(E)-5,5′-Di(thiophen-2-yl)-3,3′-bi[thiophen-3(2H)-ylidene]-2, 2′-diones - From conspicuous blue impurities to quasi-metallic golden-bronze crystals

Article abstract of DOI:10.1039/c3ob40373d

(E)-5,5′-Di(thiophen-2-yl)-3,3′-bi[thiophen-3(2H)-ylidene]-2, 2′-dione (1a) has caught the eye of several chemists who have encountered it as a trace by-product in a variety of reactions. We report here the first deliberate synthesis of this intense blue dye (λ_max 635 nm; ε_max 32400 L cm^-1 mol^-1). We prove on the basis of predictions made with TDDFT that 1a has previously been misassigned as a structural isomer and we correct the visible absorption data that was reported. Derivatives of 1a equipped with pentyl (1b) and butylsulfanyl (1c) groups are readily prepared with our divergent synthetic route (3-7% yield over five steps). Crystals of 1b and 1c exhibit an attractive quasi- metallic golden-bronze lustre. This effect is rare in molecular crystals and is observed only with the most intense dyes. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ob40373d

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
(E)-5,5’-Di(thiophen-2-yl)-3,3’-bi[thiophen-3(2H)-
ylidene]-2,2’-diones—from conspicuous blue impurities
to “quasi-metallic” golden-bronze crystals†
Cite this: DOI: 10.1039/c3ob40373d
Nicholas R. Evans* and Andrew J. P. White
(E)-5,5’-Di(thiophen-2-yl)-3,3’-bi[thiophen-3(2H)-ylidene]-2,2’-dione (1a) has caught the eye of several
chemists who have encountered it as a trace by-product in a variety of reactions. We report here the first
deliberate synthesis of this intense blue dye (λmax 635 nm; εmax 32 400 L cm−1 mol−1). We prove on the
basis of predictions made with TDDFT that 1a has previously been misassigned as a structural isomer and
we correct the visible absorption data that was reported. Derivatives of 1a equipped with pentyl (1b)
and butylsulfanyl (1c) groups are readily prepared with our divergent synthetic route (3–7% yield over
five steps). Crystals of 1b and 1c exhibit an attractive “quasi-metallic” golden-bronze lustre. This effect is
rare in molecular crystals and is observed only with the most intense dyes.
Received 20th February 2013,
Accepted 30th April 2013
DOI: 10.1039/c3ob40373d
www.rsc.org/obc
silica TLC plates gradually turned deep blue in air but not
Introduction
under argon.1 In a plausible mechanism boronic acid 2 is oxi-
The blue dye 1a first caught our eye as a trace but conspicuous dised to 5-(thiophen-2-yl)thiophen-2(3H)-one (3) or a tauto-
by-product of the preparation of 2,2′-bithiophene-5-boronic mer.2 Oxidative coupling of ketone 3 yields dimer 4 (or a
acid (2) (Scheme 1).1 Crude batches of 2 were coloured pale tautomer) which is oxidised further to blue dye 1a. This
green to deep blue and we observed that spots of purified 2 on mechanism is supported by the earlier studies reviewed below.
A similar mechanism was recently reported to be very efficient
in the case of naphthalene-1-boronic acids.3
In 1964 Hörnfeldt reported that ketone 3 is the unstable
product of reacting boronic acid 2 with hydrogen peroxide.2
Ketone 3 rapidly oxidised in air to a blue dye that Hörnfeldt
did not investigate further.2
In 1954 Kosak et al. had noted that 5-phenylthiophen-2-
(3H)-one (5) degraded to tars and blue dye 6 when solutions or
solid samples were exposed to air (Scheme 2).4 Kosak et al.
hinted at the intensity of the dye by observing that “the
Scheme 1
Department of Chemistry, Imperial College London, London SW7 2AZ,
United Kingdom. E-mail: n.evans@imperial.ac.uk
†Electronic supplementary information (ESI) available: Deconvoluted visible
absorption spectra of the blue dyes (1). DFT-optimised geometries of the fully
characterised compounds, structure 9 and blue dye 1 with R = OMe. TDDFT data
for the blue dyes (1, including R = OMe) and structure 9. Thermal ellipsoid plots
and other depictions of the crystal structures of blue dyes 1b and 1c. Proton and
carbon-13 NMR spectra of the fully characterised compounds and the insepar-
able mixtures. CCDC 921750 and 921751. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c3ob40373d
Scheme 2
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
noted the potential complication of tautomerism in these reac-
tions.6 We present here clear evidence that dimer 4 (or a tauto-
mer) is a precursor to blue dye 1a (Scheme 1) under our
reaction conditions.
We sought a rational synthesis of blue dye 1a to prove the
structure and study the intense visible absorption. Our diver-
gent synthetic route was designed not only to facilitate the
preparation of derivatives of 1a with improved physical pro-
perties but also to be an improvement on earlier synthetic
methods.6,7 We also sought to prove that 1a has previously
been misassigned5 as 9 and to correct the inaccurate character-
isation data that was published.
Finally we note that the analogous blue dyes 14a8 and 14b9
were reported recently. The visible absorptions of 14a and 14b
are similar to that of 1a except the absorption maxima occur at
shorter wavelengths.
Scheme 3
amount of oxidized material responsible for the color must be
very small.”4 The structure of blue dye 6 was supported by the
established oxidation of 5-phenylthiophen-3(2H)-one (7, a
structural isomer of 5) to dye 8 (a structural isomer of 6).4
Structure 9, analogous to that of dye 8, was assigned to a
blue by-product of the Friedel–Crafts acylation of thiophene
with succinyl chloride (Scheme 3) by Merz and Ellinger in
1991.5 Despite observing that the blue by-product was not
formed in the absence of succinyl chloride, Merz and Ellinger
proposed that 9 was a result of oxidative oligomerisation of
thiophene.5 We report here that the characterisation data for
the blue by-product isolated by Merz and Ellinger is mostly
consistent with our data for 1a. Accordingly we propose that
keto acid 10, a product of the Friedel–Crafts acylation, under-
went a Paal–Knorr thiophene synthesis4 to yield ketone 3
which was oxidised to blue dye 1a.
Some observations inconsistent with our proposed mechan-
ism were reported by Hörnfeldt in 1967.6 5-tert-Butylthiophen-
2(5H)-one (11) degrades in air to dimer 12 and not to red dye
13 (Scheme 4).6 Ketone 11 can be oxidised to red dye 13 with
potassium ferricyanide yet dimer 12 is still not oxidised under
these conditions.6 Hörnfeldt concluded that dimer 12 is not
an intermediate in the oxidation of ketone 11 to dye 13 but
Results and discussion
Our divergent synthetic route to the blue dyes (1) begins with
making the key 3,3′ C–C bond. 3,3′-Bithiophene (15) was pre-
pared in 94% yield by using a Suzuki cross-coupling method.10
Lithiation of 15 is reasonably selective for the 2,2′ positions
over the 5,5′ positions (Scheme 5).11 The 2,2′-dilithio derivative
was converted into the corresponding Grignard reagent and
quenched with tert-butyl peroxybenzoate to give di-tert-butoxy
167 in 32% yield. Conversion to the Grignard reagent is necess-
ary to prevent substitution at the peroxyester carbonyl. This
route to 16 is a significant improvement on the previously pub-
lished method.7
Functionalisation of 3,3′-bithiophenes is problematic in
general because of the instability of derivatives with common
synthetic handles.6,11 We were unable to isolate dibromide 17
from the reaction of 16 with NBS due to rapid decomposition
of the product mixture (Scheme 6). Diboronic ester 18 is more
stable and can be isolated in 73% yield after recrystallisation
from hexane. Unfortunately 18 suffers from steady
Scheme 4
Scheme 5
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
and 16 deteriorated under these conditions to several highly
coloured and markedly polar products that were easily washed
out of 1a with methanol and ethanol. In this way 1a was iso-
lated in 14% yield over three steps from 16.
Unsurprisingly 1a is very poorly soluble in common organic
solvents. Consequently we put our divergent synthetic route to
use in the preparation of derivatives equipped with 5-pentyl
(1b) and 5-butylsulfanyl (1c) groups from 2-bromothiophenes
19b and 19c respectively (Scheme 7). The syntheses of 1b and
1c were similar to that of 1a in all respects including compar-
able yields. 1b and 1c were also isolated in high purity simply
by washing the crude products with methanol and ethanol
because although 1b and 1c are quite soluble in chloroform
they have poor solubility in these alcohols.
Our synthetic route enables preparation of the blue dyes (1)
in 3–7% yield over five steps from inexpensive commercially
available starting materials. One of the limitations on the
efficiency of our route is the preparation of di-tert-butoxy 16 in
32% yield from 3,3′-bithiophene (15) (Scheme 5). The other
Scheme 6 a2-Isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane.
deboronation under any other conditions. For this reason it is and more significant limitation is the poor stability of the
more efficient not to isolate 18 but to use the crude product boronic ester groups during the second Suzuki cross-coupling
immediately in the next step.6
reaction (Scheme 7). We have proven that diboronic ester 18 is
Suzuki cross-coupling12 of crude 18 and 2-bromothiophene prepared in good yield (Scheme 6). A comparison of the esti-
(19a) was undermined by the poor stability of 18 (Scheme 7). mated yields of the quaterthiophenes (20) with the isolated
Chromatographic purification afforded the desired quaterthio- yields of the subsequent blue dyes (1) indicates that the final
phene (20a) as part of an inseparable mixture along with transformation proceeds in 67% yield or better (Scheme 7).
terthiophene 21a and bithiophene 16, the by-products of
Our characterisation data for blue dye 1a is mostly consist-
deboronation. Yields of the product and two by-products were ent with that reported by Merz and Ellinger for structure 95
estimated from the mass and proton NMR analysis of the (Scheme 3). The infrared absorption and proton NMR spectro-
mixture. Quaterthiophene 20a was present in a disappointing scopic data both match perfectly. (These data are in the Experi-
16% yield over two steps whereas terthiophene 21a was the mental section and ref. 5. 1a is not sufficiently soluble for
major species in 27% yield. Fortunately separation of the carbon-13 NMR analysis.) We were unable to reproduce any of
mixture did not prove to be necessary in order to successfully the data from the reported positive-ion mass spectrum.
complete the synthesis.
However we did find that the negative molecular ion is readily
Cleavage of the tert-butyl ethers in the mixture of 20a, 21a detected, which is not surprising given the presence of carbo-
and 16 was achieved with TFA (Scheme 7). The reaction was nyl groups in 1a. The wavelengths of the visible absorption
performed open to the air to drive oxidation of dimer 4 maxima (λmax) are in close agreement: for example 6315 nm
(Scheme 1), which we expect to be the initial product of 20a and 635 nm for the lowest-energy electronic transitions of
under these conditions, to blue dye 1a. Chlorobenzene was 9 and 1a respectively (Table 1). Unfortunately the molar
found to be the optimum solvent for accomplishing these absorption coefficients of these same maxima (εmax) disagree
transformations cleanly and in a single step. Furthermore 21a entirely: 253 0005 L cm−1 mol−1 and 32 400 L cm−1 mol−1
Scheme 7 aDicyclohexyl(2’,6’-dimethoxybiphenyl-2-yl)phosphine.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Experimental and calculated visible absorption propertiesa
Experiment (chloroform solutions)
TDDFTb (gas phase)
Dye
λmax/nm
˜νmax/cm−1
εmax/L cm−1 mol−1
˜ν0–0/cm−1
f
˜ν0–0/cm−1
f
Scaledc f
9
631d
635
660
676
15 800d
15 700
15 200
14 800
253 000d
32 400
40 200
41 900
n/ae
15 500
15 000
14 400
n/ae
0.45
0.53
0.59
20 000
15 800
15 500
14 900
0.19
0.67
0.77
0.90
0.13
0.45
0.51
0.61
1a
1b
1c
a Transitions from the ground electronic state to the first singlet excited state. b B3LYP/6-311++G**//B3LYP/6-31G*. The alkyl chains of 1b and 1c
were truncated to methyl groups. c Scale factor 0.45/0.67. d DCM solution.5 e Not reported.5
respectively. εmax reported for 9 may have gained an order of accompanying vibrational transitions (the experimentally
magnitude due to a simple error but still the values do not determined ˜ν0–0 in Table 1). (Details of the deconvolution of
agree with our data (although the two sets are proportional). these spectra are in the ESI.†) ˜ν0–0 are slightly lower than the
Fortunately the visible absorption data for derivatives 1b and corresponding wavenumbers of the maxima (˜νmax). The oscil-
1c and TDDFT predictions for the blue dyes (1) and structure 9 lator strength (f) is a measure of the intensity of an electronic
all support our data for 1a and our reassignment of the blue transition that also takes broadening of the absorption profile
by-product isolated by Merz and Ellinger as 1a.
The impact of pentyl (1b) and butylsulfanyl (1c) groups on luted spectra are also in the ESI.†) These experimentally deter-
the electronic states of 1a can be seen in the corresponding mined f are not quite proportional to the corresponding εmax
solution-phase visible absorption spectra (Fig. 1). Transitions We chose to simplify our model chemistry by calculating
into account. (Details of the calculation of f from the deconvo-
.
from the ground electronic state (S0) to the first singlet excited the absorption properties of the blue dyes (1) and structure 9
state (S1) dominate the spectra at wavelengths longer than in the gas phase rather than in a solvent reaction field. We
500 nm.13 λmax and εmax of these transitions both increase in evaluated a few functionals of the time-dependent densities
the order 1a < 1b < 1c (Table 1). However a comparison of the with a large basis set (6-311++G**14–17) and found that
spectra with predictions made with TDDFT requires the use of B3LYP18,19 was the best at reproducing the experimental data
more fundamental measures of the transition energies and for the blue dyes (1) (Table 1). (Details of DFT geometry optim-
intensities. The absorption profiles are broad and with clearly isations and TDDFT calculations are in the ESI.†) The pre-
discernible shoulders in the cases of 1a and 1b (Fig. 1) due to dicted ˜ν0–0 for S1←S0 of the blue dyes (1) are overestimates of
strong coupling of the electronic transition S1←S0 with a the experimental values by no more than 500 cm−1. ˜ν0–0 was
vibrational mode of S1. Deconvolution of the absorption pro- not reported5 for 9 but the predicted value is 4200 cm−1
files provides us with the wavenumbers of S1←S0 without greater than the reported ˜νmax. This discrepancy is much
larger than the maximum error of 500 cm−1 found for the blue
dyes (1). ˜ν0–0 for S1←S0 (i.e. the lowest-energy vibronic tran-
sition) cannot normally be greater than ˜νmax so it is reasonable
to conclude that the blue by-product isolated by Merz and
Ellinger was 1a and not 9.
The predicted f for S1←S0 of the blue dyes (1) are much
larger than the experimental values (Table 1). Fortunately the
two sets are proportional and the predicted values can be
scaled to agree well with the experimental values. This consist-
ency between our predicted and experimental data confirms
the accuracy of our experimental value for εmax of 1a over that
reported by Merz and Ellinger.5 The predicted f for S1←S0 of
structure 9 is less than a third of that for 1a. This gives us con-
fidence that 1a is the superior dye of these two structural
isomers.
We did not observe luminescence from any of the blue dyes
(1) in solution or the solid state.20 However crystals of 1b and
1c exhibited an attractive “quasi-metallic”21 golden-bronze
lustre (Fig. 2). This effect is only observed in molecular crystals
when f and the concentration of the molecules are both very
high.22 1b formed small aggregates of golden plates when a
solution was allowed to evaporate very slowly (Fig. 2a). 1c is
Fig. 1 Visible absorption spectra of the blue dyes (1) in chloroform solutions.
highly crystalline and always has a quasi-metallic lustre when
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 True-colour micrographs of typical crystalline samples of blue dyes (a) 1b and (b and c) 1c.
not in solution. 1c readily forms elongated bronze plates that
are much larger and have fewer defects (Fig. 2b and 2c) than
the aggregates of 1b.
Crystal structures of 1b and 1c were determined by means
of X-ray crystallographic analysis. At first glance, the structures
appear to be approximately isomorphous, the largest differ-
ences in their unit cell dimensions being ca. 0.64 Å in the
a axis and ca. 1.4° in the β angle; the difference in the unit cell
volumes is less than 4% (these data are in the Experimental
section). It is thus no surprise that the structures of both mole-
cules look very similar (Fig. S2 and S3 in the ESI†), with each
having a centre of symmetry in the middle of the C(3)vC(3A)
bond linking the two thiophenone units, confirming the E geo-
Fig. 3 Part of one of the stacks of molecules linked by π–π interactions
between thiophene and thiophenone rings present in the crystal of 1b. Inter-
action a has centroid⋯centroid and mean interplanar separations of ca. 3.77 Å
and 3.45 Å respectively, with the two rings inclined by ca. 3°. Adjacent mole-
cules are related by a unit cell translation along the crystallographic b axis
direction.
metry across the double bond.23 The neighbouring thiophe-
none and thiophene rings have anti configurations in both
structures, as do the terminal pentyl (1b) and butylsulfanyl
(1c) sidearms, and the four sulfur-based rings are approxi-
mately coplanar in each case, the maximum deviations from
planarity being ca. 0.06 Å and 0.05 Å in 1b and 1c respectively.
Likewise, the packing of both structures features near identical
sheets of molecules, formed by two unique pairs of Ci-related
weak C–H⋯S contacts (Fig. S6 and S7 in the ESI†).
However, when considering the π–π stacking interactions
(which form the dominant feature of the packing of both
molecules) the distinct differences between the two solid state
structures become clear. In the crystals of 1b, the unique
thiophenone and thiophene rings pack with the centrosymme-
trically related counterpart of the other (Fig. 3) with cen-
troid⋯centroid and mean interplanar separations of ca. 3.77 Å
and 3.45 Å respectively, the two rings being inclined by ca. 3°.
This extends to form a staggered stack of molecules along the
crystallographic b axis direction with adjacent molecules offset
by ca. 8.34 Å (Fig. 4), the separation between the centroids of
the C(3)vC(3A) double bonds in neighbouring molecules in
the stack being ca. 9.02 Å.24 The solid state structure of 1c has
a superficially similar arrangement (Fig. 5), but with important
differences. Here the π–π stacking is between the C(3)vC(3A)
Fig. 4 Part of one of the stacks of π–π linked molecules present in the crystal of
1b viewed perpendicular to the mean plane of the four thiophene/thiophenone
rings of the central molecule (drawn with dark bonds), showing the ca. 8.34 Å
offset of adjacent molecules.
double bond of one molecule and the thiophene ring of two two-part stack seen in 1b) generates a chain of molecules
adjacent molecules; the centroid⋯centroid and mean inter- along the crystallographic a axis direction with a much greater
planar separations are both ca. 3.36 Å, the two planes being degree of overlap between the adjacent molecules (Fig. 6), the
inclined by ca. 3°.25 This three-part stack (in contrast to the offset here being ca. 5.40 Å cf. ca. 8.34 Å in 1b; the separation
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
predictions yet is ultimately highlighted by the rare and attrac-
tive “quasi-metallic” golden-bronze lustre exhibited by crystals
of 1b and 1c.
Experimental section
Et2O was dried over 4 Å molecular sieves. Reactions were per-
formed under nitrogen unless indicated otherwise. Reaction
temperatures are bath temperatures. Silica gel was the station-
ary phase in dry flash chromatography (60 Å pore size;
40–63 μm particle size) and TLC (60 Å pore size; 9.5–11.5 μm
particle size; 0.20 mm layer thickness; fluorescent under
254 nm light). Mp are uncorrected. 1H NMR (400 MHz) and
13C NMR (100 MHz) data are for CDCl3 solutions (δH 7.26;
δC 77.0) unless indicated otherwise (CD2Cl2 δC 53.8). J are
given in Hz. IR data are for neat solid samples. IR absorptions
were assigned by comparing the wavenumbers and intensities
with those of the corresponding gas-phase B3LYP/6-31G*
vibrational modes. (Details of DFT geometry optimisations are
in the ESI.†) UV-vis data are for CHCl3 solutions.
Fig. 5 Part of one of the stacks of molecules linked by π–π interactions
between the central CvC double bond and the thiophene rings present in the
crystal of 1c. The centroid⋯centroid and mean interplanar separations for inter-
action a are both ca. 3.36 Å, with the two rings inclined by ca. 3°. Adjacent
molecules are related by a unit cell translation along the crystallographic a axis
direction.
2,2′-Di-tert-butoxy-3,3′-bithiophene (16)
16 has previously been reported as the product of a different
synthetic route and without complete characterisation.7 A solu-
tion of BuLi in hexane (1.6 M, 7.5 mL, 12 mmol, 2.0 equiv.)
was added dropwise over 8 min to a stirred suspension of
3,3′-bithiophene10 (15) (1.00 g, 6.01 mmol, 1.00 equiv.) in Et2O
(30 mL) at 0 °C. The resultant solution was stirred for 5 min at
0 °C and 1 h at rt. MgBr2·Et2O (3.11 g, 12.0 mmol, 2.00 equiv.)
was added to the reaction at 0 °C and the resultant mixture
was stirred for 20 min at 0 °C. tert-Butyl peroxybenzoate
(2.3 mL, 12 mmol, 2.0 equiv.) was added dropwise over 15 min
and the mixture was stirred for 20 min at 0 °C. The reaction
was quenched at 0 °C with water (8 mL) followed by aq HCl
(1.0 M, 12 mL, 12 mmol, 2.0 equiv.). The mixture was diluted
with CHCl3 and the aq layer was neutralised with concd aq
HCl before it was extracted with CHCl3. The combined organic
layers were washed with saturated NaHCO3, dried with brine
and Na2SO4 and evaporated under reduced pressure. Dry flash
chromatography (hexane) of the crude product recovered some
of the starting material (15) (80 mg, 8%). Repeated dry flash
chromatography (hexane–DCM 1 : 0 changing to 4 : 1) and
recrystallisation (hexane) eventually afforded 16 (604 mg, 32%)
as colourless crystals. Mp 96–98 °C (from hexane) [lit.,7 100 °C
(from EtOH)]. Rf 0.3 (hexane–DCM 3 : 1). νmax/cm−1 3095vw
(thienyl C–H), 2980w (methyl C–H), 1429 (CvC), 1140s (asym-
metric C–O–C), 1086 (thienyl C–H in-plane bend), 842s
Fig. 6 Part of one of the stacks of π–π linked molecules present in the crystal of
1c viewed perpendicular to the mean plane of the four thiophene/thiophenone
rings of the central molecule (drawn with dark bonds), showing the ca. 5.40 Å
offset of adjacent molecules.
between the centroids of the C(3)vC(3A) double bonds in
neighbouring molecules in the stack is ca. 6.36 Å cf. ca. 9.02 Å
in 1b.
It is interesting to note that despite the evidently closer
packing within the π–π stacks in the structure of 1c compared
to 1b, the packing efficiency for 1c (ca. 70.6%) is slightly lower
than that for 1b (ca. 72.1%).26
Conclusion
We have combined conventional organic synthesis with (O–tBu), 828 (C–S) and 698s (thienyl C–H out-of-plane bend).
modern computational chemistry in our contemporary δH 1.33 (18 H, s, 2 × OtBu), 6.70 (2 H, d, J 6.0, 4-H and 4′-H or
approach to resolving the assignment and characterisation of 5-H and 5′-H) and 7.40 (2 H, d, J 6.0, 4-H and 4′-H or 5-H and
1a. We have capitalised on the predictive power of modern 5′-H). δC 28.1 (OCMe3), 82.3 (OCMe3), 114.2 (5-C and 5′-C),
TDDFT in proving that 1a has previously been misassigned as 122.8 (3-C and 3′-C), 126.2 (4-C and 4′-C) and 154.2 (2-C and
a structural isomer. The marked intensity of this new family of 2′-C). m/z (EI+) 310.1065 (M+, C16H22O2S2 requires 310.1056),
+
t
blue dyes (1) has been quantified with the accustomed experi- 310 (24%, M+), 254 (20, [M − Bu + H]+), 198 (100, [M − 2tBu +
mental measurements and authenticated with modern TDDFT 2H]+) and 137 (76, [M − 2tBu − 2S + 5H]+).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
2,2′-Di-tert-butoxy-5,5′-bis(4,4,5,5-tetramethyl-1,3,2-
dioxaborolan-2-yl)-3,3′-bithiophene (18)
desired quaterthiophene (20), a terthiophene by-product (21)
and 16 that was used without further purification.
The reaction with 2-bromothiophene (19a) (0.20 mL,
2.1 mmol, 13 equiv.) afforded a pale green oil (37 mg. 1H
NMR: ca. 28 mol% 20a, 47 mol% 21a, 24 mol% 16. Estimated
yields over two steps from 16: 16% 20a, 27% 21a, 14% 16). Rf
0.4 (hexane–CHCl3 3 : 1).
A solution of BuLi in hexane (1.6 M, 0.60 mL, 0.96 mmol, 3.0
equiv.) was added dropwise over 3 min to a stirred solution of
16 (100 mg, 322 μmol, 1.00 equiv.) in Et2O (10 mL) at −78 °C.
The solution was stirred for 2 min at −78 °C and 1 h at rt.
2-Isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (0.30 mL,
1.5 mmol, 4.6 equiv.) was added to the solution at −78 °C and
the resultant suspension was allowed to warm to rt overnight.
The reaction was quenched with aq HCl (0.2 M, 10 mL,
2 mmol, 6 equiv.) and diluted with CHCl3 and water. The aq
layer was extracted with CHCl3 and the combined organic
layers were dried with MgSO4 and evaporated under reduced
pressure. The crude product was recrystallised twice from
hexane to afford 18 (132 mg, 73%) as pale yellow needles.
Mp 156 °C (from hexane) decomposition with gas evolution.
Rf 0.5 (CHCl3). νmax/cm−1 2979w (methyl C–H), 1536 (CvC),
1429s (CvC), 1349s (C–B), 1260 (asymmetric O–B–O), 1121vs
(thienyl C–H in-plane bend), 853s (C–O) and 664s (C–B(–O)2
out-of-plane deformation). δH 1.32 (24 H, s, 8 × Me), 1.34
(18 H, s, 2 × OtBu) and 7.91 (2 H, s, 4-H and 4′-H). δC
(125 MHz; CD2Cl2) 24.9 (OCMe2CMe2O), 28.3 (OCMe3), 83.4
(OCMe3), 84.2 (OCMe2CMe2O), 117.4 (br, 5-C and 5′-C), 124.4
2′′,5′-Di-tert-butoxy-2,2′:4′,3′′:5′′,2′′′-quaterthiophene (20a) δH
1.41 (18 H, s, 2 × OtBu), 6.98–7.01 (2 H, m, 4-H and 4′′′-H), 7.07
(2 H, dd, J 1.0 and 3.5, 3-H and 3′′′-H), 7.15 (2 H, dd, J 1.0 and
5.0, 5-H and 5′′′-H) and 7.61 (2 H, s, 3′-H and 4′′-H). m/z (EI+)
474.0823 (M+, C24H26O2S4 requires 474.0810), 474 (4%, M+),
+
418 (3, [M − tBu + H]+), 362 (65, [M − 2tBu + 2H]+) and 301 (15,
[M − 2tBu − 2S + 5H]+).
2′′,5′-Di-tert-butoxy-2,2′:4′,3′′-terthiophene (21a) δH 1.36 (9 H,
s, OtBu), 1.37 (9 H, s, OtBu), 6.72 (1 H, d, J 6.3, 4′′-H or 5′′-H),
6.98–7.01 (1 H, m, 4-H), 7.07 (1 H, dd, J 1.0 and 3.5, 3-H), 7.14
(1 H, dd, J 1.0 and 5.1, 5-H), 7.41 (1 H, d, J 6.3, 4′′-H or 5′′-H)
and 7.61 (1 H, s, 3′-H). m/z (EI+) 392.0944 (M+, C20H24O2S3
+
requires 392.0933), 392 (4%, M+), 336 (5, [M − Bu + H]+), 280
t
(100, [M − 2tBu + 2H]+) and 219 (29, [M − 2tBu − 2S + 5H]+).
The reaction with 2-bromo-5-pentylthiophene27 (19b)
(0.20 mL, 1.1 mmol, 6.9 equiv.) afforded a green oil (50 mg. 1H
NMR: ca. 57 mol% 20b, 37 mol% 21b, 7 mol% 16. Estimated
yields over two steps from 16: 33% 20b, 22% 21b, 4% 16). Rf
0.3 and 0.4 (hexane–CHCl3 3 : 1).
(3-C and 3′-C), 137.5 (4-C and 4′-C) and 161.2 (2-C and 2′-C).
+
m/z (CI+) 563.2865 ([M
+
H]+, C28H45B2O6S2 requires
563.2838), 563 (10%, [M + H]+), 507 (59, [M − Bu + 2H]+), 451
t
t
2′′,5′-Di-tert-butoxy-5,5′′′-dipentyl-2,2′:4′,3′′:5′′,2′′′-quaterthio-
phene (20b) δH 0.90 (6 H, t distorted by virtual coupling, J ca. 7,
2 × pentyl 5-H3), 1.31–1.42 (8 H, m, 4 × CH2), 1.39 (18 H, s,
2 × OtBu), 1.62–1.74 (4 H, m, 2 × CH2), 2.77 (4 H, t, J 7.6, 2 × pentyl
1-H2), 6.64 (2 H, d, J 3.3, 3-H and 3′′′-H or 4-H and 4′′′-H), 6.87
(2 H, d, J 3.3, 3-H and 3′′′-H or 4-H and 4′′′-H) and 7.51 (2 H, s,
(100, [M − 2tBu + 3H]+), 381 (22, [M − Bu − BO2C6H12 + 3H]+)
and 325 (24, [M − 2tBu − BO2C6H12 + 3H]+).
(E)-5,5′-Di(thiophen-2-yl)-3,3′-bi[thiophen-3(2H)-ylidene]-
2,2′-diones (1)
A solution of BuLi in hexane (1.6 M, 0.30 mL, 0.48 mmol, 3.0 3′-H and 4′′-H).
equiv.) was added dropwise over 2 min to a stirred solution of
2′′,5′-Di-tert-butoxy-5-pentyl-2,2′:4′,3′′-terthiophene (21b) δH
16 (50 mg, 0.16 mmol, 1.0 equiv.) in Et2O (5 mL) at −78 °C. 0.90 (3 H, t distorted by virtual coupling, J ca. 7, pentyl 5-H3),
The solution was stirred for 3 min at −78 °C and 1 h at rt. 1.31–1.42 (4 H, m, 2 × CH2), 1.35 (9 H, s, OtBu), 1.36 (9 H, s,
2-Isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (0.15 mL, OtBu), 1.62–1.74 (2 H, m, CH2), 2.77 (2 H, t, J 7.6, pentyl 1-H2),
0.74 mmol, 4.6 equiv.) was added to the solution at −78 °C 6.64 (1 H, d, J 3.3, 3-H or 4-H), 6.71 (1 H, d, J 6.0, 4′′-H or
and the resultant suspension was allowed to warm to rt over- 5′′-H), 6.87 (1 H, d, J 3.3, 3-H or 4-H), 7.41 (1 H, d, J 6.0, 4′′-H
night. The reaction was quenched with aq HCl (0.2 M, 5 mL, or 5′′-H) and 7.51 (1 H, s, 3′-H). m/z (EI+) 462.1739 (M+,
1 mmol, 6 equiv.) and diluted with CHCl3 and water. The aq C25H34O2S3 requires 462.1715), 462 (6%, M+), 350 (100,
+
layer was extracted with CHCl3 and the combined organic [M − 2tBu + 2H]+) and 289 (17, [M − 2tBu − 2S + 5H]+).
layers were dried with MgSO4 and evaporated under reduced
pressure. The crude product (18) (89–95 mg) was used without (19c) (0.20 mL, 1.1 mmol, 6.9 equiv.) afforded a dark green oil
further purification.
(45 mg. 1H NMR: ca. 20 mol% 20c, 46 mol% 21c, 35 mol% 16.
The reaction with 2-bromo-5-(butylsulfanyl)thiophene28
Water (0.2 mL) was added to a suspension of crude 18 Estimated yields over two steps from 16: 12% 20c, 28% 21c,
(89–95 mg), a 2-bromothiophene (19) (0.20 mL), Pd(OAc)2 21% 16). Rf 0.3 (hexane–CHCl3 3 : 1).
(2.2 mg, 10 μmol, 6.1 mol%), dicyclohexyl(2′,6′-dimethoxybi-
2′′,5′-Di-tert-butoxy-5,5′′′-bis(butylsulfanyl)-2,2′:4′,3′′:5′′,2′′′-
phenyl-2-yl)phosphine (8.2 mg, 20 μmol, 12 mol%) and K3PO4 quaterthiophene (20c) δH 0.91 (6 H, t, J 7.4, 2 × butyl 4-H3),
(86 mg, 0.41 mmol, 2.5 equiv.) in THF (2 mL). The mixture was 1.37–1.48 (4 H, m, 2 × butyl 3-H2), 1.40 (18 H, s, 2 × OtBu),
stirred vigorously for 3–5 d at rt before it was diluted with 1.57–1.66 (4 H, m, 2 × butyl 2-H2), 2.80 (4 H, t, J 7.3, 2 × butyl
CHCl3 and water. The aq layer was extracted with CHCl3 and 1-H2),6.91 (2 H, d, J 3.8, 3-H and 3′′′-H or 4-H and 4′′′-H), 6.98
the combined organic layers were dried with MgSO4 and evap- (2 H, d, J 3.8, 3-H and 3′′′-H or 4-H and 4′′′-H) and 7.56 (2 H, s,
orated under reduced pressure. The crude product was puri- 3′-H and 4′′-H).
fied using dry flash chromatography (hexane–DCM 1 : 0
2′′,5′-Di-tert-butoxy-5-(butylsulfanyl)-2,2′:4′,3′′-terthiophene (21c)
changing to 4 : 1) to afford an inseparable mixture of the δH 0.91 (3 H, t, J 7.4, butyl 4-H3), 1.36 (9 H, s, OtBu), 1.37 (9 H,
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
s, OtBu), 1.37–1.48 (2 H, m, butyl 3-H2), 1.57–1.66 (2 H, m, 87.139(19)°, V = 590.9(3) Å3, Z = 1 [Ci symmetry], Dc = 1.407 g
butyl 2-H2), 2.80 (2 H, t, J 7.3, butyl 1-H2), 6.71 (1 H, d, J 5.9, cm−3, μ(Cu-Kα) = 3.866 mm−1, T = 173 K, lustrous bronze
4′′-H or 5′′-H), 6.91 (1 H, d, J 3.5, 3-H or 4-H), 6.98 (1 H, d, plates, Oxford Diffraction Xcalibur PX Ultra diffractometer;
J 3.5, 3-H or 4-H), 7.40 (1 H, d, J 5.9, 4′′-H or 5′′-H) and 7.57 2253 independent measured reflections (Rint = 0.0500), F2
+
(1 H, s, 3′-H). m/z (EI+) 480.1289 (M+, C24H32O2S4 requires refinement,29 R1(obs)
= 0.0689, wR2(all) = 0.2082, 1864
480.1280), 480 (5%, M+), 424 (6, [M − Bu + H]+), 368 (100, independent observed absorption-corrected reflections [|Fo| >
t
[M − 2tBu + 2H]+) and 307 (12, [M − 2tBu − 2S + 5H]+).
TFA (10 drops every 30 min) was added to a stirred solution
of an inseparable mixture of 20, 21 and 16 (37–50 mg) in PhCl
(5 mL) until the solution started to turn blue. The mixture was
stirred vigorously for 2–3 d open to the air before it was con-
centrated under reduced pressure. The residue was succes-
sively washed with MeOH, EtOH and hexane. A CHCl3 solution
of the remaining solid was filtered and evaporated under
reduced pressure to afford 1.
4σ(|Fo|), 2θmax = 146°], 145 parameters; CCDC 921750.
(E)-5,5′-Bis[5-(butylsulfanyl)thiophen-2-yl]-3,3′-bi[thiophen-
3(2H)-ylidene]-2,2′-dione (1c)
(8 mg, 9% over three steps from 16) as bronze crystals. Mp
153–154 °C (from hexane–CHCl3) partial decomposition. Rf 0.3
(hexane–CHCl3 3 : 1). λmax/nm 676 (ε/L cm−1 mol−1 41 900) and
374 (17 800). νmax/cm−1 3085vw (thienyl C–H), 2925w (alkyl
C–H), 1656 (CvO), 1525s (CvC), 1404s (CvC), 1171s (thienyl
C–H in-plane bend), 830 (thienyl C–S) and 780 (thienyl C–H
out-of-plane bend). δH 0.94 (6 H, t, J 7.4, 2 × butyl 4-H3), 1.46
(4 H, sextet, J 7.4, 2 × butyl 3-H2), 1.69 (4 H, quintet, J 7.4, 2 ×
butyl 2-H2), 2.97 (4 H, t, J 7.4, 2 × butyl 1-H2), 6.99 (2 H, d,
J 3.9, 2 × thienyl 3-H or 4-H), 7.19 (2 H, d, J 3.9, 2 × thienyl 3-H
or 4-H) and 8.00 (2 H, s, 4-H and 4′-H). δC 13.6 (butyl 4-C), 21.7
(butyl 3-C), 31.3 (butyl 2-C), 37.6 (butyl 1-C), 117.1 (4-C and
4′-C), 129.8 (thienyl 3-C or 4-C), 131.1 (thienyl 3-C or 4-C),
134.5 (3-C or 5-C or thienyl 2-C or 5-C), 138.5 (3-C or 5-C or
thienyl 2-C or 5-C), 141.3 (3-C or 5-C or thienyl 2-C or 5-C),
(E)-5,5′-Di(thiophen-2-yl)-3,3′-bi[thiophen-3(2H)-ylidene]-
2,2′-dione (1a)
(8 mg, 14% over three steps from 16) as a black powder. 1a has
previously been reported but misassigned as structure 9.5 1a is
not sufficiently soluble for 13C NMR analysis. Mp > 250 °C. Rf
0.2 (hexane–CHCl3 3 : 1). λmax/nm 635 (ε/L cm−1 mol−1 32 400),
604sh (28 600), 440sh (2400), 403sh (4700), 386sh (6500), 357
(21 400), 346 (21 300) and 314sh (10 000). νmax/cm−1 3088w
(thienyl C–H), 1658s (CvO), 1532s (CvC), 1409 (CvC), 1172s
(thienyl C–H in-plane bend), 1053s (thienyl C–H in-plane
bend), 835s (C–S) and 708s (thienyl C–H out-of-plane bend). δH
7.14 (2 H, dd, J 3.9 and 5.1, 2 × thienyl 4-H), 7.39 (2 H, dd, J 1.0
and 3.9, 2 × thienyl 3-H), 7.54 (2 H, dd, J 1.0 and 5.1, 2 ×
thienyl 5-H) and 8.13 (2 H, s, 4-H and 4′-H). m/z (ESI−)
359.9394 (M−, C16H8O2S4− requires 359.9413).
144.1 (3-C or 5-C or thienyl 2-C or 5-C) and 193.9 (2-C and
2′-C). m/z (ESI−) 536.0109 (M−, C24H24O2S6
requires
ˉ
−
536.0106). Crystal: C24H24O2S6, M = 536.79, triclinic, P1 (no. 2),
a = 6.3559(4), b = 8.6356(3), c = 11.6399(6) Å, α = 78.832(4), β =
77.408(5), γ = 87.543(4)°, V = 611.71(6) Å3, Z = 1 [Ci symmetry],
Dc = 1.457 g cm−3, μ(Cu-Kα) = 5.330 mm−1, T = 173 K, dark
bronze platy needles, Oxford Diffraction Xcalibur PX Ultra
(E)-5,5′-Bis(5-pentylthiophen-2-yl)-3,3′-bi[thiophen-3(2H)-
ylidene]-2,2′-dione (1b)
diffractometer; 2354 independent measured reflections (Rint
=
0.0215), F2 refinement,29 R1(obs) = 0.0306, wR2(all) = 0.0801,
2103 independent observed absorption-corrected reflections
[|Fo| > 4σ(|Fo|), 2θmax = 145°], 146 parameters; CCDC 921751.
(18 mg, 22% over three steps from 16) as a black powder. Mp
171–173 °C (from CDCl3) partial decomposition. Rf 0.3
(hexane–CHCl3 3 : 1). λmax/nm 660 (ε/L cm−1 mol−1 40 200),
627sh (33 800), 445sh (3200), 416sh (5200), 367 (23 800), 355
(22 700) and 320sh (9800). νmax/cm−1 3083vw (thienyl C–H),
2926w (alkyl C–H), 1664 (CvO), 1506s (CvC), 1443 (CvC),
1172vs (thienyl C–H in-plane bend), 1093w (thienyl C–H in-
plane bend), 835 (C–S) and 788 (thienyl C–H out-of-plane
bend). δH 0.91 (6 H, t distorted by virtual coupling, J ca. 7, 2 ×
pentyl 5-H3), 1.30–1.42 (8 H, m, 4 × CH2), 1.66–1.77 (4 H, m,
2 × CH2), 2.85 (4 H, t, J 7.6, 2 × pentyl 1-H2), 6.81 (2 H, d, J 3.7,
2 × thienyl 3-H or 4-H), 7.19 (2 H, d, J 3.7, 2 × thienyl 3-H or
4-H) and 8.03 (2 H, s, 4-H and 4′-H). δC 14.0 (pentyl 5-C), 22.4
(pentyl 4-C), 30.7 (pentyl 1-C or 2-C or 3-C), 31.0 (pentyl 1-C or
2-C or 3-C), 31.2 (pentyl 1-C or 2-C or 3-C), 116.5 (4-C and 4′-C),
126.2 (thienyl 3-C or 4-C), 129.6 (thienyl 3-C or 4-C), 134.5 (3-C
or 5-C or thienyl 2-C or 5-C), 135.5 (3-C or 5-C or thienyl 2-C or
5-C), 142.2 (3-C or 5-C or thienyl 2-C or 5-C), 152.5 (3-C or 5-C
or thienyl 2-C or 5-C) and 194.3 (2-C and 2′-C). m/z (ESI−)
Acknowledgements
We thank Pete Haycock and Dick Sheppard for their NMR
service; Lisa Haigh for her mass spectrometry service; Nick
Brooks for assistance with optical microscopy; Paul Stavrinou
and Donal Bradley for helpful discussions; the Imperial
College High Performance Computing Service; Imperial
College for a Junior Research Fellowship (N.R.E.); the Royal
Society for a Research Grant (RG110455).
Notes and references
1 N. R. Evans, G. E. Collis, A. K. Burrell and D. L. Officer,
unpublished work, Massey University, Palmerston North,
New Zealand, 2000.
500.0981 (M−, C26H28O2S4 requires 500.0978). Crystal:
−
ˉ
C26H28O2S4, M = 500.72, triclinic, P1 (no. 2), a = 5.7198(15), b =
9.0209(19), c = 11.863(3) Å, α = 79.850(19), β = 78.78(2), γ =
2 A.-B. Hörnfeldt, Ark. Kemi, 1964, 22, 211–235.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
3 A. N. Cammidge, V. H. M. Goddard, C. P. J. Schubert, 16 T. Clark, J. Chandrasekhar, G. W. Spitznagel and
H. Gopee, D. L. Hughes and D. Gonzalez-Lucas, Org. Lett.,
2011, 13, 6034–6037.
4 The Z isomer of 6 is depicted in this reference although the
P. v. R. Schleyer, J. Comput. Chem., 1983, 4, 294–301.
17 M. J. Frisch, J. A. Pople and J. S. Binkley, J. Chem. Phys.,
1984, 80, 3265–3269.
text does not address the question of stereoisomerism: 18 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
A. I. Kosak, R. J. F. Palchak, W. A. Steele and C. M. Selwitz, 19 P. J. Stephens, F. J. Devlin, C. F. Chabalowski and
J. Am. Chem. Soc., 1954, 76, 4450–4454.
M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
20 Red dye 136 and blue dye 14a8 are reported to be
fluorescent.
21 R. R. Pennelly and C. J. Eckhardt, Chem. Phys., 1976, 12,
89–105.
22 B. G. Anex and W. T. Simpson, Rev. Mod. Phys., 1960, 32,
466–476.
23 In neither structure is there any sign of disorder in the
position of the carbonyl oxygen atom.
5 A. Merz and F. Ellinger, Synthesis, 1991, 462–464.
6 A.-B. Hörnfeldt, Acta Chem. Scand., 1967, 21, 1952–1962.
7 E. B. Pedersen and S.-O. Lawesson, Synthesis, 1969, 170.
8 T. B. Norsten, E. A. B. Kantchev and M. B. Sullivan, Org.
Lett., 2010, 12, 4816–4819.
9 E. A. B. Kantchev, T. B. Norsten, M. L. Y. Tan, J. J. Y. Ng and
M. B. Sullivan, Chem.–Eur. J., 2012, 18, 695–708.
10 15 is prohibitively expensive from commercial suppliers.
Our method is similar to one previously published: 24 The carbonyl moiety in the structure of 1b is also involved
A. Meyer, E. Sigmund, F. Luppertz, G. Schnakenburg,
I. Gadaczek, T. Bredow, S.-S. Jester and S. Höger, Beilstein
J. Org. Chem., 2010, 6, 1180–1187.
in a π–π interaction across a center of symmetry, with the
middle of the CvO bond ca. 3.24 Å away from the equi-
valent position in a neighboring molecule (the associated
O⋯O separation is slightly shorter at ca. 3.14 Å).
11 R. M. Kellogg, A. P. Schaap and H. Wynberg, J. Org. Chem.,
1969, 34, 343–346.
12 T. E. Barder, S. D. Walker, J. R. Martinelli and
S. L. Buchwald, J. Am. Chem. Soc., 2005, 127, 4685–4696.
25 The plane of the C(3)vC(3A) double bond was taken as
being the mean plane of the two atoms comprising the
bond itself plus the four atoms immediately bonded to it.
13 Time-dependent B3LYP/6-311++G**//B3LYP/6-31G* pre- 26 A. L. Spek, PLATON, A Multipurpose Crystallographic Tool,
dictions for the blue dyes (1): wavelengths of the
second (allowed) electronic transitions from S0 are all
Utrecht University, Utrecht, Netherlands, 2008. See also;
A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7–13.
shorter than 500 nm (greater than 20 000 cm−1; Table S3 in 27 A. A. Kiryanov, P. Sampson and A. J. Seed, J. Mater. Chem.,
the ESI†).
14 K. Raghavachari, J. S. Binkley, R. Seeger and J. A. Pople, 28 A. J. Seed, K. J. Toyne and J. W. Goodby, J. Mater. Chem.,
J. Chem. Phys., 1980, 72, 650–654. 1995, 5, 653–661.
15 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72, 29 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
2001, 11, 3068–3077.
5639–5648.
logr., 2008, 64, 112–122.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.