′), 1.51 + 1.71 (m, 2H, H2′), 1.51 (m, 1H, H3′), 1.71 (m,
2H, H10), 2.58 (m, 2H, H9), 3.65 (s, 3H, H13), 3.95 (m, 2H,
H1′), 3.79 + 3.78 (dd, J AB ) 9.6 Hz, 2H, H8), 4.63 + 4.55 (dd,
2H, H7), 6.80 (d, J ) 2.1 Hz, 1H, H6), 6.89 (d, J ) 3.9 Hz, 1H,
H3) ppm. IR (KBr): 2959, 2916, 2846(CH3, CH2 asymm), 1443
(C-H), 1070 (S(O)), 847 (1,4-subst. arom.) cm-1. MS (EI, m/z,
relative intensity (%)): 431 ([M + 1]+, 100), 395 ([C23H39O3S]+,
10), 325 ([C19H30ClO2]+, 30), 163 ([C10H11O2]+, 30).
Gen er a l Su lfin yl P olym er iza t ion P r oced u r e in
TH F a t Am b ien t Tem p er a t u r e: P r ep a r a t ion of P oly-
[2-(3,7-d im e t h yloct yloxy)-5-m e t h oxy-p -p h e n yle n e vi-
n ylen e] (OC1C10)P P V). Prior to reaction a 100 mL three-
neck flask with Teflon stirrer and reflux condenser was flushed
with nitrogen gas for about 30 min. The solvent (20 mL) was
degassed by bubbling N2 through it for another 15 min.
Another 5 mL of dry THF was used to rinse in 0.5 mmol
(0.2153 g) of the monomer providing a monomer concentration
of 0.02 M. In the case of polymerizations in the presence of an
additive, at this stage the additive was added to the monomer
solution in the amount of 0.5 equiv, based on the amount of
monomer (e.g., 0.25 mmol TEMPO: 0.040 g). Polymerization
was initiated by addition of 1.3 equiv (0.65 mmol, 0.073 g) of
solid potassium tert-butoxide. After 2 h at ambient tempera-
ture, pouring the mixture into 250 mL of ice-water stops the
reaction. Excess base was neutralized by addition of 1.0 M
aqueous hydrochloric acid, followed by extraction with chlo-
roform. The combined organic layers were concentrated under
reduced pressure, and the precursor polymer was redissolved
in 25 mL of toluene. Thermal elimination to the conjugated
polymer took place during 3 h of reflux. At about 40 °C a
dropwise addition of 30 mL of methanol caused the polymer
F igu r e 4. Overlay of SEC chromatograms showing suppres-
sion of tailing.
F igu r e 5. Numbering of monomers for NMR data.
afforded a mixture of BCDM (6%), the desired monomer (86%),
and the corresponding disulfoxide (8%) as a yellow oil. The
reaction mixture was purified by column chromatography

3038 Hontis et al.
Macromolecules, Vol. 36, No. 9, 2003
Ta ble 1. Resu lts of Sta n d a r d a n d Ad d itive Exp er im en ts
on th e Su lfin yl Rou te in THF a
Ta ble 2. Exp er im en ta l Resu lts on th e Gilch Rou te in
THF : After On e P u r ifica tion a
yield (%)
Mw (g/mol)
Mn (g/mol)
PD
yield (%)
M
w (g/mol)
695 000
Mn (g/mol)
PD
standardb
tBuPhCH2Cl
tBuPhCH2Br
TEMPO
35
55
53
55
34
51
0
0
6
7
7
178 000
226 000
213 000
205 000
141 000
130 000
47 300
46 500
52 100
50 300
47 300
39 600
3.7
4.9
4.1
4.0
3.0
3.3
standardb
38
23
44
56
43
39
42
50
0
66 800
71 400
55 100
58 800
52 000
34 300
72 300
57 400
10
15
10
13
18
16
18
14
1 100 000
530 000
772 000
928 000
534 000
1 280 000
781 000
tBuφCH2Cl
tBuφCH2Br
CBr4
25 300
38 500
11 900
12 400
10 500
19 400
4 900
2.4
2.0
2.4
2.7
TEMPO
CBr4
0
tBuMeOPhOH
50
32
56
336 000
266 000
427 000
29 300
43 300
56 000
11
6
8
6
4 600
a
Polymerization conditions: solvent ) dry THF; ambient
temperature; 2 h; amount of additive, 0.5 equiv; KtBuO added as
a solid. Standard: general polymerization procedure in which
no additives are used.
a
Polymerization conditions: solvent ) dry THF, ambient
temperature, 2 h, amount of additive: 0.5 equiv, KtBuO added as
a solid; purification in THF for 2 h at 68 °C. Standard: general
b
b
polymerization procedure in which no additives are used.
to precipitate (except in the case of tert-butylmethoxyphenol
addition, where precipitation by addition of methanol only
occurred after evaporation of toluene). The precipitated poly-
mer was recovered by filtration, washed with methanol and
dried under reduced pressure at room temperature. Residual
fractions were concentrated from the filtrate in vacuo. 1H NMR
(CDCl3, 300 MHz): δ 0.6-2.0 (br, 19H, H2′-H9′), 3.7-4.0 (br,
3H, H13), 4.0-4.2 (br, 2H, H1′), 7.1-7.3 (br, 2H, PhCHd
CHPh), 7.3-7.6 (br, 2H, H3 + H6) ppm. IR (KBr): ν 2957,
2925, 2860, 1510, 1469, 1395, 1217, 1028, 872 cm-1. Corre-
sponding yields and polymer characteristics like Mn, Mw, and
polydispersity values are listed in Table 1.
Gen er a l Gilch P olym er iza tion P r oced u r e in THF a t
Am bien t Tem p er a tu r e tow a r d OC1C10)P P V. The general
procedure for sulfinyl polymerization in THF was followed
except for the addition of the base and the elimination
procedure. In the case of polymerizations in the presence of
an additive, the additive was added to the monomer solution
in an amount of 0.5 equiv (e.g., 0.25 mmol tBuPhCH2Br: 0.060
g). Polymerization of 0.18 g (0.5 mmol) of solid 2,5-bis-
(chloromethyl)-1-(3,7-dimethyloctyloxy)-4-methoxybenzene was
initiated by addition of 1.3 equiv (0.65 mmol; 0.073 g) of solid
potassium tert-butoxide via an elbow. In most cases the
viscosity increased significantly while the reaction mixture
turned from colorless to yellow/orange. After 10 min, 3.3 equiv
(1.65 mmol; 0.185 g) of solid potassium tert-butoxide was added
to start elimination, and the reaction mixture became a deep
orangesexcept for the TEMPO experiments where a yellow
color with an orange tone was preserved. Stirring was con-
tinued for 2 h. For workup the solution was poured in ice-
water under vigorous stirring. Aqueous hydrochloric acid (3.5
mL of 1.0 M) and methanol (2.5 mL) were added, and the
precipitated polymer was recovered by filtration over a glass
filter. The red fibrous polymer was washed with methanol and
dried under reduced pressure at room temperature. Residual
fractions were concentrated in vacuo. 1H NMR of the polymer
was identical to the one described previously.
Ta ble 3. Resu lts of Gilch P olym er iza tion s in Dioxa n e a t
98 °C Usin g Va r iou s Rela tive Ba se Am ou n tsa
KtBuO
portion
1 (equiv)
KtBuO
portion
2 (equiv)
yield
(%)b
Mw
(g/mol)
Mn
(g/mol)
PD
1.3
2.6
3.3
2.0
47 (54)
33 (36)
24 (44)
32 (39)
805 000
748 000
903 000
707 000
77 100
92 000
80 200
76 000
10
8
11
9
a
Polymerization conditions: solvent ) dry dioxane; 98 °C; 2 h;
no additive; KtBuO added as a solid; purification in THF for 2 h
at 68 °C. Number in parentheses: yield before purification.
b
from 1.3 to 2.6 equiv (initiation) and the second KtBuO portion
from 3.3 to 2.0 equiv (elimination). The polymers were
subjected to a general purification procedure in THF.
Gen er a l Gilch P olym er iza tion P r oced u r e in 1,4-d iox-
a n e tow a r d OC1C10-P P V a t Am bien t Tem p er a tu r e. This
procedure is based on the general procedure for Gilch polym-
erization in THF, only the solvent THF was replaced by dry
1,4-dioxane. In the case of polymerizations in the presence of
an additive, the additive was added to the monomer solution
in an amount of 0.5 equiv (e.g., 0.25 mmol tBuPhCH2Cl: 0.046
g). Polymerization was initiated by addition of 1.3 equiv (0.65
mmol; 0.073 g) of solid potassium tert-butoxide via an elbow.
After 10 min, 3.3 equiv (1.65 mmol; 0.185 g) of solid potassium
tert-butoxide was added and the viscosity increased signifi-
cantly while the reaction mixture became a deep orange.
Within the hour a phase separation between a red polymer
lump and a yellowish solution was completed. Nevertheless,
stirring was continued for 2 h, and then the polymerization
temperature was raised to 98 °C (2-12 h) until a homogeneous
solution was obtained. For workup the solution was cooled to
40 °C, and during vigorous stirring 25 mL of water was added
slowly. Also 3.5 mL of 1.0 M aqueous hydrochloric acid and
2.5 mL of methanol were added, and the precipitated polymer
was recovered by filtration. It was washed with methanol and
dried under reduced pressure at room temperature. Next, the
polymer was subjected to a standard purification procedure
in THF. Information on SEC results and polymer yields are
listed in Table 6. Residual fractions were dried under reduced
pressure. In the case of the additional experiment with
TEMPO an additional amount of 0.2 equiv (0.016 g) of TEMPO
was added to the reaction mixture together with the second
portion of base (3.3 equiv added as a solid).
Gen er a l P u r ifica tion P r oced u r e in THF . The polymers
were purified using a similar procedure to that described in
ref 11. The polymer was dissolved in 25 mL of THF and heated
to 68 °C for 2 h. The solution was cooled to 40 °C and a
dropwise addition of methanol caused the solution to become
red due to precipitation of the polymer. The precipitated
polymer was recovered by filtration, washed with methanol
and dried under reduced pressure at room temperature.
Corresponding polymer yields and Mn, Mw, and polydispersity
values after purification are listed in Table 2.
Gilch P olym er iza tion s in 1,4-d ioxa n e tow a r d OC1C10
)
Ad d ition a l Exp er im en ts on Ad d itives 9, 10, a n d 11. A
100 mL three-neck flask with Teflon stirrer, reflux condenser,
and septum was flushed with N2. Two base solutions were
prepared under a nitrogen atmosphere: the first of 1.3 equiv
(0.65 mmol; 0.073 g) of KtBuO in 4 mL of dioxane, and a second
containing 3.3 equiv (1.65 mmol; 0.185 g) of KtBuO in 7 mL
of dioxane. Additives 9, 10, and 11 were added to the first base
P P V a t Eleva ted Tem p er a tu r e. This procedure is based on
the general procedure for Gilch polymerization (as reported
previously for Gilch in THF). This time dry 1,4-dioxane was
used as the polymerization solvent, and reactions were ex-
ecuted at a temperature of 98 °C using an oil bath. A variation
of base amounts was tested: changing the first KtBuO portion

Macromolecules, Vol. 36, No. 9, 2003
Radical and Anionic Polymerization Mechanisms 3039
Ta ble 4. Resu lts of Gilch P olym er iza tion s in Dioxa n e a t
products. These results indicated that additives 9 and 10 can
be partly deprotonated in these basic conditions, while the di-
tert-butylated derivative cannot.
Am bien t Tem p er a tu r ea -a t Va r iou s P u r ifica tion Levels
entry purification level yield (%) Mw (g/mol) Mn (g/mol) PD
1
Gen er a l Rem a r k s a n d In str u m en ta tion . H NMR spec-
1
2
3
4
5
94
89
74
78
74
3 160 000
2 020 000
754 000
816 000
813 000
77 000
64 300
94 000
130 000
111 000
41
32
8
6
7
tra were obtained in CDCl3 at 300 or 400 MHz on a Varian
Inova spectrometer using a 5 mm probe. Chemical shifts (δ)
are expressed in ppm relative to the TMS absorption. All
spectra were recorded at room temperature. Direct insert probe
mass spectrometry (DIP-MS) analyses were carried out on a
Finnigan TSQ 70, electron impact mode, mass range 35-550
and an interscan time of 2 s. The electron energy was 70 eV.
Fourier transform infrared spectroscopy was performed on a
Perkin-Elmer 1600 FT-IR (nominal resolution 2 cm-1, sum-
mation of 16 scans). Molecular weights and molecular weight
distributions were determined relative to polystyrene stan-
dards (Polymer Labs) with a narrow polydispersity by size
exclusion chromatography (SEC). Separation to hydrodynamic
volume was obtained using a Spectra series P100 (Spectra
Physics) equipped with two mixed-B columns (10 µm, 2 × 30
cm × 7.5 mm, Polymer Labs) and a refractive index (RI)
detector (Shodex) at 40 °C. SEC samples were normally filtered
through a 0.45 µm filter, but in the case of filtration problems,
a 1.0 µm filter was used. HPLC grade THF (p.a.) was used as
the eluent at a constant flow rate of 1.0 mL/min. Toluene was
used as flow rate marker. Thin-layer chromatography (TLC)
was carried out on Merck precoated silica gel 60 F254 plates
using chloroform as eluent.
THF, 68 °C
dioxane, 98 °C
dioxane, 98 °C
THF, 68 °C
a
Polymerization conditions: solvent ) dry dioxane; ambient
temperature; 2 h; no additive; KtBuO added as a solid.
Ta ble 5. Effect of P u r ifica tion in THF on SEC
Ch a r a cter istics of Gilch P olym er s Syn th esized in
Dioxa n ea
entry
additive
none
Mw (g/mol)
Mn (g/mol)
PD
1
2
3
4
5
6
788 000
754 000
606 000
582 000
59 000
46 000
94 000
32 000
60 000
9 600
17
8
19
10
6
tBuPhCH2Cl
CBr4
73 000
19 000
4
a
Odd entries: values after purification in dioxane and before
purification in THF at 68 °C. Even entries: values after the
purification in THF at 68 °C. Polymerization conditions: solvent
) dry dioxane; ambient temperature; 2 h; amount of additive,0.5
equiv; KtBuO added as a solid. Purification: first for 2-12 h in
dioxane at 98 °C; second in THF at 68 °C for 2 h.
Resu lts a n d Discu ssion
Ta ble 6. Resu lts of Sta n d a r d a n d Ad d itive Exp er im en ts
on th e Gilch Rou te in Dioxa n e a t Am bien t Tem p er a tu r e:a
After Th er m a l Tr ea tm en tb
A combination of additives (Figure 2) was chosen to
respond to recent publications and clarify some claimed
effects. Additive amounts of 0.5 equiv were used to
enhance the visibility of effects. Additives 9, 10, and 11
were included in one of the final stages of the research,
in response to a publication of Ferraris and co-workers.7
As a consequence, they are not used to the same extent
in every section.
The precipitation procedure in methanol following
polymerization resulted in a polymer fraction held by
the filter and a residual fraction containing the methanol-
soluble entities. The results listed in the tables under
the title “standard” refer to the results of a general
polymerization procedure in which no additives are used
(for more details see the Experimental Section).
Th e Su lfin yl Rou te in THF . The additive effects
were first studied in the sulfinyl route in THF (Table
1). Experiments were performed in duplicate and at
randomsto check for reproducibility and cancel possible
time influences. All SEC analyses present a monomodal
molecular weight distribution. All of the 1H NMR
spectra of the polymer fractions contain the same set
of signals corresponding to OC1C10-PPV; end groups
were not detected. However, a common feature of the
1H NMR analyses of all residual fractions is the
emergence of several minor signals in the region be-
tween 10.30 and 10.50 ppmspossibly originating from
carbonyl and/or carboxyl groups.11
yield (%)
Mw (g/mol)
Mn (g/mol)
PD
standardc
78 (85)
74 (78)
85 (86)
76 (77)
80 (83)
80 (83)
80 (83)
75 (100)
83 (100)
21 (33)
26 (38)
12 (30)
3 (22)
754 000
797 000
650 000
582 000
720 000
655 000
533 000
73 300
94 000
109 000
93 200
60 100
73 000
98 000
90 000
19 300
26 000
40 000
53 000
87 100
6 200
8
7
7
10
10
7
6
4
3
5
tBuPhCH2Cl
tBuPhCH2Br
CBr4
85 000
TEMPO
179 000
227 000
1 510 000
16 000
4
17
3
MeOPhOH
10 (39)
8 (14)
2 (20)
679 000
479 000
58 100
48 000
49 000
15 300
20 000
14
10
4
tBuMeOPhOH
3 (40)
190 000
10
a
Polymerization conditions: solvent ) dry dioxane; ambient
temperature; 2 h; amount of additive, 0.5 equiv; KtBuO added as
b
a solid. Purification: first for 2-12 h in dioxane at 98 °C; second
in THF at 68 °C for 2 h. c Standard: general polymerization
d
procedure in which no additives are used. Number in parenthe-
ses: yield before purification in THF.
solution in an amount of 0.5 equiv, respectively. Base solutions
were added via the septum using a glass syringe. After the
standard reaction time of 2 h, the reaction mixtures were still
mainly yellow. The temperature was raised to 98 °C for 2 h,
and the homogeneous solution was then cooled to 40 °C.
Addition of water did not result in precipitation for reaction
with additives 9 and 10, although for the latter an oily fraction
was withheld by filtration. Filtrates were extracted with
chloroform, and the combined organic layers were dried under
reduced pressure.
Dep r oton a tion Test on Ad d itives 9, 10, a n d 11. To check
if deprotonation occurs, the additives were dissolved in diethyl
ether and repeatedly washed with an aqueous solution of
NaOH. In the case of additives 9 and 10 the aqueous layer
did color slightly pink; as could be expected, no coloration is
visible in the case of additive 11. TLC analyses of the organic
layers after separation did reveal the presence of original
Addition of 0.5 equiv of TEMPO completely inhibited
polymer formation, indicating that radical processes
1
occur. H NMR analyses of the corresponding residual
fractions demonstrate that a new, and most intense
downfield signal emerges at 9.93 ppm. Signals at 7.81,
7.30, and 2.95 ppm and their integration values are
consistent with the dialdehyde identified in earlier
work.8 This dialdehyde originates from a derivative of
the initiating dimer. When the residual monomer
intensity is set at 100%, this product has a relative
intensity of 20%, which is relatively high compared to
other signals present.

3040 Hontis et al.
Macromolecules, Vol. 36, No. 9, 2003
The additive tert-butylbenzyl chloride (tBuPhCH2Cl)
did not seem to exhibit a strong effect, contradicting the
experience of Hsieh et al.6 and the occurrence of an
anionic polymerization mechanism. The addition of
carbon tetrabromine decreased the molecular weight of
the polymer by almost a factor of 10; an expression of
its activity as a chain transfer agent. At the same time
problems with precipitation after elimination occurs
the precipitate is oily instead of fibroussresulting in
very low yields of isolated polymer. The corresponding
residual fractions are reddish-brown in color, and ac-
cording to NMR analysis, still contain an amount of
polymer. Possibly, the remaining additive or side prod-
ucts hinder precipitation of the polymer by methanol
addition. In view of these carbon tetrabromine results,
the effect of tert-butylbenzyl bromide (tBuPhCH2Br) can
be attributed to a chain transfer effect caused by the
carbon-bromine bond, rather than to an anionic influ-
ence. Addition of 0.5 equiv of 2-tert-butyl-4-methoxyphe-
nol to the polymerization mixture results in a somewhat
deviating reaction workup. Although the color of the
reaction mixture, after the elimination procedure, is
(only) slightly orange, addition of methanol does not
result in precipitation. Only after evaporation of all
solvent does addition of methanol result in a yellow-
orange solution containing a small amount of finely
distributed red polymer particles, which can be recov-
ered by filtration. Clearly, both yield and molecular
weight of the polymer are seriously affected. Additive
10 probably acts just like TEMPO as a radical scaven-
ger, but is somewhat less effective. In contrast with the
CBr4 experiments the 1H NMR spectra of the corre-
sponding residual fractions do not contain any residual
polymer signals, thus confirming the low yields.
Th e Gilch Rou te in Dioxa n e. (a ) Sca n n in g for a
Rep r od u cible P olym er iza tion Meth od . In an at-
tempt to obtain reproducible results using the Gilch
route, the solvent was changed to dry 1,4-dioxanesthe
solvent used for the industrial manufacture of OC1C10-
PPVsand several other parameters, like relative base
quantities and temperature were evaluated.
(1) Va r ia tion of Ba se Con d ition s. According to the
Covion publication11 the use of dry dioxane as polym-
erization solvent is linked to working at elevated tem-
perature (98 °C) under reflux and this is applied in this
set of experiments. The effect of the change of solvent
is tested respecting the base quantities of a standard
procedure applied in previous sections: 1.3 equiv of
potassium tert-butoxide (KtBuO) to initiate polymeri-
zation, followed by the addition of another 3.3 equiv of
solid base to proceed with the conversion from the
precursor stage to the conjugated polymer. Next, a
variation in base addition is evaluated by applying the
base quantities of the publication11 (2.6 and 2.0 equiv
of solid base for polymerization and elimination, respec-
tively). Notice that the total amount of base added to
the reaction mixture is kept constant in all experiments.
All polymers are purified in THF at 68 °C prior to SEC
analysis (Table 3), suppressing tailing on the low
molecular weight side of the SEC-chromatogram (Figure
4) and lowering the polydispersity.
As is demonstrated by the results listed in Table 3
the transfer to another solvent can be regarded as a
success: performing the Gilch route in dioxane at 98
°C leads to an improved reproducibility. However, the
obtained yields are slightly lower compared to those of
the industrial process (54% after purification), possibly
due to a larger sensitivity to losses while working with
relatively small quantities. Currently, the reason for the
difference in reproducibility between the results ob-
tained in THF and dioxane is still obscure. Changing
the partial amounts of base addedsthe last two rowss
does not seem to exhibit much of an effect. The corre-
sponding deviations in molecular weight do not seem
to be satisfactorily pronounced to determine a concen-
tration effect, if any. As a consequence, the amounts of
base, 1.3 and 3.3 equiv, are set as standard conditions
for future experiments, even as the addition in solid
state. In this way, a link with the Gilch experiments in
THF as well as with the sulfinyl experiments is re-
tained.
We conclude that the sulfinyl route proceeds by a
radical polymerization mechanism, and yields high
molecular weight polymer. Figure 3 proposes a set of
elementary reactions for the radical polymerization. The
exact nature of the termination reaction remains un-
clear, but the presence of low intensity carboxyl and/or
carbonyl signals in NMR spectra indicate that oxidation
occurs. We found no evidence for an anionic polymeri-
zation mechanism.
Th e Gilch Rou te in THF . A similar series of
experiments on the Gilch route in THF showed irrepro-
ducibility, as was observed by others.12 All of the
polymers were purified by dissolution in THF at 68 °C
and precipitation with methanol at 40 °C. This purifica-
tion procedure brings polydispersities to a more accept-
able level, and already improves reproducibility, but not
satisfactory. The corresponding results after purification
are listed in Table 2. All SEC analyses present a
monomodal molecular weight distribution. For most of
the polymer solutions filtration through a 0.45 µm filter
during sample preparation is (severely) hampered,
indicating “microgel” formation.6
(2) P olym er iza tion s a t Am bien t Tem p er a tu r e.
Because the experiments in THF were performed at
ambient temperature, the effect of this condition was
checked on a Gilch polymerization in dioxane-hence
temperature was lowered from 98 to about 25 °C. The
corresponding results are listed in Table 4. During these
polymerizations a new feature arose: within 1 h of
reaction a phase separation was completed-an orange-
red colored lump of polymer was floating in a light
orange solution. Further work up resulted in complete
isolation of the crude lump, representing the major part
of reaction products (yield ) 94%). The corresponding
characterization of its molecular weight distribution is
listed in entry 1. During SEC sample preparation,
difficulties with solubility and filtration were encoun-
tered, pointing at some gel properties of the lump-
explaining or resulting in the extremely high numbers
for molecular weight and polydispersity. In an attempt
to solve the problem of the phase separation and/or
gelation, the lump of crude polymer was subjected to a
The poor reproducibility indicates that for some
reason the Gilch route in THF is much more sensitive
to reaction conditions than the sulfinyl route and limits
the number of conclusions to be drawn. Only the
complete inhibition of polymer formation by addition of
0.5 equiv of the radical scavenger TEMPO is striking
and reproducible. As this is considered to be insufficient
to comment on the nature of the polymerization mech-
anism of the Gilch route, a reliable system to investigate
the effect of additives on the Gilch route toward
OC1C10-PPV was developed.

Macromolecules, Vol. 36, No. 9, 2003
Radical and Anionic Polymerization Mechanisms 3041
purification treatment in THF at 68 °C. After a short
while the contours of the lump began to fade, and some
swelling was observed. A continuation of the treatment
for over 72 h merely resulted in a minor dissolution of
the lump; hence, a further continuation of the thermal
treatment was considered to be inadequate. From the
precipitated fraction a SEC sample was prepared,
resulting in the values listed in the second entry. The
molecular weight as well as the polydispersity is already
somewhat lowered, but still the same filtration problems
as with the previous sample are encountered.
molecular weight distribution. To demonstrate this
effect a selection of data prior to and after this second
purification is listed in Table 5 in the even and odd
numbered entries, respectively, and an overlay of the
corresponding chromatograms of entries 3 and 4 is
presented in Figure 4.
This thermal procedure gives rise to two distinct
residual fractions: before and after purification in THF.
1H NMR analysis of the former residual fraction only
shows sharp signals of low molecular weight products,
while the spectra of the latter residual fractions also
contain broader polymer or oligomer signals, indicating
that the treatment causes some polymer to stay in
solution during the precipitation procedure. From now
on, only the SEC results after the thermal treatment
in THF will be listed, as evaluations are based upon
these.
Because the primary experiments conducted at el-
evated temperature did not yield any phase separation,
dioxane was added once more to the polymer fraction
and the mixture was brought at 98 °C for 2 h. In this
relatively short time, the lump completely dissolved and
an entirely homogeneous orange solution was obtained.
This behavior is an indication that a physically cross-
linked gel is involved. Cooling down does result in an
increase in viscosity, however a phase separation no
longer occurs. During SEC sample preparation, no
significant problems with solubility or filtration are
encountered. The corresponding results are displayed
in the third entry of Table 4, and the SEC data show
good correspondence with those of the experiments at
elevated temperature (Table 3, first and second entry).
For all experiments, the amount of additive is set at
0.5 equiv to enhance the visibility of effects and main-
tain a link with previous experiments (sulfinyl and Gilch
in THF). As response to the publication of Ferraris et
al.7 additives 9 and 10 are included in the study. An
overview of the obtained results confirms that the
optimized polymerization procedure indeed shows an
acceptable reproducibility (Table 6).
All polymer fraction 1H NMR spectra are identical
To check the observed influence of temperature
another polymerization is performed at ambient tem-
perature-again resulting in a phase separationsbut
this time the temperature is raised until 98 °C im-
mediately after the standard reaction time of 2 h. In
less than 12 h, a completely homogeneous orange
solution is obtained and further work up is executed. A
sample of the red fibrous polymer is analyzed with
SECsno problems were encountered during sample
preparationsand the corresponding results are listed
in the fourth entry. They already are in good agreement
with the results displayed in entry 3. Nevertheless, the
polymer fraction is subjected to purification in THF at
68 °C, but this does not alter much (entry 5). On the
basis of these observations, it can be argued that a
polymerization at ambient temperature in dioxane
results in a physical network that can only be unraveled
by a thermal treatment at a temperature of about 98
°C. This feature of phase separation/gelation is currently
under evaluation and will be discussed in more detail
in a future paper.
and in accordance with the structure of OC1C10-PPV.
1
The complexity of the H NMR spectra of the residual
fractions does not allow a detailed characterization of
side products. It is observed that monomer and additive
signals are always present to some extent in these
residual fractions. The low intensity signals neighboring
the monomer peaks are probably originating from an
oligomer fraction. Between 10.20 and 10.40 ppm several
new minor signals emergespossibly pointing at the
presence of carbonyl and/or carboxyl groups. Only the
residual fraction of the TEMPO experiment shows an
additional signal around 9.7 ppm, possibly pointing at
a dialdehyde structure. This could be originating from
an initiating dimer,9 but its detailed structure could not
be confirmed.
The standard polymerization procedure leads to poly-
mer with a molecular weight from 700 000 to 800 000
in a yield exceeding 70% after purification. Addition of
4-tert-butylbenzyl chloride or bromide does not affect the
polymer yield and resulted in a small decrease in
molecular weight. It can be concluded that these
additivesseven in the relatively large amount of 0.5
equivsdo not exhibit the effect experienced by Hsieh
et al. and do not act as anionic initiators or end caps
for the polymerization. This is also confirmed by Ander-
sson et al.13 The observed small decrease in molecular
weight could, considering the effect of carbon tetrabro-
mine, rather be interpreted as a chain transfer effect.
The formation of stilbenes is confirmed by analysis of
the residual fractions using mass spectrometry. The
addition of carbon tetrabromine results in a serious
decrease in molecular weight and a lowering of poly-
dispersity, while the yield is not affected. This is
consistent with it acting as a chain transfer agent. The
effect of TEMPO on the polymerization is in accordance
with the inhibition of a radical polymerization mecha-
nism; the yield of polymer and the molecular weight
both decreased by a factor of about 4. Possible reasons
why polymerization is not completely inhibited are a
second initiation of polymerization by addition of the
second portion of base and a decreased efficiency of
A close comparison with the experimental results
obtained in dioxane at elevated temperature yields
comparable molecular weight distributions. However,
the ambient temperature experiments result in a sig-
nificant higher yield (less side reactions) and therefore
this room temperature is set as standard for further
experiments in order to evaluate the influence of addi-
tives on the Gilch route.
(b) Resu lts a n d Discu ssion of Ad d itive Effects.
To solve the feature of phase separation all polymers
were subjected to a two-step purification procedure. A
first step consists of the heating of the crude reaction
mixture to 98 °C immediately after polymerization
reaction until a homogeneous solution is obtained (2-
12 h). After appropriate workup, the dried polymer
fraction is subjected to a second, thermal treatment in
THF at 68 °C for 2 h. As mentioned earlier, the latter
procedure results quite often in a suppression of tailing
on the low molecular weight side of the SEC chromato-
gram, leading to a lower polydispersity value of the

3042 Hontis et al.
Ta ble 7. Resu lts of Ad d ition of 9, 10, a n d 11 in
Macromolecules, Vol. 36, No. 9, 2003
higher weight-average molecular weight (Mw) value and
an increased polydispersity in comparison to reactions
where the base is added as a solid (Table 3, entries 1
and 2).
Com bin a tion w ith th e Ba se a s Solu tion ; Com p a r ison
w ith a Refer en ce P olym er iza tion a
yield
fraction Mw (g/mol) Mn (g/mol) PD (%)
additive
referencea
During the polymerization reactions with these ad-
ditives 9, 10, and 11 no phase separation occurs, even
if the base amounts are adjusted to 2.6 and 2.0 equiv to
enhance the probability of polymerization. The reaction
mixtures do not color a deep orange, but remain yellow
to slightly orange for 9 and 10, and light orange in the
case of 11. The result of the precipitation procedure is
dependent on the additive used and will therefore be
discussed separately. In the case of the reaction with
4-methoxyphenol (9), no polymer precipitates; hence, an
extraction is performed leading to one yellow-orange
fraction containing all organic products formed. For this
reason, in Table 7, it is referred to as fraction “all” .
Precipitation of the reaction mixture resulting from
addition of 2-tert-butyl-4-methoxyphenol (10) yields
some kind of reddish “oily” fraction, instead of fibers,
which can be recovered by filtration. The remains of the
reaction mixture is subjected to further workup consist-
ing of extraction and drying under reduced pressure
(precipitation does not yield any polymer). This yellow-
orange fraction is referred to as the “residual” fraction.
Polymerization in the presence of 2,6-di-tert-butyl-4-
methoxyphenol (11) does yield an orange reaction
mixture. By precipitation in water and neutralization,
a small amount of red polymer (“PM” fraction) separates
from the solution. However, this polymer turns out to
be unmanageable, complicating its recovery by filtra-
tion. It easily spreads out over a large surface and is
very hard to remove although addition of methanol
somewhat facilitates collection (similar to the previous
experiments with carbon tetrabromine on the sulfinyl
route toward OC1C10-PPV). Although no explicit poly-
mer could be separated in the case of additives 9 and
10, SEC samples from each fraction are nevertheless
prepared and analyzed. These analyses detect the
presence of a higher molecular weight species, however,
an important remark is that the detector response for
these samples is extremely low. This is in accordance
with the NMR spectra, where at first no polymer signals
could be detected, pointing at very low corresponding
concentrations and hence low yields. In the case of
polymerizations in the presence of 9 and 10, the yield
is calculated taking relative peak areas in the corre-
sponding chromatograms together with relative amounts
of the fractions (e.g., the ratio residual over oily fraction)
into account. Although other monomer signalssincluding
those of the alkoxy side chainssare present, none of the
other spectra seem to contain remaining PhCH2Cl
signals. The residual fraction resulting from the experi-
ments with 10 do not seem to vary much from the
spectra of the corresponding oily fractions. Additive
signals all are present except for the hydroxide signal,
which is possibly hidden below other signals or its
disappearance indicates the formation of an adduct
structure. However, due to an overcrowding of the NMR
spectra between 2.5 and 4.2 ppm, the identification of
such adduct structure is hampered. New intense signals
emerge between 5.8 and 7.0 ppm, but currently no
reasonable explanation can be offered for this feature.
PMd
PM
all
all
residual
oily
1 021 000
1 189 000
63 600
66 800
50 400
2500
60 000
42 000
30 500
29 900
21 600
740
17
28
65
70
MeOPhOHb
2.1
2.2
2.3
3.4
2.4
2.3
2.6
2.7
2c
3c
2c
tBuMeOPhOHb
residual
oily
41 300
1700
17 300
740
3c
ditBuMeOPhOHb
PM
PM
47 600
51 000
18 100
18 900
8
10
a
Polymerization conditions: solvent ) dry dioxane; ambient
temperature; 2 h; no additives; KtBuO added as solution. Purifica-
tion: first for 2-12 h in dioxane at 98 °C; second in THF at 68 °C.
b Polymerization conditions: see footnote a, +0.5 equiv of additives.
c These yields are estimated values, calculated from corresponding
d
SEC chromatograms. PM: polymer fraction.
TEMPO in the present conditions. To distinguish both
possibilities, an additional experiment was set up in
which a second amount of TEMPO (0.2 equiv) was added
to the reaction mixture together with the second base
portion. In this case, no polymer was found, supporting
1
the idea of a second initiation. Corresponding H NMR
spectra similar to the other TEMPO experiments are
found.
The effects of the additives 4-methoxyphenol, 9, and
2-tert-butyl-4-methoxyphenol, 10, are more complex.
Reproducibility of molecular weight is poorsexplaining
the experiments in triplicateswhereas the yield after
purification is somewhat more consistent. No phase
separation occurred during polymerization, but the
thermal treatment in dioxane and THF was neverthe-
less completed. Before purification, the resulting poly-
mer was a wet, sticky substancesprobably because side
products are entrappedswhich complicated separation,
leading to a spread in yields. Apparently for these
reactions the molecular weight of the resulting polymer
is very sensitive to reaction conditions, leading to a large
spread in molecular weights. The yields decreased to
an extremely low level indicating a huge influence of
these additives on the reactions occurring during syn-
thesis. However, as the additive 2-tert-butyl-4-methox-
yphenol (10), which has less nucleophilic properties,
seems to exhibit a similar effect on the yield as additive
9, these results do not support the hypothesis of anionic
initiation.
To approximate the conditions of Ferraris et al.7
somewhat more, additional experiments were set up in
which the additives 9 and 10 were added together with
the base as solution, increasing the probability of them
acting as nucleophile or base. Additive 11 is included
in the study to obtain additional information on the
nature of the effects. If all three additives exhibit a
comparable effect on the polymerization, a radical
polymerization process is more likely than an anionic.
On the contrary, when large differences in effects of the
distinct additives are obtained, this does present an
important indication in favor of anionic processes. To
have some reference values two polymerizations (with-
out additives) are performed in which the base is added
as solution (in amounts of 1.3 and 3.3 equiv). The
corresponding results are listed in entries 1 and 2 of
Table 7. It can be concluded that the polymerizations
in which the base is added as solution yield a somewhat
Although the various fractions that contain polymer
to a certain extent behave differently, it can be stated
that the effect of the three additives is comparable. They
all result in a substantial decrease in yield together with

Macromolecules, Vol. 36, No. 9, 2003
Radical and Anionic Polymerization Mechanisms 3043
a significant lowering of molecular weight, pointing at
a similar interference in reactions. In combination with
the other results reported, it can be concluded that their
main effect is a strong radical inhibiting one, as describ-
ed previously for the case of quinones.14 This strongly
supports the hypothesis of a radical mechanism.
As a whole, the results of the experiments on the
Gilch route in dioxane present a strong indication that
the main polymerization mechanism of the Gilch route
is radical in nature and yields high molecular weight
polymer.
what more. All three additives result in a substantial
decrease in polymer yield together with a significant
lowering of molecular weight. This points to a similar
radical inhibiting interference in the reactions and
strongly refutes the anionic assertion claimed by others.
Taken as a whole, all results support the argument
that also the main polymerization mechanism operating
in the Gilch route is radical in nature, yielding high
molecular weight polymers, as is the case for the sulfinyl
route. Moreover, the results obtained are inconsistent
with an anionic polymerization mechanism and invite
some questions on the conclusions of others.
Con clu sion s
This knowledge of the nature of the polymerization
mechanism offers potential for an enhanced control on
the reaction processes and product handling and hence
on the molecular weight and structure of the resulting
polymer.
This study was undertaken in an attempt to elucidate
the nature of the polymerization mechanismsanionic
or radicalsof p-quinodimethane-based polymerizations.
A combination of additives was used to investigate the
polymerization mechanism by looking at their influence
on the reaction and its products.
Ack n ow led gm en t. This work was supported by the
InterUniversitaire AttractiePool (IUAP IV) Belgian
Government Services of the Prime Minister-Federal
Services for Scientific, Technical, and Cultural Affairs
(predoctoral grant L.H.). Financial support through
Bijzonder OnderzoeksFonds (BOF) and Fonds voor
Wetenschappelijk Onderzoek (FWO) is gratefully ac-
knowledged. The authors thank Dr. H. Becker and Dr.
W. Kreuder of Covion Organic Semiconductors for
helpful discussions and for donating the BCDM mono-
mer.
First of all, the effect of the additives was studied on
the sulfinyl route in THF toward OC1C10-PPV. One of
the most striking observations is the result of the
TEMPO experiments: this addition leads to a complete
inhibition of polymer formation, presenting a strong
indication that radical processes occur. The additive tert-
butylbenzyl chloride does not seem to exhibit a strong
effect, if any. Carbon tetrabromine and tert-butylbenzyl
bromine affect the molecular weight just like a chain
transfer agent would (except for the lowering of the yield
in the case of the former). It can be stated that these
results lie in the line of expectation elaborated in
previous mechanistic studies on the sulfinyl route and
present strong indications that in these experimental
conditions a radical polymerization mechanism is in-
volved that yields high molecular weight polymer.
The complexity of the NMR spectra of experiments
with additives on the Gilch route clearly indicates that
this route is even more complex and presents new
problems to solve. The Gilch route in THF is demon-
strated to yield irreproducible results, which invites
some questions on the conclusions of some publications
of others. Perhaps a reexamination and/or reinterpreta-
tion of the results are required. As a consequence of the
poor reproducibility of results in THF, the solvent was
changed to dioxane. A reproducible polymerization
procedure to synthesize OC1C10-PPV via the Gilch
route at room temperature was developed, based on the
industrial procedure, and used to investigate the influ-
ence of the various additives.
Compared to the sulfinyl route, the Gilch route is
clearly more sensitive to reaction conditions, but similar
conclusions can be drawn. Addition of TEMPO can
completely inhibit polymer formation by its function of
radical scavenger. Carbon tetrabromine acted as chain
transfer agent: molecular weight decreased while the
yield was left unchanged. The small decrease in molec-
ular weight by the additives 4-tert-butylbenzyl chloride
and bromide can in this view also be interpreted as a
chain transfer effect. No indications are found for them
acting as anionic initiators or end-caps for the polym-
erization, contradicting findings of other research groups.
Primary experiments with distinct methoxyphenols, to
elucidate the effect of the nonsubstituted derivative, led
to a poor reproducibility of molecular weights. In ad-
ditional experiments the various p-methoxyphenols
were added together with the base as solution, ap-
proximating the conditions of other publications some-
Refer en ces a n d Notes
(1) (a) Bloor, D. Chem. Br. 1995, May, 385. (b) Roth, S. One-
Dimensional Metals; VCH: Weinheim, Germany, 1995; p
209-231.
(2) (a) Halls, J . J . M.; Walsh, C. A.; Greenham, N. C.; Morseglia,
E. A.; Friend, R. H.; Moratti, S. C.; Holmes, A. B. Nature
(London) 1995, 376, 498. (b) Yu, G.; Gao, J .; Hummelen, J .
C.; Wudl, F.; Heeger, A. J . Science 1995, 270, 1789. (c)
Service, R. F. Science 1995, 269, 91. (d) Dai, L. J . M. S. Rev.
Macromol. Chem. Phys. 1999, C39 (2), 273. (e) Wallace, G.
G.; Dastoor, P. C.; Officer, D. L.; Too, C. O. Chem. Innovation
2000, April, 15. (f) van der Ent, L. Kunststoff Mag. 2001, April
(3), 38.
(3) (a) Liedenbaum, C.; Croonen, Y.; van de Weijer, P.; Vleggaar,
J .; Schoo, H. Synth. Met. 1997, 91, 109. (b) Visser, R. J .
Philips J . Res. 1998, 51 (4), 467. (c) Blom, P. W. M.; de J ong,
M. J . M. Philips J . Res. 1998, 51 (4), 479. (d) de J ong, M. J .
M.; Vissenberg, M. C. J . M. Philips J . Res. 1998, 51 (4), 495.
(e) Berntsen, A. J . M.; van de Weijer, P.; Croonen, Y.;
Liedenbaum, C. T. H. F.; Vleggaar J . J . M. Philips J . Res.
1998, 51 (4), 511. (f) Schoo, H. F. M.; Demandt, R. C. J . E.
Philips J . Res. 1998, 51 (4), 527. (g) Segura, J . L. Acta Polym.
1998, 49, 319.
(4) (a) Eur. Plast. News 1998, March, 21. (b) Hebner, T. R.; Wu,
C. C.; Marcy, D.; Lu, M. H.; Sturm, J . C. Appl. Phys. Lett.
1998, 72, 519. (c) Bharathan, J .; Yang, Y. Appl. Phys. Lett.
1998, 72, 2660.
(5) (a) Vanderzande, D. J .; Issaris, A. C.; Van Der Borght, M. J .;
van Breemen, A. J .; de Kok, M. M.; Gelan, J . M. Macromol.
Symp. 1997, 12, 189. (b) van Breemen, A. J . J . M.; Issaris,
A. C. J .; de Kok, M. M.; Van Der Borght, M. J . A. N.;
Adriaensens, P. J .; Gelan, J . M. J . V.; Vanderzande, D. J . M.
Macromolecules 1999, 32, 5728.
(6) (a) Hsieh, B. R.; Yu, Y.; VanLaeken, A. C.; Lee, H. Macro-
molecules 1997, 30, 8094. (b) Hsieh, B. R.; Yu, Y.; Forsythe,
E. W.; Schaaf, G. M.; Feld, W. A. J . Am. Chem. Soc. 1998,
120, 231.
(7) Neef, C. J .; Ferraris, J . P. Macromolecules 2000, 33, 2311.
(8) (a) Wessling, R. A. J . Polym. Sci.: Polym. Symp. 1985, 72,
55. (b) Denton, F. R., III.; Lahti, P. M.; Karasz, F. E. J . Polym.
Sci.: Polym. Chem. 1992, 30, 2223. (c) Cho, B. R.; Kim, Y.
K.; Han M. S. Macromolecules 1998, 31, 2098.
(9) (a) Issaris, A.; Vanderzande, D.; Gelan, J . Polymer 1997, 38
(1), 2571. (b) Issaris, A. Ph.D. Dissertation, Limburgs Uni-
versitair Centrum, Diepenbeek, Belgium, 1997.

3044 Hontis et al.
Macromolecules, Vol. 36, No. 9, 2003
(10) (a) Adriaensens, P.; Van Der Borght, M.; Hontis, L.; Issaris,
A.; van Breemen, A.; de Kok, M.; Vanderzande, D.; Gelan, J .
Polymer 2000, 41, 7003. (b) Van Der Borght, M.; Adriaensens,
P.; Vanderzande, D.; Gelan, J . Polymer 2000, 41, 2743. (c)
Hontis, L.; Van Der Borght, M.; Vanderzande, D.; Gelan, J .
Polymer 1999, 40, 6615.
(12) Personal communication with Dr. H. Becker from Covion
Organic Semiconductors.
(13) J ohansson, D. M.; Srdanov, G.; Yu, G.; Theander, M.; Ingana¨s,
O.; Andersson, M. R. Macromolecules 2000, 33, 2525.
(14) Finley, K. T. In Chemistry of Quinonoid Compounds; Patai,
E. S., Ed.; Wiley: New York, 1974; p 339.
(11) Becker, H.; Spreitzer, H.; Kreuder, W.; Kluge, E.; Schenk,
H.; Parker, I.; Yong Cao Adv. Mater. 2000, 12 (1), 42.
MA020992O