Circular dichroism, optical rotation and absolute configuration of 2-cyclohexenone-cis-diol type phenol metabolites: Redefining the role of substituents and 2-cyclohexenone conformation in electronic circular dichroism spectra

Article abstract of DOI:10.1039/c0ob00422g

The absolute configurations of 2-cyclohexenone cis-diol metabolites resulting from the biotransformation of the corresponding phenols have been determined by comparison of their experimental and calculated circular dichroism spectra (TDDFT at the PCM/B2LYP/Aug-cc-pVTZ level), optical rotations (calculated at the PCM/B3LYP/Aug-cc-pVTZ level) and by stereochemical correlation. It is found that circular dichroism spectra and optical rotations of 2-cyclohexenone derivatives are strongly dependent on the ring conformation (M or P sofa S(5) or half-chair), enone non-planarity and the nature and positions of the hydroxy and alkyl substituents. The effect of non-planarity of the enone chromophore, including the distortion of the CC bond, is determined for the model structures by TDDFT calculations at the PCM/B2LYP/6-311++G(2d,2p) level. Non-planarity of the CC bond in the enone chromophore is commonly encountered in 2-cyclohexenone derivatives and it is a source of significant rotatory strength contribution to the electronic circular dichroism spectra. It is shown that the two lowest-energy transitions in acrolein and 2-cyclohexenone and its derivatives are n_CO-π_CO* and π_CC-π_CO*, as expected, while the shorter-wavelength (below 200 nm) transitions are of more complex nature. In 2-cyclohexenone and its alkyl derivatives it is predominantly a mixture of π_CC-π_CC* and π_CC-σ* transitions, whereas the presence of hydroxy substituent results in a dominant contribution due to the n_OH-π_CO* transition. A generalized model for correlation of the CD spectra of 2-cyclohexenones with their structures is presented.

Full text of DOI:10.1039/c0ob00422g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Circular dichroism, optical rotation and absolute configuration of
2-cyclohexenone-cis-diol type phenol metabolites: redefining the role of
substituents and 2-cyclohexenone conformation in electronic circular
dichroism spectra†‡
Marcin Kwit,*a Jacek Gawronski,a Derek R. Boyd,*b Narain D. Sharmab and Magdalena Kaikb
Received 14th July 2010, Accepted 6th September 2010
DOI: 10.1039/c0ob00422g
The absolute configurations of 2-cyclohexenone cis-diol metabolites resulting from the
biotransformation of the corresponding phenols have been determined by comparison of their
experimental and calculated circular dichroism spectra (TDDFT at the PCM/B2LYP/Aug-cc-pVTZ
level), optical rotations (calculated at the PCM/B3LYP/Aug-cc-pVTZ level) and by stereochemical
correlation. It is found that circular dichroism spectra and optical rotations of 2-cyclohexenone
derivatives are strongly dependent on the ring conformation (M or P sofa S(5) or half-chair), enone
non-planarity and the nature and positions of the hydroxy and alkyl substituents. The effect of
non-planarity of the enone chromophore, including the distortion of the C C bond, is determined for
the model structures by TDDFT calculations at the PCM/B2LYP/6-311++G(2d,2p) level.
Non-planarity of the C C bond in the enone chromophore is commonly encountered in
2-cyclohexenone derivatives and it is a source of significant rotatory strength contribution to the
electronic circular dichroism spectra. It is shown that the two lowest-energy transitions in acrolein and
2-cyclohexenone and its derivatives are nC O–pC O* and pC C–pC O*, as expected, while the
shorter-wavelength (below 200 nm) transitions are of more complex nature. In 2-cyclohexenone and its
alkyl derivatives it is predominantly a mixture of pC C–pC C* and pC C–s* transitions, whereas the
presence of hydroxy substituent results in a dominant contribution due to the nOH–pC O* transition. A
generalized model for correlation of the CD spectra of 2-cyclohexenones with their structures is
presented.
1. Introduction
Aromatic hydroxylation is among the most common type of
enzyme-catalysed reaction of aromatic substrates (A) and thus
phenols (C), catechols (D) and hydroquinones (I) are generally
among the early intermediates formed during the biodegradation
of arenes in plants, animals and microorganisms (Scheme 1).1
The phenolic bioproducts (C, D and I) from aromatic substrates
can however be derived from transient or relatively unstable
non-aromatic initial metabolites including epoxides (E, arene
oxides and their oxepine valence tautomers), cis-dihydrodiols (B)
and trans-dihydrodiols (F) via their enzyme-catalysed or sponta-
neous isomerisation, dehydration or dehydrogenation (Scheme 1).
Recent reports on the isolation of relatively stable arene ox-
ides/oxepines (E)2 and cis-dihydrodiols (B),3–6 as initial oxidative
aDepartment of Chemistry, A. Mickiewicz University, 60 780 Poznan,
Poland. E-mail: Marcin.Kwit@amu.edu.pl; Fax: +48 628291505; Tel: +48
628291367
bSchool of Chemistry and Chemical Engineering, the Queen’s University of
Belfast, Belfast, BT9 5AG, United Kingdom. E-mail: dr.boyd@qub.ac.uk;
Tel: +44 90974687
† Dedicated to Professor Koji Nakanishi on the occasion of his 85th
birthday
Scheme 1 Formation of cis-dihydrodiol (B), trans-dihydrodiol (F), cyclo-
hexenone cis-diol (H/G), arene oxide (E) and phenol (C, D, I) metabolites
from arenes (A).
‡ Electronic supplementary information (ESI) available: Full version of
Fig. 5, experimental details, relative energies, structural details for all
calculated molecules, calculated ECD data and optical rotations for
individual conformers of 1–3 and 4–23. See DOI: 10.1039/c0ob00422g
metabolites of mono- (A) or poly-cyclic arenes, is again consistent
with the involvement of similar, but less stable, intermediates in
the formation of phenols.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5635
©

View Article Online
A new family of even more stable 2-cyclohexenone cis-diols
(H) derived from phenols (C), has recently been isolated in
these laboratories.7 Type H bioproducts could be considered as
second generation metabolites involving (i) the initial dioxygenase-
catalysed cis-dihydroxylation of an arene A (e.g. R = R¢ = Me)
to yield an unstable cis-dihydrodiol B (e.g. R = R¢ = Me), (ii)
spontaneous aromatization to form a phenol C (e.g. R = R¢ =
Me) and (iii) a second enzymatic cis-dihydroxylation to give
intermediate G (e.g. R = R¢ = Me) which prefers to exist as the
more stable 2-cyclohexenone cis-diol tautomer H (e.g. R = R¢ =
Me).7 Most members of the new cyclohexenone cis-diol family (H)
have however been produced via biotransformation of the cor-
responding substituted phenol substrates (C). 2-Cyclohexenone
cis-diols (H), obtained using whole cells of a constituent mutant
(UV4) of the soil bacterium Pseudomonas putida containing
toluene dioxygenase (TDO), were in some cases accompanied by
catechols (D).
In the preliminary report involving a very limited range of
cyclohexenone cis-diol metabolites H/G (R = Me, R¢ = H; R = R¢ =
Me; R = I, R¢ = H; R = H, R¢ = Me),7 it was found that they were
enantiopure (>98% ee) via NMR analysis of the corresponding
chiral boronate derivatives. Their absolute configurations were
mainly determined by a combination of X-ray crystallography
H/G (R = R¢ = Me; R = I, R¢ = H) and stereochemical correlation
(R = R¢ = H; R = CN, R¢ = H; R = CO2Me, R¢ = H).
A much wider range of the cyclohexenone cis-diol family (H/G,
>20 members), formed using several dioxygenase types, has now
been isolated and structurally characterised.8 Prior to further
studies of the potential of cyclohexenone cis-diols H/G as syn-
thetic precursors, it was necessary to find a more convenient and
generally applicable method for the determination of the absolute
configurations of this new class of metabolites. In this context
electronic circular dichroism (ECD) spectra and optical rotations
(OR), obtained both from experimental measurements and from
calculations, have earlier proved to be of considerable value
in the assignment of absolute configurations to cis-dihydrodiol
metabolites. These were obtained from dioxygenase-catalysed
cis-dihydroxylation of monosubstituted benzenes,9 disubstituted
benzenes10–12 and from bicyclic aromatics including substituted
naphthalene13 and quinoline14 ring systems. Based on the calcu-
lated ECD and OR values it was possible to provide a rigorous
approach to the determination of both preferred conformations
and absolute configurations of cis-dihydrodiols of type B.9–12
A major objective of the current study is to determine if a
combination of ECD and OR data, obtained by experimental
measurement and calculation, can also be applied to new members
of the 2-cyclohexenone cis-diol family (H, R, R¢ = H, alkyl)8 as a
reliable method for the assignment of absolute configuration.
A further objective is to determine which structural factors
(chromophore nonplanarity, ring conformation, substituents) are
primarily responsible for the chiroptical properties (ECD, OR)
of 2-cyclohexenone derivatives. The enone chromophore has
been a subject of extensive ECD studies in the period of the
nineteen sixties to eighties, particularly in relation to trans-enone
structures present in polycyclic molecules of isoprenoids, including
important steroidal hormones (see reviews15–19). Empirical rules
for planar20 and twisted21–24 trans-enone chromophores have been
postulated on the basis of numerous experimental ECD data,
supported by then available theoretical models.25 A more rigorous
approach to the evaluation of structural factors contributing to
the chiroptical properties of trans-enones was not available at
that time, due to structural complexity of the majority of enones
studied and the lack of a practical computational approach to
structural and spectroscopic studies of molecules of medium
complexity. With simpler 2-cyclohexenone derivatives H now
available, and DFT computational methods at various levels of
accuracy at hand, we felt confident that systematic studies of
flexible trans-enones would provide a host of information which
could be generalized and confronted with the earlier postulated
empirical rules. Such an approach has already proven successful
for 1,3-cyclohexadiene type metabolites B.9,11,12
a,b-Unsaturated ketones typically exhibit two absorption
bands: one very weak near 330 nm, which is the result of
an n–p* electronic transition (R-band) and the second around
230 nm which is due to a p–p* electronic transition (K-band).
Circular dichroism spectra of a,b-unsaturated ketones exhibit
three (sometimes four) Cotton effects between 360 and 185 nm.
The weak UV absorption at about 300 nm is responsible for the
first long-wavelength Cotton effect of moderate intensity. This
band is followed by a p–p* Cotton effect corresponding to the UV
band placed within the range 220 and 250 nm. Another Cotton
effect that does not have the corresponding UV maximum appears
at around 200 nm or at an even higher energy region and usually
exhibits high rotatory strength. The origin of this Cotton effect is
not clear. Earlier calculations by Liljefors and Allinger suggested
that this is another p–p* transition of low intensity in nearly planar
enones26a and the direction of polarization of this transition has
been determined by linear dichroism measurements.26b
During the last few decades various chirality rules have been
proposed for correlating the Cotton effects of a,b-unsaturated ke-
tones with their stereochemistry. Almost fifty years ago Djerassi,23
Whalley,24 and Snatzke,27 proposed rules for the first two Cotton
effects: positive enone helicity was correlated with a negative n–
p* and a positive p–p* Cotton effect. This rule was thereafter
extended by Kirk to s-cis and s-trans enones unperturbed by
polar substituents.17 Snatzke proposed his own sector rule, a
modification of the octant rule that correlates the sign of the n–p*
Cotton effects with the stereostructure of planar enones.20,27 The
n–p* Cotton effects of steroidal enones in oriented (anisotropic)
systems were studied by Kuball et al.28 Their studies demonstrated
that sector or helicity rules can be applied providing that vibronic
progressions originating from various conformers are taken into
account.
On the other hand, Burgstahler29 has shown for the first time the
importance of allylic axial substituents that led to a breakdown
of the enone helicity rule for the p–p* transition of some s-cis
steroid enones. The sign of the third electronic transition Cotton
effect was correlated with the helicity of the C(axial)–C4–C3 C2
bond system,15 P helicity inducing a positive Cotton effect (for
numbering see H in Scheme 1).
However, another analysis of the experimental data led to a
conclusion that it is the arrangement of the R¢–C6–C5–C4–R zig-
zag bonds in a 2-cyclohexenone ring, in relation to the enone
chromophore, that determines the sign of the short-wavelength
Cotton effect.16
To the best of our knowledge none of the proposed correlations
takes into account a contribution from the twisted C C bond to
the rotatory strength at a given wavelength. Recently Grimme30
5636 | Org. Biomol. Chem., 2010, 8, 5635–5645
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
performed a quantum chemical calculation of the rotatory
strength for the electronic transitions of twisted (-10 deg) ethylene.
The calculated rotatory strength for the p–p* transition was of the
order 75–198 ¥ 10-40 cgs units, depending on the method used.
Since non-planar ethylene generated a high rotatory strength,
neglecting such an effect in the case of p–p* electronic transitions
of non-planar enone would seem unjustified. Therefore in the
later part of our computational studies on variously substituted
2-cyclohexenone molecules we aimed at assessing (i) the rotatory
Calculations included both the ring conformers and the sub-
stituent rotamers.§ Despite the highly flexible nature of the 2-
cyclohexenone skeleton, its principal conformers could be char-
acterized by just a few torsion angles. As shown in Scheme 2
for the assumed 4R,5S absolute configuration, each ring con-
formation could be described as either M (torsion angles C3–
C4–C5–C6 negative, C4–C5–C6–C1 positive, (pseudo)axial C4–
OH, (pseudo)equatorial C5–OH or P (torsion angle C3–C4–C5–
C6 positive, C4–C5–C6–C1 negative, (pseudo)equatorial C4–OH,
(pseudo)axial C5–OH). Furthermore, the important enone torsion
angle w (O C1–C2 C3) can be either very close to 180◦ (planar
enone chromophore) or out of planarity; the same applies to the
C1–C2 C3–C4 torsion angle t.
strength generated by a twisted enone and/or a twisted C
C
bond in the enone moiety; (ii) the contribution of the saturated
C4–C5–C6 bond chain to the rotatory strength of planar and
twisted 2-cyclohexenones; (iii) the contributions due to non-polar
(methyl) and polar (hydroxy) substituents in various positions of
the 2-cyclohexenone skeleton.
2. Results and discussion
2.1 Structures of 2-cyclohexenone cis-diols 1–3 from DFT
calculations
In order to determine the chiroptical properties of enones
1–3 (Chart 1) with a sufficient level of confidence, struc-
tures and conformational equilibria of cyclohexenones 1–3 were
calculated at the PCM/MP2/Aug-cc-pVTZ//PCM/B3LYP/
6-311++G(2d,2p) level (see Experimental for details). Additional
molecules 4–23 were used in this study as model structures to
help determine the effect of hydroxy and methyl substituents
on the chiroptical properties of 2-cyclohexenones. In the case of
compounds 4–23 we used a P-helical 2-cyclohexenone as the ring
structure, where descriptor P indicates the arrangement of the
C4–C5–C6 bridge.
Scheme 2 Diastereomeric P and M conformers of cis-ketodiols 1–3 and
the definition of torsion angles a, b, w, t, j and j¢.
Calculated structures of individual low-energy conformers of
enones 1–3 (for details see Tables A and B in Supplementary
Information) are best described as sofa S(5) of either M or P
helicity, with various degrees of distortion from the ideal form,
in the direction of a distorted half-chair (HC) conformer. In fact,
nearly ideal M-S(5) is found in the M4 conformer of 1b and M4a
conformer of 1c while a distorted HC structure can be ascribed to
M1a conformer of 2c (Fig. 1).
Interestingly, the calculated most stable conformation of the
parent 2-cyclohexenone molecule (4) was found to be a distorted
S(5) structure, with torsion angles w = 180◦ and t close to 0◦
(see Table B2, Supplementary Information). The energy of the
molecule showed little sensitivity to C C bond twisting (t 5◦
to -5◦), as long as the torsion angle w was 180◦ (Fig. 1(a),
Supplementary Information).
The sofa S(5) structure is characterized by a planar enone
chromophore, i.e. torsion angles w (O C1–C2 C3) and t
(C1–C2 C3–C4) ideally are correspondingly 180◦ and 0◦. In a
distorted HC conformer the torsion angle w may be much smaller
§ The use of molecular dynamics for small systems does not provide better
results compared to the use of systematical conformational search. In the
case of 1e the number of individual conformers obtained with the use of
MD was 5, i.e. less by 2 to that obtained by systematical conformational
search.
Chart 1
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5637
©

View Article Online
Fig. 1 (a) Structures of sofa S(5) and half-chair HC conformers of the M
series; (b) ring torsion angles of an ideal sofa conformer of 1b(M4) and a
distorted half-chair conformer of 2c(M1a).
Fig. 2 Populations of main conformers (>10%) of keto-cis-diols 1–3 cal-
culated at (PCM)MP2/Aug-cc-pVTZ//(PCM)B3LYP/6-311++G(2d,2p)
level for DE values. Population of conformers with ethyl and isopropyl
substituents is the sum of their rotamers.
(down to 168◦) and the distortion of the C C bond (angle t)
may reach 4◦. It is of interest to note that the sum of the torsion
angles w and t is positive for P conformers and negative for
M conformers. Thus the sign of overall distortion of the enone
chromophore from planarity conforms to the helicity descriptor
P or M.
stabilizes the P conformers so that no significant proportions of
conformers M can be found by calculation. Another interesting
result of calculations is an evident effect of medium polarity on
the conformer population. A simulated polar solvent environment
(acetonitrile or methanol) brings about the equilibrium shift from
P1 to P3 and P4, most evidently seen in the case of enones 1a–1d
and 2a–2c (Fig. 2). Conformers P3 and P4, with an equatorial
hydroxy group in the allylic position, are apparently better suited
for intermolecular hydrogen bonding with the solvent, compared
to P1 in which the allylic hydroxy group is involved as a donor for
intramolecular hydrogen bonding.
As in the case of arene-cis-dihydrodiol metabolites9 the number
of available conformers is not limited to M and P diastereoiso-
meric structures. Intramolecular hydrogen bond patterns of the
hydroxy groups further increase the number of distinct stereoiso-
meric structures (Scheme 2). Thus, either C4–OH or C5–OH can
be a hydrogen bond donor and the orientation of the O–H bond
against the vicinal C–H bond can be either syn or anti (as shown by
the corresponding torsion angles a and b). As shown in Scheme 2,
up to eight conformers (M1–M4, P1–P4) are available but this
number can still be greater if one includes the rotamers due to
alkyl substituents in the ring, e.g. ethyl (1c, 2c, 3c) and isopropyl
(1d). Torsion angles a and b fall into the range. (147◦–179◦) for
anti rotamers and the range is still wider for syn rotamers (5◦–
72◦). For the majority of cases the intramolecular hydrogen bond
2.2 Absolute configurations of 2-cyclohexenone-cis-diols 1–3
from circular dichroism and optical rotation data
Earlier studies have shown that the enantiomeric excess values for
each of the cyclohexenone cis-diols 1a–1e and 2a–2c was >98%
and that absolute configurations for compounds 1a, 2a and 2b had
been unequivocally assigned by stereochemical correlation and X-
ray crystallography.7,8 CD spectra of substituted 2-cyclohexenones
1–3, of assumed 4R,5S absolute configuration, were calculated at
the PCM/TDDFT/B2LYP/Aug-cc-pVTZ level of theory for each
conformer of the relative energy within the 0–2 kcal mol-1 window
(see Supplementary Information). After Boltzmann averaging
and Gaussian curve fitting31 the calculated CD spectra of 1–3
were compared with the experimental ones, as shown in Fig. 3
and 4.
The calculated CD curves were wavelength-corrected to match
the experimental and calculated maximum of the strong p–
p* absorption at ca. 220 nm (the scaling factors are ranging
from 1.08 for 1a to 1.10 for 2b). This procedure still left the
calculated CD curves 10 nm blue-shifted. This is mainly due to an
apparent blue shift of the experimental UVmax position on the slope
of the strong short-wavelength background absorption (Figure
B, Supplementary Information). Otherwise, the calculated CD
spectra of the parent cyclohexenone cis-diol 1a and metabolites
1b–1d were very similar to the experimental ones and also quite
similar to each other, indicating only a small effect of the alkyl
substituents at C3. This confirmed the absolute configuration
of 1a and 1b as 4R,5S, as it was previously established by
˚
O–H ◊ ◊ ◊ O is weak (calculated length 2.50 0.15 A). However, in
several conformers of alkylated enones 1c–1e, 2a–2c and 3c (M1,
˚
M2, P1, P2), this bond is substantially shorter (2.18–2.31 A). In
these conformers the equatorial hydroxy group is a donor to an
axial oxygen atom.
The distribution of conformers for each ketodiol 1–3 has been
calculated for a free molecule in the gas phase or in a polar
medium i.e. acetonitrile or methanol. Due to the large number
of conformers involved, further analysis was limited to those
conformers having contributions over 10% at equilibrium. Fig. 2
shows the calculated distribution of the main conformers in these
three different environments.
2-Cyclohexenone-cis-diols of 4R,5S configuration are prefer-
entially of ring P helicity in which the allylic hydroxy group
is equatorial. Depending on the calculation method used, some
conformers were not found as stable structures (for example
conformer P3 of 1e). Significant amounts of M conformers appear
in 3-alkylated enones 1b–1e and 3a–3c, since there is a destabilizing
interaction between the 3-alkyl and 4-equatorial hydroxy groups.
Calculations indicate a high population of M4 conformer in the
case of the 3-tert-butyl derivative 1e. In the case of compounds 3a–
3c, the equilibrium is additionally driven toward M conformers by
the equatorial position of the 6-alkyl group. For the same reason,
the 6-alkyl substituent in epimeric molecules 2a–2c additionally
5638 | Org. Biomol. Chem., 2010, 8, 5635–5645
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 4 CD spectra of cis-ketodiols ent-2a–2c experimental (in acetoni-
trile solutions, solid lines), DEMP2 Boltzmann averaged calculated at
PCM/B2LYP/Aug-cc-pVTZ level for 2a–2c (dashed lines) and 3a–3c
(dash-dot-dot lines). Rotatory strengths R were calculated in dipole-ve-
locity representation. Inserts show nC O–pC O* transition region of the
CD spectra. All calculated spectra were wavelength-corrected to match
the experimental short-wavelength UV lmax at ca. 220 nm.
environment. This is in sharp contrast to the results from
structure calculations of the lower homologues 1a–1d for which
P-type conformers are the most abundant (up to 81% for 1a
in the acetonitrile continuous solvent model). In fact, a better
agreement between the calculated and experimental CD curves
was obtained if a higher ratio (> 70 : 30) of conformer P4 to M4
was assumed (these two conformers are the representative ones for
acetonitrile solution), see Figure C, Supplementary Information.
We noted that the method of geometry optimization has little
effect on the further calculated ECD spectra.¶ In the case of
1a, 1b and 1e the geometries of individual conformers were
re-optimized at the PCM/MP2/Aug-cc-pVTZ//PCM/MP2/6-
311G(d,p) level of theory (see Tables B3 and B4 in Supple-
mentary Information), followed by calculations of ECD spectra
at the PCM/TDDFT/B2LYP/Aug-cc-pVTZ level of theory.
Except for numerical values the results thus obtained were
not different from those obtained with PCM/MP2/Aug-cc-
pVTZ//PCM/B3LYP/6-311++G(2d,2p) method.
Fig. 3 CD spectra of cis-ketodiols 1a–1e, experimental (in acetonitrile
solutions, solid lines) and DEMP2 Boltzmann averaged calculated at
PCM/B2LYP/Aug-cc-pVTZ level (dashed lines). Rotatory strengths R
were calculated in dipole-velocity representation. All calculated spectra
were wavelength-corrected to match the experimental short-wavelength
UV lmax at ca. 220 nm.
X-ray determination and chemical transformation of the 3-iodo
derivative 1 (X = I)7 into enone 1a and established the absolute
configuration of 1c and 1d.
The case of 3-tert-butyl substituted 2-cyclohexenone-cis-diol 1e
is apparently more complex, since the experimental CD spectrum
is not reproduced well by computation for either enantiomer
(Fig. 3). As is usually the case, the discrepancy between the
experiment and the computation of chiroptical properties results
from the incorrect computational reproduction of conformer
population. The calculations show a low population (11%) of
P conformers in the equilibrium with M conformers (89%)
for molecules of 1e in vacuo and even less (5–6%) in a polar
¶ During the work on the cis-ketodiols and other arene metabolites we
tested various methods for calculation of molecular structure. These in-
clude different density functionals (with or without dispersion correction)
and MO methods. Based on these results we did not observe any significant
effect of long-range interactions on the conformation of molecules studied.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5639
©

View Article Online
Table 1 Specific optical rotations for cis-ketodiols 1–3, measured
in methanol solution and Boltzmann averaged calculated at the
PCM/B3LYP/Aug-cc-pVTZ level (values in parentheses)
on 2-cyclohexenones (1c and 1d). In some of the calculated
conformations the alkyl group attached to a sp2 carbon atom
generates a rotamer in which the methyl group is oriented
perpendicularly to the C2 C3 bond (torsion angle j 90–92◦,
conformers M1b, M2b, M3b, M4b of 1c and j 113–116◦ for the
same conformers of 1d, see Table B, Supplementary Information).
Therefore the pseudoaxial methyl group of the substituent at the
non-stereogenic C3 carbon atom contributes to the rotational
strength of the n–p* and p–p* transitions in an opposite direction
to that of the hydroxy substituent in C4 position, resulting in
a change of the sign pattern of the dihydroxyenone Cotton
effects.
A similar effect of conformation of the C3-alkyl substituent is
seen in calculated optical rotations. In Fig. 5 is shown a selected
example: ethyl-substituted keto-cis-diol 1c (full version on Fig. 5 is
deposited in Supplementary Information). Whereas OR values for
the majority of conformers of 1a–1d is negative (regardless whether
P or M conformation), it is strongly positive for conformers
mentioned above, i.e. M1b, M2b, M3b and M4b of 1c and 1d.
Thus simple rotation of the ethyl or isopropyl group at the non-
stereogenic carbon atom (C3) in 1c or 1d can dramatically change
the chiroptical properties of the molecule.
Diol
589 nm
578 nm
546 nm
436 nm
1a
1b
1c
1d
1e
2a
2b
2c
3a
3b
3c
-217 (-255)
-119 (-103)
-103 (-82)
-79 (-72)
-9 (+51)
-227 (-329)
-124 (-140)
-107 (-88)
-82 (-62)
-9 (+63)
-261 (-383)
-141 (-162)
-123 (-101)
-94 (-71)
-10 (+74)
+273a (-399)
+116a (-161)
+128a (-168)
(-387)
-465 (-716)
-237 (-287)
-201 (-168)
-150 (-112)
-17 (+151)
+540a (-814)
+237a (-352)
+259a (-377)
(-677)
+223a (-264)
+94a (-105)
+104a (-96)
(-239)
+234a (-340)
+98a (-135)
+109a (-144)
(-337)
(-121)
(-108)
(-185)
(-151)
(-211)
(-343)
(-171)
(-248)
a Measured for enantiomer.
The results of measurement and calculation of optical rotation
for 1a–1d are in agreement with the assignments obtained by
ECD. Thus, the calculated and measured signs and magnitudes
of specific optical rotations at four different wavelengths for 1a–1d
are in agreement with the assumed 4R,5S absolute configuration
(Table 1, see also Table C in Supplementary Information).
On the other hand, no such agreement is seen for tert-butyl
substituted 2-cyclohexenone-cis-diol 1e, for which the measured
optical rotation is weakly negative, whereas the calculated ro-
tations are positive. Since the calculated optical rotations of P
conformers of 1e are negative, the correct sign of optical rotation
is obtained with over 50% population of these conformers in the
equilibrium.
Thus, for 1e, the discrepancy between the experimental and
calculated OR data can be rationalised in the same manner as
discussed above for the CD spectra of this compound.
For metabolites 2a–2c having an additional alkyl substituent
at C6, calculated and experimental CD spectra are of mirror-
image type in the spectral region below 300 nm, indicating that
the absolute configuration is reverse to that assumed, i.e. it is
4S,5R. This is indeed in agreement with the absolute configuration
determined for the monocamphanate derivative of 2b, using the
X-ray diffraction method.7
Dihydroxyenones 3a–3c, epimeric at C6, were used for com-
putational assessment of CD spectra in order to differentiate
between 2a–2c and 3a–3c. Unlike 2a–2c, molecules of 3a–3c are
not restricted to conformers of P type (Fig. 2). Nevertheless, their
CD spectra below 300 nm are deceptively similar to those of 2a–2c
(Fig. 4). The only difference between the calculated CD spectra of
epimers 2a–2c and 3a–3c is the opposite sign of the n–p* Cotton
effects at above 300 nm for each pair of epimers and slightly weaker
positive Cotton effects at ca.: 210 nm for compounds 3a–3c.
This computational study shows that the CD spectra of
dihydroxyenones 1–3 reflect primarily the absolute configuration
of these molecules at C4 and C5 and to a much lesser extent
the conformation of the molecules. Namely, for 4R,5S absolute
configuration a characteristic pattern of the Cotton effects is as
follows: (+) n–p*, (-) p–p*, (+) third Cotton effect. Thus, the CD
spectroscopy can be regarded as a reliable method for determining
absolute configuration of 4,5-dihydroxy-2-cyclohexenones with-
out resorting to computations. The only difficult cases are those
having spherically unsymmetrical 3-alkyl (Et, i-Pr) substituents
Fig. 5 Effect of rotation of hydroxy and ethyl substituents on the
calculated (PCM/B3LYP/Aug-cc-pVTZ level) optical rotation at 589 nm
for low-energy conformers of 1c.
These findings clearly indicate that, in order to clarify the role of
various constituents of a chiral enone structure in determining its
chiroptical properties, calculations for simpler model compounds
are necessary.
5640 | Org. Biomol. Chem., 2010, 8, 5635–5645
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
3. Analysis of the calculated circular dichroism spectra of
acrolein and model 2-cyclohexenones 4–23
a reverse order of the two transitions is suggested by calculations
for 6-hydroxy substituted 2-cyclohexenones. The third electronic
transition at ca. 180 nm is a mixture of pC C–pC C* transition
3.1 Nature of the low-energy transitions
(HOMO–LUMO+1, >80% contribution), pC C–sC–C* and pC
–
C
s
C–H* transitions (HOMO–LUMO+2, minor contributions) of
Acrolein (CH2 CHCHO) is structurally the most simple model
of an a,b-unsaturated carbonyl compound. The molecule can
be chiral if it is nonplanar. Since the structures of acrolein and
2-cyclohexenone (4) differ significantly, the origin of calculated
rotatory strengths may be different for both model compounds.
We performed analysis of the electronic structures of acrolein
and a number of 2-cyclohexenones by the natural bond orbital
(NBO) method.32 This method has a unique feature of analyzing
the electron density by using sophisticated basis sets and methods
in terms that are familiar to a chemist, such as Lewis structure,
lone pairs, bonds, antibonds and Rydberg orbitals.
For planar acrolein the origin of the first two low-energy singlet
transitions is clear: the long-wavelength electronic transition
involves HOMO-1 and LUMO orbitals, whereas in the case
of the next electronic transition the major contribution comes
from the transition between HOMO and LUMO orbitals. The
higher energy region is more complex, since there are three
electronic transitions of similar energy at around 160 nm. These
include HOMO-2–LUMO and HOMO-1–LUMO+2 transitions.
low oscillator strength. This is a new finding on the origin of
the short-wavelength transition and it is apparently in line with
the previously postulated16 sensitivity of the short wavelength
transition in 2-cyclohexenone derivatives to the pattern of the
C–C bonds in the vicinity of the enone C C bond.
It should be noted that for the majority of methyl-substituted
model compounds 5–23 (with a planar chromophore) the same
order of interacting orbitals for the first two electronic transitions
is retained, as is the case for 2-cyclohexenone (4). The third
electronic transition in the case of 4,5-dimethyl derivatives 18
involves additionally HOMO-2 and LUMO orbitals (pC O–pC
*
O
transition). However, in the cases of all hydroxy substituted 2-
cyclohexenones the origin of the third electronic transition of
low intensity, appearing at ca. 190 nm, is quite different and
it involves the lone pair of the hydroxy group (HOMO-2) and
LUMO orbitals, see Fig. 6(c), therefore constituting (typically
in ca. 80%) the nOH–pC O* transition. This new finding suggests
that in the cases of hydroxylated 2-cyclohexenone derivatives
the sign of the short wavelength electronic transition Cotton
effect may be dependent on the helicity of the O–C–C O bond
system. Indeed, for epimeric 4-hydroxy-2-cyclohexenones (4S)-7
and (4R)-7 (a known phenol metabolite)7 strong, opposite sign
short-wavelength Cotton effects are calculated (see Figure D,
Supplementary Information).
According to the calculation the first electronic transition is nC
–
O
pC O*, the second pC C–pC O* and the third band is composed of
pC O–pC O* and nC O–sC(O)-H* transitions (Fig. 6(a)).
3.2 Contribution due to non-planarity of the enone chromophore
As it was mentioned earlier, a non-planar enone is an inherently
dissymmetric (helical) chromophore. In addition, planar s-trans
enones (torsion angle w = 180◦) may be chiral due to non-planarity
of the C C bond or chiral disposition of the rest of the molecule.
This is the case of 2-cyclohexenone (4), where at least one of the
carbon atoms of the C4–C5–C6 bridge is out of the plane generated
by the planar enone chromophore (w = 180◦, t = 0◦).
A good model for the enone chromophore can be acrolein. The
molecule can be made helical by changing torsion angles w and t
(the latter defined in this case as (O)C–C C–H angle, where H is
syn to (O)C–C bond) to form (+/+ or -/-) or ( or -/+) structures.
For calculations the torsion angles w and t were varied in the range
5◦, in one degree steps. For all conformers the rotatory strengths
were calculated at PCM/B2LYP/6-311++G(2d,2p) level. The
results are shown in diagrams representing the calculated rotatory
strengths for the first three electronic transitions as the functions
of torsion angles w and t (see Figure E in Supplementary
Information). Fig. 7 shows the cases where either the enone or
the C C bond is planar.
Fig. 6 Molecular orbitals obtained with the use of the NBO method
for representative planar enones: acrolein (a) and P-type conformers
of substituted 2-cyclohexenones 4 (b) and (4S)-7 (c). I nC O–pC O*, II
As expected, for a planar C C bond, the calculated rotatory
pC C–pC O*, IIIa pC O–pC O*, IIIb nC –sC(O)-H*, IIIc pC C–pC C*, IIId
strengths support the enone helicity rule, for both nC O–pC
*
O
O
pC –sC–C*/sC–H*, IIIe nOH–pC O*.
C
and pC C–pC O* electronic transitions – a positive w angle results
in a negative nC O–pC O* and a positive pC C–pC O* rotatory
strength. The short-wavelength electronic transition generates
rotatory strength of opposite sign to the sign of the rotatory
strength of the pC C–pC O* transition. It should be noted that
unlike nC O–pC O* and pC C–pC O* transitions, the calculated
In the case of 2-cyclohexenone (4) with either a planar or a
twisted chromophore the first two electronic transitions, calculated
at 286 nm and 211 nm, respectively, originate (in >80%) from the
same orbitals as in the case of acrolein (see Fig. 6(b)). However,
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5641
©

View Article Online
Table 2 Averaged contributions of the rotatory strength (R¥10-40 erg
esu cm Gauss-1) to (P)-2-cyclohexenone (4) Cotton effects due to non-
planarity of the enone chromophore
nC O–pC
*
pC C–pC
*
O
O
w = 175◦
w = -175◦
t = 2◦
t = -2◦ to 2◦
-1.5
2
2
-2
9
12
t = -2◦ to 2◦
-8
w = -175◦ to 175◦
w = -175◦ to 175◦
w = -175◦ to 175◦
w = -175◦ to 175◦
-17
t = -2◦
t = 5◦
17
-48
55
t = -5◦
-7
was positive for all combinations of the enone and C C twist
angles in P-type conformers. For the reasons mentioned in the
preceding section, analysis of the rotatory strength of the third
electronic transition as a function of w and t torsion angles was
not pursued.
The contribution of the C4–C5–C6 chain to the calculated
rotatory strength of the nC O–pC O* and pC C–pC O* transitions
in (P)-2-cyclohexenone with a planar enone chromophore is
estimated correspondingly as 4.3 ¥ 10-40 and -5.6 ¥ 10-40 cgs units.
The effect of non-planarity of the enone chromophore in 2-
cyclohexenone 4 and its monosubstituted derivatives 5–11 could
be assessed by subtracting the rotatory strength calculated for a
molecule with a planar chromophore (w = 180◦, t = 0◦) from that
of a molecule for which either w or t is varied within 5◦ range
(Table D, Supplementary Information). As in the case of acrolein,
changes of either w or t resulted in significant changes of the
rotatory strengths, shown in Table 2.
Comparable distortions of the C C bond (angle t) and
enone chromophore (angle w) resulted in a more significant
rotatory strength change in the former case, the signs following
the pattern determined for acrolein. The low-energy conformers
of 2-cyclohexenone 4 with a twisted enone chromophore have
torsion angles w and t of opposite sign (Table E, Supplementary
Information), hence the contributions to the rotatory strengths
due to non-planarity of the chromophore are partially cancelling.
A similar effect can be observed for 3-methyl-2-cyclohexenone 5
and (4S)-4-hydroxycyclohex-2-enone 7 (Figures H and I, Supple-
mentary Information). In the latter case the rotational strengths
of the low energy transitions are strongly dependent on angle t,
however due to the presence of the 4-hydroxy substituent, the signs
of the rotatory strengths for pC C–pC O* and nOH–pC O* transitions
remain unchanged (correspondingly negative and positive).
Fig. 7 Rotatory strengths R of the three low-energy electronic transitions
of acrolein as a function of torsion angles w and t, calculated at
PCM/B2LYP/6-311++G(2d,2p) level (left panels t = 0◦, right panels w =
180◦).
rotatory strength of the third electronic transition is not a simple
function of angle w.
On the other hand, for w = 180◦, even a small deviation from
planarity of the C C bond generates rotatory strengths of the
nC O–pC O* and pC C–pC O* transitions of much higher intensity
(up to 10 times) compared to those calculated for a twisted acrolein
model with a planar C C bond. The sign dependence of the
rotatory strengths of the nC O–pC O* and pC C–pC O* transitions
on the sign of angle t is opposite to that of angle w, i.e. for a
positive angle the t calculated rotatory strengths are positive for
the nC O–pC O* and negative for the pC C–pC O* transitions. Thus,
one can conclude that the contribution to the calculated rotatory
strength due to non-planarity of the C C bond can override the
contribution due to non-planarity of the enone chromophore. The
acrolein model obeys the enone helicity rule when w and t angles
are of opposite sign.
Similar analysis was performed for 2-cyclohexenone (4) (see
Figure F in Supplementary Information). This case is more
complex due to the presence of a non-planar saturated carbon
chain by which the chromophore is embedded into the ring. An
approximate contribution from that chain to the rotatory strength
of a given electronic transition was obtained by subtraction of
the rotatory strength calculated for acrolein from that obtained
for conformers of 2-cyclohexenone 4 (see Figure G in Supplemen-
tary Information). The presence of the saturated carbon chain
3.3 Contributions due to the substituents
The investigated cis-diols 1–3 contain polar as well as alkyl groups
around their cyclohexenone skeleton that may modify the CD
spectra significantly. To determine the effect of the substituents
we replaced systematically the hydrogen atoms at C3–C6 in 2-
cyclohexenone 4 with either a polar hydroxy group or a non-
polar methyl group (molecules 5–23). The conformation of the
hydroxy group was arbitrarily fixed antiperiplanar to the C*–H
bond (see Figure J and Table D in Supplementary Information),
except for molecules having an equatorial OH group at C6 where
the H–O/C–H torsion angle was fixed at -108◦. The structures
of all molecules were fixed to have a planar chromophore. To
estimate the effect of the substituent, the calculated rotatory
strengths of these model compounds were divided by those
significantly affected the rotatory strengths of the pC C–pC
*
O
transition, especially for the structures where the enone was twisted
“unnaturally” in P-type conformers, i.e. when w was negative.
Magnitudes, but not the signs, of the calculated rotatory strengths
for the nC O–pC O* transition were also affected by the presence
of the C4–C5–C6 chain. The contribution of the C4–C5–C6 chain
5642 | Org. Biomol. Chem., 2010, 8, 5635–5645
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
calculated for a planar unsubstituted 2-cyclohexenone (Table
F, Supplementary Information). Alternatively, the substituent
effect could be determined by subtracting the calculated rotatory
strength for a given transition of 4 from that of substituted 2-
cyclohexenone.
most significantly to the rotatory strength of the pC C–pC
transition. The contributions from non-planarity of the C
*
C
O
bond and from the helicity of the HO–C–C C system are of
the same order of magnitude and much stronger than that due to
non-planarity of the enone chromophore. The contribution due
to an axial hydroxy group at C5 is also significant. A further
possibility is that a better overlap of the equatorial oxygen at C4
by HOMO results in a higher value of the pC C–pC O* transition
rotatory strength, compared to the methyl-substituted analogue.
Thus, our study shows for the first time that the sign of the principal
pC C–pC O* transition Cotton effect is less dependent on the enone
non-planarity (angle w) while being more dependent on the non-
planarity of the C C bond and the presence of a substituted
helical 2-cyclohexenone ring. By way of example we compare the
rotatory strengths of the nC O–pC O* and pC C–pC O* transitions,
calculated either as a sum of contributions of structure elements
or calculated directly for conformer P1 of dihydroxyenone 2a.
This conformer is representative as it contributes 70% to confor-
mational equilibrium. The contributions to the rotatory strengths
of the two lowest-energy transitions are correspondingly +5, -11
(2-cyclohexenone ring, w = 175◦, t = 1.6◦), 0, -68 (4eq OH group),
-3, -33 (5ax OH group), -1, +9 (6eq Me group), data taken
from Tables A, B and D, Supplementary Information). The sum
of contributions, +1, -103 compares well with the data for this
conformer +0.3, -113. Note that the relatively weak contribution
of a helical P type 2-cyclohexenone ring is due to the cancelling
effect of the twist of the enone and the C C bond.
Among monosubstituted 2-cyclohexenones 6–11, substitution
with either a methyl or a hydroxy group provides a negative (or very
small positive) contribution to the rotatory strength of the nC
–
O
pC O* transition, regardless of the position of the substituent. The
most significant effects are noted for the hydroxy groups in 4-
axial, 5-equatorial and 6-equatorial positions and for the methyl
substituent in a 5-axial position.
As expected for the pC C–pC O* transition, the most significant
changes to the calculated rotatory strengths are due to the
substituents in the C4 (allylic) position. The effects due to
the equatorial and axial substituents are of opposite sign. An
equatorial substituent gives a strong negative contribution, in
contrast to a positive contribution of the axial substituent in the
P conformers, in effect revealing the helicity of the R–Callyl–C
C
system as negative for the equatorial and positive for the axial
substituent. The effect of the allylic axial substituent appears
insensitive to its polarity. In contrast, replacing an equatorial
methyl by a hydroxy group (as in the cases of (4S)-6 and (4S)-
7), increases twofold the calculated rotatory strength of the 2-
cyclohexenone derivative. This is apparently due to a better overlap
of the polar C–O bond by HOMO, especially when the OH group
is in the allylic equatorial position.
Substituents at C5 affect the CD spectrum strongly if they
occupy an axial position. However, the contributions calculated
for the methyl group (in (5S)-8 and (5R)-8) and the hydroxy
group (in (5S)-9 and (5R)-9), are in pairs of opposite sign,
adding complexity to the general scheme. Going further to the C6
position, the calculated rotatory strength for the axially methyl-
substituted molecule (6S)-10 is higher than that calculated for the
equatorial epimer (6R)-10, while an opposite relationship holds
for the hydroxy-substituted 2-cyclohexenones (6R)-11 and (6S)-
11. In addition, contributions due to the axial and equatorial
substituents are mutually of opposite sign.
4. Conclusions
We have shown that the absolute configurations of 2-
cyclohexenone cis-diols 1 and 2, bacterial metabolites of phenols,
can be determined by comparison of their experimental and
calculated ECD spectra and optical rotations, despite the highly
flexible nature of the molecules. By proper choice of the TDDFT
method (PCM/B2LYP/Aug-cc-pVTZ), excellent agreement was
obtained for most of the cases, despite the significant number of
conformers and structural variables involved in the analysis (ring
torsion angles, rotamers of the hydroxy and alkyl groups). It is
shown that both the ECD spectra and optical rotations (including
the sign) are also significantly affected by rotation of the ethyl or
isopropyl substituent at C3.
Based on the calculated rotatory strengths for the pC C–pC
*
O
transition of monosubstituted (P)-2-cyclohexenones 5–11, the
substituent contributions, including signs, follow the sequence:
4eq-OH (-), 4ax-OH (+), 4ax-Me (+) > 4eq-Me (-), 5ax-OH (-),
6eq-OH (+) > 5ax-Me (+), 6ax-Me (-) > 5eq-OH (+), 5eq-Me
(-), 6eq-Me (+), 6ax-OH (+). The dominant role of substituent at
C4 over substituents either in C5 or C6 positions is clearly visible,
regardless of their electronic nature and orientation (see Table F,
Supplementary Information).
The validity of the earlier postulated semiempirical helicity
rules has been examined by calculations of the ECD spectra of
a series of model enones using the above mentioned DFT method.
Calculations confirmed that the two lowest-energy transition
Cotton effects are predominantly of the nC O–pC O* and pC
–
C
Specific optical rotations were calculated at the same level of
theory for several monosubstituted 2-cyclohexenones 6–10 and for
which experimental data are available in the literature (see Table
G, Supplementary Information). Except for 2-cyclohexenone 10,
optical rotations at 589 nm, both calculated and experimental, are
negative. This is apparently due to the substituent effect shown
above i.e. negative contributions due to 4eq-Me, 4eq-OH, 5eq-Me
or 5ax-OH, positive due to 6eq-Me.
In summary, going back to molecules 1 and 2, and taking into
account that their most dominant conformers are of P-helicity
with the hydroxy groups occupying 4-equatorial and 5-axial
positions, we can now order the structural factors contributing
pC O* type and that their rotatory strengths are indeed dependent
on the non-planarity of the enone chromophore (angle w), as it
was previously formulated within the trans-enone helicity rule.
However, the present study shows that chiroptical behavior of 2-
cyclohexenones is far more complex. The cyclohexenone ring is of
C1 symmetry and it contributes to the rotatory strengths according
to its helicity (P or M). The C C bond (angle t) is not planar in
low-energy conformers of substituted 2-cyclohexenones and this
is a source of substantial contribution to the rotational strength
of the two lowest-energy electronic transitions.
For the majority of low-energy conformers this contribution is
of opposite sign to that due to non-planarity of the enone group.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5643
©

View Article Online
Table 3 Signs of contributions to the rotatory strengths of the nC
pC O* and pC C–pC O* Cotton effects of (P)-2-cyclohexenones
–
establish (i) if opposite absolute configurations are consistently
found for metabolite types 1 and 2 and (ii) if a predictive model
for TDO-catalysed cis-dihydroxylation of phenols can also be
developed.
O
Acknowledgements
Contribution
nC O–pC
*
pC C–pC
*
O
O
This work was supported by a grant No. N N204 056335 from
the Ministry of Science and Higher Education Fund and by an
EU Marie Curie Host Fellowship for the Transfer of Knowledge,
NOVACAT, Project No. 29846 (to MK). All calculations were
performed at Poznan´ Supercomputing Center.
P-helicity ring
(-)-w
+
—
—
weak +
—
—
(-)-t
+ to ++
+ +
+ +
- -
- -
+
- -
—
+
—
4ax Me
4ax OH
4eq Me
4eq OH
5ax Me
5ax OH
5eq Me
5eq OH
6ax Me
6ax OH
6eq Me
6eq OH
—
ca. 0
ca. 0
—
—
—
Notes and references
1 B. L. Goodwin, in Handbook of Biotransformations of Aromatic
Compounds, CRS, Press, Boca Raton, 2005.
—
ca. 0
ca. 0
—
2 D. R. Boyd, N. D. Sharma, J. S. Harrison, J. F. Malone, W. C.
McRoberts, J. T. G Hamilton and D. B. Harper, Org. Biomol. Chem.,
2008, 6, 1251.
ca. 0
+
+ +
—
3 R. A. Johnson, Org. React., 2004, 63, 117.
4 D. R. Boyd and T. D. H. Bugg, Org. Biomol. Chem., 2006, 4, 181.
5 K. A. B. Austin, M. Matveenko, T. A. Reekie and M. G. Banwell, Chem.
Aust., 2008, 75, 3.
Furthermore, strong contributions to the rotatory strength of the
pC C–pC O* transition are associated with a substituent (OH, Me)
in positions 4 and 5. This is summarized in Table 3.
6 T. Hudlicky and J. W. Reed, Synlett, 2009, 685.
7 D. R. Boyd, N. D. Sharma, J. F. Malone and C. C. R. Allen, Chem.
Commun., 2009, 3633.
Another striking result of our study is the determination
of the nature of the third band in the ECD spectra of 2-
cyclohexenones. Calculations show that its nature is dependent
on the presence of substituents. In 2-cyclohexenone and its alkyl
8 D. R. Boyd, N. D. Sharma, P. Stevenson, C. C. R Allen, L. Kulakov,
H. Mundi, W. C. McRoberts and J. T. G. Hamilton, manuscript in
preparation.
9 J. K. Gawronski, M. Kwit, D. R. Boyd, N. D. Sharma, J. F. Malone
and A. Drake, J. Am. Chem. Soc., 2005, 127, 4308.
10 D. R. Boyd, N. D. Sharma, G. N. Coen, P. Gray, J. F. Malone and J.
Gawronski, Chem.–Eur. J., 2007, 13, 5804.
11 M. Kwit, N. D. Sharma, D. R. Boyd and J. Gawronski, Chem.–Eur. J.,
2007, 13, 5812.
12 M. Kwit, N. D. Sharma, D. R. Boyd and J. Gawronski, Chirality, 2008,
20, 609.
13 M. Kwit, J. Gawronski, D. R. Boyd, N. D. Sharma, M. Kaik, R. A.
More O’Ferrall and J. S. Kudavalli, Chem.–Eur. J., 2008, 14, 11500.
14 M. Kwit, J. Gawronski, L. Sbircea, N. D. Sharma, M. Kaik and D. R.
Boyd, Chirality, 2009, 21, E37.
15 R. D. Burnett and D. N. Kirk, J. Chem. Soc., Perkin Trans. 1, 1981,
1460.
derivatives it is predominantly a mixture of pC C–pC C* and pC
–
C
s* transitions, whereas substitution with hydroxy groups results
in a dominant contribution due to the nOH–pC O* transition. Thus,
although Cotton effects associated with these transitions may
provide useful stereochemical information, their interpretation
in any particular case will require computational analysis of the
electronic transitions involved.
The single absolute configuration of enantiopure cis-
dihydrodiol metabolites consistently obtained from TDO-
catalysed cis-dihydroxylation of monosubstituted,9 and ortho- or
meta-disubstituted benzene substrates,12 can now be rationalised
on the basis of a simple model.9,12 This contrasts with the enan-
tiocomplementarity observed when enantiopure cyclohexenone
cis-diols 1a–1e or 2a–2c of opposite configuration were formed
using TDO and the corresponding phenols (Scheme 3). Further
studies are in progress to identify more members of this new
cyclohexenone cis-diol family and to use CD spectroscopy to
16 J. K. Gawronski, Tetrahedron, 1982, 38, 3.
17 N. Kirk, Tetrahedron, 1986, 42, 777.
18 J. Gawronski, Conformations, Chiroptical and Related Spectral Prop-
erties of Enones, Chapter 3 in The Chemistry of Enones, S. Patai,
Z. Rappoport, ed., J. Wiley, 1989.
19 D. A. Lightner and J. E. Gurst, Organic Conformational Analysis and
Stereochemistry from Circular Dichroism Spectroscopy, Wiley-VCH,
2000.
20 G. Snatzke, Angew. Chem., Int. Ed. Engl., 1979, 18, 363.
21 G. Snatzke, in Fundamental Aspects and Recent Developments in Optical
Rotatory Dispersion and Circular Dichroism, F. Ciardelli, P. Salvadori,
ed., Heyden, London, 1973, pp. 109–124.
22 (a) G. Snatzke, Tetrahedron, 1965, 21, 413; (b) G. Snatzke, Tetrahedron,
1965, 21, 421.
23 C. Djerassi, R. Records, E. Bunnenberg, K. Mislow and A. Moscowitz,
J. Am. Chem. Soc., 1962, 84, 870.
24 W. B. Whalley, Chem. Ind., 1962, 1024.
25 W. Hug and G. Wagniere, Helv. Chim. Acta, 1971, 54, 633.
26 (a) T. Liljefors and N. L. Allinger,, J. Am. Chem. Soc., 1976, 98, 2745;
(b) J. K. Gawronski, T. Liljefors and B. Norde´n, J. Am. Chem. Soc.,
1979, 101, 5515.
27 G. Snatzke, Tetrahedron, 1965, 21, 439; G. Snatzke, in Optical Rotatory
Dispersion and Circular Dichroism in Organic Chemistry, G. Snatzke,
Ed. Sadtler Research Labs, Inc, Philadelphia, 1967, pp. 208–223.
28 (a) H.-G. Kuball, S. Neubrech and A. Scho¨nhofer, Chem. Phys., 1992,
163, 115; (b) H.-G. Kuball, B. Schultheis, M. Klasen, J. Frelek and A.
Scho¨nhofer, Tetrahedron: Asymmetry, 1993, 4, 517; (c) J. Frelek, W. J.
Scheme 3
5644 | Org. Biomol. Chem., 2010, 8, 5635–5645
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Szczepek, H. P. Weiss, G. J. Reiss, W. Frank, J. Brechtel, B. Schultheis
and H.-G. Kuball, J. Am. Chem. Soc., 1998, 120, 7010; (d) J. Frelek,
W. J. Szczepek, S. Neubrech, B. Schultheis, J. Brechtel and H.-G. Kuball,
Chem.–Eur. J., 2002, 8, 1899.
32 (a) F. Weinhold, in, Encyclopedia of Computational Chemistry, P. v.
R. Schleyer, N. L. Allinger, T. Clark, J. Gasteiger, P. A. Kollman,
H. F. Schaefer III and P. R. Schreiner, ed., John Wiley & Sons,
Chichester, UK, 1998, Vol. 3, pp. 1792–1811; (b) A. E. Reed, L. A.
Curtiss and F. Weinhold, Chem. Rev., 1988, 88, 899; (c) F. Weinhold
and J. E. Carpenter, in, The Structure of Small Molecules and Ions,
R. Naaman and Z. Vager, ed., Plenum, New York, 1988, pp. 227–236;
(d) F. Weinhold, Computational Methods in Organic Photochemistry:
Molecular and Supramolecular Photochemistry, A. G. Kutateladze, Ed.,
Taylor & Francis/CRC Press, Boca Raton FL, 2005, pp. 393–476.
29 A. W. Burgstahler and R. C. Barkhurst, J. Am. Chem. Soc., 1970, 92,
7601.
30 A. Diedrich and S. Grimme, J. Phys. Chem. A, 2003, 107, 2524.
31 (a) N. Harada and P. Stephens, Chirality, 2010, 22, 229; (b) N. M.
O’Boyle, A. L. Tenderholt and K. M. Langner, J. Comput. Chem.,
2008, 29, 839.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5635–5645 | 5645
©