Self-association in {2-[3,4-alkylenedioxy-5-(3-pyridyl)]-thienyl}alkanols: An NMR, IR, and single-crystal X-ray study

Article abstract of DOI:10.1002/poc.1166

syn-2,2,4,4-Tetramethyl-3- {2-[3,4-alkylenedioxy-5-(3-pyridyl)]thienyl} pentan-3-ols self-associate both in the solid state and in solution. Single-crystal X-ray diffraction study of the 3,4-ethylenedioxythiophene (EDOT) derivative shows that it exists as a centrosymmetric head-to-tail, syn dimer in the solid state. The IR spectra of the solids display only a broad OH absorption around 3300 cm^-1, corresponding to a hydrogen-bonded species. ¹H Nuclear Overhauser Effect Spectroscopy (NOESY) NMR experiments in benzene reveal interactions between the fert-butyl groups and the H2 and H6 protons of the pyridyl group. Two approaches have been used to determine association constants of the EDOT derivative by NMR titration, based on the concentration dependence of (i) the synl anti ratio and (ii) the OH proton shift of the syn rotamer. Reasonably concordant results are obtained from 298 to 323 K (3.6 and 3.9 M^-1, respectively, at 298 K). Similar values are obtained from the syn OH proton shift variation for the 3,4- methylenedioxythiophene (MDOT) derivative. Concentration-dependent variation of the anti OH proton shift in the latter suggests that the anti isomer associates in the form of an open, singly hydrogen-bonded dimer, with a much smaller association constant than the syn rotamer. Self-association constants for 3-pyridyl-EDOT-alkanols with smaller substituents vary by a factor of 4 from (i-Pr)₂ up to (CD₃)₂, while the hetero-association constants for the same compounds with pyridine vary slightly less. Copyright

Full text of DOI:10.1002/poc.1166

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2007; 20: 410–421
Published online 29 March 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/poc.1166
Self-association in {2-[3,4-alkylenedioxy-5-(3-pyridyl)]-
thienyl}alkanols: an NMR, IR, and single-crystal
X-ray study
1
1
1
1
2
*
John S. Lomas, Christine Cordier, Alain Adenier, Franc¸ois Maurel and Jacqueline Vaissermann
1
`
´
Interfaces, Traitements, Organisation et Dynamique des Systemes, Universite de Paris 7, CNRS UMR 7086, 1 rue Guy de la Brosse,
75005 Paris, France
2
´
´
´
Laboratoire de Chimie Inorganique et Materiaux Moleculaires, Universite de Paris 6, CNRS UMR 7071, Case 42, 4 place Jussieu,
75252 Paris Cedex 05, France
Received 4 August 2006; accepted 15 January 2007
ABSTRACT: syn-2,2,4,4-Tetramethyl-3-{2-[3,4-alkylenedioxy-5-(3-pyridyl)]thienyl}pentan-3-ols self-associate
both in the solid state and in solution. Single-crystal X-ray diffraction study of the 3,4-ethylenedioxythiophene
(EDOT) derivative shows that it exists as a centrosymmetric head-to-tail, syn dimer in the solid state. The IR spectra of
the solids display only a broad OH absorption around 3300 cmꢀ1, corresponding to a hydrogen-bonded species.
1H Nuclear Overhauser Effect Spectroscopy (NOESY) NMR experiments in benzene reveal interactions between the
tert-butyl groups and the H2 and H6 protons of the pyridyl group. Two approaches have been used to determine
association constants of the EDOT derivative by NMR titration, based on the concentration dependence of (i) the syn/
anti ratio and (ii) the OH proton shift of the syn rotamer. Reasonably concordant results are obtained from 298 to 323 K
(3.6 and 3.9 Mꢀ1, respectively, at 298 K). Similar values are obtained from the syn OH proton shift variation for the
3,4-methylenedioxythiophene (MDOT) derivative. Concentration-dependent variation of the anti OH proton shift in
the latter suggests that the anti isomer associates in the form of an open, singly hydrogen-bonded dimer, with a much
smaller association constant than the syn rotamer. Self-association constants for 3-pyridyl-EDOT-alkanols with
smaller substituents vary by a factor of 4 from (i-Pr)2 up to (CD3)2, while the hetero-association constants for the same
compounds with pyridine vary slightly less. Copyright # 2007 John Wiley & Sons, Ltd.
Supplementary electronic material for this paper is available in Wiley Interscience at http://www.interscience.
wiley.com/jpages/0894–3230/suppmat/
KEYWORDS: hydrogen bonding; rotamers; self-association; hetero-association
INTRODUCTION
and interact in a 1:1 ratio: AꢁꢁꢁD. Such 1:1 hydro-
gen-bonded complexes are intrinsically unstable,
because of the weakness of the hydrogen bond, and
are easily disrupted by solvation of the acceptor and/or
donor sites. Multiplication of the number of hydrogen
bonds by the inclusion of several sites leads to more
complicated bonding patterns and enhanced stability.2
The extreme situation is that of water, where the entire
molecular surface consists of donor and acceptor
groups, the cooperative effect of which leads to stability
out of all proportion to the strength of the hydrogen bond
in the water dimer.3
In this work we shall consider O—HꢁꢁꢁN bonding in
species where the donor and the acceptors are present in
the same molecule, giving rise potentially to a centro-
symmetric array. This is a key step towards the cons-
truction of supramolecular architectures based on a
relatively uncommon type of hydrogen bonding. Previous
work on self-association by O—HꢁꢁꢁN bonding in
pyridine-substituted alcohols is limited to three stu-
The hydrogen bond is a non-covalent force arising from
the electrostatic interaction between an electronegative
donor group, D, consisting of a hetero-atom, usually
oxygen or nitrogen, bearing a covalently bound
hydrogen atom, and an electron-acceptor, A, again O
or N, in another or the same molecule. This type of
bonding, albeit much weaker than covalent bonding, is
one of the most important factors determining the
three-dimensional structures of supramolecular assem-
blies in chemical and biochemical systems.1 The very
simplest bonding situation is that where two different
molecules carry a single donor and a single acceptor,
*Correspondence to: J. S. Lomas, Interfaces, Traitements, Organis-
`
ation et Dynamique des Systemes, Universite de Paris 7, 1 rue Guy de la
Brosse, 75005 Paris, France.
´
E-mail: lomas@itodys.jussieu.fr
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
411
dies.4–6 Early NMR4 and IR5 spectroscopic studies on
RESULTS AND DISCUSSION
Synthesis
(2-pyridyl)alkanols reveal intermolecular as well as
intramolecular hydrogen bonding due to the proximity of
the OH group and the nitrogen atom. More generally,
pyridyl derivatives of simple alkanols, X-pyridyl
(CH2)nOH, where X ¼ 2, 3, or 4 and n ¼ 1, 2, or 3,
corresponding to methanol, ethanol, or propanol, respect-
ively, form hydrogen-bonded self-association complexes,
probably open and/or cyclic dimers as well as cyclic
trimers and tetramers.6 In the sterically hindered
2,2,4,4-tetramethyl-3-(X-pyridyl)pentan-3-ols, the pyri-
dyl group and the encumbered OH hydrogen are
paradoxically too close for cyclic dimers or higher
n-mers to be formed, though NMR studies are consistent
with association (not X ¼ 2), presumably as open dimers
or n-mers.6 Our aim was to find a rigid spacer
2-(X-Pyridyl)-3,4-alkylenedioxythiophenes, 4–6 (X ¼ 2
or 3), were prepared by reaction of the appropriate
2-(tri-n-butylstannyl)thiophene with X-bromopyridine in
DMF at 120 8C catalyzed by bis(triphenylphosphine)palla-
dium(II) chloride (Stille reaction).11 The 3,4-substituted
2,2,4,4-tetramethyl-3-{2-[5-(3-pyridyl)]thienyl}pentan-3-
ols, 1, 2a, and 3, were synthesized from the above by
lithiation with n-BuLi/TMEDA in diethyl ether at room
temperature followed by reaction with 2,2,4,4-tetra-
methylpentan-3-one. Less encumbered derivatives 2b–d
(R ¼ i-Pr, Et, or Me) were synthesized in the same way
from 5 at ꢀ75 8C.
which would prevent intramolecular O—HꢁꢁꢁN hydrogen
bonding, at the same time as it reduced the
impact of steric hindrance and would thereby allow
self-association of such alcohols. A suitable system
was suggested by work on the 3-alkoxy,7 3,4-dialkoxy,7
and 3,4-alkylenedioxy7–10 derivatives of 2,2,4,4-tetra-
methyl-3-(2-thienyl)pentan-3-ols, which exist in two
rotameric forms, where the hydroxy hydrogen is either
intramolecularly hydrogen-bonded (anti) or ‘free’ (syn).
Hydrogen-bond-acceptor solvents such as pyridine and
DMSO favor the syn form in which the OH group is
hydrogen-bonded to the solvent (AꢁꢁꢁD hetero-associat-
ion). If an acceptor group is introduced into such a
molecule with a suitable geometry, self-association
should occur. To test this prediction we have synthesized
two 5-(3-pyridyl) derivatives of 2,2,4,4-tetramethyl-
3-[2-(3,4-alkylenedioxy)thienyl]pentan-3-ols, 1 and 2a.
Their IR and NMR spectroscopic behavior has been
investigated and the single-crystal X-ray diffraction
structure of 2a determined. Equilibrium constants for
self- and hetero-association (with pyridine) of 2a and less
encumbered species 2b–d (t-Bu replaced by i-Pr, Et, or
Me), determined by NMR titration, establish that steric
effects on association are small.
Single-crystal X-ray diffraction structure
of syn-2-2,4,4-tetramethyl-3-{2-[3,4-
ethylenedioxy-5-(3-pyridyl)]thienyl}pentan-
3-ol, 2a
The unit cell of compound 2a contains a centrosymmetric
pair of molecules associated by two hydrogen bonds
between the OH hydrogens and the pyridine nitrogens, the
distance between the oxygen and nitrogen atoms being
˚
2.86 A. The —C(t-Bu)2OH group is in the syn orientation,
and the pyridine ring is at about 208 to the thiophene ring
(Fig. 1). Selected bond lengths, bond angles, and torsion
angles are given in Supplementary Material Table S1 with
a labeled CAMERON diagram.
Although O—HꢁꢁꢁN hydrogen bonding occurs in a
large number of structures determined by X-ray diffrac-
tion – a search for crystallographic data with ConQuest (v.
1.8, CDDC, Cambridge, UK) gave about 1200 hits with
˚
HꢁꢁꢁN distances between 1.5 and 2.5 A – only two other
structures appear to involve similar centrosymmetric
pairs joined through weak hydrogen bonds, with NꢁꢁꢁO
12
distances of 3.03 and 2.95 A. One of these involves a
˚
pyridine ring and a phenolic OH,12a the other an
imidazole ring and an alcoholic OH.12b
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

412
J. S. LOMAS ET AL.
of about 45% of the anti rotamer, and the apparently low
association constant (Supplementary Material Figure S1).
Weak correlations are observed for the tert-butyl signal of
the anti rotamer with the {H2, H4} pyridyl protons. For
the syn isomer of 2a there are correlations between the
tert-butyl group and only the {H2, H6} protons. An un-
expected feature of the NOESY spectrum is a cross-peak
corresponding to the tert-butyl group of the syn
isomer and the OH group of the anti isomer. This is
attributed to fast chemical exchange of the OH group
protons between the different rotamers during the mixing
time, an explanation confirmed by a 1D homonuclear
decoupling experiment, in which irradiation at either OH
proton frequency leads to a marked reduction in the signal
of the other OH proton (Supplementary Material
Figure S2).
IR Spectroscopy
On the basis of their wavenumbers and shapes, nOH bands
are assigned to the different alcohol isomers in carbon
tetrachloride (1 and 2a–d) and to solid samples deposited
from solution onto a KBr plate (1, 2a, and 3) (Table 1).
The strong, very broad signals at low wavenumbers for 1
and 2a in the solid state are consistent with a structure in
which the OH group is hydrogen-bonded to the pyridyl
nitrogen in an associated form of the syn rotamer. For the
2-pyridyl-EDOT derivative, 3, there is no association
band whatsoever in the cast spectrum, simply two
absorptions, corresponding to the syn and anti rotamers.
Previous work7 on analogs lacking the 3-pyridyl
substituent facilitates the assignment of the free syn and
intramolecularly hydrogen-bonded anti OH vibrations in
1, 2a, and 3 around 3600 cmꢀ1, but for 2b–c the
wavenumber differences are smaller and the assignments
less reliable. Alcohol 2d gives only a single broad
absorption in this region. For alcohol 1 there is a very
broad band at 3310 cmꢀ1, with a red-shift of about
310 cmꢀ1 relative to the free syn, while for 2a the
Figure 1. CAMERON diagram for syn-2-2,4,4-tetramethyl-
3-{2-[3,4-ethylenedioxy-5-(3-pyridyl)]thienyl} pentan-3-ol,
2a, showing 30% probability displacement ellipsoids. Hydro-
gen atoms, except OH, have been omitted for clarity
1H NOESY NMR experiments on 2,2,4,4-
tetramethyl-3-{2-[3,4-alkylenedioxy-5-
(3-pyridyl)]thienyl}pentan-3-ols
As compound 1 exists very predominantly as the syn
rotamer, no cross-peaks corresponding to the anti isomer
are detected. The 2D 1H Nuclear Overhauser Effect
Spectroscopy (NOESY) NMR spectrum of 1 in benzene
displays correlations between the tert-butyl groups of the
syn rotamer and the {H2, H6} pyridine protons, the
correlations with H6 being stronger. This correlation is
particularly significant, since H6 is the most remote from
the tert-butyl groups of the same sub-unit in both
rotamers, regardless of the orientation of the pyridyl
group. Furthermore, there are also correlations between
the syn OH proton and the same pyridyl protons, though
that with H6 is weak (Fig. 2). According to the X-ray
crystallographic structure of 2a there are four small
distances between tert-butyl group protons and the
red-shifts of the two weak bands are 410 and 330 cmꢀ1
.
That both are weak is due to the poor solubility of this
material and the fact that nearly half is in the anti form
(for 1 there is only 10% which is not syn – see below). For
2b–d the absorptions of the free syn and anti rotamers are
accompanied in all cases by strong bands attributed to an
associated form of the syn rotamer, with the red-shift
increasing from about 330 to 350 to 370 cmꢀ1 as the
substituent becomes smaller.
˚
pyridyl H6 (2.45, 3.53, 3.73, and 3.91 A); there is only
˚
one similar distance for H2 (3.50 A). Crystallography
˚
indicates that for 2a the OHꢁꢁꢁH2 distance (3.05 A) is
˚
greater than OHꢁꢁꢁH6 (2.88 A). Nevertheless, in 1, which
would be expected to have a very similar structure, the
correlation peak for OHꢁꢁꢁH2 appears to be more intense
than that for OHꢁꢁꢁH6. These contacts are consistent with
the presence of an associated species in which the
tert-butyl groups of one sub-unit are brought close to the
pyridyl substituent of a second sub-unit.
Self-association constants for {2-[3,4-
alkylenedioxy-5-(X-pyridyl)]thienyl}-alkanols
The NOESY experiment on 2a is complicated by the
fact that the effective concentration of the self-associated
species is small, owing to the low solubility, the presence
1H NMR titration. Benzene solutions of 2,2,4,4-
tetramethyl-3-{2-3,4-alkylenedioxy-5-(3-pyridyl)]thi-
enyl} pentan-3-ols,
1
and 2a, and of 2,2,4,
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
413
Figure 2. 2D NOESY spectrum of 2,2,4,4-tetramethyl-3-{2-[3,4-methylenedioxy-5-(3-pyridyl)]thienyl}pentan-3-ol, 1, in
benzene at 293 K
4-tetramethyl-3-{2-[3,4-ethylenedioxy-5-(2-pyridyl)]thi-
enyl}pentan-3-ol, 3, at various concentrations were
studied by NMR spectroscopy. For compound 1 the
shifts of the OH proton, the methylene, and the tert-butyl
group protons in the very predominant syn isomer
(ca. 91%) increase with concentration; that of the anti
OH proton increases very slightly, but the other signals
of the anti isomer are difficult to differentiate and,
depending on the temperature, are partially or totally
coalesced with those of the syn isomer (Supplementary
Material Table S2). For 2a the shifts of the OH proton
and the tert-butyl group protons of the syn isomer
also increase with concentration, while the corres-
ponding shifts for the anti isomer are virtually constant
within the experimental error (Supplementary Material
Table S3).
Table 1. IR OH stretching vibrations (nOH in cmꢀ1) for
{2-[3,4-alkylenedioxy-5-(X-pyridyl)]thienyl}alkanols 1, 2a–d,
and 3
The variations in the shifts of certain protons in
compounds 1 and 2a are consistent with the formation
of a self-associated species of the syn conformer.
While the antiÐsyn rotation (equilibrium constant K1)
is slow on the NMR time-scale and, consequently,
gives rise to well defined peaks for each rotamer,
the freeÐassociated interconversion (equilibrium const-
ant K2) is rapid and does not give separate signals but a
single somewhat broadened signal (Scheme 1). In view of
the results of the X-ray diffraction study it seems
Cpd. nOH Free syn nOH Bound anti nOH Bound syn n-mer
1a
—
—
3300
3280
—
2aa
3a
3562
3570
3613
3576
3598
3601
3600
3605
3624
3604, 3627
3619
3610
1
3310
2a
2b
2c
2d
3220, 3300
3285
3250
3600
3235
a Solid deposited onto KBr plate from CH2Cl2 solution; all others in CCl4.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

414
J. S. LOMAS ET AL.
Scheme 1
reasonable to assume that in solution the associated
species is also a cyclic dimer.
MA01060, USA), which uses the Levenberg–Marquardt
algorithm.
In what follows, S, A, and SS refer to the syn rotamer,
the anti rotamer, and the syn dimer, respectively; [ROH]o
is the analytical concentration of the alcohol. Taking
We then assume that hetero-association of the alcohol
with the solvent mimics self-association in benzene, that
the effects of inter- and intramolecular hydrogen bonding
are similar, that is, that dSpy, the OH proton shift in the
pyridine complex, can be taken as dSS. The self-
(dOH)
syn, the chemical shift of the syn OH, as the weighted
average of the shifts of the monomer, dS, and of the dimer,
dSS, following Chen and Shirts,13 we have Equation (1):
association constant, K2, is then found by fitting the
syn
experimental values of (dOH
)
versus [S]o to the
Chen–Shirts equation [Equation (1)]. Results are listed
in Tables 2 and 3. The very small standard deviations on
dS and K2 indicate that the Chen–Shirts equation fits the
data very well with this approximation.
If self-association involved only a single hydrogen
bond, then the syn dimer could be assumed to contain one
hydrogen-bonded OH group, with shift dSS, and one free
OH group with the same shift, dS, as in the monomer,
which gives:
1=2
ð1 þ 8K2½SꢂoÞ ꢀ 1
syn
ðdOH
Þ
¼ dSþðdSS ꢀ dSÞ
ð1Þ
1=2
ð1 þ 8K2½SꢂoÞ þ 1
where [S]o is the analytical concentration of the syn
rotamer. If R is the syn/anti ratio, [S]o/[A]o, where
[S]o ¼ [S] þ 2[SS], and [A]o is the concentration of the
anti rotamer, Equation (2) applies:
R½ROHꢂo
½Sꢂo ¼
(2)
1 þ R
2
2
½SSꢂ ¼ K3½Sꢂ and ½Sꢂo ¼ ½Sꢂ þ 2K3½Sꢂ
Since measurements are only possible at low monomer
and, therefore, low dimer concentration, the value of the
chemical shift of the OH proton in the associated form,
dSS, is poorly defined by parameter optimization through
Equation (1).14 This difficulty is overcome by determin-
ing the hetero-association constant with pyridine in
benzene and, at the same time, dSpy by means of Equation
(3):15
whence: ½Sꢂ ¼ ðB1=2 ꢀ 1Þ=4K3, where: B ¼ 1 þ 8K3½Sꢂo.
This leads to Equation (4):
1=2
ð1 þ 8K3½SꢂoÞ ꢀ 1
syn
ðdOH
Þ
¼ dS þ 0:5ðdSS ꢀ dSÞ
1=2
ð1 þ 8K3½SꢂoÞ þ 1
(4)
n
o
ðdSpy ꢀ dSÞ
2½Sꢂo
syn
1=2
ðdOH
Þ
¼ dS þ
B ꢀ ðB2 ꢀ 4½Sꢂo½pyꢂoÞ
Solving this equation for dS and K3, using as dSS and
dSpy values determined for hetero-association, gives for
alcohol 2a the data listed in Supplementary Material
Table S4, corresponding to reaction enthalpy and entropy
of ꢀ11.6 kcal molꢀ1 and ꢀ34 cal molꢀ1 Kꢀ1, respectively
(1 cal ¼ 4.184 J). Both terms are unusually high for the
formation of a single O—HꢁꢁꢁN hydrogen bond. For 2a a
second approach to the determination of the association
constant, described below, makes it possible to eliminate
the eventuality of an open dimer.
(3)
where B¼ ½Sꢂoþ½pyꢂoþ1=K; [py]o is the analytical
pyridine concentration. In order to minimize the
effect of self-association these measurements were
made at low alcohol concentration: at the lowest
pyridine concentration not more than 1% of the
alcohol is self-associated. Values of K, dSpy, and dS
are determined by fitting the experimental values of
syn
The use of an extended Chen–Shirts equation with
terms for both open and cyclic dimers [Equation (5)] leads
to particularly ill-defined association constants, one of
which becomes negative at the highest temperatures
(dOH
)
to those of [S]o and [py]o by means of the
non-linear least-squares curve fitting option of the Origin
program (Microcal Software, Inc., now OriginLab
Corporation, One Roundhouse Plaza, Northampton,
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
415
Table 2. Hetero- and self-association constants for 2,2,4,4-tetramethyl-3-{2-[3,4-methylenedioxy-5-(3-pyridyl)]thienyl}pentan-3-ol,
1, in benzene
b
dS /ppm
b
c
T/K
dSpya/ppm
Ka/Mꢀ1
K2 /Mꢀ1
dAc/ppm
K6 /Mꢀ1
298
303
308
313
318
323
6.548 ꢃ 0.045
6.508 ꢃ 0.013
6.480 ꢃ 0.016
6.468 ꢃ 0.013
6.375 ꢃ 0.013
6.323 ꢃ 0.015
1.10 ꢃ 0.03
0.99 ꢃ 0.01
0.88 ꢃ 0.01
0.76 ꢃ 0.01
0.71 ꢃ 0.01
0.64 ꢃ 0.01
1.969 ꢃ 0.008
1.963 ꢃ 0.008
1.958 ꢃ 0.007
1.958 ꢃ 0.007
1.958 ꢃ 0.006
1.960 ꢃ 0.005
2.57 ꢃ 0.05
2.08 ꢃ 0.03
1.69 ꢃ 0.03
1.38 ꢃ 0.03
1.17 ꢃ 0.02
0.98 ꢃ 0.02
3.010 ꢃ 0.002
3.012 ꢃ 0.001
3.016 ꢃ 0.001
3.020 ꢃ 0.001
3.023 ꢃ 0.001
3.025 ꢃ 0.002
0.28 ꢃ 0.01
0.26 ꢃ 0.01
0.24 ꢃ 0.01
0.21 ꢃ 0.01
0.21 ꢃ 0.01
0.20 ꢃ 0.02
a Hetero-association with pyridine.
b Self-association, from concentration dependence of syn OH proton shift, taking dSpy as dSS
c Self-association, from concentration dependence of anti OH proton shift.
.
Table 3. Hetero- and self-association constants for 2,2,4,4-tetramethyl-3-{2-[3,4-ethylenedioxy-5-(3-pyridyl)]thienyl}pentan-3-ol,
2a, in benzene
T/K
dSpya/ppm
Ka/Mꢀ1
K2 /Mꢀ1
b
K2 /Mꢀ1
c
b
dS /ppm
c
K1
298
303
308
313
318
323
6.518 ꢃ 0.009
6.477 ꢃ 0.010
6.445 ꢃ 0.009
6.400 ꢃ 0.010
6.296 ꢃ 0.012
6.296 ꢃ 0.011
1.03 ꢃ 0.01
0.93 ꢃ 0.01
0.83 ꢃ 0.01
0.74 ꢃ 0.01
0.68 ꢃ 0.01
0.59 ꢃ 0.01
2.008 ꢃ 0.012
2.007 ꢃ 0.010
2.009 ꢃ 0.007
2.013 ꢃ 0.005
2.016 ꢃ 0.003
2.019 ꢃ 0.002
3.88 ꢃ 0.11
3.01 ꢃ 0.08
2.37 ꢃ 0.06
1.89 ꢃ 0.04
1.55 ꢃ 0.02
1.24 ꢃ 0.02
1.064 ꢃ 0.006
1.061 ꢃ 0.004
1.052 ꢃ 0.007
1.037 ꢃ 0.004
1.043 ꢃ 0.005
1.042 ꢃ 0.003
3.55 ꢃ 0.20
2.88 ꢃ 0.14
2.54 ꢃ 0.21
2.43 ꢃ 0.12
1.78 ꢃ 0.15
1.59 ꢃ 0.08
a Hetero-association with pyridine.
b Self-association, from concentration dependence of syn OH proton shift, taking dSpy as dSS
c From concentration dependence of the syn/anti ratio (R).
.
(Table S4). This is not, therefore, a convincing model.
ð½ROHꢂo ꢀ ½AꢂÞ=½Aꢂ and rearranging give Equation (6):
1=2
ð1 þ 8K4½SꢂoÞ ꢀ 1
2
R2 þ Rð1 ꢀ K1Þ ¼ K1 þ 2½ROHꢂoðK1Þ K2
(6)
syn
ðdOH
Þ
¼ dS þ 0:5ðdSS ꢀ dSÞ
1=2
ð1 þ 8K4½SꢂoÞ þ 1
The values of K1 and K2 are obtained by solving the
quadratic equation for R and using the non-linear
least-squares procedure to fit the curve of R versus
[ROH]o. The similarity of the two sets of K2 values for 2a
(Table 3) indicates that our estimate, based on the
assumption that 2a forms a cyclic dimer in which the OH
shift is approximately the same as in the corresponding
association complex with pyridine, is satisfactory. The
greatest discrepancies occur at the highest temperatures,
1=2
ð1 þ 8K5½SꢂoÞ ꢀ 1
þ ðdSS ꢀ dSÞ
(5)
1=2
ð1 þ 8K5½SꢂoÞ þ 1
The syn/anti ratio. In the case of the EDOT derivative,
2a, where the syn/anti ratio is not far from unity and
the rotation barrier sufficiently high, R can be measured
in benzene with an accuracy of 1–2% by integration
of the tert-butyl group signals. This is also concentra-
tion-dependent, increasing with the concentration
(Supplementary Material Tables S2 and S5). We have
then:
where both methods are less reliable, the changes in both
syn
R and (dOH
)
being smaller than at low temperatures.
Mean K2 values for the whole temperature range are
2.46 Mꢀ1 (based on the syn/anti ratio) and 2.32 Mꢀ1
(based on (dOH)
syn), whereas for an open dimer it would
be 6.92 Mꢀ1 (Table S4). This second approach involves
no prior assumption about the shift of the associated syn
form, nor about the structure of the dimer. However, it is
limited to situations where the signals of the syn and anti
rotamers are well separated, and the syn/anti ratio can be
measured accurately. For 1 only NMR titration could be
used.
½Sꢂ ¼ K1½Aꢂ
and, assuming that SS is any sort of dimer:
2
2
½SSꢂ ¼ ðK1Þ K2½Aꢂ
2
2
whence:½ROHꢂo¼½AꢂþK1½Aꢂþ2ðK1Þ K2½Aꢂ . Putting R ¼
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

416
J. S. LOMAS ET AL.
Neither the shifts nor the
R
values for 2,
barriers for the —CR2OH group, and the syn and anti
isomers cannot be distinguished by 200 MHz 1H NMR at
298 K. A simpler form of the Chen–Shirts equation13 then
applies [Equation (7)]:
2,4,4-tetramethyl-3-{2-[3,4-ethylenedioxy-5-(2-pyridyl)]-
thienyl}pentan-3-ol, 3, in benzene at 298 K show
any great concentration dependence (Supplementary
Material Table S6). The very slight variation of the syn
OH shift corresponds to an association constant of
(1.7 ꢃ 0.2) ꢄ 10ꢀ2 Mꢀ1, the assumed dSS value being that
of dSpy determined by hetero-association with pyridine.
1=2
ð1 þ 8K2½ROHꢂoÞ ꢀ 1
dOH ¼ dM þ ðdD ꢀ dMÞ
(7)
1=2
ð1 þ 8K2½ROHꢂoÞ þ 1
where dM and dD are the shifts of the non-associated and
associated forms of the alcohol, respectively. As
described above for dSpy in the case of 1 and 2a, the
chemical shift of the dimer dD was attributed the value of
dMpy obtained by NMR titration of the alcohol against
pyridine in benzene. Furthermore, in order to compare 2a
with alcohols 2b–d, the weighted average OH shift was
calculated from the shifts of the two rotamers and the syn/
anti ratio, and was used with [ROH]o in Equation (7). This
leads to association constants which are necessarily
smaller than those based on the syn isomer alone.
Incidentally, this treatment reduces the self-association
constant for alcohol 3 in benzene at 298 K to zero
(Table 4).
The results (Table 4; full details in Supplementary
Material Table S7) show that the lowest hetero-association
constant is for the tert-butyl derivative, 2a, but that the
isopropyl derivative, 2b, is the least self-associated, there
being a slight increase for the ethyl and deuteriomethyl
compounds, 2c and 2d. The overall range is, however, very
small, being a factor of 2.9 (2d/2a) for hetero-association at
298 K and a little higher, 3.8 (2d/2b) for self-association. A
greater difference might have been expected, since two
hydrogen bonds are involved in self-association and only
one in hetero-association.
There is a marked increase in the shift of the associated
form, dMpy, at 298 K for 2c and 2d, 6.94 and 7.52 ppm,
respectively, whereas 2a and 2b are very similar at 6.40
and 6.50 ppm, respectively (Table 4). The shift of the
non-associated alcohols, on the other hand, falls in the
series 2a–d. Given that the value of 3.50 ppm for 2a is
the mean of the contributions of the syn and anti rotamers,
this variation implies that intramolecular hydrogen
bonding is less important for the alcohols with smaller
substituents. Qualitative comparison of the IR spectra in
carbon tetrachloride suggests an increase in the import-
ance of the syn isomer on going from 2a to 2c; the single
peak at 3600 cmꢀ1 for 2d makes it impossible to
determine whether this trend continues for the smallest
substituent, but overall the NMR data in benzene and the
IR data in carbon tetrachloride are compatible.
anti-2,2,4,4-Tetramethyl-3-{2-[3,4-methylene-
dioxy-5-(3-pyridyl)]thienyl}pentan-3-ol, 1. The OH
proton shift of the anti rotamer of the MDOT derivative,
1, is concentration-dependent in benzene (Table S2),
whereas that for the EDOT derivative, 2a (Table S3), is
constant. This suggests that the weakly intramolecularly
hydrogen-bonded OH hydrogen7 in 1 is able to
hydrogen-bond with the pyridine nitrogen of a second
molecule, but this can only be an open dimer with a single
hydrogen bond. In this case both homo- and hetero-
dimers are possible where, in the latter, an anti ‘head’ (the
OH group) is associated with a syn ‘tail’ (the pyridyl
group). To simplify the mathematics it is assumed that the
self-association of the syn form is unaffected, that is, that
the association of the syn rotamer with the anti rotamer
has an insignificant effect upon its concentration. The
treatment, detailed in the Supplementary Material,
assumes the same association constant, K6, for the
association of anti with anti or syn. Values of dA, the shift
of the free anti rotamer, and K6 are determined by
non-linear curve fitting, the chemical shift of the dimer
being attributed, for lack of a better estimate, the value of
dSpy used for the syn isomer (Table 2). At 298 K K6 is
small (0.31 Mꢀ1), about nine times lower than K2
(2.57 Mꢀ1). Because of the difference in the temperature
dependence this ratio falls to about 5 at 323 K but, since
the concentration of the anti rotamer is less than a tenth
that of syn, the concentration of associated species
involving the anti rotamer will never exceed 2% that of
the syn dimer. This justifies a posteriori the assumption
regarding the self-association of the syn isomer. The van’t
Hoff plot gives rough DH8 and DS8 values of
ꢀ2.7 kcal molꢀ1 and ꢀ11.6 cal molꢀ1 Kꢀ1, respectively,
rather smaller than those found for the singly hydro-
gen-bonded species obtained by hetero-association of
encumbered alcohols with pyridine.16
Substituent effects on self-association. Steric
effects on self-association are generally small.16 In the
context of the present work it was of interest to examine
the effect of replacing the bulky tert-butyl groups in 2a by
isopropyl, ethyl, or methyl to give alcohols 2b–d
(R ¼ i-Pr, Et, or Me). Hexadeuterioacetone was used in
preference to unlabeled acetone so as to reduce the extent
of hydrogen abstraction in the reaction with
3-pyridyl-EDOT-lithium and, thus, to improve the alcohol
yield. The labeled methyl group is slightly smaller than
the unlabeled one.17 These alcohols have low rotation
Reaction enthalpies and entropies. In all cases, for
both hetero- and self-association, the free energy
difference, DG8 (¼ ꢀRT ꢅlnK), varies linearly with the
temperature. For self-association of the EDOT derivative,
2a, the reaction enthalpy, DH8, and the reaction entropy,
DS8, are ꢀ8.7 kcal molꢀ1 and ꢀ26 cal molꢀ1 Kꢀ1
,
respectively (Table 5). As expected, the enthalpy term
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
417
Table 4. Hetero- and self-association constants for {2-[3,4-alkylenedioxy-5-(X-pyridyl)]thienyl}alkanols, 1, 2a–d, and 3 in
benzene at 298 K
a
dMpy (ppm)
b
Cpd.
Ka/Mꢀ1
dMb/ppm
K2 /Mꢀ1
1
6.548 ꢃ 0.045
6.402 ꢃ 0.028
6.496 ꢃ 0.021
6.943 ꢃ 0.010
7.518 ꢃ 0.025
6.249 ꢃ 0.080
1.10 ꢃ 0.03
0.59 ꢃ 0.01
0.83 ꢃ 0.01
1.27 ꢃ 0.01
1.72 ꢃ 0.01
0.46 ꢃ 0.03
2.064 ꢃ 0.007
3.498 ꢃ 0.007
2.453 ꢃ 0.002
2.082 ꢃ 0.002
2.063 ꢃ 0.009
3.895 ꢃ 0.002
2.14 ꢃ 0.03
1.10 ꢃ 0.01
0.76 ꢃ 0.01
1.70 ꢃ 0.01
2.90 ꢃ 0.04
ꢀ0.01 ꢃ 0.01
2a
2b
2c
2d
3
a Hetero-association with pyridine.
b Self-association, from concentration dependence of OH proton shift, taking dMpy as dD.
Table 5. Reaction enhalpies and entropies for hetero- and self-association of {2-[3,4-alkylenedioxy-5-(3-pyridyl)]thienyl}alkanols,
1 and 2a–d
Hetero-association
Self-association
Cpd.
ꢀDH8/kcal molꢀ1
ꢀDS8/cal molꢀ1 Kꢀ1
ꢀDH8/kcal molꢀ1
ꢀDS8/cal molꢀ1 Kꢀ1
1a
4.18 ꢃ 0.14
4.18 ꢃ 0.14
—
4.22 ꢃ 0.15
5.02 ꢃ 0.21
4.21 ꢃ 0.09
4.30 ꢃ 0.15
4.61 ꢃ 0.07
13.8 ꢃ 0.5
13.8 ꢃ 0.5
—
14.1 ꢃ 0.5
17.9 ꢃ 0.7
14.5 ꢃ 0.3
13.9 ꢃ 0.5
14.4 ꢃ 0.2
7.37 ꢃ 0.10
7.35 ꢃ 0.10
2.69 ꢃ 0.33
8.66 ꢃ 0.09
10.42 ꢃ 0.42
5.30 ꢃ 0.05
6.11 ꢃ 0.11
7.05 ꢃ 0.11
22.9 ꢃ 0.3
23.2 ꢃ 0.3
11.6 ꢃ 1.1
26.4 ꢃ 0.3
34.8 ꢃ 1.3
18.3 ꢃ 0.2
19.5 ꢃ 0.3
21.6 ꢃ 0.4
1b
1c
2aa
2ab
2bb
2cb
2db
a In terms of [S]o.
b In terms of [ROH]o.
c Association of anti isomer.
is favorable, corresponding to the formation of two
hydrogen bonds, but this is to a large extent compensated
by the entropy term which reflects the difficulty of
bringing together two molecules into a single unit. For
comparison, corresponding DH8 and DS8 values for the
hetero-association of 2a with pyridine are very close to
half of those for self-association (see also Ref. 16). For
both the hetero- [Equation (3)] and self-association
[Equation (6)] of 2a, replacing the treatment of the syn
OH proton shift in terms of [S]o by that of the average OH
proton shift in terms of [ROH]o results in greater DH8 and
DS8 values which, nevertheless, still follow the same
ratio. Given that hetero- and self-association involve one
and two hydrogen bond(s), respectively, this appears to
show that enthalpy and even entropy terms are additive:
two AꢁꢁꢁD pairs contribute twice as much as one, and
cooperativity is negligible since the pairs are effectively
isolated by the intervening structure. However, further
inspection of the data suggests that this result may be
fortuitous.
values are surprisingly slightly smaller than for 2a when
calculated on the basis of [ROH]o rather than [S]o.
However, the ‘[ROH]o’ data for 2a are considerably less
reliable than the ‘[S]o’ set, and the exceptionally high DS8
value may be an artefact. Reaction enthalpies and
entropies for the less encumbered alcohols, 2b–d, are
much smaller than for 1 and 2a, increasing in absolute
magnitude as substituent size decreases, and bearing no
simple relationship to the hetero-association values. This
suggests that entropic factors, related to conformational
changes involved in self-association or solvation phenom-
ena, are very different in this set, and that they are
correlated with changes in the reaction enthalpy. For 1
and 2a–d, DH8 is a linear function of DS8 with an
isokinetic temperature of 292 ꢃ 26 K. Early IR spectro-
scopic work by Findlay and Kidman18 on the hetero-
association of the two propanols and two butanols with
pyridine in carbon tetrachloride shows comparable
variations in reaction enthalpy (ꢀ4.1 to ꢀ6.1 kcal molꢀ1
)
)
1) and reaction entropy (ꢀ11.9 to ꢀ18.9 cal molꢀ1 Kꢀ1
Values of DH8 and DS8 for the MDOT derivative, 1, are
somewhat less than double the values for hetero-
association. For the hetero-association of 2b–d with
pyridine the reaction enthalpy varies from ꢀ4.2 to
ꢀ4.6 kcal molꢀ1, apparently increasing as the steric
demand of the substituent falls (Table 5). All these
and a similar DH8ꢀDS8 correlation. The DH8 values and
the DH8–DS8 correlations showed the same scatter in the
IR and NMR spectroscopic data of simple alcohols cited
19
by Kluk et al. It is, therefore, not clear whether such
compensation is significant at the molecular level or is the
result of statistical artefacts and/or experimental error.20
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

418
J. S. LOMAS ET AL.
Quantum mechanical DFT/ONIOM calculations
deviations in bond lengths, bond angles, and torsion
˚
angles for this selection are 0.038 A, 1.48, and 128,
Quantum mechanical calculations on large molecules can
be facilitated by treating different parts of the molecule at
different levels. In the present case, the entire molecule
was treated by DFT but a more extensive basic set was
used for the interacting COH and pyridine systems
than for the rest of the molecule (ONIOM).21 Preliminary
tests were successfully run on monomer species of
{2-[3,4-ethylenedioxy-5-(3-pyridyl)]thienyl}methanol,
3-py-EDOT-CH2OH, 2e, to check that this approximation
has no serious effects on the relative energies of the anti
and syn rotamers. Full ab initio calculations at the
B3LYP/6-311þG(d) level on 2e for the four conformers,
taking into account the orientations of both the pyridyl
ring and the CH2OH group, give values ranging over
0.3 kcal molꢀ1. The ONIOM values for 2e at the B3LYP/
6-311þG(d):B3LYP/3-21G level differ from the ab initio
values by no more than 0.2 kcal molꢀ1 (Supplementary
Material Table S8). For 2a, conformations with the
pyridyl ring in the anti orientation are favored by about
1 kcal molꢀ1, whereas the opposite is true for 2d (for
simplicity calculations were run with R ¼ CH3). For 2a
the anti-syn and anti-anti conformations differ by less
than 0.2 kcal molꢀ1, which is in good agreement with the
equilibrium constant of close to unity (K1 in Table 2)
found in a non-hydrogen-bonding solvent, such as
benzene or chloroform.
The dimers of 2a and 2d give, after correction for the
basis set superposition error,22 interaction energies of
ꢀ9.4 and ꢀ10.6 kcal molꢀ1, respectively (Supplementary
Material Table S9). The former is of the same order of
magnitude as the experimental reaction enthalpy,
ꢀ10.4 kcal molꢀ1. That the value for the dimethyl
compound, 2d, should be greater is intuitively appealing,
but unfortunately conflicts with the experimental data,
which appear to indicate that the small enhancement of
association in 2d as compared to 2a is due to a much more
favorable entropy factor. A possible explanation is that
the less encumbered monomer is more solvated, even in
benzene, and that dimer formation is accompanied by
release of solvent molecules, resulting in smaller reaction
enhalpy and entropy terms. Nevertheless, it is not clear
how 2b and 2c fit into this scheme, since the substituents
are bigger than in 2d but the reaction enhalpy and entropy
terms even smaller.
respectively. The bond length deviation is to a large extent
due to overestimation of the C—S and C—O distances in
the calculations, no doubt due to the use of a small basis
set for the relevant parts of the molecule. Corresponding
rms values from a molecular mechanics calculation
(MMFF94 in Sybyl 7.0 from Tripos, St. Louis, MO,
USA)23 are 0.016 A, 1.88, and 108, respectively, but the
˚
˚
NꢁꢁꢁO distance is seriously overestimated at 3.12 A
(details not given). Interaction energies are similar to
those found by the ONIOM approach: 10.5 and
11.4 kcal molꢀ1 for 2a and 2d, respectively.
CONCLUSION
In the solid state 2,2,4,4-tetramethyl-3-{2-[3,4-alkyl-
enedioxy-5-(3-pyridyl)]thienyl}pentan-3-ols, 1 and 2a,
occur as the syn rotamer and are predominantly or
completely associated, as demonstrated by IR spec-
troscopy and a single-crystal X-ray diffraction study of
2a. In solution anti and syn rotamers are in equilibrium,
and in benzene the syn rotamers self-associate with
relatively small equilibrium constants for the formation of
cyclic dimers. The concentration dependence of the syn/
anti ratio in the case of the EDOT derivative confirms this
model. For these alcohols the rotation barrier about the
C—C(t-Bu)2OH bond is so low that the syn!anti
conversion accompanying dissolution is imperceptible.
We shall report in a subsequent paper on slightly more
sterically hindered analogs, where the rate of this
conversion can be measured directly.
In one isolated case, 2a, the reaction enthalpy and
entropy for self-association (involving two hydrogen
bonds) are close to twice the values for hetero-association
(involving only one hydrogen bond) of the same alcohol
with pyridine, but such a simple relationship does not
exist for the other alcohols, possibly due to solvation
effects. In a series comprising tert-butyl, isopropyl, ethyl,
and deuteriomethyl substituents, 2a–2d, self-association
constants vary slightly more than hetero-association
constants, but the total range is very small in both cases.
There is nothing which can be attributed to a cooperative
effect.
The geometry of the dimer of syn 2a, as determined by
ONIOM calculations (Supplementary Material Figure
S3), despite the difference in phase, is not very different
from that determined by X-ray crystallography. The
distance between the oxygen and nitrogen atoms is
EXPERIMENTAL
Routine 1H and 13C NMR spectra of all new compounds
were determined in chloroform-d at 298 K on a Bruker
AC 200 spectrometer with a spectral resolution of 0.001
ppm/point. Chemical shifts (in ppm) are referenced to
TMS at 0.000 ppm (1H) or to solvent peak at 77.0 ppm
(13C). Wherever the NMR data are not fully assigned, this
is because signals overlap. The pyridine 1H NMR signals
were assigned and coupling constants (J in Hz, not
˚
˚
2.82 A, as against the measured 2.86 A, and the pyridine
ring is at 108 to the thiophene ring, as against 208 in the
crystal. A more general comparison of selected internal
coordinates from the gas-phase calculations and the
solid-state X-ray data is given in the Supplementary
Material Table S1. Overall the root mean square
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
419
signed) determined by spectrum simulation using gNMR
(version 4.1, Adept Scientific, Letchworth, UK). Most
1H NMR shifts in alcohols, particularly those of the OH
protons, are concentration-dependent. The corresponding
13C signals were assigned by XHCORR experiments;
Cq’s were identified by comparison with spectra of
3,4-alkylenedioxythiophenes,7 pyridylthiophenes,24 and
2,2,4,4-tetramethyl-3-(3,4-alkylenedioxythienyl)pentan-
3-ols,7 it being assumed that substituent effects are
additive. IR transmission spectra were recorded on a
Nicolet Magna 860 FTIR spectrometer with 4 cmꢀ1
resolution (Table 1). Drops of saturated solutions of
alcohols 1, 2a, and 3 in dichloromethane were allowed to
evaporate on KBr plates and the spectra recorded at
298 K. Solutions of 1 and 2a–d in carbon tetrachloride
were examined in 1 mm glass cells. Details on how the 2D
NOESY NMR spectra were recorded and the association
constants measured are given in the Supplementary
Material.
7.26 (H5-py, J 0.9, 4.8, 8.1), 7.97 (H4-py, J 1.6, 2.4, 8.1),
8.43 (H6-py, J 0.3, 1.6, 4.8), and 8.94 (H2-py, J 0.3, 0.9,
2.4).
2-(2-Pyridyl)-3,4-ethylenedioxythiophene, 6. Yield
25%; mp 79 8C (Found: C, 60.0; H, 4.1. C11H9NO2S
requires C, 60.26; H, 4.14%). dC 64.3 (CH2), 64.9 (CH2),
101.2 (C5-th), 118.6 (C2-th), 120.2 (C3-py), 120.9
(C5-py), 136.3 (C4-py), 139.9 (C3-th), 142.0 (C4-th),
149.1 (C6-py), and 151.5 (C2-py); dH 4.24 (CH2, J 2.2,
6.1, 11.6), 4.35 (CH2, J 2.2, 6.1, 11.6), 6.42 (s, H5-th),
7.05 (H5-py, J 1.1, 4.9, 7.5), 7.64 (H4-py, J 1.9, 7.5, 8.1),
7.92 (H3-py, J 1.0, 1.1, 8.1), and 8.51 (H6-py, J 1.0, 1.9,
4.9).
Synthesis of 2,2,4,4-tetramethyl-3-{2-[5-(X-
pyridyl)]thienyl}pentan-3-ols
To a mixture of the appropriate 2- or 3-pyridylthiophene
(x mmol, x generally 2–5) and TMEDA (x mmol) in
diethyl ether (10–15 ml) under argon at room temperature
was added a solution of n-butyl-lithium in hexane (1.6 M,
x mmol). After 30 min stirring, 2,2,4,4-tetramethylpen-
tan-3-one (x 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 gave a residue from which the alcohol was
isolated by chromatography on alumina or by trituration
in light petroleum-diethyl ether mixtures. Less encum-
bered alcohols, 2b–d, were synthesized by the same
procedure except that n-BuLi and ketone were added
successively at ꢀ75 8C, after which the reaction mixture
was allowed to warm slowly to room temperature before
work-up.
Synthesis of pyridylthiophenes, 4–6
2-(Tri-n-butylstannyl)thiophenes were prepared by reac-
tion of tri-n-butyltin chloride with the organolithium
reagent prepared by reaction of the parent thiophene with
n-butyl-lithium/TMEDA in diethyl ether under argon at
room temperature, followed by aqueous work-up.
Treatment of the appropriate 2-(tri-n-butylstannyl)
thiophene with 2- or 3-bromopyridine in DMF at 115–
1258C under argon for 2–5 h in the presence of
bis(triphenylphosphine)palladium(II) chloride gave 2-
or 3-pyridylthiophene.11 The crude product was first
extracted into 1 N aqueous HCl, the aqueous layer rinsed
with diethyl ether to remove neutrals, and the bases
regenerated by treatment with 1 N NaOH, followed by
diethyl ether extraction. The residue after solvent
evaporation was purified by chromatography on alumina
in light petroleum-diethyl ether mixtures.
2,2,4,4-Tetramethyl-3-{2-[3,4-methylenediox-
y-5-(3-pyridyl)]thienyl}pentan-3-ol, 1. Yield 25%;
mp 178 8C (dec.). (Found: C, 65.9; H, 7.3; N,
3.9. C19H25NO3S requires C, 65.68; H, 7.25; N,
4.03%). syn: dC 28.9 (CH3), 42.7 (Cq), 85.0 (COH),
104.5 (C5-th), 107.3 (CH2), 121.3 (C2-th), 123.4 (C5-py),
128.5 (C3-py), 131.7 (C4-py), 140.3 (C3-th), 144.2
(C4-th), 146.1 (C2-py), and 146.8 (C6-py); dH 1.17 (s,
6 CH3), 2.71 (s, OH), 6.20 (s, CH2), 7.23 (H5, J 0.8, 4.9,
8.0), 7.79 (H4, J 1.7, 2.4, 8.0), 8.38 (H6, J 0.7, 1.7, 4.9),
and 8.82 (H2, J 0.7, 0.8, 2.4).
2-(3-Pyridyl)-3,4-methylenedioxythiophene, 4.
Yield 45%; mp 119 8C (Found: C, 58.5; H,
3.4. C10H7NO2S requires C, 58.52; H, 3.44%). dC 91.5
(C5-th), 106.8 (C2-th), 109.4 (CH2), 123.6 (C5-py), 128.3
(C3-py), 132.1 (C4-py), 145.5 (C3-th), 146.0 (C2-py),
147.3 (C6-py), and 148.3 (C4-th); dH 6.07 (s, H5-th), 6.32
(s, CH2), 7.27 (H5-py, J 0.9, 4.8, 8.0), 7.80 (H4-py, J 1.6,
2.4, 8.0), 8.45 (H6-py, J 0.3, 4.8), and 8.82 (H2-py, J 0.3,
0.9, 2.4).
2,2,4,4-Tetramethyl-3-{2-[3,4-ethylenedioxy-5-(3-
pyridyl)]thienyl}pentan-3-ol, 2a. Yield 42%; mp
176 8C (dec.). (Found: C, 66.4; H, 7.6; N,
3.9. C20H27NO3S requires C, 66.45; H, 7.53; N, 3.87).
anti: dC 29.1 (CH3), 42.9 (Cq), 63.6 (CH2), 64.6 (CH2),
87.0 (COH), 110.0 or 110.2 (C5-th), 121.5 (C2-th), 123.3
(C5-py), 129.0 or 129.7 (C3-py), 132.5 (C4-py), 137.3
(C4-th), 139.0 (C3-th), 146.9 (C2-py), and 147.1 (C6-py);
2-(3-Pyridyl)-3,4-ethylenedioxythiophene, 5. Yield
61%; mp 63 8C (Found: C, 60.3; H, 4.2. C11H9NO2S
requires C, 60.26; H, 4.14%). dC 64.0 (CH2), 64.4 (CH2),
98.5 (C5-th), 113.1 (C2-th), 123.0 (C5-py), 129.1
(C3-py), 132.3 (C4-py), 138.9 (C3-th), 141.9 (C4-th),
146.4 (C2-py), and 146.8 (C6-py); dH 4.25 (CH2, J 2.2,
6.2, 13.2), 4.32 (CH2, J 2.2, 6.2, 13.2), 6.36 (s, H5-th),
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

420
J. S. LOMAS ET AL.
dH 1.18 (s, 6 CH3), 4.30 (m, 2 CH2), 4.76 (s, OH), 7.25
(H5, J 0.9, 4.8, 8.1), 7.95 (H4, J 1.6, 2.4, 8.1), 8.41 (H6, J
0.2, 1.6, 4.8), and 8.95 (H2, J 0.2, 0.9, 2.4). syn: dC 29.1
(CH3), 42.6 (Cq), 64.2 (CH2), 64.3 (CH2), 86.0 (COH),
110.0 or 110.2 (C5-th), 123.2 (C5-py), 128.2 (C2-th),
129.0 or 129.7 (C3-py), 132.4 (C4-py), 135.1 (C4-th),
138.7 (C3-th), 146.6 (C6-py), and 146.9 (C2-py); dH 1.17
(s, 6 CH3), 2.50 (s, OH), 4.23 (m, 2 CH2), 7.22 (H5, J 0.9,
4.8, 8.1), 7.97 (H4, J 1.6, 2.4, 8.1), 8.37 (H6, J 0.2, 1.6,
4.8), and 8.97 (H2, J 0.2, 0.9, 2.4).
149.2 (C6-py); syn: dC 29.1 (CH3), 42.5 (Cq), 63.6 (CH2),
64.4 (CH2), 86.0 (COH), 120.4 (C3-py), 120.4 (C5-py),
136.2 (C4-py), and 149.2 (C6-py). Cq’s at 114.9, 115.2,
124.1, 130.8, 134.9, 138.1, 138.8, 139.4, 151.4, and 152.0
not assigned; anti and syn: dH 1.17 (syn), 1.19 (anti) (s,
6 CH3), 2.25 (s, 0.4H, syn-OH), 4.2–4.4 (m, CH2), 4.79 (s,
0.6H, anti-OH), 7.01–7.08 (m, H5), 7.60–7.68 (m, H4),
7.87–7.94 (m, H3), and 8.49–8.53 (m, H6).
X-Ray crystallography: syn-2,2,4,4-
Tetramethyl-3-{2-[3,4-ethylenedioxy-
5-(3-pyridyl)]thienyl}pentan-
3-ol, 2a, C20H27NO3S
2,4-Dimethyl-3-{2-[3,4-ethylenedioxy-5-(3-pyrid-
yl)]thienyl}pentan-3-ol, 2b. Yield 41%; mp 170 8C.
(Found: C, 65.1; H, 7.0; N, 4.1. C18H23NO3S requires C,
64.84; H, 6.95; N, 4.20%). dC 16.6 (CH3), 17.8 (CH3),
34.5 (CH), 64.2 (CH2), 64.5 (CH2), 82.3 (COH), 110.6
(C5-th), 121.8 (C2-th), 123.3 (C5-py), 129.5 (C3-py),
132.4 (C4-py), 136.8 (C3-th), 138.6 (C4-th), and 146.8
(C2-py and C6-py); dH 0.92 (2 CH3, J 6.8), 0.96 (2 CH3, J
6.8), 2.36 (2 CH, J 6.8), 2.75 (br s, OH), 4.21 (CH2, J 2.2,
6.1, 11.3), 4.30 (CH2, J 2.2, 6.1, 11.3), 7.23 (H5, J 0.9,
4.8, 8.0), 7.95 (H4, J 1.6, 2.4, 8.0), 8.38 (H6, J 1.6, 4.8),
and 8.94 (H2, J 0.9, 2.4).
˚
Crystal data. M ¼ 361.51. Triclinic a ¼ 8.074(1) A,
˚
˚
b ¼ 8.694(1) A, c ¼ 13.748(2) A. a ¼ 80.73(1), b ¼
3
˚
87.50(1), g ¼ 79.34(1), V ¼ 935.9(2) A (refined from
75 reflections, 4 < u< 248), space group P-1, Z¼ 2,
Dx ¼ 1.28 g cmꢀ3. Colorless prismatic crystals, m(Mo Ka)¼
0.191 mmꢀ1.
Data collection and processing. At 295 K, Enraf-
Nonius Kappa CCD diffractometer, v mode, graphite-
˚
monochromated Mo Ka radiation (l ¼ 0.71073 A). 7706
reflections measured (2.4 ꢆ u ꢆ 27.58), 4089 unique,
giving 2614 with I > 3s(I).
3-{2-[3,4-Ethylenedioxy-5-(3-pyridyl)]thienyl}pe-
ntan-3-ol, 2c. Yield 9%; mp 141 8C. (Found: C, 63.0;
H, 6.5; N, 4.6. C16H19NO3S requires C, 62.93; H, 6.27; N,
4.59%). dC 8.0 (CH3), 33.6 (CH2), 64.3 (CH2), 64.6
(CH2), 77.2 (COH), 110.1 (C5-th), 123.3 (C5-py), 123.7
(C2-th), 129.6 (C3-py), 132.4 (C4-py), 136.1 (C3-th),
139.0 (C4-th), and 146.7 (C2-py and C6-py); dH 0.92
(2 CH3, J 7.4), 1.83 (2 CH, J 7.4, 13.9), 1.97 (2 CH, J 7.4,
13.9), 2.87 (br s, OH), 4.23 (CH2, J 2.2, 6.1, 10.9), 4.30
(CH2, J 2.2, 6.1, 10.9), 7.23 (H5, J 0.9, 4.8, 8.0), 7.95 (H4,
J 1.6, 2.4, 8.0), 8.38 (H6, J 1.6, 4.8), and 8.93 (H2, J 0.9,
2.4).
Structure analysis and refinement. Full-matrix
least-squares refinement with all non-hydrogen atoms
anisotropic; hydrogen H11 located from Fourier differ-
ence map and not refined; all other hydrogens placed in
geometrically calculated positions; only one overall
isotropic thermal parameter refined (227 refinable
parameters). Absorption corrections were determined
using DIFABS.25 Final R and Rw (Chebyshev series)
values are 0.056 and 0.64. Programs used were the PC
version of CRYSTALS26 for refinements and
CAMERON27 for views. Selected bond lengths, bond
angles, and torsion angles are listed in Supplementary
Material Table S1.
2-{2-[3,4-Ethylenedioxy-5-(3-pyridyl)]thienyl}pro-
pan-2-ol-d6, 2d. Yield 20%; mp 148 8C. (Found: C,
59.2; N, 4.9; S, 11.5. C14H9D6NO3S requires C, 59.34; N,
4.94; S, 11.31%). dC 64.4 (CH2), 64.6 (CH2), 70.9 (COH),
109.6 (C5-th), 123.3 (C5-py), 125.5 (C2-th), 129.5
(C3-py), 132.6 (C4-py), 136.4 (C3-th), 139.2 (C4-th),
and 146.9 (C2-py and C6-py); dH 2.80 (br s OH), 4.28
(CH2, J 2.2, 6.1, 11.5), 4.31 (CH2, J 2.2, 6.1, 11.5), 7.24
(H5, J 0.9, 4.8, 8.1), 7.94 (H4, J 1.6, 2.4, 8.1), 8.40 (H6, J
1.6, 4.8), and 8.921 (H2, J 0.9, 2.4).
Atomic coordinates, bond lengths and angles, and
thermal parameters have been deposited at the Cambridge
Crystallographic Data Centre (CCDC 289515). Any
request to the CCDC for this material should quote the
full literature citation and the reference number.
2,2,4,4-Tetramethyl-3-{2-[3,4-ethylenedioxy-5-(2-
pyridyl)]thienyl}pentan-3-ol, 3. Yield 42%; mp
138 8C (dec.). (Found: C, 66.3; H, 7.6; N,
3.8. C20H27NO3S requires C, 66.45; H, 7.53; N, 3.87%).
anti: dC 29.2 (CH3), 42.9 (Cq), 64.4 (CH2), 64.5 (CH2), 87.2
(COH), 120.6 (C3-py), 120.8 (C5-py), 136.3 (C4-py), and
Quantum mechanical DFT/ONIOM calculations
Details of this approach along with data (Tables S1, S8,
and S9), structure diagrams for the dimers of 2a and 2e
(Figure S3), and full references are given in the
Supplementary Material.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc

SELF-ASSOCIATION IN (3-PYRIDYL)THIENYLALKANOLS
421
Acknowledgements
7. Lomas JS, Adenier A, Gao K, Maurel F, Vaissermann J. J. Chem.
Soc., Perkin Trans. 2 2002; 216–224.
8. Lomas JS. J. Chem. Soc., Perkin Trans.2 2001; 754–757.
9. Lomas JS, Cordier C. J. Phys. Org. Chem. 2003; 16: 361–368.
10. Lomas JS, Vaissermann J. J. Phys. Org. Chem. 2004; 17: 102–112.
11. (a) Stille JK. Angew. Chem. Int. Ed. 1986; 25: 508–524; (b)
McKean DR, Parrinello G, Renaldo AF, Stille JK. J. Org. Chem.
1987; 52: 422–424.
We are indebted to P. Herson (Univ. Paris 6) for help with
the crystallographic data, and to Dr A. Yassar (Univ. Paris
7) for advice on the synthesis of pyridylthiophenes. We
thank a referee for suggesting the ONIOM calculations.
12. (a) Larson SB, Simonsen SH, Martin GE, Puig-Torres S, Smith K.
Acta Crystallogr., Sect. C 1986; 42: 766–768; (b) Walton AR,
Jenkins TC. Acta Crystallogr., Sect. C 1990; 46: 614–617.
13. Chen JS, Shirts RB. J. Phys. Chem. 1985; 89: 1643–1646.
14. Fielding L. Tetrahedron 2000; 56: 6151–6170.
15. (a) Pineau P, Fuson N, Josien ML. J. Chim. Phys. 1958; 55:
464–469; (b) Johnston MD, Gasparro FP, Kuntz ID. J. Am. Chem.
Soc. 1969; 91: 5715–5724.
16. Lomas JS. J. Phys. Org. Chem. 2005; 18: 1001–1012.
17. Carter RE, Melander L. Adv. Phys. Org. Chem. 1973; 10: 1–27.
18. Findlay TJV, Kidman AD. Aust. J. Chem. 1965; 18: 521–530.
´
19. Kluk H, Skawinska G, Kluk E. Acta Phys. Pol. A 1976; 50:
647–659.
REFERENCES
1. (a) Pimental GC, McClellan AL. The Hydrogen Bond. Freeman:
San Francisco, 1960; (b) Vinogradov SN, Linell RH. Hydrogen
Bonding. Van Nostrand: New York, 1971; (c) Joesten MD, Schaad
LJ. Hydrogen Bonding. M. Dekker: New York, 1974; (d) Jeffrey
GA, Saenger W. Hydrogen Bonding in Biological Structures.
Springer: Berlin, 1991; (e) Jeffrey GA. An Introduction to Hydro-
gen Bonding. Oxford University Press: New York, 1999; (f)
Desiraju GR, Steiner T. The Weak Hydrogen Bond in Structural
Chemistry and Biology. Oxford University Press: New York, 1999;
(g) Lehn JM. Supramolecular Chemistry: Concepts and Perspect-
ives. VCH: Weinheim, 1995; (h) Steed JW, Atwood JL. Supra-
molecular Chemistry. Wiley: Chichester, 2000; (i) Schnieder HJ,
Yatsimirski AK. Principles and Methods in Supramolecular
Chemistry. Wiley: New York, 2000; (j) Reinhoudt DN, Crego-
Calama M. Science 2002; 295: 2403–2407; (k) Ikkala O, ten
Brinke G. Science 2002; 295: 2407–2409; (l) Kato T. Science
2002; 295: 2414–2417; (m) Whitesides GM, Grzybowski B.
Science 2002; 295: 2418–2421.
2. (a) Schmuck C, Wienand W. Angew. Chem. Int. Ed. 2001; 40:
4363–4369; (b) Archer EA, Sochia AE, Krische MJ. Chem. Eur. J.
2001; 7: 2059–2065; (c) Stubbs LP, Weck M. Chem. Eur. J. 2003;
9: 992–999; (d) Lafitte VGH, Aliev AE, Horton PN, Hursthouse
MB, Bala K, Golding P, Hailes HC. J. Am. Chem. Soc. 2006; 128:
6544–6545.
3. Fayereisen MW, Feller D, Dixon DA. J. Phys. Chem. 1996; 100:
2993–2997.
4. Brown JN, Jenevein RM, Stocker JH, Trefonas LM. J. Org. Chem.
1972; 37: 3712–3718.
20. (a) Exner O. Prog. Phys. Org. Chem. 1973; 10: 411–482; (b) Ford
DM. J. Am. Chem. Soc. 2005; 127: 16167–16170.
21. (a) Dapprich S, Komaromi I, Byun KS, Morokuma K, Frisch MJ.
J. Mol. Struct. Theochem. 1999; 461-462: 1–21; (b) Vreven T,
Morokuma K, Farkas O, Schlegel HB, Frisch MJ. J. Comput.
Chem. 2003; 24: 760–769; (c) Vreven T, Morokuma K. J. Comput.
Chem. 2000; 21: 1419–1432; (d) Vreven T, Frisch MJ, Kudin KN,
Schlegel HB, Morokuma K. Mol. Phys. 2006; 104: 701–714; (e)
Vreven T, Byun KS, Komaromi I, Dapprich S, Montgomery JA,
Morokuma K, Frisch MJ. J. Chem. Theory Comput. 2006; 2:
815–826.
22. Boys SF, Bernardi F. Mol. Phys. 1970; 19: 553–566.
23. (a) Halgren TA. J. Comput. Chem. 1996; 17: 490–519, 520–552,
553–586, 616–641; (b) Halgren TA, Nachbar RB. J. Comput.
Chem. 1996; 17: 587–615; (c) Halgren TA. J. Comput. Chem.
1999; 20: 720–729, 730–748.
24. Jones RA, Civcir PU. Tetrahedron 1997; 53: 11529–11540.
25. Walker N, Stuart D. Acta Crystallogr., Sect. A 1983; 39: 158–166.
26. Watkin DJ, Prout CK, Carruthers JR, Bettridge PW, Cooper RI.
CRYSTALS Issue 11, 2001; Chemical Crystallography Laboratory,
Oxford, UK.
5. Kuhn LP, Wires RA, Ruoff W, Kwart H. J. Am. Chem. Soc. 1969;
91: 4790–4793.
6. Lomas JS, Adenier A, Cordier C. J. Phys. Org. Chem. 2006; 19:
295–307.
27. Watkin DJ, Prout CK, Carruthers JR, Pearce LJ. CAMERON, 1996;
Chemical Crystallography Laboratory, Oxford, UK.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 410–421
DOI: 10.1002/poc