Synthesis of novel poly(dithienylpyridines)

Article abstract of DOI:10.1039/b201229d

This paper describes the chemical and electrochemical synthesis of novel copolymers of thiophene and pyridine. Di-iodination of 3-hydroxypyridine 12 followed by O-substitution gave a series of ethers 14b-d and esters/carbamates 15a-d which were reacted with the stannylated bithiophene derivative 17 in a Stille cross-coupling reaction yielding poly (1b-d) and poly (2a-d) respectively. These chemical polymerisation reactions generally resulted in highly insoluble materials which were difficult to characterise. Ethers 14b-d and esters/carbamates 15a-d gave O-substituted 3-hydroxy-2,6-di(2-thienyl)pyridines 1b-d and 2a-d respectively in Stille cross-coupling reactions with the stannylated thiophene 16. Ethers 1b-d underwent electrochemical polymerisation allowing the synthesis of O-alkylated polymers, poly (1b-d), with electrochemical band-gaps of 1.4 to 1.6 eV. In contrast, the esters/carbamates 2a-d could not be electropolymerised.

Full text of DOI:10.1039/b201229d

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis of novel poly(dithienylpyridines){
Glen M. Chapman,a Stephen P. Stanforth,*a Rory Berridge,b Cristina Pozo-Gonzalob
and Peter J. Skabarab
aSchool of Applied and Molecular Sciences, University of Northumbria, Newcastle upon Tyne,
UK NE1 8ST. E-mail: steven.stanforth@unn.ac.uk.
bDepartment of Chemistry, University of Manchester, Oxford Road, Manchester,
UK M13 9PL. E-mail: peter.skabara@man.ac.uk.
Received 4th February 2002, Accepted 12th April 2002
First published as an Advance Article on the web 30th May 2002
This paper describes the chemical and electrochemical synthesis of novel copolymers of thiophene and pyridine.
Di-iodination of 3-hydroxypyridine 12 followed by O-substitution gave a series of ethers 14b–d and
esters/carbamates 15a–d which were reacted with the stannylated bithiophene derivative 17 in a Stille
cross-coupling reaction yielding poly (1b–d) and poly (2a–d) respectively. These chemical polymerisation
reactions generally resulted in highly insoluble materials which were difficult to characterise. Ethers 14b–d and
esters/carbamates 15a–d gave O-substituted 3-hydroxy-2,6-di(2-thienyl)pyridines 1b–d and 2a–d respectively in
Stille cross-coupling reactions with the stannylated thiophene 16. Ethers 1b–d underwent electrochemical
polymerisation allowing the synthesis of O-alkylated polymers, poly (1b–d), with electrochemical band-gaps of
1.4 to 1.6 eV. In contrast, the esters/carbamates 2a–d could not be electropolymerised.
(NBMO). In contrast, the covalent isomer 7 of the mesomeric
Introduction
betaine 5 is associated with an energetically lower HOMO.
Thus, the NBMO–LUMO energy difference in mesomeric
betaine 5 is significantly smaller than the HOMO–LUMO
energy difference in 3-methoxypyridine 7. The relatively low
energy NBMO–LUMO is a general property of heterocyclic
mesomeric betaines derived from odd-alternant hydrocarbon
anions of which compound 5 is just one of many examples.6
Since the NBMO coefficients of mesomeric betaine 5 are
located at carbon atoms 2,4 and 6 of the pyridine ring, the sub-
stitution of two thienyl groups at any of these positions will
lead to polymers in which the positive and negative charges can
be delocalised through the polymer backbone. We have chosen
the 2,6-disubstitution pattern for our studies i.e. monomer 3
from which the corresponding poly (3) compounds are derived.
Pyridine N-oxide 8 is also associated with a HOMO which
has similar characteristics to a NBMO. Thus, our interest in
dithienylpyridine copolymers also includes the polymers poly
(4a,b) which possess a pyridine N-oxide moiety as the electron-
rich pyridine component. In view of the novelty of poly (1) and
(2), we were also interested in preparing and studying these
materials.
Related work in this area of copolymer chemistry includes
Yamamoto and co-workers’ studies of thienylpyridine copo-
lymers 9 in which the pyridine unit is substituted at the 2 and 5
positions by 2-thienyl moieties.8–11 These copolymers 9 behave
as ‘charge transfer’ materials because of the alternating motif
of electron-rich and electron-deficient heterocyclic rings. This
work has been extended to include copolymers 10 of thiophene
and pyrido[3,4-b]pyrazine.11,12 Significant differences would
be expected between the electronic properties of poly (3, 4)
systems and copolymers 9 because of the different nature of
their HOMOs. Higgins and Crayston have prepared polymer
11 by electropolymerisation of 2,6-bis(2-thienyl)pyridine.13
Their work did not include pyridines with electron-releasing
substituents.
Recently, there has been considerable interest in the synthesis
of organic polymers which might act as conductors in either the
doped or un-doped state and would therefore have the
potential to function as organic metals in electronic devices.
One of the main advantages of organic polymers as organic
metals over inorganic materials is their potential for processing.
Polythiophenes and substituted polythiophenes1–3 have been
extensively studied as conducting polymers and their syn-
thesis and properties continue to attract considerable interest.
Although polythiophene has the disadvantage of being inso-
luble which limits its potential for processing, substituted
polythiophenes possessing aliphatic or polar substituents fre-
quently have improved solubility characteristics.
One fundamental challenge in polythiophene chemistry has
been to prepare low band-gap materials. This is often achieved
by introducing pendant substituents onto the polymer back-
bone in order to increase the energy of the highest occupied
molecular orbital (HOMO) or reduce the energy of the lowest
unoccupied molecular orbital (LUMO). Band-gap energies can
be readily determined from the low energy absorption edge
of their electronic spectra or by cyclic voltammetry. Band-gaps
in polythiophenes are typically about 2 eV although some
polymers have been reported with significantly lower band-
gaps, for example the band-gap in regioregular polyalkylthio-
phenes was determined as 1.7 eV.4 We have been interested in
preparing a series of polythiophene analogues [poly (1)–(4)]
which possess an electron-rich pyridine ring as part of the
polymer backbone. In particular, we anticipated that the band-
gaps in poly (3) might be significantly lower than the band-gaps
in poly (1) and (2) for the reasons which are now outlined. The
pyridinium-3-olate 5 is a mesomeric betaine5–7 in which the
HOMO is distributed over the carbon atoms at the 2,4 and
6-positions and the exocyclic oxygen atom as shown in formula
6. The HOMO of mesomeric betaine 5 is therefore topologically
and energetically similar to a non-bonding molecular orbital
This paper describes our preliminary studies of potential
low band-gap polythiophene analogues in which the polymer
backbone consists of both thiophene and electron-rich pyridine
{Electronic supplementary information (ESI) available: further experi-
mental details. See http://www.rsc.org/suppdata/jm/b2/b201229d/
2292
J. Mater. Chem., 2002, 12, 2292–2298
This journal is # The Royal Society of Chemistry 2002
DOI: 10.1039/b201229d

View Article Online
Petroleum ether refers to the fraction bp 40–60 uC. The
synthesis of the monomers 1b–1d used in electropolymerisation
reactions and the precursors 14c and 14d of these monomers
is described below. The synthesis of all other compounds are
reported in the electronic supplementary information.{
Synthesis of 2,6-diiodopyridines
2,6-Diiodo-3-hexyloxypyridine 14c. A mixture of potassium
carbonate (3.98 g; 28.8 mmol), hexyl iodide (3.36 g; 15.8 mmol),
2,6-diiodo-3-hydroxypyridine 13 (5.00 g; 14.0 mmol) and dry
propanone (40 mL) was heated at reflux for 8 hours. The
mixture was allowed to cool to room temperature and the
solvent was evaporated under reduced pressure. Dichloro-
methane (20 mL) was added to the residue and the mixture was
then washed with distilled water (3 6 20 mL). The organic
layer was dried (MgSO4), filtered, and evaporated under
reduced pressure. The product was then recrystallised from
ethanol giving compound 14c (4.88 g; 79%) as white crystals,
mp 52–54 uC. IR (KBr): nmax 3438, 2927, 2855, 1654, 1551,
1535, 1466, 1417, 1383, 1278, 1227, 1143, 1094, 1072 and
1049 cm21; 1H NMR (CDCl3): d 7.52 (1H, d, J ~ 8 Hz, ArH),
6.66 (1H, d, J ~ 8 Hz, ArH), 3.99 (2H, t, J ~ 6 Hz, -OCH2-),
1.84 (2H, m, J ~ 6 Hz, -OCH2CH2-), 1.56–1.45 (2H, broad m,
-OCH2CH2CH2-), 1.43–1.24 (4H, broad m, -CH2CH2CH3) and
0.91 (3H, t, J ~ 7 Hz, -CH3) ppm; 13C NMR (CDCl3): d 155.4,
134.1, 119.7, 111.1, 103.1, 69.7, 31.4, 28.7, 25.6, 22.5 and
14.0 ppm; elemental analysis: C11H15I2NO requires C, 30.65;
H, 3.5; N, 3.25. Found C, 30.6; H, 3.25; N, 3.3%; MS (ES)
calculated mass 431.9321, measured mass 431.9321.
2,6-Diiodo-3-benzyloxypyridine 14d
A mixture of propanone (40 mL), potassium carbonate (8.07 g;
58.0 mmol), benzyl bromide (5.42 g; 32.0 mmol) and 2,6-
diiodo-3-hydroxypyridine 13 (10.00 g; 29.0 mmol) was heated
at reflux for 8 hours. The reaction mixture was allowed to cool
to room temperature and the solvent was evaporated under
reduced pressure. Dichloromethane (20 mL) was added to the
residue and the mixture was washed with distilled water (3 6
20 mL). The organic layer was dried (MgSO4), filtered and
evaporated under reduced pressure. The residue was extracted
with three portions of hot hexane to remove any remaining
benzyl bromide. The crude product was then recrystallised
from ethanol giving compound 14d (5.00 g; 40%) as white
crystals, mp 112–114 uC. IR (KBr): nmax 3414, 1637, 1617, 1544,
1498, 1455, 1411, 1376, 1341, 1278, 1226, 1099, 1056 and
1019 cm21; 1H NMR (CDCl3): d 7.51 (1H, d, J ~ 8 Hz, ArH),
7.46–7.31 (5H, m, ArH), 6.73 (1H, d, J ~ 9 Hz, ArH) and 5.15
(2H, s, -CH2-) ppm; 13C NMR (CDCl3): d 155.2, 135.3, 134.4,
129.0, 128.9, 127.4, 120.9, 111.5, 104.2 and 71.5 ppm; elemental
analysis: C12H9I2NO requires C, 33.0; H, 2.1; N, 3.2. Found C,
33.0; H, 1.9; N, 3.1%; MS (ES) calculated mass 437.8852,
measured mass 437.8859.
rings. Poly (1–4) have been represented as regioregular
polymers for simplicity but the polymers we describe in this
paper are probably regio-irregular.
Synthesis of 2,6-bis(2-thienyl)pyridines
3-Methoxy-2,6-bis(2-thienyl)pyridine 1b. Under a nitrogen
atmosphere a mixture of magnesium (0.34 g; 14.0 mmol),
2-bromothiophene (2.29 g; 14 mmol), a crystal of iodine and
dry tetrahydrofuran (10 mL) was stirred and warmed gently
until the Grignard reagent had formed. The solution was then
cooled to 25 uC and a mixture of 2,6-diiodo-3-methoxy-
pyridine 14b (2.00 g; 5.5 mmol), Ni(dppe)Cl2 (0.054 g;
0.098 mmol) in dry tetrahydrofuran (10 mL) was added
dropwise with stirring. The mixture was stirred and slowly
warmed to reflux. The mixture was heated at reflux for
20 hours, allowed to cool to room temperature, poured onto ice
and extracted with dichloromethane (3 6 30 mL). The organic
layers were combined, dried (MgSO4) and evaporated under
reduced pressure. The dark solid was then purified by column
Experimental
General
The following compounds were prepared by literature pro-
cedures: 13,14 14b,15 15a,15 17,18 21,19 2220 and 2-ethynyl-
thiophene.16 NMR spectra were determined at 270 MHz.
J. Mater. Chem., 2002, 12, 2292–2298
2293

View Article Online
chromatography over silica gel (eluent; ethyl acetate : petro-
leum ether, 3 : 7) giving compound 1b (0.70 g; 46%) as yellow
crystals, mp 95–96 uC. IR (KBr): nmax 3440, 3100, 2938, 1654,
1566, 1465, 1437, 1388, 1334, 1281, 1265, 1202, 1178, 1138,
Results and discussion
Monomer synthesis
Our synthetic strategy is outlined in Scheme 1 and uses
commercially available and inexpensive 3-hydroxypyridine 12
as the starting material. Our strategy was chosen so that the
required polymers could be synthesised from appropriate
monomer precursors using either electrochemical or cross-
coupling methodology. Di-iodination of 3-hydroxypyridine 12
using a literature procedure14 gave 2,6-diiodo-3-hydroxypyr-
idine 13 from which the ether derivatives 14b–d and ester/
carbamate derivatives 15a–d were prepared by alkylation or
acylation respectively.15 Stille coupling of compounds 14b–d
and 15a–d with the stannylated thiophene 16 gave the
corresponding teraryl derivatives 1b–d and 2a–d. The crude
yields of teraryls from these Stille reactions were generally
moderate and products were purified by chromatography and
recrystallisation. Compound 1b was also prepared from
compound 14b and 2-thienylmagnesium bromide in a nickel
catalysed Kharasch type reaction with an improved yield.
Compound 1a was prepared by hydrolysis of compound 2a.
The alkyne analogue 18 of compound 1d was also prepared
(62% yield) from compound 14d and 2-thienylethyne16 in a
Sonogashira type reaction (Scheme 2).
1082, 1054, 1025 and 1011 cm21 1H NMR (CDCl3): d 8.02
;
(1H, dd, J ~ 1 and 4 Hz, ArH), 7.47 (1H, dd, J ~ 1 and 4 Hz,
ArH), 7.44–7.39 (2H, m, ArH), 7.32 (1H, dd, J ~ 1 and 5 Hz,
ArH), 7.19 (1H, d, J ~ 9 Hz, ArH), 7.12 (1H, dd, J ~ 4 and
5 Hz, ArH), 7.06 (1H, dd, J ~ 4 and 5 Hz, ArH) and 3.93 (3H,
s, -OCH3) ppm; 13C NMR (CDCl3): d 150.9, 144.9, 144.3,
142.0, 141.4, 128.2, 127.9, 127.8, 127.7, 126.6, 123.3, 119.3,
117.4 and 55.7 ppm; elemental analysis: C14H11NOS2
requires C, 61.5; H, 4.1; N, 5.1. Found C, 61.3; H, 3.85; N,
5.1%; MS (EI) calculated mass 273.0282, measured mass
273.0279.
Stille reactions: general method
Under a nitrogen atmosphere a solution of the appropriate
2,6-diiodopyridine or 2,6-dibromopyridine derivative (1.00 g),
2.2 mol equivalents of 2-(tributylstannyl)thiophene, toluene
(40 mL) and 0.11 mol equivalents of tetrakis(triphenyl-
phosphine)palladium(0) was stirred at reflux for 24 hours.
The solution was cooled to room temperature and was filtered
to remove any precipitated palladium. The mother liquor was
evaporated under reduced pressure, dichloromethane (40 mL)
was added to the residue and the mixture was washed with 2 M
potassium fluoride solution (2 6 40 mL) and then with distilled
water (1 6 40 mL). The crude product was purified by column
chromatography over silica gel (eluent: petroleum ether to
remove the tin residues and then ethyl acetate : petroleum ether,
3 : 7 to elute the product). Using this method the following
compounds were prepared.
We have attempted to prepare the mesomeric betaine
derivative 19 as shown in Scheme 3. Thus, when the acetate
derivative 15a was treated with dimethyl sulfate a mixture of
1
products was obtained by H NMR spectroscopy. N-Methyl-
ation and deacylation had apparently occurred (no carbonyl
group was observed in the 13C NMR spectra) but satisfactory
microanalytical data and high resolution mass spectra could
3-Hexyloxy-2,6-bis(2-thienyl)pyridine 1c. The crude product
was recrystallised from ethanol giving compound 1c (0.25 g;
31%) as light yellow crystals, mp 84 uC (with decomposition).
IR (KBr): nmax 3446, 3066, 2916, 2849, 1564, 1460, 1426, 1390,
1337, 1268, 1201, 1135, 1057 and 1022 cm21 1H NMR
;
(CDCl3): d 8.02 (1H, dd, J ~ 1 and 4 Hz, ArH), 7.49 (1H, dd,
J ~ 1 and 4 Hz, ArH), 7.45 (1H, d, J ~ 8, ArH), 7.41 (1H, dd,
J ~ 1 and 5 Hz, ArH), 7.32 (1H, dd, J ~ 1 and 5 Hz, ArH),
7.22 (1H, d, J ~ 8 Hz, ArH), 7.13 (1H, dd, J ~ 4 and 5 Hz,
ArH), 7.08 (1H, dd, J ~ 4 and 5 Hz, ArH), 4.11 (2H, t, J ~
7 Hz, -OCH2-), 1.94 (2H, m, -OCH2CH2-), 1.55 (2H, broad m,
-OCH2CH2CH2-), 1.44–1.32 (4H, broad –CH2CH2CH3) and
0.92 (3H, t, J~ 6 Hz, -CH3) ppm; 13C NMR (CDCl3): d 150.2,
144.9, 144.0, 141.9, 141.2, 127.9, 127.8, 127.7, 127.5, 126.4,
123.0, 129.7, 117.2, 68.9, 31.5, 29.1, 25.8, 22.5 and 14.0 ppm;
elemental analysis: C19H21NOS2 requires C, 66.4; H, 6.2; N,
4.1. Found C, 66.35; H, 6.2; N, 4.0%; MS (EI) calculated mass
343.1065, measured mass 343.1077.
3-Benzyloxy-2,6-bis(2-thienyl)pyridine 1d. The crude pro-
duct was recrystallised from ethanol to give compound 1d
(0.15 g; 19%) as white crystals, mp 107–109 uC. IR (KBr): nmax
3442, 3029, 2883, 1552, 1450, 1377, 1276, 1197, 1141 and
1000 cm21; 1H NMR (CDCl3): d 8.01 (1H, dd, J ~ 1 and 4 Hz,
ArH), 7.43–7.33 (8H, m, ArH), 7.30 (1H, dd, J ~ 1 and 5 Hz,
ArH), 7.16 (1H, d, J ~ 9 Hz, ArH), 7.11–7.02 (2H, m, ArH)
and 5.12 (2H, s, -OCH2-) ppm; 13C NMR (CDCl3): d 149.8,
144.9, 144.6, 142.1, 141.6, 128.8, 128.4, 128.4, 128.0, 127.9,
127.8, 127.8, 127.6, 126.7, 123.5, 120.7, 117.3 and 70.9 ppm;
elemental analysis: C20H15NOS2 requires C, 68.7; H, 4.3; N,
4.0. Found C, 69.1; H, 4.3; N, 4.0%; MS (EI) calculated mass
349.0595, measured mass 349.0602.
Scheme 1 Monomer and polymer synthesis. Reagents and conditions: i,
aq. Na2CO3–I2; ii, RX–base; iii, ClCOR–base; iv, 16–Pd(0); v, 17–Pd(0);
vi, cyclic voltammetry.
2294
J. Mater. Chem., 2002, 12, 2292–2298

View Article Online
precluded an accurate evaluation of the polymers’ optical
band-gaps. The polymers were relatively soluble in trifluoro-
acetic acid and the UV spectra of poly (1c and 2b) were also
determined in this solvent. These polymers showed lmax at
531 nm and 494 nm respectively but we must bear in mind that
in trifluoroacetic acid solution we are recording the spectra of
protonated materials. The longest wavelength absorption edge
in these two protonated polymers were well defined at 613 nm
and 578 nm respectively indicating band-gaps of 2.0 eV for poly
Scheme 2 Synthesis of compound 18.
1
(1c) and 2.15 eV for poly (2b). The H NMR spectra for the
series of polymers (1b–d and 2a–d) could be obtained in
deuterated trifluoroacetic acid solution which confirmed their
proposed structures. Microanalytical data on the chemically
prepared polymers generally indicated lower than calculated
percentages of carbon and nitrogen and this might reflect the
presence of relatively short chains, possibly terminated with
iodine.
Scheme 3 Attempted synthesis of betaine 19.
When the N-oxides 22 and 23 were reacted with the
stannylated bithiophene 17 only highly insoluble materials
that gave UV spectra without well defined long wavelength
1
absorption edges could be obtained. The H NMR spectra of
these materials in deuterated trifluoroacetic acid showed
the presence of aliphatic protons indicating the presence of
tributyltin residues and hence low molecular weight products
are probably formed. This has been attributed to the lower
reactivity of the bromo substituents in the Stille coupling
reactions and, in the case of the N-oxide 23, steric hindrance
at the 2-position. These materials were not investigated further.
In view of the insolubility of chemically prepared polymers
our attention was directed at preparing the target polymers
electrochemically.
Scheme 4 Synthesis of N-oxides 4a and 4b.
not be obtained. Iodination of the heterocyclic mesomeric
betaine 5 also failed to give compound 19.
The N-oxide derivatives 4a and 4b were prepared as outlined
in Scheme 4. Thus, treatment of 2,6-dibromopyridine 20 with
aqueous hydrogen peroxide in hot trifluoroacetic acid afforded
2,6-dibromopyridine N-oxide 22 which underwent a Stille
reaction with the stannylated thiophene 16 giving the desired
product 4a. By an analogous procedure, N-oxide 4b was
obtained from 2,6-dibromo-3-methoxypyridine 21 via 2,6-
dibromo-3-methoxypyridine N-oxide 23.
Electrochemistry
The redox behaviour of compounds 1a–d, 2a–d and 18 were
firstly characterised by cyclic voltammetry (vs. Ag/AgCl), using
dichloromethane as the solvent and tetrabutylammonium
hexafluorophosphate (0.1 M) as the supporting electrolyte at
a scan rate of 200 mV s21. For the teraryl derivatives, two
irreversible oxidation waves were observed between 11.09 and
11.99 V (Table 1), and there was a further single electron
irreversible reduction process in the range 21.00 to 21.32 V.
Whilst there was no apparent relationship between E1red and
the nature of the substituents, there was a pronounced
difference in the oxidation potentials between compounds
1b–d bearing ether groups (E1ox ~ 11.34 to 11.42 V) and
those compounds 2a–d possessing ester/carbamate functiona-
lities (E1ox ~ 11.59 to 11.73 V). The shift towards more
positive potentials for compounds 2a–d was attributed to a
combination of: (i) steric hindrance between the bulky ester/
carbamate group and the adjacent thiophene ring, which forces
the heterocyclic rings out of coplanarity, and (ii) reduced
electron donating effect of the ester/carbamate functionality
Chemical polymerisation reactions
Compounds 14b–d and 15a–d were reacted with the stanny-
lated bithiophene derivative 17 to yield poly (1b–d) and poly
(2a–d) respectively. The crude polymers obtained from these
reactions were extracted with hot toluene (Soxhlet) to remove
lower molecular weight material but the remaining higher
molecular weight materials proved to be highly insoluble in a
variety of solvents, even when they possessed long side chains
e.g. poly (1c and 2b). The UV spectra (ethanol) of the polymers
were not particularly informative because there was not a clear
cut-off point on the long wavelength absorption edge and this
Table 1 UV and electrochemical data for compounds 1a–d, 2a–d and 18.
Electrochemistrya
Absorption maxima/nm
Band-gap
Compound
E1ox
E2ox
E1red
Monomerb
Polymerc
Optical
Electrochemical
Redox peaks, polymer/V
1a
1b
1c
1d
2a
2b
2c
2d
18
11.09
11.37
11.42
11.34
11.65
11.59
11.60
11.73
11.54
11.56
11.90
11.75
11.46
11.94
11.99
11.95
11.97
11.92
21.32
21.14
21.01
21.13
21.04
21.16
21.00
21.08
21.11
351
355
355
356
337
338
337
338
354
412, 602, 787
448
383, 408, 442
452
—
—
—
—
—
—
1.63 eV
1.52 eV
1.92 eV
—
—
—
—
—
c
—
11.14d (76), 21.13
1.39 eV
1.63 eV
1.55 eV
—
—
—
11.10d (50), 21.11
11.14d (213), 21.35
11.10d (143), 21.21
—
—
—
—
—
—
—
aExperiments carried out in dichloromethane, values given in V. bIn dichloromethane. Film on ITO glass. dReversible or quasi-reversible wave,
E1/2 given in V, Epa 2 Epc (in mV) given in parentheses.
J. Mater. Chem., 2002, 12, 2292–2298
2295

View Article Online
compared with that of the ether groups in compounds 1b–d.
Both these factors result in a less efficient p-electron donor
system for derivatives 2a–d and these electrochemical results
concur with data obtained from electronic absorption studies:
the values of lmax for the ether-containing compounds 1b–d are
ca. 18 nm higher than those of the ester/carbamate analogues
2a–d (Table 1). Interestingly, the cyclic voltammogram of the
hydroxy system 1a gave lower oxidation potentials compared
with the alkoxy analogues 1b–d. A reduction in steric hindrance
in this case, coupled to an increase in planarity of the molecule,
can not be invoked, since the absorption maximum for 1a
(351 nm) is the lowest in series 1. It is possible that compound
1a exists partially as its mesomeric betaine tautomer and
consequently pyridoxy radicals are formed on oxidation.
The bis(thienylacetylene)pyridine species 18, showed similar
redox behaviour to the substituted dithienylpyridine deriva-
tives. Thus, compound 18 gave two irreversible oxidation peaks
at 11.54 V and 11.92 V, together with a single irreversible
reduction signal at 21.11 V. Notably, the acetylene derivative
18 is a significantly weaker donor than the analogous
dithienylpyridine 1d. The difference between the values for
E1ox and E2ox is 200 mV and 460 mV, respectively, yet the
absorption maxima for the two compounds differ by only 2 nm.
The N-oxide derivatives 4a and 4b were not electroactive
within the limits of the solvent.
Fig. 2 Cyclic voltammogram of poly (1b) in monomer-free acetonitrile
solution as a thin film on a Au working electrode (Ag/AgCl reference
electrode, Pt counter electrode, 0.1 M tetrabutylammonium hexa-
fluorophosphate, scan rate 100 mV s21).
similar redox behaviours: reversible/quasi-reversible oxidation
waves were observed between 11.10 and 11.14 V, and
irreversible reduction processes took place within the range
21.11 and 21.35 V. Fig. 2 depicts a representative cyclic
voltammogram for the series. The relationship between the
scan rate and peak current (Eox) for all four polymers gave a
linear fit (R w 0.999; Fig. 3). This indicated the stability of the
polymers towards p-doping and confirmed that charge
transport through the film is not diffusion limited (i.e. the
electroactive polymer film behaves identically to a monolayer-
derivatised electrode).17
Electropolymerisation
Electropolymerisation experiments were attempted for all
compounds shown in Table 1 (dichloromethane, 1023
M
The absorption maxima for poly (1a), poly (1b) and poly (1d)
were red-shifted, compared to those of the corresponding
monomers by 61 nm, 93 nm and 96 nm respectively. The
difference is indicative of an increase in conjugation length of
the polymer, compared to the monomer unit, which is typical
for polythiophene species. The absorption spectrum for poly
(1c) gave a major absorption peak at 383 nm, together with
small shoulder peaks at 408 and 442 nm. Also, there was only a
substrate, 0.1 M TBAPF6, Ag/AgCl reference electrode). The
procedure involved repetitive scanning over the second and/or
first oxidation waves at a scan rate of 100 mV s21, using a gold
disk or an indium–tin oxide (ITO) working electrode. The
cyclic voltammograms representing polymer growth for each
of the compounds in series 1 are shown in Fig. 1. None of the
ester/carbamate derivatives 2a–d or the acetylenic system 18
showed any signs of electrodeposition. Poly (1a–d) all exhibited
Fig. 1 Electrochemical polymerisation of 1a–d on Au disk electrodes (Ag/AgCl reference electrode, Pt counter electrode, 0.1 M tetra-
butylammonium hexafluorophosphate, scan rate 100 mV s21).
2296
J. Mater. Chem., 2002, 12, 2292–2298

View Article Online
given above. As expected, the as-grown polymers were
obtained in the oxidised state. The doped state of poly (1a)
proved to be persistent, even after repetitive cycling over the
range 0.1 V to 20.2 V for several hours. Fig. 4 shows the
absorption spectra for the monomer 1a and poly (1a) during
the dedoping process. The maximum at 787 nm is indicative
of chain oxidation, whereas that at 602 nm could be due
to intramolecular charge transfer involving the mesomeric
pyridinolate structure and the thiophene rings. Such a ’push–
pull’ mechanism would stabilise the oxidised state, giving good
reason for the failure to dedope the polymer. After dedoping
the remaining polymers (repetitive cycling over the range 0.1 V
to 20.2 V for 2–5 hours), the absorption spectra were obtained
for poly (1b–d) (Fig. 5). The band-gap (Eg) values are given in
Table 1. Due to the doped state of poly (1a), the Eg for this
material could not be deduced; however, it was found that the
band-gap of poly (1b) was 0.29 eV lower than that of poly (1d),
which is attributed to the increased steric effect of the bulkier
benzyl ether substituent. Because of the postulated high poly-
dispersity of poly (1c), the precise value of the absorption
edge in this case is probably meaningless, since the longest
chain system in the film is unlikely to represent a significant
proportion of the material.
Fig. 3 Plot of current vs. scan rates for 1a–d.
For each dedoped polymer, the electrochemical band-gap
can be determined by the difference between the onset poten-
tials for p- and n-doping (Table 1). Although the values are
somewhat lower than the optical band-gaps, it is evident from
the electrochemical data that poly (1b) bears a higher level of
conjugation and coplanarity than poly (1d). The value for the
electrochemical band-gap of poly (1c) when compared to the
optical band-gap does not follow the same trend as for poly
(1b and 1d); for the reasons given above, this behaviour cannot
be compared directly to the results obtained from poly (1b
and 1d).
The reasons for the relatively low band-gaps of poly (1b–1d)
compared with polythiophenes are intriguing and may be a
consequence of the delocalisation depicted in Fig. 6. These
polymers might therefore be considered as ‘charge transfer’
polymers in which the oxygen lone pair is mesomerically
associated with a pyridine ring further down the chain.
Fig. 4 Electronic absorption spectra of 1a (—) (in dichloromethane)
and poly (1a) (on ITO glass) after dedoping for 2 hours (- - -) and 5
hours (????).
Conclusions
In conclusion, poly (1b–d) have been successfully prepared
from teraryl precursors 1b–d using electrochemical oxidation
and deposited on ITO glass. These electrochemically prepared
polymer films all exhibit lower band-gaps than typical poly-
thiophenes indicating that an electron-rich pyridine ring can
therefore usefully replace a thiophene ring in polythiophene
analogues.
Fig. 5 Electronic absorption spectra of poly (1b) (—), poly (1c) (- - -)
and poly (1d) (????) on ITO glass after dedoping.
Acknowledgement
small shift in lmax for poly (1c) (28 nm); therefore, the polymer
is most likely to be a short chain material with a high
polydispersity.
The optical band-gap of a conjugated polymer can be
deduced from the absorption spectrum by identifying the
longest wavelength absorption edge. Electrodeposition on ITO
glass of the four polymers was achieved under the conditions
We thank the EPSRC for a studentship (to GMC) and the
EPSRC mass spectrometry service, Swansea, for high resolu-
tion mass spectra. We thank the EPSRC for funding (RB)
under grant GR/R23053. We also thank Seal Sands Chemicals
Ltd. for providing 3-hydroxypyridine and their interest in this
work.
Fig. 6 Delocalisation in poly (1b–1d).
J. Mater. Chem., 2002, 12, 2292–2298
2297

View Article Online
T. Maruyama, Y. Nakamura, T. Fukuda, B.-L. Lee, N. Ooba,
S. Tomaru, T. Kurihara, T. Kaino, K. Kubota and S. Sasaki,
J. Am. Chem. Soc., 1996, 118, 10389.
References
1
2
J. Roncali, Chem. Rev., 1997, 97, 173 and references therein.
G. Schopf and G. Koßmehl, ‘Polythiophenes-Electrically Conductive
Polymers’, Springer-Verlag, 1997.
11 T. Kanbara, Y. Miyazaki and T. Yamamoto, J. Polym. Sci., Part
A: Polym. Chem., 1995, 33, 999.
12 B.-L. Lee and T. Yamamoto, Macromolecules, 1999, 32, 1375.
13 S. Higgins and J. A. Crayston, Synth. Met., 1993, 55–57, 879.
14 V. Koch and S. Schnatterer, Synthesis, 1990, 497.
15 H. Bojavska-Dahlige and A. Swirska, Roczniki Chem., 1953, 27,
258 (Chem. Abstr., 49, 10287h).
3
G. Tourillon, ‘Handbook of Conducting Polymers’, Ed. T. A.
Skotheim, Marcel Dekker, 1986, 1, 293.
T.-A. Chen and R. D. Rieke, Synth. Met., 1993, 60, 175.
N. Dennis, A. R. Katritzky and Y. Takeuchi, Angew. Chem., Int.
Ed. Engl., 1976, 15, 1.
4
5
16 A. G. Mal’Kina, L. Brandsma, S. F. Vasilevsky and B. A.
Trofimov, Synthesis, 1996, 589.
17 Charge Percolation in Electroactive Polymers, Chapter in
Electroactive Polymer Chemistry, Part 1 Fundamentals, Ed.
M. E. G. Lyons, Plenum Press, New York, 1994.
18 L. L. Miller and Y. Yu, J. Org. Chem., 1995, 60, 6813.
19 G. J. Clark and L. W. Deady, Aust. J. Chem., 1981, 34, 927.
20 U. Neumann and F. Vo¨gtle, Chem. Ber., 1989, 122, 589.
6
7
C. A. Ramsden, Adv. Heterocycl. Chem., 1976, 19, 1.
W. D. Ollis, S. P. Stanforth and C. A. Ramsden, Tetrahedron,
1985, 41, 2239.
T. Yamamoto, M. Shimura, K. Osakada and K. Kubota, Chem.
Lett., 1992, 1003.
8
9
T. Yamamoto, Z.-H. Zhou, T. Maruyama and T. Kanbara, Synth.
Met., 1993, 55–57, 1209.
10 T. Yamamoto, Z.-H. Zhou, T. Kanbara, M. Shimura, K. Kizu,
2298
J. Mater. Chem., 2002, 12, 2292–2298