The through-bond interaction of a sulfur lone pair with oxygenated substituents in the thiacyclohexane framework

Article abstract of DOI:10.1071/CH05073

Low-temperature X-ray crystal structures were determined on a range of derivatives of 4-thiacyclohexanol 5a of varying electron demand with a view to finding evidence for a through-bond interaction between the sulfur lone pair and the oxygenated substituent. In contrast to earlier suggestions, plots of C-OR bond distance versus pK_a (ROH) showed that any interaction between the sulfur and the OR group is unlikely to be of a through-bond origin. Furthermore, unimolecular solvolysis rate measurements on the nosylate ester derivative 5g showed that the sulfur actually retards the reaction slightly in comparison with the corresponding sulfur-free analogue 6. CSIRO 2005.

Full text of DOI:10.1071/CH05073

Full Paper
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2005, 58, 531–534
The Through-Bond Interaction of a Sulfur Lone Pair with Oxygenated
Substituents in the Thiacyclohexane Framework
Laura AndrauA and Jonathan M. WhiteA,B
A School of Chemistry, University of Melbourne, Parkville VIC 3010, Australia.
B Corresponding author. Email: whitejm@unimelb.edu.au
Low-temperature X-ray crystal structures were determined on a range of derivatives of 4-thiacyclohexanol 5a of
varying electron demand with a view to finding evidence for a through-bond interaction between the sulfur lone
pair and the oxygenated substituent. In contrast to earlier suggestions, plots of C OR bond distance versus pKa
(ROH) showed that any interaction between the sulfur and the OR group is unlikely to be of a through-bond origin.
Furthermore, unimolecular solvolysis rate measurements on the nosylate ester derivative 5g showed that the sulfur
actually retards the reaction slightly in comparison with the corresponding sulfur-free analogue 6.
Manuscript received: 11 March 2005.
Final version: 21 April 2005.
O
Introduction
O
O
Interactions between donor and acceptor groups within
molecules can have a huge impact on both their ground state
and reactivity properties. The interactions can occur directly
through space, by transmission through the carbon frame-
work (through bond), or sometimes by a combination of the
two.[1,2] For example, a through-space interaction between
the nitrogen lone-pair and the transannular carbonyl group in
clivorine[3] 1(Scheme1)resultsinaclosecontactbetweenthe
amino nitrogen and the carbonyl carbon. On the other hand, a
through-bond interaction between the nitrogen lone pair and
the developing carbocation orbital is believed to be respon-
sible for the hugely different reactivities of the bicyclic chlo-
rides 2 and 3 (Scheme 2) towards unimolecular solvolysis.[4]
Whichever of the above type of interaction occurs, a strong
dependence of the magnitude of the interaction upon the
relative orientation of the donor and acceptor fragments is
observed. The results of these interactions are often referred
to as stereoelectronic effects, a description which reflects
their electronic basis and recognizes their dependence upon
stereochemistry.
O
O
O
N
O
1
Scheme 1.
N
N
Cl
Cl
2
3
Scheme 2.
O
S
NH
NH
O
NH2
4
Scheme 3.
The present study was prompted by a theoretical investi-
gation into the mechanism of the carboxyl transfer reaction,
which in living systems is facilitated by the coenzyme biotin
4 (Scheme 3).[5]
It was suggested that a through-bond interaction develops
between the sulfur p-type lone pair and the ureidyl moiety
when the biotin molecule is twisted towards a transition-state
geometry.[6] This interaction, which results in donation of
electron density from the sulfur atom towards the uriedyl
moiety would be expected to increase the nucleophilicity of
the urea nitrogen, and therefore, improve the efficiency of the
carboxyl transfer step.
As part of our general interest in the structural effects of
donor–acceptor interactions,[7] particularly as they manifest
themselves in the cyclohexane framework, we chose to inves-
tigate whether a through-bond interaction between the sulfur
lone pair and the σC∗–O orbital in ester and ether derivatives of
thiocyclohexanol 5 (Fig. 1) leads to any observable effects on
the ground state structures or reactivity of these derivatives.
Such an interaction might be expected to result in the
lengthening of the C OR bond distance in comparison to
similar model compounds that lack the interaction and also
lead to increased rates of unimolecular solvolysis.A powerful
© CSIRO 2005
10.1071/CH05073
0004-9425/05/070531

532
L. Andrau and J. M. White
OR
ϪOR
OR
S
S
S
ϩ
5a R ϭ H
5
5b R ϭ p-NO2C6H4
5c R ϭ m-NO2C6H4CO
5d R ϭ p-NO2C6H4CO
5e R ϭ o-NO2C6H4CO
5f R ϭ 2,4-NO2C6H3CO
5g R ϭ p-NO2C6H4SO2
Fig. 1. The through-bond nS–σC∗–O interaction.
SiMe3
O
OR
But
OR
RЈ
OR
Scheme 5.
6
7
8
could not be used in this study because they do not meet the
stereoelectronic requirements for through-bond interaction
with the sulfur lone pair.
The X-ray structures of 5a, 5b, and 5e–g were determined
at 130 K to minimize the unwanted effects of thermal motion.
Selected bond distances are presented inTable 1 and a thermal
ellipsoid plot for one of the structures (5g) is presented in
Fig. 2.
ExaminationofthedistancesinTable1showsthat(a)there
is excellent agreement between those bonds which are related
by the approximate local plane of symmetry defined by S,
C4, and O1; and (b) there is a clear relationship between the
C1 O1 bond distance and the electron demand of the oxygen
substituent [as estimated by the pKa(ROH)]. In particular, the
C1 O1 bond distance increases as the electron demand of the
oxygensubstituentincreases.Thus, theC4 O1bonddistance
in the weakly electron demanding alcohol 5a is 1.431(2) Å
and increases to 1.487(2) for the strongly electron demanding
nosylate derivative 5g. This data is presented graphically in
Fig. 3 and gives rise to the following relationship (Eqn 4):
Scheme 4.
X-ray structural technique for detecting the presence of inter-
actions between donor groups and oxygenated substituents is
known as the variable oxygen probe.[8] The variable oxy-
gen probe was first introduced by Kirby and coworkers who
established that the C O bond distance in the C OR frag-
ment increases as the electron demand of the OR substituent
increases, and thereby, reflects an increasing contribution of
the C+ −OR valence bond form to the ground state structure.
If the electron demand of a substituent (OR) is quantified as
the pKa value for the parent acid (ROH), then a plot of C OR
bond distance versus pKa(ROH) will be linear, and the slope
of the resultant plot will be sensitive to the effects of electron
donation into the C OR σ∗ antibonding orbital.The presence
of good donor orbitals vicinal and antiperiplanar to the C O
bond results in a strong response of the C OR distance to
the electron demand of OR, and is reflected by the increased
stabilization of the cation part of the valance bond form
C+ −OR. For example, plots of C OR bond distance [Å]
versus pKa(ROH) constructed for 6,[9] 7,[10] and 8[11]
(Scheme 4) gave the following relationships (Eqns 1–3):
rC–O = 1.48 − 2.76 × 10−3 pKa(ROH) R2 = 0.99 (4)
6 rC–O = 1.493 − 6.49 × 10−3 pKa(ROH) R2 = 0.985 (1)
7 rC–O = 1.502 − 5.30 × 10−3 pKa(ROH) R2 = 0.986 (2)
8 rC–O = 1.48 − 2.77 × 10−3 pKa(ROH) R2 = 0.976 (3)
This relationship is essentially identical to that obtained
for the derivatives of cyclohexanol 8 (Eqn 3) above, which
suggests that the proposed through-bond interaction between
the sulfur lone pair and the oxy substituent is either not
present, or is simply too small to be detected by this method.
Stereoelectronic effects on ground states of molecules are
relatively small in comparison to the effects on transitions
states. Therefore, in order to magnify the effects of the pro-
posed through-bond interaction we determined the rate of
unimolecular solvolysis of the nosylate derivative 5g so that
this could be compared with that which has been previously
reported for the equatorial cyclohexyl nosylate derivative 6
(Scheme 6).[15] The relative reactivities of 5g and 6 would
provide a measure of the relative stabilities of the carbenium
ion intermediates 7 and 8, and hence, provide a measure of
the through-bond stabilization of 8 by the sulfur lone pair.
The solvolysis of 5g was carried out in 97%
CF3CD2OD/D2O in the presence of 2,6-lutidine at 40◦C
using the NMR method reported by Creary and Jiang.[16]
Absolute and relative rates of solvolysis of 5g and 6 are
presented in Table 2.
A strong response of C OR bond distance to the electron
demand of OR is demonstrated for 6, which has an oxygen
lone pair (nO) orbital antiperiplanar to the OR substituent
(this is the basis of the well known anomeric effect).[12]
A
strong response is also observed for 7, which has a C Si
bond antiperiplanar to the OR substituent (this is the basis
of the silicon β-effect).[13] However, a weaker response is
obvious in 8, which has a σC–C bonding orbital, and is the
result of a weaker donor orbital being situated antiperipla-
nar to the OR bond. Thus, the slope of the plot of C OR
bond distance versus the pKa(ROH) for 5a and its deriva-
tives was expected to provide a measure of the extent of the
through-bond interaction between the sulfur lone pair and the
oxygenated substituent.
Results and Discussion
Alcohol 5a was prepared from thiocyclohexanone and con-
verted into the crystalline ester and ether derivatives 5b–g
(Scheme 5) covering the pKa range +16 to −3.18. As the
m-nitrobenzoate 5c and p-nitrobenzoate 5d derivatives sur-
prisingly crystallized with the cyclohexane ring adopting a
conformation in which the ester groups were axial,[14] they
Surprisingly, nosylate 5g underwent solvolysis at a rate
of more than one order of magnitude slower than the cor-
responding sulfur cyclohexyl ester 6. Thus, the presence of
sulfur appears to destabilize the intermediate carbenium ion 8
in comparison to 7. This result suggests that there is either no

Through-Bond Interaction of a Sulfur Lone Pair
533
Table 1. Selected bond distances for compounds 5a, 5b, and 5e–g
Compound
pKa
S–C2
S–C6
C2–C3
C3–C4
C4–C5
C5–C6
C4–O1
5a; M1
M2
M3
5b
5e
5f
16
–
–
7.15
2.17
1.4
−3.8
1.792(2)
1.803(2)
1.796(2)
1.809(3)
1.809(2)
1.801(3)
1.808(2)
1.798(2)
1.799(2)
1.793(2)
1.810(3)
1.810(2)
1.800(3)
1.806(2)
1.516(3)
1.523(3)
1.500(3)
1.526(4)
1.526(2)
1.527(4)
1.525(2)
1.510(3)
1.512(3)
1.513(3)
1.513(4)
1.515(2)
1.517(4)
1.517(2)
1.504(3)
1.505(3)
1.500(3)
1.513(4)
1.516(2)
1.504(4)
1.519(2)
1.521(3)
1.526(3)
1.520(3)
1.524(4)
1.530(2)
1.520(4)
1.524(2)
1.442(3)
1.428(3)
1.423(3)
1.460(3)
1.470(2)
1.471(3)
1.487(2)
5g
ϩ
O5
ϩ
But
But
OSO2PhNO2
S
C11
6
7
8
N1
O4
C12
Scheme 6.
Table 2. Solvolysis rates for 5g and 6
C10
k [s−1
]
k(rel)
C7
C9
Compound
O3
5g
6
1.5 × 10−6
1.0
ca. 0.1
C8
O1
C6
approx. 1.9 × 10−7
S2
C4
C3
C5
O2
N
S1
Nϩ
ClϪ
Cl
C2
2
9
Fig. 2. Thermal ellipsoid plot for ester 5g. Ellipsoids are at the 20%
probability level. The numbering scheme for the thiacyclohexane ring
is employed for all structures.
Scheme 7.
CF3CD2OD/D2O
OSO2PhNO2
S
S
1.49
1.48
1.47
1.46
1.45
1.44
1.43
5g
10
Scheme 8.
carbenium ion 8 prefers to be stabilized by hyperconjugation
with the neighbouring axial hydrogens.
Conclusions
We have examined the through-bond interaction between a
sulfur lone pair and oxygenated substituents in the unstrained
thiacyclohexane framework in terms of both ground state
and reactivity effects. Our data provides evidence that biotin
is unlikely to facilitate the carboxyl transfer reaction by a
through-bond interaction.
Ϫ10
0
10
20
pK (ROH)
a
Experimental
General experimental details have been published elsewhere.[11]
Fig. 3. r(C–O) versus pKa(ROH) relationship for the derivatives 5a,
5b, and 5e–g.
General Procedure for the Preparation of Ester Derivatives 5e–g
through-bond interaction between the sulfur lone pair and the
carbenium ion p-orbital, or that perhaps it is more than com-
pensated for by an unfavourable inductive or field effect of the
sulfur atom. Evidence for through-bond participation during
solvolysis reactions can be provided by analysis of the reac-
tion products. For example, solvolysis of the bicyclic amine
2 leads exclusively to products derived from the iminium ion
9 (Scheme 7), and indicates the participation of the nitrogen
lone pair in the displacement of the chloride ion.
In contrast, solvolysis of 5g leads exclusively to 4-
thiocyclohexene 10 (Scheme 8) through a simple E1 elim-
ination. This provides no evidence that the sulfur lone pair
is involved in the reaction, but rather, that the intermediate
To a solution of tetrahydrothiopyran-4-ol (0.102 g, 0.87 mmol) in pyri-
dine (1.5 mL) was added 4-nitrobenzenesulfonyl chloride (0.222 g,
1.00 mmol). Upon completion of the reaction (formation of salts), the
mixture was quenched with a few drops of water, extracted with EtOAc
(3 × 20 mL), and washed with NaHCO3 (10 mL). The organic layer
was then evaporated to give 5g, which was crystallized from pentane,
mp 121–123◦C. δH (CDCl3) δ 8.41 (2H, d, J 9.0), 8.12 (2H, d, J 9.0),
4.75–4.69 (1H, m), 2.84–2.78 (2H, m), 2.55–2.48 (2H, m), 2.14–2.07
(2H, m), 2.04–1.96 (2H, m). δC (CDCl3) 150.6, 143.0, 128.9, 124.5,
80.8, 33.2, 25.0.
4-Thiacyclohexyl 2,4-Dinitrobenzoate 5f
From diethyl ether/pentane, mp 101–103◦C. δH (CDCl3) 8.81 (1H, d,
J 2.0), 8.54 (1H, dd, J 8.4 and 2.2), 7.95 (1H, d, J 8.2), 5.18–5.12 (1H, m),

534
L. Andrau and J. M. White
2.83–2.77 (2H, m), 2.69–2.62 (2H, m), 2.29–2.22 (2H, m), 2.05–1.97
(2H, m). δC (CDCl3) 162.9, 148.9, 148.0, 133.0, 131.3, 127.5, 119.6,
74.3, 31.9, 25.6.
79.971(4), γ86.087(4)◦, V 605.5(3) Å3, Z 2, Dc 1.466 mg M−3, µ
(MoKα) 0.273 mm−1, F(000) 280, crystal size 0.3 × 0.3 × 0.15 mm3,
5520 reflections measured, 2708 independent reflections (Rint 0.0228),
the final R was 0.0410 [I > 2σ(I)] and wR(F2) was 0.1082 (all data).
4-Thiacyclohexyl 2-Nitrobenzoate 5e
From diethyl ether/pentane, mp 93–95◦C. δH (CDCl3) 7.93−7.90 (1H,
m), 7.75 (1H, dd, J 7.5 and 1.7), 7.71–7.62 (2H, m), 5.16–5.10 (1H, m),
2.25–2.18 (2H, m), 2.04–1.96 (2H, m). δC (CDCl3) 164.6, 148.3, 132.9,
131.8, 129.9, 127.8, 123.9, 73.0, 32.0, 25.7.
Crystal Data for 5f
C12H12N2O6S, M 312.30, T 130.0(1) K, λ 0.71069, orthorhombic,
space group Pbca, a 7.4594(15), b 8.921(2), c 40.313(9) Å, V 2682 Å3,
Z 8, Dc 1.547 mg M−3, µ(MoKα) 0.272 mm−1, F(000) 1296, crystal
size 0.3 × 0.2 × 0.2 mm3, 17813 reflections measured, 2365 indepen-
dent (Rint 0.0874), the final R was 0.0476 [I > 2σ(I)] and wR(F2) was
0.1346 (all data).
4-Thiacyclohexyl p-Nitrophenyl Ether 5b
To a solution of tetrahydrothiopyran-4-ol (0.100 g, 0.85 mmol) in THF
(5 mL) was added NaH (50%, 0.05 g, 1.04 mmol) at 0◦C and the mix-
ture was stirred for 45 min at room temperature. p-Fluoronitrobenzene
(0.125 g, 0.89 mmol) was then added and the mixture was stirred
overnight. The resultant mixture was subsequently treated with H2O
(20 mL) and extracted with diethyl ether (3 × 10 mL). The combined
organic layers were washed with H2O (2 × 10 mL), dried (Na2SO4), fil-
tered, and concentrated to give the desired compound as an oil which
slowly crystallized. Recrystallization from MeOH gave 5b, mp 116–
118◦C. δH (CDCl3)8.19(2H, d, J9.5), 6.94(2H, d, J9.2), 4.53–4.48(1H,
m), 2.95–2.89 (2H, m), 2.64–2.57 (2H, m), 2.27–2.20 (2H, m), 2.10–
2.02 (2H, m). δC (CDCl3) 162.3, 141.4, 126.0, 115.3, 73.8, 32.1, 25.2.
Crystal Data for 5g
C11H13NO5S2, M 303.34, T 130.0(1) K, λ 0.71069, triclinic, space
¯
group P1, a 6.8688(7), b 10.059(1), c 10.294(1) Å, α 113.164(2), β
96.939(2), γ 91.788(2)◦, V 646.6(1) Å3, Dc 1.558 mg M−3, µ(MoKα
)
0.427 mm−1, F(000) 316, crystal size 0.4 × 0.2 × 0.1 mm3, 4637 reflec-
tions measured, 2900 independent reflections (Rint 0.0555), the final R
was 0.0362 [I > 2σ(I)] and wR(F2) was 0.0966 (all data).
References
[1] K. Bowden, E. J. Gibbs, Chem. Soc. Rev. 1996, 171.
doi:10.1039/CS9962500171
[2] R. Hoffmann, Acc. Chem. Res. 1971, 4, 1. doi:10.1021/
AR50037A001
[3] K. B. Birnbaum, Acta Crystallogr. 1972, B28, 2825.
[4] C. A. Grob, Angew. Chem., Int. Ed. Engl. 1969, 8, 535.
doi:10.1002/ANIE.196905351
[5] A. S. Grant, J. Mol. Struct. THEOCHEM 1998, 422, 79.
doi:10.1016/S0166-1280(97)00091-2
[6] A. S. Grant, K. Chaudhary, L. Stewart, A. Peters, S. Delisle,
A. Decken, Tetrahedron Lett. 2004, 45, 1777. doi:10.1016/
J.TETLET.2003.12.087
[7] J. M. White, C. I. Clark, Topics in Stereochemistry (Ed.
S. E. Denmark) 1999, Vol. 22, Ch. 3 (John Wiley & Sons:
New York, NY).
[8] A. J. Briggs, R. Glenn, P. G. Jones, A. J. Kirby, P. Ramaswamy,
J. Am. Chem. Soc. 1984, 106, 6200. doi:10.1021/JA00333A014
[9] R. D. Amos, N. C. Handy, P. G. Jones, A. J. Kirby, J. K. Parker,
J. M. Percy, M. D. Su, J. Chem. Soc., PerkinTrans. 2 1992, 4, 549.
doi:10.1039/P29920000549
[10] A. J. Green, J. Giordano, J. M. White, Aust. J. Chem. 2000, 53,
285. doi:10.1071/CH99138
[11] M. Spiniello, J. M. White, Org. Biomol. Chem. 2003, 1, 3094.
doi:10.1039/B303453D
[12] (a) A. J. Kirby, The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen 1983 (Springer: Berlin).
(b) The Anomeric Effect and Associated Stereoelectronic Effects
(Ed. G. R. W. Thatcher) in ACS Symposium Series 539 1993
(American Chemical Society: Washington, DC).
[13] J. B. Lambert, Tetrahedron 1990, 46, 2677. doi:10.1016/S0040-
4020(01)88362-9
[14] L. Andrau, J. M. White, Acta Crystallogr. 2004, E60, o134.
[15] A. J. Green, T. Pigdon, J. M. White, J.Yamen, J. Org. Chem. 1998,
63, 3943. doi:10.1021/JO9723265
[16] X. Creary, Z. Jiang, J. Org. Chem. 1994, 59, 5106.
doi:10.1021/JO00096A073
[17] C. H. Chen, G. A. Reynolds, N. Zumbulyadis, J. A. Van Allan,
J. Heterocycl. Chem. 1978, 15, 289.
Kinetics
2,6-Lutidine was stirred over KOH for 24 h before distillation from
CaH2 under nitrogen. [D4]Trifluoroethanol and D2O were used without
purification.
The method used was based on that reported by Creary and Jiang.[17]
The substrate (1.5 × 10−5 mol) was dissolved in a solution of 0.04 M
2,6-lutidine in 97% [D4]trifluoroethanol/D2O (0.5 mL). The solution
was maintained at 40◦C using a Laude constant temperature bath, and
1H NMR spectra were measured every 5–6 h over a period of several
days.The reaction was essentially complete after 24 days, as at this stage
the only species present were the lutidinium nosylate salt and signals at
δH 2.30 (2H, m), 2.70 (2H, t), 3.10 (2H, m), and 5.87 (2H, m), which
correspond to 4-thiocyclohexene.[17]
Crystallography
Intensity data were collected with a Bruker SMARTApex CCD detector
using MoKα radiation (graphite crystal monochromator, λ 0.71073).[18]
Data were reduced using the program SAINT and corrected for absorp-
tion where appropriate (SADABS). Structures were solved by direct
methods and difference Fourier synthesis using the SHELX [19] suite
of programs as implemented by the WINGX [20] software. Crystallo-
graphic information files (CIFs) for compounds 5a, 5b, and 5e–g have
been deposited at the Cambridge Crystallographic Data Centre, CCDC
numbers 258422–258426; www.ccdc.cam.ac.uk/data_request/cif.
Crystal Data for 5a
C5H10SO, M 118.19, T 130.0(1) K, λ 0.71069, monoclinic, space
group P21/c, a 6.492(1), b 16.408(3), c 17.390(3) Å, β 90.637(4)◦, V
1852.3 Å3, Z 12, Dc 1.271 mg M−3, µ(MoKα) 0.401 mm−1, F(000) 768,
crystal size 0.05 × 0.05 × 0.01 mm3, 9283 reflections measured, 3257
independent reflections (Rint 0.07), the final R was 0.0393 [I > 2σ(I)]
and wR(F2) was 0.0825 (all data).
Crystal Data for 5b
C11H13NO3S, M 239.28, T 130.0(1) K, λ 0.71069, monoclinic,
space group Cc, a 6.541(1), b 23.824(4), c 7.240(1) Å, β 92.451(3)◦, V
1127.1(3) Å3, Z 4, Dc 1.410 mg M−3, µ(MoKα) 0.278 mm−1, F(000)
504, crystal size 0.3 × 0.15 × 0.05 mm3, 3856 reflections measured,
2238 independent reflections (Rint 0.0386), the final R was 0.0437
[I > 2σ(I)] and wR(F2) was 0.0871 (all data).
[18] SMART, SAINT, SADABS 1999 (SiemensAnalytical X-ray Instru-
ments Inc.: Madison, WI).
[19] G. M. Sheldrick, SHELX97 (Includes SHELXS97, SHELXL97)—
Programs for Crystal Structure Analysis (Release 97–2) 1998
(Universität Göttingen: Göttingen).
Crystal Data for 5e
[20] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837. doi:10.1107/
S0021889899006020
C12H13NO4S, M 267.29, T 130.0(1) K, λ 0.71069, triclinic, space
group P1, a 6.891(2), b 7.887(2), c 11.711(3) Å, α 75.111(4), β
¯