Hydrogen bonding and steric effects on rotamerization in 3,4-alkylenedioxy-, 3-alkoxy- and 3,4-dialkoxy-2-thienyldi(tert-butyl)-methanols: An NMR, IR and X-ray crystallographic study

Article abstract of DOI:10.1039/b109612p

The equilibrium constant for the anti ? syn rotamerization (anti: intramolecularly hydrogen-bonded hydroxy group; syn: "free" hydroxy group) of 3,4-alkylenedioxy-, 3-alkoxy- and 3,4-dialkoxy-2-thienyldi(tert-butyl)methanols depends on the 3,4-alkylenedioxy or alkoxy group(s) and the solvent, hydrogen-bonding solvents such as DMSO and pyridine favouring the syn isomer. Equilibrium constants ([syn]/[anti]) in chloroform and benzene decrease in the order: 3,4-OCH₂O-, 3,4-O(CH₂)₂O-, 3-OMe, 3-OEt, 3,4-(OMe)₂ ≈ 3-Oi-Pr, 3,4-(OEt)₂, ranging over about 2.5 orders of magnitude. Variations in the IR OH stretching frequencies and the NMR OH proton shifts for the anti isomer indicate that intramolecular hydrogen bonding increases in roughly the same order. The syn → anti rotation barrier in DMSO increases with substituent size and number. The 3,4-methylenedioxythienyl derivative has a rather lower barrier (17.5 kcal mol_-1) than all the others (21.0-22.3 kcal mol_-1). The syn → anti rotation barrier is largely determined by steric effects but intramolecular hydrogen bonding in the anti isomer contributes to the variation of the anti → syn rotation barrier. A single crystal X-ray diffraction study of the anti-3,4-diethoxy derivative shows that the orientation of the 3-alkoxy group is very different from that in anti-3-methoxy-2-thienyldi-(1-adamantyl)methanol. Molecular mechanics and quantum mechanical calculations are used in an attempt to rationalize the equilibrium data.

Full text of DOI:10.1039/b109612p

View Article Online / Journal Homepage / Table of Contents for this issue
Hydrogen bonding and steric effects on rotamerization in 3,4-
alkylenedioxy-, 3-alkoxy- and 3,4-dialkoxy-2-thienyldi(tert-butyl)-
methanols: an NMR, IR and X-ray crystallographic study†
2
John S. Lomas,*a Alain Adenier,a Kun Gao,a François Maurela and Jacqueline Vaissermannb
a ITODYS, Université de Paris 7, associé au C.N.R.S., 1 rue Guy de la Brosse, 75005 Paris,
France
b Laboratoire de Chimie des Métaux de Transition, Université de Paris 6, associé au C.N.R.S.,
Case 42, 4 place Jussieu, 75252 Paris Cedex 05, France
Received (in Cambridge, UK) 22nd October 2001, Accepted 26th November 2001
First published as an Advance Article on the web 10th January 2002
The equilibrium constant for the anti ↔ syn rotamerization (anti: intramolecularly hydrogen-bonded hydroxy group;
syn: “free” hydroxy group) of 3,4-alkylenedioxy-, 3-alkoxy- and 3,4-dialkoxy-2-thienyldi(tert-butyl)methanols
depends on the 3,4-alkylenedioxy or alkoxy group(s) and the solvent, hydrogen-bonding solvents such as DMSO
and pyridine favouring the syn isomer. Equilibrium constants ([syn]/[anti]) in chloroform and benzene decrease in the
order: 3,4-OCH2O-, 3,4-O(CH2)2O-, 3-OMe, 3-OEt, 3,4-(OMe)2 ≈ 3-Oi-Pr, 3,4-(OEt)2, ranging over about 2.5 orders
of magnitude. Variations in the IR OH stretching frequencies and the NMR OH proton shifts for the anti isomer
indicate that intramolecular hydrogen bonding increases in roughly the same order. The syn
in DMSO increases with substituent size and number. The 3,4-methylenedioxythienyl derivative has a rather lower
barrier (17.5 kcal molϪ1) than all the others (21.0–22.3 kcal molϪ1). The syn
anti rotation barrier is largely
determined by steric effects but intramolecular hydrogen bonding in the anti isomer contributes to the variation
of the anti syn rotation barrier. A single crystal X-ray diffraction study of the anti-3,4-diethoxy derivative
anti rotation barrier
shows that the orientation of the 3-alkoxy group is very different from that in anti-3-methoxy-2-thienyldi-
(1-adamantyl)methanol. Molecular mechanics and quantum mechanical calculations are used in an
attempt to rationalize the equilibrium data.
the equilibrium constant for anti ↔ syn rotamerization depend-
Introduction
ing on the solvent, hydrogen-bonding solvents favouring the syn
isomer.9 The rotation barrier, about 26 kcal molϪ1 (1 cal = 4.184
J), is too small for the syn isomer to be isolated free of anti.
3,4-Ethylenedioxythiophene (EDOT), 1 (n = 2), and its deriv-
atives have been and continue to be the subject of many studies
devoted almost exclusively to their polymerization and to the
optoelectronic properties of the materials obtained.1–3 Other
thiophenes with oxygen functions at the 3- or 3- and 4- posi-
tions, such as 3,4-alkylenedioxythiophenes, 1 (n = 1, 3, 4),2,4,5 3-
alkoxythiophenes, 2-R,6 and 3,4-dimethoxythiophene, 3-Me,6c,7
have been less studied, the qualities of the polymers being
poorer than those of polyEDOT (PEDOT).
The 3-alkoxythienyl derivatives, 4-R, can also adopt the syn
form but the rotation barriers are much higher and the equi-
librium constant favours the anti isomer.10 This must reflect
differences in hydrogen bonding and steric interactions in the 3-
alkoxythienyl and EDOT systems. When the adamantyl groups
are replaced by the smaller, less rigid tert-butyl groups anti ↔
syn rotamerization becomes fast at room temperature. This
offers us the opportunity to compare 3,4-alkylenedioxythio-
phenes, 3-alkoxythiophenes and 3,4-dialkoxythiophenes from a
chemical standpoint. It was hoped that the results would give
useful insights into the difference between EDOT and the other
oxygen-substituted thiophenes.
Addition of 3-alkoxy-2-thienyl-lithium compounds to a
highly congested ketone, di(1-adamantyl) ketone, gives the anti
rotamers (sulfur remote from the OH group), 4A-R (R = Me, Et
and i-Pr), in which the hydroxy proton is hydrogen-bonded to
the alkoxy oxygen.8 However, the corresponding EDOT deriv-
ative, 5, exists in two readily interconvertible forms, 5A and 5S,
† Electronic supplementary information (ESI) available: NMR data;
activation parameters for rotation; MMFF94 steric energies and alkoxy
group geometries; thermodynamic data; quantum mechanical calcula-
tions of geometries; bond lengths, bond angles and torsion angles of
8A-Et; NMR and IR data on new compounds. See http://www.rsc.org/
suppdata/p2/b1/b109612p/
Results and discussion
Synthesis
Apart from EDOT, 1 (n = 2), 3,4-alkylenedioxythiophenes with
n = 3 and 4 have been described.2,4 Electropolymerization of the
216
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b109612p

View Article Online
methylene derivative, 1 (n = 1), has been reported5 but the prep-
aration and characteristics of the monomer were not published.
The pentylene derivative, 1 (n = 5), has not been reported.
3,4-Diethoxythiophene, 3-Et, another new compound, was
prepared as 3,4-dimethoxythiophene, 3-Me.11 All alcohols [6
(n = 1–5), 7-R (R = Me, Et and i-Pr) and 8-R (R = Me and Et)]
were prepared by lithiation of the 3-alkoxy-,12 3,4-dialkoxy-
or 3,4-alkylenedioxythiophene by n-butyllithium–TMEDA in
diethyl ether at room temperature, followed by reaction with
di(tert-butyl) ketone.
Table 1 Equilibrium constants ([syn]/[anti]) for 3,4-alkylenedioxy-,
6(n), 3-alkoxy-, 7-R, and 3,4-dialkoxy-2-thienyldi(tert-butyl)methanols,
8-R, in hydrogen-bonding and non-hydrogen-bonding solvents at 298 K
Compound
K(CDCl3)
K(benzene)
K(pyridine)
K(DMSO)
6 (n = 1)
6 (n = 2)
6 (n = 3)
6 (n = 4)
6 (n = 5)
7-Me
9.2
7.7
0.91a(0.94)
0.24
0.87a(0.84)
0.22
9.2
2.5
2.3
1.6
1.85
1.3
0.64
0.88
0.41
25a(18)
5.45
5.2
4.8
4.0
3.3
1.5
2.1
1.0
0.18
0.18
0.18
0.14
0.071
0.069
0.032
0.20
0.15
0.20
0.15
0.078
0.090
0.037
7-Et
7-i-Pr
8-Me
8-Et
a This work; other data for 6 (n = 2), ref. 9.
oxythienyl derivatives can be correlated by a second-order
polynomial, which goes through the points for the syn isomers
(Fig. 1). Previous IR–NMR correlations of this type have
IR Spectroscopy
The IR spectra of the various alcohols were determined in
carbon tetrachloride. With the notable exception of the 3,4-
methylenedioxythienyl and EDOT derivatives, 6 (n = 1) and 6
(n = 2), (vide infra) there is a strong, structure-dependent OH
absorption at around 3500 cmϪ1, due to the intramolecularly
hydrogen-bonded anti form, and weaker absorptions at higher
wavenumbers, close to 3606 and 3628 cmϪ1 regardless of struc-
ture, due to the syn isomer. The considerable difference between
the frequencies for the anti forms of 6 (n = 2) (3568 cmϪ1) and of
all the others indicates weaker hydrogen bonding in the former
case.
More extreme is the 3,4-methylenedioxythienyl derivative, 6
(n = 1), whose IR absorption consists of a broad band peaking
at about 3623 cmϪ1 with a shoulder to lower wavenumbers.
Decomposition of this band into Lorentzian components is not
unambiguous. Three-peak decomposition gives strong bands at
3624 and 3608 cmϪ1, with a very weak band (about 3% of the
integrated intensity) at 3554 cmϪ1. However, it can be repre-
sented equally well by four bands: 3624 (62%), 3605 (21%),
3612 (11%) and 3561 cmϪ1 (4%), though only three are expected
for syn (two) and anti (one) isomers. Comparison with data for
the other 3,4-alkylenedioxythienyl derivatives suggests that the
two strongest bands can be associated with the syn isomer, that
at 3612 cmϪ1 with the anti, and the weakest band unattributed.
Fig. 1 Non-linear correlation between IR absorption frequency, νOH
(CCl4/cmϪ1) and 1H NMR hydroxy proton shift δOH (CDCl3/ppm).
been linear.13,14 Both data reveal small differences in hydro-
gen bonding in the 3-alkoxy, 3,4-dialkoxy and long-chain
3,4-alkylenedioxy (n ≥ 3) derivatives, weaker bonding in
the EDOT compound and very weak bonding in the 3,4-
methylenedioxythienyl analogue.
NMR spectroscopy: hydrogen-bonding solvents
1
In non-hydrogen-bonding solvents the H NMR OH proton
signal in the syn rotamer is considerably upfield of that in the
anti isomer. However, DMSO and pyridine hydrogen-bond
strongly with the syn hydroxy proton, causing marked down-
field displacement of this signal; pyridine brings it to within
0.2–0.3 ppm of the signal for the anti isomer while in DMSO it
is at about 4.7 ppm regardless of the structure (Supplementary
Material Table S1). The temperature dependence (Table S1) of
the shift indicates whether the proton corresponding to a given
signal is intramolecularly hydrogen-bonded (low coefficient) or
is hydrogen-bonded to the solvent (high coefficient).14
NMR spectroscopy: non-hydrogen-bonding solvents
In chloroform or benzene, the hydroxy proton shift for the syn
rotamer ranges from 2.0 to 2.2 ppm, while that for the anti
isomer is at much lower field. Within the 3-alkoxy-2-thienyl
series, 7-R, there is a small increase in the shift as methoxy is
replaced by ethoxy and then isopropoxy (0.3 ppm overall).
Shifts for the 3,4-dialkoxythienyl derivatives, 8-R, are similar to
those for 7-R. In the 3,4-alkylenedioxythienyl series, 6(n), the
OH proton shift for the anti isomer varies from 2.7 ppm (n = 1)
to about 6.0 ppm (n = 5). The former value is particularly low
for an intramolecularly hydrogen-bonded proton. The NMR
and the IR data for the 3,4-alkylenedioxythienyl and 3-alk-
The most important effect of transferring these alcohols to a
hydrogen-bonding solvent is the increase in the equilibrium
concentration of the syn isomer. This will be discussed in the
following paragraph.
Rotamer equilibria at room temperature
These di(tert-butyl) derivatives equilibrate rapidly at room
temperature in solution (Table 1). In DMSO, the stronger
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224
217

View Article Online
hydrogen-bonding solvent, the equilibrium lies further in favour
of the syn rotamer than in pyridine. New measurements on the
EDOT derivative, 6 (n = 2), in DMSO at lower temperatures
than in the previous study9 give a slightly higher value for the
equilibrium constant, ([syn]/[anti]), and a revised free energy vs.
temperature correlation in better agreement with data for the
other derivatives (vide infra).
As the size of the alkoxy group in the 3-alkoxythienyl series,
7-R, increases the proportion of anti rotamer rises. In the same
way, in the 3,4-alkylenedioxythienyl series, 6(n), the proportion
of the anti isomer increases markedly from 6 (n = 1) to 6 (n = 2)
to 6 (n = 3) and then remains practically constant. For the 3,4-
methylenedioxythienyl derivative, 6 (n = 1), even in chloroform
or benzene the major isomer is syn and in the hydrogen-
bonding solvents no anti isomer can be detected at all. The
values of the equilibrium constant are not very different from
that for 2-thienyldi(1-adamantyl)methanol (7.0),14 a further
indication that intramolecular hydrogen bonding to oxygen in
the anti isomer, 6A (n = 1), is very weak.
The positions of the 3,4-dimethoxy and 3,4-diethoxy deriv-
atives, 8-Me and 8-Et, between 7-Et and 7-i-Pr and beyond the
3-isopropoxy compound, 7-i-Pr, respectively, are somewhat
surprising. The NMR shifts of the hydroxy protons in these
compounds are similar to those for the corresponding 3-alkoxy
compounds and, if this is taken as an indicator of hydrogen
bond strength, there is no reason for the anti isomer to be
favoured. However, part of the variation of the equilibrium
constant is due to changes in the relative steric energies of the
two rotamers as the effective bulk of the 3-substituent is varied.
This may be increased by the buttressing effect of the 4-alkoxy
group on the 3-alkoxy, and be decreased when the methylene
group is more or less tied back by a short alkylene chain (n = 1
or 2). Molecular mechanics and quantum mechanical calc-
ulations on the energies and structures of 6(n), 7-R and 8-R will
be presented below.
Fig. 2 (a) log–log correlation of equilibrium constants for alcohols in
chloroform and benzene; (b) log–log correlation of equilibrium
constants for alcohols in pyridine and DMSO.
Correlation of the equilibrium constants [log K(1) vs. log
K(2)] for the 9 or 10 compounds in the different solvents taken
pairwise gives gradients not far from unity in all cases, indicat-
ing that the solvent effect is virtually structure-independent.
Not surprisingly, the best correlations are those for solvents of
the same type: chloroform vs. benzene (Fig. 2a, slope 1.07
0.04; corr. coeff. 0.9956) and pyridine vs. DMSO (Fig. 2b, slope
0.97 0.04; corr. coeff. 0.9935).
alcohols except 6 (n = 1) are listed in Table 2. The enthalpy term
favours the syn isomer but the entropy term the anti. This indi-
cates that the syn rotamer is favoured by hydrogen bonding
to the solvent but that this involves a more ordered structure,
presumably due to solvent organization.9,16 For pyridine the
entropy terms are very similar for all the compounds investi-
gated, covering a rather narrower range than for DMSO. In
particular, for the 3-alkoxythienyl derivatives, 7-R, the entropy
terms are virtually identical, the change in the equilibrium con-
stant being due to the progressive fall in the enthalpy term on
going from methoxy to ethoxy to isopropoxy.
In previous work15 it was found that there was a good corre-
lation between solvent effects (8 solvents) on rotamer equilibria
of 2-anisyl(isopropyl)(tert-butyl)methanol, 10-Me, and EDOT-
di(tert-butyl)methanol, 6 (n = 2). For the set of alcohols con-
sidered in this work mean solvent effects on the free energy
difference, ∆GЊ, relative to chloroform are: benzene (0.03
0.09 kcal molϪ1); pyridine (1.38
0.08 kcal molϪ1); DMSO
0.09 kcal molϪ1). These values are quite similar to
Rotamerization kinetics
(1.92
those for both 2-anisyldi(tert-butyl)methanol, 9-Me, and
2-anisyl(isopropyl)(tert-butyl)methanol, 10-Me, at the same
temperature (298 K).15
The rotation barriers (Table 3) for most of the various substi-
tuted 2-thienyldi(tert-butyl)methanols fall in a range where they
1
can only be studied by dynamic H NMR in DMSO, the only
solvent which has a high boiling point and where the rotamer
ratio is sufficiently close to unity for reliable measurement. The
EDOT derivative, 6 (n = 2), was previously studied in nitro-
benzene9 but we have now succeeded in obtaining data in
DMSO. The 3,4-methylenedioxythienyl derivative, 6 (n = 1),
was studied in chloroform and benzene, the only solvents where
a detectable amount of the anti isomer is present. Simulation of
the exchange spectra using the gNMR programme17 provides
rate constants for the syn
anti and anti
syn intercon-
version processes, from which are calculated the rotation
barriers. Estimates based on different parts of the spectra for
the 3-alkoxythienyl derivatives, 7-R, give closely similar results.
Nevertheless, the derived thermodynamic data (Supplemen-
tary Material Table S2) cannot be considered very reliable. In
particular, there is considerable scatter in the activation entro-
Temperature dependence of equilibrium constants in hydrogen-
bonding solvents
In pyridine or DMSO as the temperature is increased the
amount of anti isomer increases at the expense of the syn. The
free energy difference ∆GЊ(anti Ϫ syn) varies linearly with tem-
perature, allowing the evaluation of ∆HЊ and ∆SЊ. Data for all
pies, which for the syn
anti reaction in DMSO range from 3
218
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224

View Article Online
Table 2 Temperature dependence of equilibrium constants ([syn]/[anti]) for 3,4-alkylenedioxy-, 6(n), 3-alkoxy-, 7-R, and 3,4-dialkoxy-2-
thienyldi(tert-butyl)methanols, 8-R, in hydrogen-bonding solvents [∆GЊ(anti Ϫ syn) = ∆HЊ Ϫ T ∆SЊ; ∆HЊ in kcal molϪ1; ∆SЊ in cal molϪ1 KϪ1
]
Pyridine
DMSO
Compound
Temp. range/K
∆HЊ
∆SЊ
Temp. range/K
∆HЊ
∆SЊ
6 (n = 2)
6 (n = 3)
6 (n = 4)
6 (n = 5)
7-Me
298–328a
298–348
298–343
298–333
298–348
298–368
298–368
298–368
298–368
4.31 0.08
3.21 0.11
3.78 0.04
3.51 0.08
3.12 0.06
2.89 0.05
2.45 0.05
2.42 0.08
2.37 0.11
10.1 0.3
9.0 0.3
11.0 0.1
10.9 0.2
9.3 0.2
9.2 0.2
9.2 0.1
8.5 0.2
9.7 0.3
298–348
298–383
298–358
298–358
298–358
298–358
298–388
298–388
298–388
4.77 0.10
3.15 0.09
3.64 0.06
3.71 0.05
3.09 0.02
3.57 0.09
2.35 0.04
2.55 0.03
2.30 0.02
9.6 0.3
7.2 0.3
8.9 0.2
9.3 0.2
7.6 0.1
9.6 0.3
7.1 0.1
7.1 0.1
7.7 0.1
7-Et
7-i-Pr
8-Me
8-Et
a Ref. 9.
Table 3 Activation energies for syn
anti and anti
syn rotation in 3,4-alkylenedioxy-, 6(n), 3-alkoxy-, 7-R, and 3,4-dialkoxy-2-thienyldi(tert-
butyl)methanols, 8-R, in DMSO [mean ∆G≠ in kcal molϪ1 at T (mean) in K]
Compound
Temp. range/K
T (mean)
∆G≠(syn
anti)
∆G≠(anti
syn)
6 (n = 1)a
6 (n = 1)b
6 (n = 2)
6 (n = 3)
6 (n = 4)
6 (n = 5)
7-Me
298–323c
298–328c
353–398d
378–423d
373–428d
383–428d
378–423d
378–408e
368–413d
368–398f
368–403d
358–393e
358–393f
358–398c
378–428d
311
313
374
401
401
406
400
391
393
384
388
379
377
378
404
17.4
17.5
21.0g
22.0
22.2
22.0
21.4
21.5
21.8
21.9
21.9
21.9
21.95
22.0
22.3
16.3
16.1
19.8g
21.7
22.0
22.0
21.4
21.35
21.85
21.9
22.3
22.2
22.3
22.1
23.05
7-Et
7-i-Pr
8-Me
8-Et
a In chloroform. b In benzene. c Hydroxy group. d Aromatic(s). e Methyl group(s). f tert-Butyl. g In nitrobenzene at 365–400 K: 20.8 kcal molϪ1, ref. 9.
to Ϫ7 cal molϪ1
K
Ϫ1, with that for the anti
syn reaction on
3-alkoxythienyl compounds, 7-R, give very similar rotation
barriers, with a small increase on going from methoxy to ethoxy
and isopropoxy. The 3,4-dialkoxy compounds, 8-R, are associ-
ated with barriers about 0.5 kcal molϪ1 higher than those of the
corresponding 3-alkoxy derivatives.
average about 5 cal molϪ1 KϪ1 more negative. The differences in
the activation entropies are doubtless more reliable than the
individual values but it should be noted that they are smaller
than the ∆SЊ values for the same solvent, albeit in a lower tem-
perature range. Likewise, the differences in the activation
enthalpies are smaller than the ∆HЊ values. This implies that the
∆GЊ plot at high temperature deviates somewhat from that at
lower temperature. Comparison of the two sets of ∆GЊ values
suggests that they are parts of a continuous curve and that
either the ∆CpЊ = 0 assumption18 is incorrect or that there is a
temperature-dependent change in the nature of the solvation of
the OH group.
The differences between the 3-alkoxy groups and between
the 3,4-dialkoxy and 3-alkoxy derivatives are somewhat more
marked for the anti
syn rotation barrier. The reason is that
this reaction involves the breaking of an intramolecular hydro-
gen bond, whose strength varies significantly over the range of
compounds studied, and the formation of a solvent-hydrogen-
bonded transition state, the energetic contribution of which is
relatively independent of structure.
There are few comparable data on rotation barriers involv-
ing thiophene derivatives: 2-thienyldi(1-adamantyl)methanes
and -methanols have barriers ranging from about 20
(2-thienyldiadamantylmethane) to 38 kcal molϪ1 (3-methyl-2-
thienyldiadamantylmethane)20 while values for anti-3-alkoxy-2-
thienyldi(1-adamantyl)methanes rise from 28.4 to 30.7 kcal
molϪ1 on going from methoxy to isopropoxy.8 For 2-
alkoxyphenyl(isopropyl)(tert-butyl)methanols, 10-R, the span
is only 0.5 or 0.9 kcal molϪ1, depending on the direction of
rotation.16 Similar small effects are observed when methoxy
is replaced by ethoxy in 2-alkoxyphenyldi(tert-butyl)methanols,
9-R.15 These results suggest that the effect of varying the alkoxy
group is conditioned by the size of the alkyl substituents to the
C–OH carbon. 1-Adamantyl groups, whose rigidity reduces
their ability to distort,21 enhance rotation barriers much more
than tert-butyls, which means that rotation barrier differences
should be amplified in the corresponding 3- or 3,4-substituted
2-thienyldi(1-adamantyl)- and 2-thienyl(1-adamantyl)(tert-
butyl)methanols. This will be the subject of further work.
Structural effects on rotation barriers
In previous work2,15,19 we have shown that the activation energy
for rotation from the form with a “free” hydroxy group,
through a transition state with again a “free” hydroxy group, to
the form with an intramolecularly hydrogen-bonded hydroxy
group, is virtually solvent-independent. For the reverse reac-
tion, the barrier is solvent-dependent, falling for solvents in
which the transition state is specifically solvated by hydrogen
bonding with the hydroxy group.
The syn
anti rotation barrier for the 3,4-methylene-
dioxythienyl derivative, 6 (n = 1) (17.4 and 17.5 kcal molϪ1 in
chloroform and benzene, respectively) is much lower than that
measured for the EDOT analogue, 6 (n = 2), in nitrobenzene
9
(about 20.8 kcal molϪ1
)
or DMSO (21.0 kcal molϪ1), and
clearly shows the reduced steric effect of the tied-back oxygen
atom at the 3-position; further lengthening of the chain results
in only a modest increase to 22.0–22.2 kcal molϪ1. The three
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224
219

View Article Online
Separation of hydrogen bonding and steric effects
reasons, our previous attempt to correct for steric effects in
2-alkoxyphenyl(isopropyl)(tert-butyl)methanols15 is a measure
of the error in the MM calculations rather than of the variation
in the hydrogen bond energy.
The variations in the equilibrium constants and the rotation
barriers clearly result from the interplay of steric and hydrogen
bonding effects, which appear to vary approximately in parallel,
and the question arises as to their relative importance. There is
no reliable basis in the literature for quantifying the IR and
NMR spectroscopic data in terms of energy changes. Corre-
lations between hydrogen bond formation enthalpies and IR
frequency shifts, the Badger–Bauer rule,22 mainly for inter-
molecular hydrogen bonding, have gradients ranging from
0.004–0.04 kcal molϪ1 cm,13d,e,23 with a preference for values
around 0.01 kcal molϪ1 cm.23 It is clear from the variation of the
The steric energies of both rotamers of 6 (n = 1–4), 7-R and
8-R were calculated by means of the MMFF94 force field24
contained in the Sybyl package:25 data refer to gas-phase ener-
gies. In the 3,4-alkylenedioxythiophene derivatives, 6(n), the
orientation of the O–CH2 bond at C3 is largely determined by
the requirements of the chain connecting it to C4; however, the
longer the chain the more difficult it is to determine its con-
formation. For the syn rotamers of 7-R the C2–C3–O–C tor-
sion angle is close to 180Њ regardless of R, i.e. the alkoxy carbon
is practically in the plane of the thiophene ring. In the anti
rotamers, it deviates by 12Њ, 15Њ and 38Њ as R goes from Me to
Et to i-Pr. In the 3,4-dialkoxy derivatives, 8-R, the 3-alkoxy
carbon is about 15Њ and 67Њ from the plane in the syn and
anti isomers, respectively, regardless of R. Details are given in
Supplementary Material Table S3.
Except for the two 3,4-dialkoxythienyl derivatives, 8-R, where
the steric energies of the rotamers are very similar, the syn iso-
mer is calculated to be the more stable, the difference, ∆SE/H,
ranging from 0.6 (7-i-Pr) to 2.1 [6 (n = 1)] kcal molϪ1. The data
(Table 4) indicate that ∆SE/H overestimates the stability of the
syn isomer of 6 (n = 1) and 6 (n = 2) by nearly 1 kcal molϪ1, and
that of 7-Et, 7-i-Pr and 8-Et by about 2 kcal molϪ1. Since ∆SE/
H is always greater than ∆GЊ this would make SHB(syn) a
negative quantity and variable, whereas we have defined it as
positive and believe it to be constant. The correlation of the
mean values of ∆GЊ for chloroform and benzene against ∆SE/
H is rather poor (corr. coeff. 0.9140; slope 1.42 0.24) (Fig. 3).
syn
anti rotation barrier that the steric effect of the substi-
tuted thienyl group increases through the different series, but it
is not obvious how much this contributes to the equilibrium
constants. The rotation barrier is the difference between the
ground state and rotation transition state energies, and though
it generally increases with steric strain in the ground state,21a it
says nothing about the difference in the steric energies of the
two rotamers.
Considering only steric energies and hydrogen bonding, we
can write:
∆G≠(syn) = SE(TS) Ϫ SE(syn) Ϫ SHB(TS) ϩ SHB(syn) (1)
where SE(TS) is the steric energy of the transition state,
SHB(TS) its hydrogen bonding solvation energy, and SE(syn)
and SHB(syn) are the corresponding terms for the syn isomer.
If SHB(TS) ≈ SHB(syn) we obtain:
∆G≠(syn) = SE(TS) Ϫ SE(syn)
(2)
which expresses the fact that syn
anti rotation barriers are
solvent-independent.
For rotation in the other direction, we have:
∆G≠(anti) =
SE(TS) Ϫ SE(anti) Ϫ SHB(TS) ϩ IHB(anti) (3)
where IHB(anti) is the intramolecular hydrogen bond energy.
Subtraction of (3) from (1) gives eqn. (4):
∆G≠(syn) Ϫ ∆G≠(anti) =
SE(anti) Ϫ SE(syn) ϩ SHB(syn) Ϫ IHB(anti) (4)
where the left hand side is, of course, the same as ∆GЊ(anti Ϫ
syn), whence:
Fig. 3 Correlation of ∆GЊ for equilibrium in benzene and chloroform
against ∆SE/H calculated by MMFF94 (both in kcal molϪ1).
∆GЊ(anti Ϫ syn) =
∆SE(anti Ϫ syn) ϩ SHB(syn) Ϫ IHB(anti) (5)
The relative lack of success of this approach is not altogether
surprising. Molecular mechanics force fields are not well par-
ametrized for heteroatomic and particularly heterocyclic sys-
tems, and the parameter set for these thiophenes includes many
lacunes which are filled with “default” and “estimated” para-
meters.24 Since the number, the nature and the importance of
these parameters differ from one set of compounds to another,
6(n), 7-R or 8-R, considerable discrepancies are to be expected.
Another way of tackling this problem is to assume that the
hydrogen bond enthalpy is correlated with the IR frequency
shift relative to the mean value of 3617 cmϪ1 for the syn
rotamers, IR being chosen rather than NMR for the simple
reason that there are more data on this type of correlation. In
particular, there is a range of values for the ∆H/∆ν gradient.
The question arises as to whether it is legitimate to work with
free energies, which are known with greater precision, rather
than with enthalpies. The ∆SЊ values reported in Table 2 repre-
sent the entropy change associated with the equilibrium
If it can be assumed that SHB(syn) is a constant regardless of
structure, which is supported by the equilibrium data, there are
only two variables in this equation: the steric energy difference
between the two rotamers and the intramolecular hydrogen
bond energy for the anti rotamer. Unfortunately, in modern
molecular mechanics force fields hydrogen bonding is treated as
an electrostatic interaction, and its contribution is therefore
included in the steric energy, so that we have:
∆GЊ(anti Ϫ syn) = ∆SE/H(anti Ϫ syn) ϩ SHB(syn) (6)
where ∆SE/H(anti Ϫ syn) includes the intramolecular hydrogen
bonding term.
Insofar as the presence of this particular interaction deter-
mines in part, or even to a large extent (since it can represent
several kcal molϪ1), the optimum geometry, it affects the magni-
tude of all other deformations and interactions. For these
220
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224

View Article Online
Table 4 MMFF94 steric energy differences [∆SE/H(anti Ϫ syn) in kcal molϪ1], free energy differences [∆GЊ(anti Ϫ syn), mean for benzene and
chloroform in kcal molϪ1], IR OH frequencies (anti in CCl4/cmϪ1) and hydrogen bond enthalpy contributions (in kcal molϪ1
)
Compound
∆SE/H
∆GЊ
νOH
∆Ha
∆SEa
∆Hb
∆SEb
6 (n = 1)
6 (n = 2)
6 (n = 3)
6 (n = 4)
7-Me
2.07
0.86
0.77
0.87
0.72
0.92
0.59
0.00
1.26
Ϫ0.07
Ϫ0.87
Ϫ0.98
Ϫ0.98
Ϫ1.14
Ϫ1.54
Ϫ1.50
Ϫ1.99
3612
3568
3507
3505
3516
3500
3490
3502
3486
0.05
0.49
1.10
1.12
1.01
1.17
1.27
1.15
1.31
1.31
0.42
0.23
0.14
0.03
0.07
0.64
1.43
1.46
1.31
1.52
1.65
1.50
1.70
1.33
0.57
0.56
0.48
0.33
0.38
0.11
0.00
Ϫ0.29
7-Et
0.03
7-i-Pr
8-Me
8-Et
Ϫ0.27
Ϫ0.35
Ϫ0.68
Ϫ0.08
a Gradient ∆H/∆ν = 0.010 kcal molϪ1 cm. b Gradient ∆H/∆ν = 0.013 kcal molϪ1 cm.
Table 5 Semi-empirical (AM1) heats of formation differences [∆∆HfЊ(syn Ϫ anti) in kcal molϪ1], HF/3-21G and HF/6-31G* potential energy
differences [∆E in kcal molϪ1]; calculated H ؒ ؒ ؒ O and O ؒ ؒ ؒ O distances (in Å)
Compound
H ؒ ؒ ؒ Oa
O ؒ ؒ ؒ Oa
∆∆HfЊa
H ؒ ؒ ؒ Ob
∆Eb
H ؒ ؒ ؒ Oc
∆Ec
6 (n = 1)
6 (n = 2)
6 (n = 3)
6 (n = 4)
7-Me
2.340
2.196
2.139
2.089
2.160
2.145
2.130
2.075
1.980
2.835
2.59
2.55
2.62
2.51
2.53
2.66
2.65
2.66
1.98
2.92
4.48
4.62
3.53
4.18
4.83
6.31
6.24
1.895
1.749
Ϫ0.05
1.97
2.132
1.894
Ϫ1.03
Ϫ0.15
1.699
1.774
1.768
1.766
2.52
4.10
4.43
3.91
1.831
1.886
1.883
1.920
Ϫ0.02
1.82
2.39
7-Et
7-i-Pr
8-Me
8-Et
3.06
a AM1. b HF/3-21G. c HF/6-31G*.
between two conformers, where one has a hydrogen bond to
solvent (if at all) and the other an intramolecular hydrogen
bond. Given the similarity of the structures and the constancy
of the solvation effects, it is reasonable to assume that the
entropy of formation of the hydrogen bond to solvent is
roughly constant. The fact that ∆SЊ values in DMSO and pyr-
idine vary by little more than the experimental error suggests
that the intramolecular hydrogen bond entropy is also constant.
By the same reasoning, the fact that the equilibrium constants
for 6 (n = 2) in non-hydrogen-bonding solvents are almost
temperature-independent implies that intramolecular hydro-
gen bond entropies are moreover close to zero. This means that
free energy changes correspond to enthalpy differences, i.e.
IHB(anti) ≈ ∆H. We shall assume that SHB(syn) is negligible
for chloroform and benzene.
calculated using the semi-empirical AM1 parametrization.27
Calculations with the PM3 parametrization gave less satisfac-
tory results and will not be discussed here. Some of the struc-
tures were also investigated by the ab initio approach with small
basis sets.
In all cases the AM1 parametrization gives the anti rotamer
as the more stable, to a much greater extent than is observed
in even a non-hydrogen-bonding solvent. The results (Table 5
and Supplementary Material Table S4) show, however, sharp
increases in the heat of formation difference, ∆∆HfЊ, on going
from 6 (n = 1) to 6 (n = 2) to 6 (n = 3, 4), and a steady increase
through the 3-alkoxythiophene derivatives, 7-R, while there is
apparently little difference between 8-Me and 8-Et. The overall
pattern is roughly consistent with the observed changes in the
equilibrium constant, but the range of ∆∆HfЊ for rotamer pairs
is substantially greater than that of ∆GЊ or ∆HЊ, and the corre-
lation with ∆GЊ is poor (slope 1.32 0.25; corr. coeff. 0.8957).
Because of the difficulty of locating the overall minimum in
the case of the flexible 7- and 8-membered rings, ab initio calc-
ulations of the potential energies at the HF/3-21G and HF/6-
31G* levels were run on a reduced selection of 6 structures.
Though the ∆E values for 6 (n = 1) and 6 (n = 2) with the more
extended basis set are close to the experimental values of ∆GЊ
the overall ∆E/∆GЊ correlation is again poor (slope 1.30 0.38;
corr. coeff. 0.8604). Surprisingly, for the smaller set the corre-
lation is better but the energy range is even higher (slope 1.52
0.21; corr. coeff. 0.9636).
The AM1-calculated H ؒ ؒ ؒ O and O ؒ ؒ ؒ O distances for the
anti rotamer vary roughly as expected, decreasing through each
series (Table 5 and Supplementary Material Table S5). The
values of 1.98 and 2.66 Å for 8-Et are in very good agreement
with the crystallographic data (vide infra). Other aspects of this
structure, in particular the disposition of the ethoxy groups, are
also very well reproduced by the AM1 calculation. The H ؒ ؒ ؒ O
and O ؒ ؒ ؒ O distances calculated by the ab initio approach are
less satisfactory, those given by the smaller basis set being par-
ticularly low. A corollary of this strong interaction appears to
be that in 7A-Me and 7A-Et the alkoxy group is swung out of
the plane of the thiophene ring.
We now need to choose the ∆H/∆ν gradient according to the
qualitative criteria that the steric effect as originally defined in
eqn. (5), ∆SE(anti Ϫ syn), should decrease as we descend each
family, 6(n), 7-R or 8-R, and that the greatest ∆H value should
be compatible with literature data for intramolecular hydrogen
bonds of this type. These can be reasonably well satisfied, apart
from a problem with 7-Me, by using small values of the gradi-
ent between 0.010 and 0.013 kcal molϪ1 cm, the lower of these
values ordering 6(n) the best but somewhat reducing the highest
∆H value below those for 1,ω-methoxy-alcohols (up to 2.7 kcal
molϪ1 26a and 1,ω-diols (up to 2.00 kcal molϪ1).26b At the lower
)
limit we have 1.3 kcal molϪ1 of the overall 3.27 kcal molϪ1 of
free energy variation due to change in hydrogen bond strength
and 2.0 kcal molϪ1 due to changes in steric effects; at the upper
limit, 1.6 and 1.6 kcal molϪ1, respectively (Table 4).
Semi-empirical quantum mechanical calculations
Though quantum mechanics make no distinction between
hydrogen bonding and steric effects, we thought it interesting to
examine to what extent readily available and comparatively
economical calculations are capable of describing our results.
Fully optimized gas-phase heats of formation and geometries
for all but one of the isomer pairs discussed in this work were
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224
221

View Article Online
Single crystal X-ray diffraction study of anti-3,4-diethoxy-2-
thienyldi(tert-butyl)methanol, 8A-Et
intrinsic contribution of this bond. Neither molecular mech-
anics nor quantum mechanics distinguishes intramolecular
hydrogen bonding from other electrostatic interactions, making
the separation of steric and hydrogen bonding effects not only
difficult but to a certain extent artificial. Our attempt to do so,
in spite of this observation, suggests that in the oxygen-
substituted thiophene series investigated these two factors are of
roughly equal importance. Quantum mechanical calculations,
whether semi-empirical or ab initio, tend to overestimate the
difference between the extremes. The AM1 parametrization
accurately reproduces the hydrogen bond geometry of 8A-Et in
the solid state, ab initio calculations suggesting much tighter
hydrogen bonds.
The reason for the popularity of EDOT as a monomer for
electrochemical and chemical polymerization lies in the fact
that it is readily available on a large scale and that its polymers
are electrochemically stable, highly conductive, transparent in
the doped form and have a low band gap. Early studies showed
that 1 (n = 3) gave shorter, more soluble chains, presumably
because of steric hindrance to coupling.2 If this is the case, one
cannot but wonder why more attention has not been paid to 1
(n = 1). However, the effective conjugation length is shorter in
PMDOT than in PEDOT, possibly because of ring strain.5
From a chemical standpoint, as probed by studying the equi-
libria between intramolecularly hydrogen-bonded and solvent-
hydrogen-bonded 2-thienyldi(tert-butyl)methanol rotamers,
EDOT is situated on a continuum between 3,4-methylene-
dioxythiophene and other 3,4-alkylenedioxy-, 3-alkoxy- and
3,4-dialkoxythiophenes.
Insofar as tert-butyl and 1-adamantyl have the same symmetry
and despite the fact that the latter is more rigid and, con-
sequently, more space-demanding, it is instructive to compare
the structure of this new product with that previously deter-
mined for anti-3-methoxy-2-thienyldi(1-adamantyl)methanol,
4A-Me.8 In the diethoxy compound the hydroxy proton is 1.91
Å from the 3-ethoxy oxygen, the O ؒ ؒ ؒ O distance 2.66 Å and
the O–H ؒ ؒ ؒ O angle 138Њ. The distances are somewhat greater
than those for the 3-methoxy derivative (1.78 and 2.62 Å,
respectively) and the angle somewhat smaller (156Њ). Such
O ؒ ؒ ؒ O distances are associated with much stronger hydrogen
bonding in the methanol and acetic acid dimers.28 It is interest-
ing to note that the two ethoxy groups adopt the TGGЈ con-
formation, that which is preferred in 1,2-dimethoxyethane.29 All
other features (Fig. 4, Supplementary Material Table S6) are
Experimental
General methods have been described in previous papers.8,20,30f
NMR chemical shifts of hydroxy protons in deuteriochloro-
form at 298 K are given in ppm (reference value: δH = 7.26 ppm
with respect to TMS). NMR data for hydroxy proton shifts
in other solvents and temperature coefficients are given in
Supplementary Material Table S1. Full details of the 1H
and 13C NMR spectra of all new compounds in chloroform are
given in Supplementary Material Table S7.
Fig. 4 CAMERON diagram for anti-3,4-diethoxy-2-thienyldi(tert-
butyl)methanol, 8A-Et, showing 30% probability displacement
ellipsoids. Hydrogen atoms have been omitted for clarity.
typical of aryl- and heteroaryldiadamantylmethyl derivatives
previously studied.30 The major difference between 4A-Me and
8A-Et is in the orientation of the 3-alkoxy group, the O–CH2
bond being at about 70Њ to the thiophene plane (relative to C2)
in 8A-Et, whereas in the 3-methoxy derivative the methyl group
is almost coplanar (10Њ). These values are close to those calc-
ulated by MMFF94 [68 and 12Њ, respectively, both for the
di(tert-butyl) derivative] and by AM1 (69 and 7Њ, respectively).
The 4-ethoxy group, on the other hand, is in the plane of the
ring, whereas the MM calculation puts it at about 60Њ to the
plane.
3,4-Alkylenedioxythiophenes, 1
Prepared from diethyl 3,4-dihydroxythiophene-2,5-dicarboxy-
late by reaction with 1,ω-dibromoalkanes in DMF at 90–100 ЊC
in the presence of potassium carbonate,4a,31 hydrolysis to the
diacid, followed by decarboxylation and purification as for
3,4-diethoxythiophene (vide infra).11
Conclusion
3,4-Methylenedioxythiophene, 1 (n ؍ 1). Yield, from the
dicarboxylate, 13%; oil (Found: C, 46.8; H, 3.3; S, 24.5.
C5H4O2S requires C, 46.86; H, 3.15; S, 25.02%).
In solution at room temperature 2-thienyldi(tert-butyl)-
methanols, with alkoxy substituent(s) at the 3- or 3- and 4-
positions or with 3,4-alkylenedioxy chains, exist as a mixture of
two conformers in equilibrium. As the size and number of the
substituent(s) or the length of the alkylene chain is increased
the equilibrium favours the anti rotamer, that in which there is
intramolecular hydrogen bonding between the oxygen at the 3-
position and the hydroxy proton. Much the same trend is fol-
lowed by the activation energies for the interconversion of the
two rotamers, the barrier height increasing in the same order as
the equilibrium constant. IR and NMR spectroscopic evidence
indicates that hydrogen bond strength increases on going from
3,4-methylenedioxy to longer chains and through the 3-alkoxy
series, increasing with the size of the alkoxy group, but that
the 3,4-dialkoxy derivatives differ little from the 3-alkoxy
compounds.
3,4-Propylenedioxythiophene, 1 (n ؍ 3). Yield, from the
dicarboxylate, 20%; mp 81–82 ЊC (lit.4a 82–84 ЊC).
3,4-Butylenedioxythiophene, 1 (n ؍ 4). Yield, from the
dicarboxylate, 21%; oil.
3,4-Pentylenedioxythiophene, 1 (n ؍ 5). Yield, from the
dicarboxylate, 4%; oil (Found: C, 58.8; H, 6.7. C9H12O2S
requires C, 58.67; H, 6.56%).
3,4-Diethoxythiophene, 3-Et
Prepared from the dipotassium salt of diethyl 3,4-
dihydroxythiophene-2,5-dicarboxylate by reaction with diethyl
sulfate in toluene catalysed by [18]crown-6, as for 3,4-
dimethoxythiophene, 3-Me,11 followed by hydrolysis to the
diacid and subsequent decarboxylation in quinoline for 2 h at
The failure of the Badger–Bauer rule22 for intramolecular
hydrogen bonding has been attributed to “conformation” and
“geometric” effects, meaning that changes in conformational
energy associated with hydrogen bonding tend to obscure the
222
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224

View Article Online
170 ЊC with copper chromite as catalyst; purified by chrom-
atography on alumina in light petroleum (b. range 35–60 ЊC)–
diethyl ether mixtures (yield, from the dipotassium salt, 30%):
mp 42 ЊC (Found: C, 55.6; H, 7.2; S, 19.1. C8H12O2S requires C,
55.78; H, 7.02; S, 18.62%).
were linear over the temperature ranges indicated for each
compound in a given solvent. The further the equilibrium
constant from unity, the lower the accuracy of its measurement:
at the extremes values estimated from different parts of the
1H NMR spectrum may vary by up to 20%, which represents
0.08 l.u. or 0.11 kcal molϪ1 at 298 K.
The temperature dependence of the equilibrium constant for
6 (n = 2) was also studied in chloroform and benzene. Variations
in chloroform (298–328 K) were random and no greater than
1%; in benzene (298–343 K) the variation was greater (3%),
[syn]/[anti] tending to increase slightly with temperature.
Chloroform: ∆GЊ(anti Ϫ syn) = 0.028 0.042 Ϫ (0.3 0.1)T ;
benzene: ∆GЊ (anti Ϫ syn) = 0.088 0.026 Ϫ (0.6 0.1)T .
Alcohol synthesis
To a mixture of the appropriate thiophene derivative (5 mmol)
and TMEDA (0.75 cm3, 5 mmol) in diethyl ether (15 cm3) under
argon at room temperature was added a solution of n-
butyllithium in hexane (1.6 M, 3.2 cm3, 5 mmol). After 30 min
stirring di(tert-butyl) ketone (0.78 g, 5.5 mmol) was added. The
mixture was stirred for a further 30 min, then quenched with
water and the organic materials extracted with diethyl ether.
Washing with water, drying and evaporation of solvent
and excess ketone gave an oily or solid residue from which the
alcohol was isolated by chromatography on alumina in light
petroleum (b. range 35–60 ЊC)–diethyl ether mixtures.
Rotation kinetics
Dynamic NMR was used. The 1H NMR spectrum of a solution
of the compound in DMSO-d6 was recorded at temperatures
ranging from 298 to over 400 K. Exchange generally occurred
at a significant rate from about 350 K onwards. The 3,4-
methylenedioxythienyl derivative, 6 (n = 1), was studied in
chloroform and benzene up to 328 K only. Simulation of
the tert-butyl, methyl, hydroxy or aromatic proton signals by
gNMR17 gives the exchange rate and the relative concentrations
of the two species from which rate constants and the rotation
barriers are calculated. To obtain an Eyring plot with the points
all within 0.02 kcal molϪ1 of the line the rate constants must be
known to better than 3%. The activation energies (∆G≠ in kcal
molϪ1) listed are the means of 6–12 self-consistent data points
(i.e. following a roughly linear Eyring plot) for the mean
temperature (in K) at which the corresponding rate data
were recorded (Table 3). Activation enthalpies and entropies are
listed in Supplementary Material Table S2. These data are
somewhat sensitive to errors in the assumed line-width for
spectra in the absence of exchange.
3,4-Methylenedioxy-2-thienyldi(tert-butyl)methanol, 6 (n ؍
1). Yield 64%: mp 64 ЊC (Found: C, 62.2; H, 8.4; S, 11.9.
C14H22O3S requires C, 62.19; H, 8.20; S, 11.86%). anti: νOH/cmϪ1
(CCl4) see text; δOH 2.70. syn: νOH/cmϪ1 (CCl4) see text; δOH 2.18.
3,4-Propylenedioxy-2-thienyldi(tert-butyl)methanol, 6 (n ؍ 3).
Yield 31%: mp 81 ЊC (Found: C, 64.6; H, 8.8; S, 10.8. C16H26O3S
requires C, 64.39; H, 8.78; S, 10.74%). anti: νOH/cmϪ1 (CCl4)
3507; δOH 5.96. syn: νOH/cmϪ1 (CCl4) 3606, 3628; δOH 2.19.
3,4-Butylenedioxy-2-thienyldi(tert-butyl)methanol, 6 (n ؍ 4).
Yield 56%: oil (Found: C, 65.3; H, 8.9; S, 10.5. C17H28O3S
requires C, 65.34; H, 9.03; S, 10.26%). anti: νOH/cmϪ1 (CCl4)
3505; δOH 5.87. syn: νOH/cmϪ1 (CCl4) 3605, 3628; δOH 2.15.
3,4-Pentylenedioxy-2-thienyldi(tert-butyl)methanol, 6 (n ؍ 5).
Yield 61%: mp 57 ЊC (Found: C, 66.2; H, 9.1; S, 10.0. C18H30O3S
requires C, 66.22; H, 9.26; S, 9.82%). anti: νOH/cmϪ1 (CCl4)
3496; δOH 6.06. syn: νOH/cmϪ1 (CCl4) 3606, 3629; δOH 2.16.
Molecular mechanics calculations
Molecular mechanics calculations were performed using the
MMFF94 force field with the MMFF94 charge model in the
Sybyl 6.7 package.25 Steric energies (kcal molϪ1) and torsion
angles for the alkoxy substituents in the most stable conform-
ations are given in Supplementary Material Table S3.
3-Methoxy-2-thienyldi(tert-butyl)methanol, 7-Me. Yield 64%:
oil (Found: C, 65.4; H, 9.6; S, 12.4. C14H24O2S requires C, 65.58;
H, 9.43; S, 12.50%). anti: νOH/cmϪ1 (CCl4) 3516; δOH 5.90. syn:
νOH/cmϪ1 (CCl4) 3607, 3630; δOH 2.24.
Quantum chemical calculations
3-Ethoxy-2-thienyldi(tert-butyl)methanol, 7-Et. Yield 79%: oil
(Found: C, 66.4; H, 9.9; S, 12.0. C15H26O2S requires C, 66.62;
H, 9.69; S, 11.86%). anti: νOH/cmϪ1 (CCl4) 3500; δOH 6.12. syn:
νOH/cmϪ1 (CCl4) 3606, 3630; δOH 2.24.
Energies and geometries of the syn and anti rotamers of all
alcohols except 6 (n = 5) were determined by fully optimized
semi-empirical AM1 quantum mechanical calculations, using
the Spartan package.27 Data for the most stable conformations
are reported (Supplementary Material Tables S4 and S5).
Ab initio calculations (HF/3-21G and HF/6-31G*) were run on
6 (n = 1, 2), 7-R and 8-Me only.
3-Isopropoxy-2-thienyldi(tert-butyl)methanol, 7-i-Pr. Yield
59%: oil (Found: C, 67.6; H, 10.0; S, 11.4. C16H28O2S requires C,
67.56; H, 9.92; S, 11.27%). anti: νOH/cmϪ1 (CCl4) 3490; δOH 6.21.
syn: νOH/cmϪ1 (CCl4) 3606, 3630; δOH 2.24.
X-Ray crystallography‡
3,4-Dimethoxy-2-thienyldi(tert-butyl)methanol, 8-Me. Yield
52%: mp 112 ЊC (Found: C, 62.9; H, 9.3; S, 11.5. C15H26O3S
requires C, 62.90; H, 9.15; S, 11.19%). anti: νOH/cmϪ1 (CCl4)
3502; δOH 5.87. syn: νOH/cmϪ1 (CCl4) 3606, 3626; δOH 2.20.
anti-3,4-Diethoxy-2-thienyldi(tert-butyl)methanol.
C17H30O3S, 8A-Et. Crystal data. M = 314.5. Monoclinic,
a = 12.367(3), b = 11.139(2), c = 14.127(3) Å, β = 112.72(7)Њ,
V = 1795(7) Å3 (by least squares refinement on diffractometer
angles for 25 automatically centred reflections, λ = 0.71069 Å),
space group P21/c, Z = 4, Dx = 1.16 g cmϪ3. Colourless prismatic
3,4-Diethoxy-2-thienyldi(tert-butyl)methanol, 8-Et. Yield
47%: mp 101 ЊC (Found: C, 64.8; H, 9.8; S, 10.0. C17H30O3S
requires C, 64.93; H, 9.62; S, 10.20%). anti: νOH/cmϪ1 (CCl4)
3486; δOH 6.09. syn: νOH/cmϪ1 (CCl4) 3605, 3629; δOH 2.20.
crystals, ν(Mo-Kα) = 1.8 cmϪ1
.
Data collection and processing. At 295 K, Enraf-Nonius
MACH-3 diffractometer, ω/2θ mode with ω scan width = 0.8 ϩ
0.345 tan θ, graphite-monochromated Mo-Kα radiation. 3898
reflections measured (1 ≤ θ ≤ 26Њ), 3523 unique, giving 2271 with
I > 3σ(I ).
Equilibrium constants for anti ↔ syn rotamerization
Samples of the various alcohols (ca. 10 mg) were made up in
deuteriated chloroform, benzene, pyridine or DMSO (0.5 cm3).
Except in the case of 6 (n = 1) and 6 (n = 2), the first two solvents
were studied only at 298 K, the others over a temperature range
depending on the boiling point of the solvent. Free energy plots
‡ CCDC reference number 172908. See http://www.rsc.org/suppdata/
p2/b1/b109612p/ for crystallographic files in .cif or other electronic
format.
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224
223

View Article Online
Chem., 1970, 74, 3535; ( f ) M. C. R. Symons, Chem. Soc. Rev., 1983,
12, 1.
14 J. S. Lomas, A. Adenier, C. Cordier and J. C. Lacroix, J. Chem. Soc.,
Perkin Trans. 2, 1998, 2647.
15 J. S. Lomas and A. Adenier, J. Chem. Soc., Perkin Trans. 2, 2001,
1051.
16 P. Hodgson, G. C. Lloyd-Jones, M. Murray, T. M. Peakman and
R. L. Woodward, Chem. Eur. J., 2000, 6, 4451.
17 gNMR (version 4.1) from Adept Scientific plc, Amor Way,
Letchworth SG6 1ZA.
18 S. W. Benson, Thermochemical Kinetics, Wiley, New York, 2nd Edn.,
1976, p. 20.
19 See also: H. Suezawa, H. Wada, H. Watanabe, T. Yuzuri, K.
Sakakibara and M. Hirota, J. Phys. Org. Chem., 1997, 10, 925.
20 J. S. Lomas, J. C. Lacroix and J. Vaissermann, J. Chem. Soc., Perkin
Trans. 2, 1999, 2001.
21 (a) J. S. Lomas and J. E. Dubois, Tetrahedron, 1981, 37, 2273; (b) A.
Kolocouris, J. G. Outeiriño, J. E. Anderson, G. Fytas, G. B. Foscolos
and N. Kolocouris, J. Org. Chem., 2001, 66, 4989.
Structure analysis and refinement. Full-matrix least-squares
refinement with all non-hydrogen atoms anisotropic; hydrogens
placed in geometrically calculated positions and refined with one
overall isotropic thermal parameter (192 refinable parameters).
No absorption correction. Final R and Rw (Chebyshev series)
values are 0.055 and 0.074. Programmes used were the PC
version of CRYSTALS32 for refinements and CAMERON33 for
views. Selected bond lengths, bond angles and torsion angles
are listed in Supplementary Material Table S6.
Atomic coordinates, bond lengths and angles and thermal
parameters have been deposited at the Cambridge Crystallo-
graphic Data Centre (CCDC). For details of the deposition
scheme, see ‘Instructions for Authors’ (http://www.rsc.org/
authors). Any request to the CCDC for this material should
quote the full literature citation and the reference number
172908.
22 R. M. Badger and S. H. Bauer, J. Chem. Phys., 1937, 5, 839.
23 M. D. Joesten and L. J. Schaad, Hydrogen Bonding, Marcel Dekker,
New York, 1974, chap. 4 and 5.
24 T. A. Halgren, J. Comput. Chem., 1996, 17, 490, 520, 553 and 615;
T. A. Halgren and R. B. Nachbar, J. Comput. Chem., 1996, 17, 587;
T. A. Halgren, J. Comput. Chem., 1999, 20, 720.
25 Sybyl 6.7 from Tripos Inc., 1699 South Hanley Road, St. Louis,
MO 63144–2917, USA.
26 (a) L. P. Kuhn and R. A. Wires, J. Am. Chem. Soc., 1964, 86, 2161;
(b) E. Fishman and T. L. Chen, Spectrochim. Acta, Part A, 1969, 25,
1231.
References
1 For a recent review on polyEDOT, see L. Groenendaal, F. Jonas, D.
Freitag, H. Pielartzik and J. R. Reynolds, Adv. Mater., 2000, 12, 481.
2 G. Heywang and F. Jonas, Adv. Mater., 1992, 4, 116; M. Dietrich,
J. Heinze, G. Heywang and F. Jonas, J. Electroanal. Chem., 1994,
369, 87.
3 A. K. Mohanakrishnan, A. Hucke, M. A. Lyon, M. V.
Lakshmikantham and M. P. Cava, Tetrahedron, 1999, 55, 11745.
4 (a) A. Kumar, D. M. Welsh, M. C. Morvant, F. Piroux, K. A.
Abboud and J. R. Reynolds, Chem. Mater., 1998, 10, 896; (b) D. M.
Welsh, A. Kumar, M. C. Morvant and J. R. Reynolds, Synth. Met.,
1999, 102, 967.
27 PC Spartan from Wavefunction, Inc., 18401 Von Karman Avenue,
Suite 370, Irvine, CA 92612, USA.; AM1: M. J. S. Dewar, E. G.
Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am. Chem. Soc., 1985,
107, 3902.
5 H. J. Ahonen, J. Kankare, J. Lukkari and P. Pasanen, Synth. Met.,
1997, 84, 215.
6 (a) G. Dian, G. Barbey and B. Decronix, Synth. Met., 1986, 13, 281;
(b) S. Tanaka, M. Sato and K. Kaeriyama, Synth. Met., 1988, 25,
277; (c) T. Yamamoto, A. Kashiwazaki and K. Kato, Makromol.
Chem., 1989, 190, 1649; (d ) See also L. L. Miller and Y. Yu, J. Org.
Chem., 1995, 60, 6813 and references therein.
7 T. Hagiwara, M. Yamaura, K. Sato, M. Hirasaka and K. Iwata,
Synth. Met., 1989, 32, 367.
8 J. S. Lomas, E. Vauthier and J. Vaissermann, J. Chem. Soc., Perkin
Trans. 2, 2000, 1399.
9 J. S. Lomas, J. Chem. Soc., Perkin Trans. 2, 2001, 754.
10 J. S. Lomas, unpublished work.
11 A. Merk and C. Rehm, J. Prakt. Chem., 1996, 338, 672.
12 M. A. Keegstra, T. H. A. Peters and L. Brandsma, Tetrahedron,
1992, 48, 3633.
13 (a) L. W. Reeves, E. A. Allan and K. O. Strømme, Can. J. Chem.,
1960, 38, 1249; (b) I. Gränacher, Helv. Phys. Acta, 1961, 34, 272; (c)
J. R. Merrill, J. Phys. Chem., 1961, 65, 2023; (d ) G. O. Dudek, J. Org.
Chem., 1965, 30, 548; (e) A. D. Sherry and K. F. Purcell, J. Phys.
28 S. N. Vinogradov and R. H. Linnell, Hydrogen Bonding, van
Nostrand-Reinhold, New York, 1971; see also: R. Taylor and O.
Kennard, Acc. Chem. Res., 1984, 17, 320; F. Hibbert and J. Emsley,
Adv. Phys. Org. Chem., 1990, 26, 255.
29 S. Tsuzuki, T. Uchimaru, K. Tanabe and T. Hirano, J. Phys. Chem.,
1993, 97, 1346.
30 (a) J. S. Lomas and J. Vaissermann, J. Chem. Soc., Perkin Trans. 2,
1996, 1831; (b) J. S. Lomas and J. Vaissermann, Bull. Soc. Chim. Fr.,
1996, 133, 25; (c) J. Vaissermann and J. S. Lomas, Acta Crystallogr.,
Sect. C, 1997, 53, 1341; (d ) J. S. Lomas and J. Vaissermann, J. Chem.
Soc., Perkin Trans. 2, 1997, 2589; (e) J. S. Lomas and J. Vaissermann,
J. Chem. Soc., Perkin Trans. 2, 1998, 1777; ( f ) J. S. Lomas and
J. Vaissermann, J. Chem. Soc., Perkin Trans. 2, 1999, 1639.
31 See also: F. Dallacker and V. Mues, Chem. Ber., 1975, 108, 569.
32 D. J. Watkin, J. R. Carruthers and P. W. Bettridge, CRYSTALS. User
Guide, Chemical Crystallography Laboratory, University of Oxford,
1988.
33 J. Pearce, D. J. Watkin and D. P. Prout, CAMERON. A Program
for Plotting Molecular Structures, Chemical Crystallography
Laboratory, University of Oxford, 1992.
224
J. Chem. Soc., Perkin Trans. 2, 2002, 216–224