"fibonacci's route" to regioregular oligo(3-hexylthiophene)s

Article abstract of DOI:10.1021/ja4057932

We describe a new synthetic approach to regioregular monodisperse oligo(3-alkylthiophene)s allowing for simple separation of regioregular material in gram quantities. The number of repeat units follows the Fibonacci numbers up to a length of 21. In a small adaption of this approach, introduction of a protecting group was used to synthesize an oligo(3-hexylthiophene) with 36 repeating units, the longest regioregular 3-hexylthiophene oligomer synthesized to date.

Full text of DOI:10.1021/ja4057932

Communication
pubs.acs.org/JACS
“Fibonacci’s Route” to Regioregular Oligo(3-hexylthiophene)s
Felix P. V. Koch,† Paul Smith,†,‡ and Martin Heeney*,‡,§
†Department of Materials, Eidgenossische Technische Hochschule (ETH) Zurich, Wolfgang-Pauli-Strasse 10, 8093 Zurich,
̈
̈
̈
Switzerland
‡Centre for Plastic Electronics, Imperial College London, Exhibition Road, London SW7 2AZ, U.K.
§Department of Chemistry, Imperial College London, Exhibition Road, London SW7 2AZ, U.K.
S
* Supporting Information
by activation of the newly formed chain end. Often a protecting
group is used to block one oligomer end and enforce
regioselective functionalization.8−11 Immobilization of the
growing oligomer on a solid-phase support has also been
reported to simplify purification,8,9,11−13 enabling the synthesis
of a dodecamer with high regioregularity, albeit on the milligram
scale.8 A major drawback of these approaches for long oligomers
is the growth by only one or two thiophene units per cross-
coupling step, necessitating numerous steps. Into the bargain, the
separation of oligomers with length n from unreacted oligomers
with length n − 1 or n − 2 is increasingly difficult for longer
chains.
ABSTRACT: We describe a new synthetic approach to
regioregular monodisperse oligo(3-alkylthiophene)s allow-
ing for simple separation of regioregular material in gram
quantities. The number of repeat units follows the
Fibonacci numbers up to a length of 21. In a small
adaption of this approach, introduction of a protecting
group was used to synthesize an oligo(3-hexylthiophene)
with 36 repeating units, the longest regioregular 3-
hexylthiophene oligomer synthesized to date.
ithin the class of conjugated polymeric materials, poly(3-
Whexylthiophene) (P3HT) has been one of the most
widely investigated materials because of its relative ease of
synthesis combined with good processability and promising
performance in a range of applications. The synthesis of P3HT
has been reported by a number of routes, mainly based around
transition-metal-catalyzed cross-coupling methodologies.1,2
However, despite this interest, the molecular-weight dependence
of the characteristics of P3HT and its structure−property
relationships are still not fully understood. As is the case more
often than not, this is due to, among other things,
regioirregularity of the solubilizing hexyl group, polydispersities
not equal to 1, inconsistent end-group chemistry, and ill-defined
processing schemes.
The availability of precisely defined oligomers is well-known to
enable the development of structure−property studies to
understand and control the crystal structure, self-assembly, and
performance of these materials. Previous work has demonstrated
the relevance of this approach, with molecularly defined
oligomers being produced to study, for instance, various crystal
structures and the onset of chain folding in poly(ethylene),
among other materials.3,4
In the synthesis of well-defined oligomers of monomers with
C2v symmetry, as in the case of 3,3′-dihexyl-2,2′-bithiophene, the
major challenges to overcome are selective monofunctionaliza-
tion and the separation of side products with similar molecular
weights to yield monodisperse materials.5 Oligomers of such C2v-
symmetric monomers with up to 96 repeating units have already
been prepared.6
For monomers with Cs symmetry such as 3-hexylthiophene,
the synthesis is far more challenging because of the additional
issue of regioregularity.7 Most approaches have been based on an
iterative synthesis involving stepwise extension of the oligomer
by one or two monomer units per cross-coupling step followed
An alternative approach is based upon sequential dimeriza-
tions to afford a geometric sequence of oligomers. This can afford
reasonable quantities of long oligomers in a short number of
steps and is well-suited to monomers of C2v symmetry. However,
for oligomers of Cs symmetry based upon 3-alkylthiophenes, it
has been reported that significant amounts of regioirregular
byproducts are formed by oxidative dimerization14 or homo-
coupling that occurs during cross-coupling.15 Such byproducts
are very difficult to remove, as the regioregular and irregular
materials have the same molecular weight and molecular weight
makes the highest impact on purification properties in oligomers.
For example, in our own attempts to make the octamer of 3-
hexylthiophene by geometric Stille coupling of the brominated
and stannylated mono-, di-, and tetramers, the undesired
byproducts with tail-to-tail miscoupling were consistently
formed in 5−15% yield. These could be separated from the
desired head-to-tail product only by preparative-recycling gel-
permeation chromatography (GPC) after a few tens of cycles (ca.
45 h runtime) on the milligram scale (Figure S1 in the
Supporting Information). The significant impact that regioirre-
gular impurities can have on the physical properties is apparent
from a simple examination of the melting points. The purely
regioregular hexadecamer has a sharp melting point at 126 °C,16
while that of the impure material is ca. 50 °C lower (78 °C).15
A synthetic strategy leading to high-purity monodisperse
regioregular oligo(3-hexylthiophene)s in useful quantities should
therefore avoid the need to separate regioirregular isomers of the
same molar mass. Our solution, as described forthwith, is based
upon the coupling of oligomers having different molecular
weights following the Fibonacci sequence. We found that such an
approach allows the facile separation of the regioirregular side
Received: June 10, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4057932 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
a
products formed by homocoupling or oxidative dimerization of
educts as a result of the significant difference in molecular weight
relative to that of the desired product, allowing the synthesis of
highly regioregular oligomers with up to 36 repeating units.
To allow easy identification of the discussed materials, the
nomenclature for oligo(3-hexylthiophene) derivatives shown in
Figure 1 will be used. A substituent at the so-called “2-position”
Scheme 1. “Fibonacci’s Route” to (3HT)n
Figure 1. Nomenclature for oligo(3-hexylthiophene)s (3HT)n and
substituted derivatives thereof.
(or “head position”) is given as a prefix, as in Br-(3HT)n. A
substituent at the “5-position” (or “tail position”) is given as a
suffix, as in (3HT)n-SnBu3.
The synthetic strategy is shown in Scheme 1. It is based on the
cross-coupling of regioselectively brominated and metalated
oligomers in solution. Bromination of the oligomers at the “2-
position” was readily achieved using N-bromosuccinimide
(NBS) at reduced temperature,8,10 while the “5-position” was
selectively stannylated by treatment with lithium diisopropyla-
mide (LDA) at cryogenic temperature followed by addition of
tributyltin chloride.17
Thus, 3,4′-dihexyl-2,2′-bithiophene, (3HT)2, was synthesized
in good yield by cross-coupling of 2-bromo-3-hexylthiophene
with 5-tributylstannyl-3-hexylthiophene using tetrakis-
(triphenylphosphine)palladium according to the reported
protocol (Scheme 1).18 Here the resulting regioirregular side
products originating from homocoupling and oxidative dimeri-
zation were removed by vacuum distillation.15 The dimer was
subsequently brominated or stannylated to yield Br-(3HT)2 and
(3HT)2-SnBu3, respectively. Coupling of Br-(3HT)2 with the
shorter (3HT)-SnBu3 afforded (3HT)3 in a yield of 73%. The
resulting homocoupled products, 3,4′,3″,3‴-tetrahexyl-
2,2′:5′,2″:5″,2‴-terthiophene [HH-(3HT)4] and 4,4′-dihexyl-
2,2′-bithiophene [TT-(3HT)2], with higher and lower molecular
weight, respectively, relative to the desired HT-(3HT)3, were
readily removed by column chromatography on a large scale,
since the influence of molecular weight on the retention factor
exceeds the influence of regioregularity.
Similarly following analogous reactions for subsequent steps,
the trimer was brominated and subsequently cross-coupled with
(3HT)2-SnBu3 to yield (3HT)5 in 77% yield following
chromatographic purification. Following Scheme 1, Br-(3HT)5
and (3HT)3-SnBu3 were reacted to give (3HT)8, and Br-(3HT)8
and (3HT)5-SnBu3 were combined to give (3HT)13. Finally, Br-
(3HT)13 and (3HT)8-SnBu3 reacted to give (3HT)21, the longest
regioregular oligo(3-hexylthiophene) reported to date.
a
Conditions: (i) NBS, −20 °C to rt; (ii) THF, LDA, Bu3SnCl, −78
°C; (iii) Pd(PPh3)4, toluene, 90 °C.
oligomers of equal length). In the latter, the synthesis of (3HT)8
is achieved in nine steps from 3HT, whereas the “Fibonacci
synthesis” requires 11 steps. However, the significantly easier
purification for high regioregularity and monodispersity is clearly
worth the additional effort. This strategy allows the synthesis of
(3HT)8 in large quantities (e.g., 5 g) and easy purification by
reprecipitation from CHCl3/acetone and column chromatog-
raphy. Admittedly, the purification of longer oligomers by
column chromatography becomes more difficult but can readily
be performed by simple preparative GPC without excessive run
times.
The synthesized oligomers were pure with respect to
molecular weight [polydispersity index (PDI) = 1.00], as
evidenced by matrix-assisted laser desorption ionization time-
of-flight mass spectrometry (MALDI-TOF-MS) measurements
(Figure 2), clearly indicating their “monodispersity”. In
comparison, typical “low-molecular-weight polymers” prepared
by Kumada polycondensation have PDIs of 1.15−1.2519 and
show a broad distribution over the different chain lengths in
In general, then, the synthesis of the oligomer (3HT)n is
i
achieved by the coupling of Br-(3HT)n and (3HT)n -SnBu3.
i−1
i−2
The progression of the oligomer length can be described by ni =
ni−1 + ni−2. Therewith, the presented synthetic strategy follows
the Fibonacci numbers, yielding (3HT)n with n = (1,) 1, 2, 3, 5, 8,
13, and 21.
The progress of the oligomer growth in this presented strategy
is slower than in a geometric sequence (i.e., coupling of
B
dx.doi.org/10.1021/ja4057932 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
guished.21 The NMR spectra of all oligomers showed no such
signals of significant intensity.
The attempted synthesis of oligomers longer than (3HT)21
was complicated because of the low solubility of (3HT)13 in
tetrahydrofuran (THF) at −78 °C. The solubility of (3HT)8 was
found to be quite poor already, requiring low concentrations in
the metalation step. The isomerization of thienyllithium
derivatives between the 2- and 5-position at elevated temper-
ature22,23 prohibited performing the reaction at higher temper-
atures (e.g., room temperature) where improved solubility in
THF was observed.
Therefore, to allow for regioselective metalation of (3HT)13
and longer oligomers, a bulky silyl protecting group was
introduced at the 2-position of the oligothiophene. We found
this to force deprotonation to occur solely in the corresponding
unprotected 5-position. This additionally allowed for the use of
excess LDA to drive the reaction to completion. A number of
different silyl derivatives were explored. As reported previously
by Spivey et al.,13 the trimethylsilyl (TMS) and triethylsilyl
(TES) groups were quite labile and therefore not suitable as a
first choice.13 By contrast, the tert-butyldimethylsilyl (TBDMS)
and (2,3-dimethylbutan-2-yl)dimethylsilyl (TEXDMS) groups
were both very stable under the required reaction and
purification conditions and could easily be cleaved off afterward
with tetra-n-butylammonium fluoride (TBAF). We favored the
TEXDMS group over the TBDMS group because its steric
demand is slightly higher and it could also conveniently be
removed using TBAF in THF at room temperature.
Figure 2. MALDI-TOF mass spectra of oligo-3-hexylthiophenes.
(3HT)n with n > 8 feature minor signals in the low-molecular-weight
regime up to m/z 1000 due to fragments and aggregates of the matrix
employed.
MALDI-TOF-MS (Figure S2). To be able to appreciate the
meaning of such a PDI in the low-molecular-weight regime, a
simple example is illustrative. A theoretical sample containing the
same number of molecules of each oligomer (3HT)n with n = 5,
6, ... 21 would have a number-average molecular weight (Mn) of
2.16 kg/mol with a PDI of 1.14 and generally would be
considered nearly “monodisperse”.
The >99% regioregularity of the oligomers and the
regioselectivity of the bromination and stannylation were proven
1
The synthesis of the longer oligomers utilizing the protecting
group is shown in Scheme 2. The silyl group was introduced into
by NMR spectroscopy. The signals in the H NMR spectrum
were assigned to a regioregular oligomer in accordance with the
literature, as shown in Figure 3 for (3HT)8.12,20,21 Regioregular
oligomers can be discerned from regio-irregular ones by both the
Scheme 2. Introduction of a Silyl Protecting Group at the
Stage of (3HT)5 and (3HT)8 Followed by the Synthesis of
(3HT)n, n = 14, 22, 36
1
aromatic and benzylic H signals. In the HH configuration,
benzylic methylene signals occur at 2.45−2.55 ppm, whereas in
TT they are at 2.57−2.61 ppm. For the aromatic protons, the four
possible triads HT−HT (6.98 ppm), TT−HT (7.00 ppm), HT−
HH (7.02 ppm), and TT−HH (7.05 ppm) can be distin-
the oligomers by Stille coupling of TEXDMS-(3HT)-SnBu3 with
Br-(3HT)5 and Br-(3HT)8 to yield TEXDMS-(3HT)6 and
TEXDMS-(3HT)9, respectively, which were subsequently
regioselectively stannylated. The resulting stannyl monomers
were both contaminated with significant amounts of starting
materials and were used crude in both cases. We note that the
impurities were readily removed following cross-coupling as a
result of the significant differences in molecular weight relative to
1
Figure 3. Assignment of the signals in the H NMR spectrum (700
MHz, CDCl3) of (3HT)8. indicating high isomeric purity.
C
dx.doi.org/10.1021/ja4057932 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
the desired product. TEXDMS-(3HT)6-SnBu3 was cross-
coupled with Br-(3HT)8 to yield TEXDMS-(3HT)14, which
was then regioselectively stannylated at 0 °C. The crude
TEXDMS-(3HT)9-SnBu3 was cross-coupled with Br-(3HT)13
to yield TEXDMS-(3HT)22 and, after deprotection and
bromination, Br-(3HT)22. Finally, the reaction of Br-(3HT)22
and TEXDMS-(3HT)14-SnBu3 under Pd(PPh3)4 catalysis
yielded (3HT)36 in 34% yield after deprotection and purification.
The conjugation of the P3HT backbone is often thought to
cause a semistiff conformation, resulting in an increased
hydrodynamic radius and therefore an overestimation of the
molecular weight in GPC in comparison with polystyrene
standards.6,24 As the synthesized oligomers were monodisperse
and had a precise molecular weight, they allowed verification of
this molecular-weight overestimation by comparison of MALDI-
TOF-MS and GPC data (Figure 4). In agreement with
AUTHOR INFORMATION
Corresponding Author
m.heeney@imperial.ac.uk
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Prof. A. D. Schluter (ETH Zurich) and his group for
̈
̈
support and use of equipment, the Mass Spectrometry Service of
the Laboratories of Organic Chemistry at ETH Zurich for
performing the MALDI-TOF-MS measurements, and Moreno
Aitor and Rene Verel (ETH Zurich) for their assistance with
NMR measurements. We acknowledge Natalie Stingelin
(Imperial College London) for her stimulating interest and C.
Luscombe (University of Washington, USA) for providing the
low-molecular-weight P3HT sample.
̈
̈
REFERENCES
■
(1) Handbook of Oligo- and Polythiophenes; Fichou, D., Ed.; Wiley-
VCH: Weinheim, Germany, 1999.
(2) Guillerez, S.; Bidan, G. Synth. Met. 1998, 93, 123.
(3) Ungar, G.; Stejny, J.; Keller, A.; Bidd, I.; Whiting, M. C. Science
1985, 229, 386.
(4) Craig, S. R.; Hastie, G. P.; Roberts, K. J.; Sherwood, J. N. J. Mater.
Chem. 1994, 4, 977.
(5) (a) Di Maria, F.; Gazzano, M.; Zanelli, A.; Gigli, G.; Loiudice, A.;
Rizzo, A.; Biasiucci, M.; Salatelli, E.; D’ Angelo, P.; Barbarella, G.
Macromolecules 2012, 45, 8284. (b) Zhang, L.; Colella, N. S.; Liu, F.;
Trahan, S.; Baral, J. K.; Winter, H. H.; Mannsfeld, S. C. B.; Briseno, A. L.
J. Am. Chem. Soc. 2012, 135, 844.
(6) Izumi, T.; Kobashi, S.; Takimiya, K.; Aso, Y.; Otsubo, T. J. Am.
Chem. Soc. 2003, 125, 5286.
(7) McCullough, R. D. Adv. Mater. 1998, 10, 93.
Figure 4. (left) Overlay of GPC traces of oligomers n = 5 to 21, shown in
black for the ones following the Fibonacci numbers (5, 8, 13, 21) and in
gray for some intermediate oligomers (6, 10, 11, 16, 18). (right)
Comparison of the molecular weights determined by MALDI-TOF-MS
(Mn) and GPC peak molecular weights (Mp) calibrated with polystyrene
standards.
(8) Kirschbaum, T.; Briehn, C. A.; Bauerle, P. J. Chem. Soc., Perkin
̈
Trans. 1 2000, 1211.
(9) Turner, D. J.; Anemian, R.; Mackie, P. R.; Cupertino, D. C.; Yeates,
S. G.; Turner, M. L.; Spivey, A. C. Org. Biomol. Chem. 2007, 5, 1752.
(10) Kirschbaum, T.; Azumi, R.; Mena-Osteritz, E.; Bauerle, P. New J.
Chem. 1999, 23, 241.
̈
measurements reported by other authors, the GPC molecular
weight was overestimated by a factor of 1.67 relative to the ones
measured by MALDI-TOF-MS6 and NMR spectroscopy.23
In conclusion, the approach presented in this work, to
synthesize regioregular oligo(3-hexylthiophene)s via “Fibonac-
ci’s route”, allows for the synthesis of large quantities of long
(3HT)n [e.g., >1 g for (3HT)13] with high regioregularity. This
route requires a marginally higher number of synthetic steps to
reach a certain oligomer length compared with an approach
following, for example, a geometric sequence, but it simplifies the
purification significantly. With the presented method, regiore-
gular oligomers up to a length of 21 repeat units were
synthesized. A small adaption of this method, namely, the
introduction of a silyl protecting group at the oligomers’ “2-
position”, allowed stannylation of, for example, (3HT)14 by
enforcing deprotonation selectively in the “5-position”. By these
means, a regioregular oligomer with 36 repeat units was
successfully synthesized. The thermal and structural character-
istics of the compounds synthesized are described in an
accompanying paper.16
(11) Kirschbaum, T.; Bauerle, P. Synth. Met. 2001, 119, 127.
̈
(12) Ferrari, M.; Mucci, A.; Schenetti, L.; Malmusi, L. Magn. Reson.
Chem. 1995, 33, 657.
(13) Spivey, A. C.; Turner, D. J.; Turner, M. L.; Yeates, S. Synlett 2004,
111.
(14) Barth, M.; Guilerez, S.; Bidan, G.; Bras, G.; Łapkowski, M.
Electrochim. Acta 2000, 45, 4409.
(15) Krebs, F. C.; Spanggaard, H. Sol. Energy Mater. Sol. Cells 2005, 88,
363.
(16) Koch, F. P. V.; Heeney, M.; Smith, P. J. Am. Chem. Soc. 2013,
DOI: 10.1021/ja405792b.
(17) (a) Mishra, S. P.; Palai, A. K.; Srivastava, R.; Kamalasanan, M. N.;
Patri, M. J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 6514. (b) Biniek,
L.; Fall, S.; Chochos, C. L.; Anokhin, D. V.; Ivanov, D. A.; Leclerc, N.;
́ ̂
Leveque, P.; Heiser, T. Macromolecules 2010, 43, 9779.
(18) Gondo, S.; Goto, Y.; Era, M. Mol. Cryst. Liq. Cryst. 2007, 470, 353.
(19) Sheina, E. E.; Liu, J.; Iovu, M. C.; Laird, D. W.; McCullough, R. D.
Macromolecules 2004, 37, 3526.
(20) (a) Barbarella, G.; Bongini, A.; Zambianchi, M. Macromolecules
1994, 27, 3039. (b) Bras, J.; Pep
39, 57.
́
in-Donat, B. Magn. Reson. Chem. 2001,
(21) Chen, T.-A.; Wu, X.; Rieke, R. D. J. Am. Chem. Soc. 1995, 117, 233.
(22) Ritter, S. K.; Noftle, R. E. Chem. Mater. 1992, 4, 872.
ASSOCIATED CONTENT
(23) Rieke, R. D.; Kim, S.-H.; Wu, X. J. Org. Chem. 1997, 62, 6921.
(24) Zhang, R.; Li, B.; Iovu, M. C.; Jeffries-El, M.; Sauve, G.; Cooper, J.;
́
Jia, S.; Tristram-Nagle, S.; Smilgies, D. M.; Lambeth, D. N.;
■
S
* Supporting Information
Experimental procedures and additional data. This material is
McCullough, R. D.; Kowalewski, T. J. Am. Chem. Soc. 2006, 128, 3480.
available free of charge via the Internet at http://pubs.acs.org.
D
dx.doi.org/10.1021/ja4057932 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX