How an early or late transition state impacts the stereoselectivity of tetrahydropyran formation by intramolecular oxa-Michael addition

Article abstract of DOI:10.1039/c9ob00750d

The intramolecular oxa-Michael addition giving tetrahydropyrans has been examined experimentally using both acidic and basic catalysis. With acidic catalysis, the diequatorial product is exclusively obtained in a kinetically controlled reaction in all cas

Full text of DOI:10.1039/c9ob00750d

View Article Online
View Journal
Organic &
Biomolecular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: D. Csokas, A. X. Y.
Ho, R. Ramabhadran and R. W. Bates, Org. Biomol. Chem., 2019, DOI: 10.1039/C9OB00750D.
This is an Accepted Manuscript, which has been through the
Volume 14 Number
1 7 January 2016 Pages 1–372
Royal Society of Chemistry peer review process and has been
accepted for publication.
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-0520
COMMUNICATION
Takeharu Haino et al.
Solvent-induced emission of organogels based on tris(phenylisoxazolyl)
benzene
rsc.li/obc

Page 1 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
View Article Online
DOI: 10.1039/C9OB00750D
Organic and Biomolecular Chemistry
ARTICLE
How an Early or Late Transition State Impacts the
Stereoselectivity of Tetrahydropyran Formation by Intramolecular
oxa-Michael Addition
Dániel Csókás,a Annabel Xuan Ying Ho,a Raghunath O. Ramabhadranb* and Roderick W. Batesa*
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
The intramolecular oxa-Michael addition giving tetrahydropyrans has been examined experimentally using both acidic and
basic catalysis. With acidic catalysis, the diequatorial product is exclusively obtained in a kinetically controlled reaction in all
cases. Under basic conditions at low temperature, the reaction is again under kinetic control, but formation of the axial-
equatorial isomer is generally favoured with an (E)-Michael acceptor, although isomerisation to the diequatorial isomer is
observed at higher temperatures. Computationally, it is found that the acid catalysed reaction has a late transition state and
the kinetic favouring of the diequatorial isomer has a steric explanation. In contrast, under strongly basic conditions, an early
transition state is found. Electrostatic effects are likely to be the main contributor to the stereoselectivity for the (E)-isomer
and steric interactions for the (Z)-isomer.
www.rsc.org/
for example, in Scheme 2.3a When the activating group is a
ketone, then this isomer is always formed under acid catalysis.
Introduction
The intramolecular oxa-Michael reaction is a widely used
method for the formation of the tetrahydropyran (THP) ring
(Scheme 1).1 This method has gained particular utility with the
advent of the alkene cross-metathesis reaction. There are
numerous examples in the literature,2 including a number from
this laboratory.3 The electron withdrawing group employed is
often an ester, a ketone or a thioester. While the reaction is
often base-catalysed, acid catalysis may also be used when the
alkene is activated by a ketone group.3a,c,d,f In general esters do
not cyclise under mild acidic conditions, unless forced.2b
Scheme 2. Acid catalysed formation of the 2,6-cis isomer in a synthesis of diospongin A.
In complete contrast, when the starting Michael acceptor has E-
stereochemistry and basic catalysis is applied at a reduced
temperature, then the major product is likely to be the 2,6-trans
isomer 3a (Scheme 3), an observation first made by Banwell et
al.2k Upon prolonged exposure to base, or use of higher
temperatures, epimerisation occurs via
a retro-Michael-
Michael mechanism, and the product is the 2,6-cis isomer 2a.
This strategy has been commonly used.4 Banwell also made the
contrasting observation that the corresponding Z-isomer 1b of
the Michael acceptor behaves differently, giving the cis isomer
2a of the tetrahydropyran under basic conditions.2k
Scheme 1. The intramolecular oxa-Michael reaction; EWG = electron withdrawing group.
An important question concerns the stereochemical outcome.
If the substrate 1 has a secondary alcohol as the nucleophilic
group, either the 2,6-cis isomer 2 or the 2,6-trans isomer 3 may
be formed. The 2,6-cis isomer 2, with diequatorial substituents
will, of course, be the thermodynamically favoured product as,
a. Division of Chemistry and Biological Chemistry, School of Physical and
Mathematical Sciences, Nanyang Technological University, 21 Nanyang Link,
Singapore 637371.
b. Indian Institute of Science Education and Research (IISER) Tirupati, Transit
Campus - Sree Rama Engineering College, Mangalam P/O, Tirupati - 517507,
Andhra Pradesh, India.
Scheme 3. Base catalysed formation of either isomer
† Footnotes relating to the title and/or authors should appear here.
Electronic Supplementary Information (ESI) available: experimental procedures for
compounds 8-12 and 14-25; 1H and 13C NMR spectra; X-ray crystallographic
structures for compounds cis-20, trans-19 and trans-21; comprehensive literature
survey; computational details. See DOI: 10.1039/x0xx00000x
It is pertinent to ask why the formation of the 2,6-trans isomer
from the E-alkene should be kinetically favoured under basic
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Organic & Biomolecular Chemistry
Please do not adjust margins
Page 2 of 11
ARTICLE
Chemical Science
conditions. Chelation of the counter ion between the alkoxide addressing the question of the effect of the counter cation
View Article Online
DOI: 10.1039/C9OB00750D
oxygen atom and an oxygen atom of the ester group has been under basic conditions.
suggested although such a mechanism would invoke a ten
membered ring containing a trans double bond.5 On the other
hand, in a synthesis of Curvulone B 7, Takahashi employed an
Results and discussion
oxa-Michael addition to an α,β-unsaturated ketone of an
alkoxide generated in situ by cleavage of a TBS ether 6 with
TBAF (scheme 4). This process, in which chelation cannot be a
factor, favoured the formation of the trans isomer and,
therefore, required a second base catalysed equilibration step
(in contrast to a closely related acid catalysed cyclisation3f). One
of the first computational studies of this reaction was focused
on a single complex polycyclic system and, therefore, may not
be general.6 Another early study examined a reaction which can
also be an SN’ process.7 Two studies are remarkably contrasting.
In one of them, by Martín, “net TSs were not detected when the
free enolates were used”,5 thus precluding proper
consideration of whether the counter cation has a role. In the
second, however, the role of the cation was disregarded!2i In
this latter study, the lower energy of the transition state leading
to the trans isomer was attributed to a more favourable angle
of nucleophilic attack. Clarke et al. reported a thorough
experimental and computational study involving a hydroxylated
thioether system.8 Similarly, a careful study conducted as a part
Synthesis of substrates
For the base catalyzed cyclisations, we selected a methyl ester
group to activate the alkene as a Michael acceptor. For the acid
catalyzed cyclisation reactions, we chose a p-bromobenzoyl
group, partly in the hope that the compounds would display
enhanced crystallinity, allowing the use of X-ray
crystallography. For the substrates that would lead to the 2,6-
disubstitution pattern, 1-phenylhex-5-en-1-ol 9 was subjected
to cross metathesis with methyl acrylate and with p-
bromophenylvinylketone 8b to give the two desired substrates,
10a and 10b, without complication (Scheme 5).
Scheme 5. Synthesis of substrates leading to 2,6-disubstitution. Ar = p-BrC6H4.
of
a synthesis of herboxidiene focused on a related
hydroxylated system.2a The -hydroxyl group was found to play
an essential role through hydrogen bonding in both studies.
Another early synthetic study concentrated on substrates with
an oxygen substituent in the allylic position.9 Thus, combined
experimental and computational studies to probe the
stereochemical outcome of the simplest form of this reaction
are yet to be reported and the question of the cause of the trans
selectivity remains open. In addition, to the best of our
knowledge there has been no explanation of the behaviour of
the Z-isomer and, to the best of our knowledge, only a single
study of the acid catalysed reaction.8
For the substrates that would lead to the 2,5-pattern, 2-
phenylhex-5-en-1-ol was prepared in modest yield by ring
opening of styrene oxide with the Grignard reagent derived
from 4-bromobutene (Scheme 6).11 While cross-metathesis
with methyl acrylate proceeded uneventfully, cross-metathesis
with enone 8b provided a very low yield of the desired enone
12b. An alternative and more conventional pathway involving
protection, ozonolysis and a Wittig reaction delivered the
corresponding TBS ether 12c.12
Scheme 4. Oxa-Michael addition in the synthesis of curvulone B with a non-
coordinating counter ion.
The 2,6-substitution pattern has been the most widely
employed. We believe that this reflects the ubiquity of this
structure amongst THP containing natural products, itself a
consequence of biosynthesis from acetate and propionate.
However, in the era of diversity oriented synthesis,10 the
question of whether the same phenomena arise with other
substitution patterns is worthy of investigation. Thus, we set
out to study both the acid and base catalysed reactions using
appropriate substrates and accessing all four substitution
patterns, both experimentally and computationally, as well as
Scheme 6. Synthesis of substrates leading to 2,5-disubstitution. Ar = p-BrC6H4.
To access the substrates that would lead to the 2,4-pattern, allyl
cinnamyl ether 13 was subjected to Wittig rearrangement,
followed by reduction to give alcohol 14a via the corresponding
aldehyde 14c (Scheme 7).13 Again, cross-metathesis with methyl
acrylate proceeded smoothly giving ester 15a, while cross-
metathesis with enone 8b proceeded in disappointing yield. An
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Chemical Science
ARTICLE
ozonolysis-Wittig procedure provided an alternative access to Similar treatment of ketones 12b, 12c, 15b, 17b and 17c gave
View Article Online
the corresponding TBS ether 15c.
the oxa-Michael products (Scheme 9D).OI:T1H0P.103t9r/aCn9sO-1B9007w50aDs
determined to be the 2,5-trans isomer by observation of 10 Hz
and 13 Hz coupling constants in the signals of H6 and H2
respectively. THP cis-20 was found to be the 2,4-cis isomer by
observation of a 15 Hz coupling constant between H2 and H3
and between H3 and H4. Indeed, H3, was found to be a 15 Hz
triplet. THP trans-21 was found to be the 2,3-trans isomer by
observation of a 10 Hz coupling constant between H2 and H3.
The stereochemistry of tetrahydropyrans trans-19, cis-20 and
trans-21
was
subsequently
confirmed
by
X-ray
crystallography.15 All four of these tetrahydropyrans have a
diequatorial substitution pattern and, thus, are in each case the
thermodynamically more stable of the two possible products.
Scheme 7. Synthesis of substrates leading to 2,4-disubstitution. Ar = p-BrC6H4.
Finally, access to the substrates for the 2,3-pattern were
obtained by oxidative ring opening of 4-phenylcyclohexanone
(Scheme 8).14 This yielded 4-phenylhex-5-enoic acid which
could be reduced to the alcohol 16a. Cross metathesis of alcohol
16a with methyl acrylate yielded ester 17a without any
problems, despite the extra hindrance at the allylic position.
Cross-metathesis with enone 8b, however, gave only a poor
yield of the desired ketone 17b. An ozonolysis-Wittig sequence
was employed, as before, to give the corresponding TBS ether
17c.
Scheme 9. Acid catalyzed cyclisation reactions. All reactions were carried out at
room temperature. Ar = p-BrC6H4.
Even though these are the more stable products, it does not
mean that they are formed for this reason. To test this, a 2:1
mixture of the 2,6-cis and 2,6-trans isomers of tetrahydropyran
18 was prepared by base catalysed cyclisation (Scheme 10).
When this mixture was exposed to the same acidic conditions
used for the cyclisation reactions, no change was observed.
Therefore, these molecules do not equilibrate under these
acidic conditions and the preference for the diequatorial
products in the oxa-Michael addition is kinetic.
Scheme 8. Synthesis of substrates leading to 2,3-disubstitution. Ar = p-BrC6H4.
The Acid-catalysed oxa-Michael cyclisation
As a control experiment, ketone 10b was exposed to acid
(Amberlyst-15) in methanol and the product was found to give
exclusively the 2,6-cis isomer cis-18, as expected (Scheme 9).
The stereochemistry was determined by the observation of
typical axial-axial coupling constants of 11 Hz in the signals of
both H2 and H6, the protons α to the ring oxygen atom, in the
1H NMR spectrum.
Scheme 10. Non-equilibration under acidic conditions. Ar = p-BrC6H4.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Organic & Biomolecular Chemistry
Please do not adjust margins
Page 4 of 11
ARTICLE
Chemical Science
Only partial conversion was observed even afterVaienw AerxtictleenOdnleinde
reaction time. This is likely to be due to tDhOeIl:o10w.1s0o39lu/Cb9ilOitBy0o0f75t0hDe
alkoxide in toluene. Using 18-crown-6 with KOt-Bu in toluene
gave complete conversion, and some slight erosion of selectivity
(entry 4). When the base was changed to NaOt-Bu in THF, the
result was very similar to that obtained with KOt -Bu (entry 5).
In contrast however, the combination of NaOt -Bu and 15-
crown-5 gave an increased proportion of the cis isomer (entry
6).17 Use of sodium isopropoxide (entry 7) resulted in a similar
level of stereoselectivity to NaOt-Bu, but a very slow reaction.
No reaction was observed using NaOi-Pr at -78°C.
An obvious set of experiments involve lithium t-butoxide, as Li+
is the most Lewis acidic cation in the group, and thus likely to
promote chelation. When LiOt-Bu in THF was employed, we
found that the reaction temperature had to be raised to -60°C
(entry 8).18 Even at this temperature, the reaction was slow
compared to comparable reactions using K+ and Na+ counter
ions. The low reactivity may be due to the tight binding of the
lithium cation to the alkoxide oxygen.19 The stereoselectivity
was poor, although still favouring the trans isomer. Surprisingly,
no reaction was observed using LiOt-Bu in the presence of 12-
crown-4 (entry 9).
Strong base-catalysed oxa-Michael cyclisation
As Banwell pointed out, “a more extended survey of metal-ion
effects seems warranted”.2k Thus, for the base catalysed
reaction, the role of the counter ion must first be addressed. If
chelation is an important factor, then it would be expected that
the stereoselectivity would be affected by the choice of cation.
Therefore, a range of bases with varied counterions was tested.
Alkoxide bases – principally t-butoxides – were used (Table 1).16
In addition, some of these experiments were then repeated in
the presence of excess of a crown ether. All initial studies were
carried out using ester 10a, giving the cis and trans
tetrahydropyrans 22 (Scheme 11).
The stereochemistry could be determined by analysis of the
1
coupling constants in the H NMR spectra, as for the ketone
analogs. For the isomer cis-22 , the equatorial position of 2- and
6-substituents was apparent from the coupling constants of 11
Hz of the corresponding protons. The coupling constants of H2
in trans-22 were observed to be 3 Hz and 7 Hz, while H6 was
observed to be an apparent triplet with the coupling constant
of 5.7 Hz. These values are consistent with trans
stereochemistry.
With KOt-Bu in THF was used at -78°C, a ratio a little better than
5:1 in favour of the trans isomer was obtained (entry 1). A very
similar result was obtained when this reaction was run in the
presence of 5 equivalents (relative to K+) of 18-crown-6 (entry
2). When the solvent was changed to toluene, a similar result
was again obtained: KOt-Bu in toluene gave the same
stereoselectivity, but a much slower reaction rate (entry 3).
Scheme 11. Base catalyzed formation of the 2,6-disubstitution pattern.20
Table 1. The Intramolecular oxa-Michael addition (10a  22) using alkoxide basesa,b
entry
base
solvent
T/ °C
t/ h
conversion
selectivity
additive
trans:cis
1
2
3
4
5
6
7
8
9
KOt -Bu
KOt -Bu
KOt -Bu
KOt -Bu
NaOt -Bu
NaOt -Bu
NaOi-Pr
LiOt -Bu
LiOt -Bu
THF
THF
toluene
toluene
THF
THF
THF-i-PrOH
THF
-78
-78
-78
-78
-78
-78
-78
-60
-78-RT
2
2
15
15
2
100%
100%
27%
100%
100%
100%
35%
5:1
5:1
5:1
3.3:1
5:1
2:1
5:1
1.4:1
NA
none
18-crown-6c
none
18-crown-6c
none
15-crown-5c
none
none
12-crown-4c
2
15
15
15
78%
0%
THF
a) 0.2 equivalents of the base was used; b) reactions were repeated in triplicate; c) 5 equivalents of crown ether relative to the base was used
counter ion, yet the opposite is observed in comparison of entry
The results thus far indicate no significant chelation effect. Most 8 with entries 1 and 5.
significant is the comparison of entries 1 and 2, showing almost
no difference with and without addition of a crown ether.
Mild base-catalysed oxa-Michael cyclisation
Comparison of entries 1 and 5 shows no difference on changing We were also interested to determine if other bases could be
from K+ to Na+. The results with a lithium counter ion are harder used (Table 2). K2CO3 and Cs2CO3 gave no conversion at RT; a
to interpret as
a different temperature is required. modest conversion was observed using K2CO3 in THF at reflux,
Nevertheless, the chelation argument would demand a higher slightly favouring the cis isomer (entry 1). No reaction was
proportion of the trans isomer with the more Lewis acidic observed using Et3N, pyridine, TMEDA, i-Pr2NEt, proton sponge
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Chemical Science
ARTICLE
or imidazole, even at reflux in THF for up to 15h. Only three oxa-Michael cyclisation of substrates with varied subVsiteiwtuAtrtiioclne Online
DOI: 10.1039/C9OB00750D
nitrogen bases gave
a respectable conversion. Use of pattern
tetramethylguanidine (TMG) in THF resulted in no reaction at
RT, but slow cyclisation at reflux giving a mixture modestly
favouring the trans isomer (entry 2). Employing the
phosphazene P2-t-Bu base, after 15h at -78°C, a 50% conversion
was observed, giving significant trans selectivity (entry 3). With
DBU at RT in either THF or toluene, poor conversion was
observed, but the cis isomer was now the major product
(entries 4 and 5). Using DBU in refluxing THF or in refluxing
It was of interest to see what happened if the substitution
pattern was varied, as was done above in the ketone series. A
series of experiments were carried out to investigate this (Table
3, Scheme 12).
toluene gave complete conversion (entries
6 and 7).
Interestingly, all of these experiments employing DBU delivered
almost the same stereoselectivity with a modest favouring of
the cis isomer. It may be noted that when the trans isomer was
resubmitted to these conditions, no equilibration was observed.
Hence the partial selectivity for the cis isomer using DBU is
kinetic. There is, therefore, a significant difference between
basic conditions using alkoxides and basic conditions using
amidines. The phosphazene, however, behaves more like an
alkoxide. It is notable that, in the case of the phosphazene base,
chelation is once again not feasible, yet the trans isomer is the
major product. This result provides further support for the
suggestion that chelation is not important. The results do
suggest, however, that the transition state involving DBU is
distinctly different. Given the lower basicity of DBU it may be
suggested that no free alkoxide is formed and the DBU molecule
remains closely associated.
Scheme 12. Base catalysed formation with the other disubstitution patterns.20
In the case of ester 12a leading to the 2,5-isomer 23, cyclisation
using either TBAF or KOt-Bu was slower than for the 2,6-isomer.
At -78°C, use of KOt-Bu gave significant selectivity in favour of
the cis isomer (entry 1). As expected, the trans (diequatorial)
isomer become the major product when the reaction was
conducted at RT overnight (entry 2). A similar pattern was
observed with TBAF: the cis isomer was by a small margin the
major product at RT (entry 3); the trans isomer was the
exclusive product when the reaction was conducted at reflux
(entry 4). In trans-23, the equatorial position of the 2-
substituent was verified by a coupling constant of 11 Hz for the
corresponding proton. The equatorial position of the phenyl
group was confirmed by observation of a 11 Hz coupling
constants between the benzylic proton H5 and the axial proton
at C6.
Table 2 The Intramolecular oxa-Michael addition (10a  22) using non-alkoxide basesa,b
entry
base
solvent
T/ °C
t/ h
Conv-
ersion
ND
29%
50%
14%
10%
100%
100%
100%
100%
selectivity
trans:cis
0.7:1
1.4:1
2.5:1
2.5:1
2:1
2:1
2.5:1
2:1
1
2
3
4
5
6
7
8
9
K2CO3
TMG
P2-t-Bu
DBU
DBU
DBU
DBU
TBAF
TBAF
MeOH
THF
THF
THF
toluene
THF
toluene
THF
THF
reflux
reflux
-78
rt
rt
reflux
reflux
rt
15
15
15
15
15
15
15
15
2.5
For ester 15a, leading to the 2,4-isomer 24, use of KOt-Bu in THF
at -78°C gave the trans isomer exclusively (entry 5) – the highest
selectivity that we have observed in the kinetically controlled
series. At RT, the cis product became the major product (entry
6). With TBAF, the cis isomer was the major product at both RT
and at reflux (entries 7 and 8). Turning to ester 17a, leading to
the 2,3 tetrahydropyrans 25, use of KOt-Bu at low temperature
gave useful selectivity in favour of the cis isomer (entry 9). At
RT, the cis isomer remained the major isomer, but by a much
smaller margin. With TBAF, cyclisation was very slow and both
reflux and a higher catalyst loading were required. Again, the cis
isomer was the major product (entry 10). It appears that the
greater steric hindrance provided by having the phenyl group
adjacent to the reacting centre slows down the chemistry.
The observation of a coupling constant of 10.4 Hz indicated the
equatorial position of the 2-substituent in cis-24. Coupling
constants 9.8 Hz and 2.3 Hz showed that the phenyl group was
reflux
cis only
a) 1 equivalent of the base was used; b) reactions were repeated in triplicate.
Mindful of Takahasi’s curvulone synthesis,2e we examined the
use of TBAF (entries 8 and 9). Complete cyclisation was
observed in THF at room temperature, modestly favouring the
trans isomer (entry 8). On the other hand, when this reaction
was carried out in THF at reflux, the cis isomer was exclusively
formed (entry 9). When the trans isomer was re-exposed to
these conditions, no change was observed at room
temperature, but complete conversion to the cis isomer was
observed at reflux. The mixture obtained at RT, therefore,
represents the kinetic distribution of products while fluoride is
sufficiently basic to cause complete equilibration at reflux. The
fact that the use of TBAF gives rise to the trans isomer as the
major product is significant, as, again, no chelation is possible
with the tetra-n-butylammonium counter ion.
1
also in an equatorial position. The H spectrum trans-24 could
not be entirely analysed as the two axial protons flanking the
ring oxygen atom overlapped. However, the equatorial position
of the phenyl group could be confirmed by the characteristic
coupling constants of 8.5 Hz and 3.8 Hz.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Organic & Biomolecular Chemistry
Please do not adjust margins
Page 6 of 11
ARTICLE
Chemical Science
To verify that the TSs obtained indeed do describe the actual
View Article Online
process, the following verifications wereDdOoIn: 1e0:.1039/C9OB00750D
(a) Animation of the imaginary frequency in order to verify that
they genuinely describe the bond formation process for the
oxa-Michael cyclisation.
Table 3 The base catalysed intramolecular oxa-Michael addition for other substitution
patterns
entry
SM,
base
T/ °C
time/ h
selectivity
trans:cis
0.3:1
2.5:1
0.8:1
pattern
12a 2,5
12a 2,5
12a 2,5
12a 2,5
15a 2,4
15a 2,4
15a 2,4
15a 2,4
17a 2,3
17a 2,3
17a 2,3
(b) The connection between the reactant and the product
through a corresponding transition state was further confirmed
by optimizing slightly altered bond distance between the
nucleophilic oxygen and the electrophilic -carbon along the
two directions which were associated with the imaginary
frequency found in the vibrational frequency calculation.
(c) IRC calculations (starting from the TS) which confirmed that
the desired internal cyclisation indeed did happen. These
calculations thus aid us in ascertaining that the transition states
correspond to the investigated intramolecular cyclisation.
Followed by geometry optimization, single point energies were
calculated for all the structures applying the larger 6-
311++G(3df,3pd) basis-set. The ultrafine integration grid was
used to enhance the accuracy of the numerical integration in all
the DFT calculations. The enthalpy was calculated according to
the following formula:
1
2
3
4
5
6
7
8
9
KOt -Bua
KOt -Bua
TBAFb
-78
RT
RT
reflux
-78
RT
RT
reflux
-78
RT
reflux
2
overnight
overnight
overnight trans only
TBAFb
KOt -Bua
KOt -Bua
TBAFb
2
2
trans only
0.7:1
0.7:1
0.5:1
0.1:1
overnight
TBAFb
2
2
3
KOt -Bua
KOt -Bua
TBAFb
10
11
0.6:1
0.6:1
48
a) 0.2 eq. of base; b) 1 eq. of base
The coupling constant of 3.5 Hz was conclusive to verify the Heq-
Hax relationship between H2 and H3 attached to Ph in cis-25.
The coupling constant of 4.4 Hz associated with H3 was
observed as a weighted average of conformers. In the case of
the trans-25, H2 and H3 were found to be axial by observation
of a 10 Hz coupling constant with one another.
Thus, a consistent pattern emerges that for all substitution
patterns, under kinetic conditions, the isomer that is formed is
the axial-equatorial isomer. Under forcing conditions, the major
product is the diequatorial product, due to equilibration. As the
earlier experiments (Tables 1 and 2) indicated that chelation is
not a factor, we were interested to know what was causing the
selectivity under kinetic conditions and, thus, turned to
computational methods.
H(corrected)=E(b97xd/6-311++G(3df,3pd)/SMD)+Thermal
correction to Enthalpy (b97xd/6-311+G(d,p)/SMD). Thermal
corrections to enthalpy were computed at 195.15 K for the
base-catalysed cyclisation and at 298.15 K for the acid-catalysed
cyclisation. Molecular structures were visualized using
CYLview.25
An important technical point to note in this context is, for the acid
catalysed reactions, the computed selectivities using the Gibbs free
energies of activation are in qualitative agreement with those
obtained experimentally i.e. the 2,6-cis isomer is the major product.
However, the computed selectivities for the base catalysed reactions
using the Gibbs free energies of activation results in selectivities
opposite to that obtained experimentally. For instance, the
enthalpies of activation and the free energies of activation in the
stereoselectivity determining 1st step in Figures 1 and 2 (vide infra)
are given below:
Computational Studies
We began the calculations by first obtaining the lowest energy
conformers of the acyclic reactant. For the conformational
analysis of the starting acyclic molecule, either protonated or
deprotonated, we applied Monte Carlo sampling using the
standard OPLS_2005 force field, which was carried out in the
MacroModel software.21 Next, all the distinct conformers within
10 kcal/mol were optimised using Density Functional Theory
(DFT) with the Gaussian 09e suite of programmes.22 The lowest-
energy structure was used as a reference point for constructing
the thermodynamic profile of the cyclisation. After due
calibration using several other density functionals (see S.I. for
more details), the B97X-D density functional was chosen in
conjunction with the 6-311+G(d,p) basis-set.23 The SMD implicit
solvation model was used to model the global solvation
effects.24 Harmonic vibrational frequency calculations were
performed to confirm the nature of the obtained structures. No
imaginary frequencies were found for all the minima reported.
For the transition states (TSs) only one imaginary frequency was
found.
Acid-catalyzed cyclisation:
dr (experimentally): 1:0
H# (computed) = 0.7 kcal/mol  dr (computed): 1:0.25
G# (computed) = 0.4 kcal/mol  dr (computed): 1:0.4
Base-catalyzed cyclisation:
dr (experimentally): 1:0.25
H# (computed) = 0.3 kcal/mol  dr (computed): 1:0.5
G# (computed) = -0.3 kcal/mol  dr( computed): 0.5 1
This does not surprise us given that the error-bars associated with
DFT energies are about 2-3 kcal/mol (the error-bar in energies even
for the celebrated and computationally highly demanding
CCSD(T)/CBS method is about 1 kcal/mol). The computation of
enthalpy and free-energies involves even larger errors. Between
enthalpy and free energies themselves, the error-bar is larger for the
free-energies because of the additional approximations involved in
the computation of the entropy. We suggest that such
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Chemical Science
ARTICLE
approximations affect our computed selectivities more severely in (optimised minima at the (b97xd/6-311+G(d,p)/SMD) level of
View Article Online
the case of the base-catalyzed reaction whose potential energy theory, the O(H)-C bond distance in the DTHOIP: 1b0e.1i0n3g9/1C.95O6BÅ0)0.7T50hDe
surface is flatter relative to the acid catalysed one. Therefore, when final products, 26-cis and 26-trans, are obtained via a simple
calculating selectivities (which are obtained by exponentiating the proton transfer from the THP oxygen to, most likely, a solvent
energies or free-energies or enthalpies), whose magnitudes are molecule (methanol).26 The relative enthalpies of the final
similar to those experimentally obtained in this work, it is to be products (neutral) relative to the reactant (positively charged)
expected that DFT calculations sometimes give values not in plus the solvent molecule (methanol) are -7.0 and -5.2 kcal/mol
concordance with the experiments. This forms the basis for why we (for neutral cis and trans 18 respectively). Note that in Figure 1
present the enthalpy profile in Figures 1 and 2, and do not use the we only show the stereoselectivity determining step (1st step)
free-energies.
leading to the formation of 1,3-cis and 1,3-trans, and we do not
include how the final neutral products are formed. This is
because there are several possibilities for the associated proton
transfer/keto enol tautomerism, and computations cannot help
here in deciphering how exactly they proceed nor would they
be expected to have any impact on the step in which the
stereochemical outcome is determined. Regardless, the main
Calculations for the acid catalysed cyclisation
The activation enthalpies of the acid-catalysed cyclisation were
calculated to be 4.8 kcal/mol for the transition state 1.1-trans
leading to the trans isomer and 4.1 kcal/mol for the
corresponding transition state 1.1-cis, leading to the cis isomer
(Figure 1). This is consistent with the observed experimental
diastereoselectivity. We found that the O-protonated products
formed, 1.2-cis and 1.2-trans, were stable intermediates
objective of our work
-
trying to understand the
stereoselectivities is unambiguously shown in Figure 1.
Figure 1. Enthalpy profile for the acid-catalysed oxa-Michael cyclisation for the formation of 1,3-cis and 1,3-trans.
Inspection of TS structures, 1.1-cis and 1.1-trans, reveals
that the forming rings in both TSs for the acid-catalyzed
cyclisation are orderly with chair-like conformations. Note
that the intramolecular approaches of the nucleophiles
follow very similar trajectories. In the case of both transition
states, the trajectories are not very different from the Bürgi–
Dunitz trajectory (about 2 degrees departure from the
tetrahedral angle in 1.1-trans and 5 degrees in 1.1-cis –
Figure 1). Hence, the stereoselectivities cannot be
rationalized on the basis of the Bürgi–Dunitz trajectories
alone. More significantly, in both transition states, 1.1-cis
and 1.1-trans, the OH…C(β) distance was shortened down to
1.9 Å, starting from a considerably larger distance in the
starting material, whereas, in the final enol structures, the
OH…C(β) distances are around 1.56 Å. This reveals that the
acid-catalyzed cyclisation takes place via a late-transition
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Organic & Biomolecular Chemistry
Please do not adjust margins
Page 8 of 11
ARTICLE
Chemical Science
state as opposed to the base-catalyzed cyclisation (vide
infra). At the TS, the carbon atom  to the carbonyl group
was found to display extensive pyramidalization (16.4° for
1.1-cis).27 This is also consistent with a late transition state,
as the -carbon approaches sp3 hydridisation.
unable to handle the co-ordination environment with
precision.
Calculations for the base catalysed cyclisation
View Article Online
DOI: 10.1039/C9OB00750D
As our experimental results have clearly shown that the role
of the cation is not significant in controlling the
stereochemistry (vide supra), no cations were included in
our computations. As can be seen in Figure 2, the transition
state 2.1-trans is 0.3 kcal/mol lower than the transition state
2.1-cis, which adequately reflects the observed
diastereoselectivity. Once bond-formation took place and
the products were formed the relative stability became the
opposite. In contrast to the trend in the TSs, enolate 2.2-cis
was found to be more stable than enolate 2.2-trans by 0.9
kcal/mol. This is aligned with general chemical intuition and
can be accounted for by the resultant 1,3-diaxial repulsion in
the 2,6-trans THP ring. This also indicates that the applied
computational method adequately predicts the relative
stability of the entities in the case of base-catalyzed
cyclisation as well. In this context, it may be noted that the
geometry of the Michael acceptor greatly influences the
order of the TSs. It was found that the transition states 3.1-
cis and 3.1-trans in which the alkene and carbonyl group
have an s-cis arrangement were lower in energy by 0.2 and
0.8 kcal/mol, respectively, compared to the corresponding
transition states with an s-trans relationship. This may be
due to 1,3-allylic strain between the methoxy group and the
proton at the -position (Figure 3).
As Clarke et al. have pointed out,8 there has been no tangible
explanation for the cis preference in the acid-catalysed oxa-
Michael cyclisation. In their study, the energy difference
between the trans and the cis TS was rationalized to be a
result of an increased pseudo-1,3 diaxial steric clash
between the hydrogen and the electrophilic Michael
acceptor in the trans TS. This is in full agreement with our
computational findings and explanation, however, this is
only the case because of the relatively short bond-forming
distance in the TSs and the consequent importance of the
late-transition state. This, we believe, is the essence of
understanding the origin of the stereocontrol for the acid
catalysed reaction.
A noteworthy comment in this context is that, the terms
“late” or “early” in this paper are meant to describe only the
stereoselctivity determining step, and not the overall energy
profile including proton-transfers and tautomerizations.
Lastly, we must also point out a possible limitation of the
theoretical model we use: experimentally, it is evident that
the acid catalysed reaction (Scheme 9) is under kinetic
control. However, the difference in the computed activation
enthalpies in the stereoselectivity determining step (Figure
1) is only 0.7 kcal/mol (which in principle is expected to be
much large). This could perhaps be due to our model being
Figure 2. Enthalpy profile for the base-catalysed cyclisation for the formation of 2.2-cis and 2.2-trans
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
View Article Online
DOI: 10.1039/C9OB00750D
Organic and Biomolecular Chemistry
ARTICLE
Notably, the C•••O distances in all the transition states
herein are long: 2.3-2.4 Å. This indicates an early transition
state (bond length(TS-product) = 0.9 Å, see Figure 2), which is
consistent with the previously reported transition states.2i, 6
Modest pyramidalization (7.3° in 2.1-trans) of the carbon
atom  to the carbonyl group in the TS is consistent with such
an early TS. In comparison with the products, the early TSs in
the base-catalyzed cyclisation are evidently much farther
from the ideal chair conformation and also from the
conformations identified in the acid-catalyzed TSs (Figure 1).
Based on the long O(-)…C(β) distance and the experimental
outcome, the 1,3-diaxial effect is arguably insignificant
under kinetic control in the base catalysed cyclisation. This
early transition state clearly explains why the cis product is
not kinetically favoured. The question of why the trans
product is favoured, however, still remains.
trans remained the lowest energy transition state (Figure 4).
Thus, the computational and experimental results are in
agreement that the cation itself, which was suggested to be
the reason for the trans selectivity due to chelation effects,
does not play a crucial role in the selectivity.
Figure 4. 2,6-cis and 2,6-trans transition states in the chelation model with
potassium and the enthalpy difference (in kcal/mol) with respect to one
other.
In a final attempt to understand the reason for trans
selectivity under basic conditions, we examined electrostatic
factors. In the transition states, 2.1-cis and 2.1-trans, leading
to both the cis and trans isomers, the alkoxide oxygen atom
and the carbon  to the carbonyl group bear charges of -
0.88 and -0.45 or -0.46 respectively (NPA charges) (Figure 5).
An electrostatic repulsion is, therefore, to be expected. The
only significant difference between the two transition states
appears to be the distance between the alkoxide oxygen and
the -carbon. In the TS leading to the trans isomer, 2.1-
trans, this distance is 0.1 Å longer than in 2.1-cis, implying
correspondingly less destabilisation of this transition state
by electrostatic repulsion. The electrostatic repulsion
between the carbonyl oxygen (NPA charges on oxygen: 2.1-
cis: -0.68, 2.1-trans: -0.67) and the alkoxide oxygen also
seems to play a role in a similar manner as, again, the atoms
in question are further apart in the transition state leading
to the trans product (4.06 vs 3.98 Å).29 Thus, we propose that
this electrostatic interaction may be one factor in the kinetic
preference for the trans isomer, although additional factors,
too small to be apparent, may also contribute.
Figure 3. Transition states with s-trans Michael acceptors having 1,3-
allylic strain and their relative enthalpies (in kcal/mol) with respect to SM-
anion.
As part of answering this question, we realized that the
Bürgi–Dunitz trajectory is closer to the tetrahedral angle in
the transition state 2.1-cis leading to the cis isomer.28 This
stands in contrast to the findings of the Schneider2i and the
Yonemitsu6 groups. We also examined the possibility of
hyperconjugative interaction between the equatorial C()-H
and the double bond in both TSs.9 However, we did not find
any significant change in the bond length neither in
comparison to the other two equatorial C-H bonds, nor
between one another.
To see if any local orbital interactions played a role in
determining selectivity, we performed NBO calculations.
However, the second-order perturbative estimates (E(2)
energy) of donor-acceptor interactions from NBO
computations showed no tangible difference in the two
investigated TSs, thereby ruling out such effects.
Wondering if, perhaps, these observations are mere
artefacts of the fact that we did not include any cation in our
calculations, we included the alkali metal cation also in our
calculations. We found that the 2,6-trans arrangement 4.1-
Figure 5. Electrostatic interactions in the TS of the base catalysed reaction. Distances
in Å.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Organic & Biomolecular Chemistry
Please do not adjust margins
Page 10 of 11
ARTICLE
Chemical Science
Note that we did not perform a similar electrostatics-based
analysis for the acid-catalyzed transition states. That is
because, with the acid catalyzed reactions, the classic Barton
model30 (invoking 1,3-diaxial interactions), adequately
explains the selectivity, whereas it is with the base catalysed
reaction that the Barton model is inadequate in explaining
the selectivities.
View Article Online
DOI: 10.1039/C9OB00750D
Further, given the ~13 kcal/mol energy barrier of the reverse
reaction, it is very unlikely the isomerisation into the cis
isomer noticeably affects the overall selectivity at -78°C
within the timeframe of our experiments. However, the
isomerization reaction, which arises from the collapse of the
enolate to afford the acyclic form, becomes rapidly
predominant at higher temperature resulting in reverse
selectivity. As can be seen in the enthalpy profile, the overall
thermodynamic product of the cyclisation is the 2,6-cis
enolate 2.2-cis, which is 0.9 kcal/mol lower in energy in than
the corresponding 2,6-trans isomer 2.2-trans. These
experimental and computational results are in accord with
the idea that the cyclisation is under kinetic control at low
temperature, particularly at -78°C, whereas thermodynamic
control governs the stereochemical outcome at higher
temperature, commonly above 0°C.
Such a change in the selectivity pattern at elevated
temperature is easy to understand as occurring due to the
unfavourable 1,3-diaxial interaction present in the trans
product, for instance in the case of 2,6-tetrahydropyrans.
This seems to be the major driving force for the
isomerisation. The most striking aspect of this phenomenon
is that it appears general irrespective of other substituents
on the rest of the ring when a strong base is used (see SI).
Figure 6. Transition states in the base catalysed cyclisation of the (Z)-Michael
acceptor.
Conclusions
In the case of the cyclisation under mildly acidic conditions,
the high preference for the formation of the 2,6-cis isomer is
a kinetic phenomenon that arises from a late transition state
with a relatively short C-O distance. The same concept holds
true for the other substitution patterns, each delivering the
diequatorial product.
Base catalysed cyclisation is more complex. The preference
for formation of the 2,6-trans isomer from the (E)-Michael
acceptor at lower temperatures is not due to chelation.
Experiments involving varying the alkali metal cation and
adding crown ethers, as well as employing TBAF and
phosphazene bases, rule this out. That the 2,6-cis isomer is
not favoured, as it is in the acid catalysed cyclisation, is
explained by the early transition state with a relatively long
C-O bond eliminating the significance of the classical 1,3-
interactions. From inspection of the transition states it is not
possible to identify a single effect that clearly controls the
stereoselectivity of the reaction. Thus, the stereoselctivity
appears to arise from a combination of minor factors
working in concert, difficult to ascertain by theory or
experiment. Amongst these effects, electrostatic repulsion
between the -carbon and the alkoxide oxygen, based upon
the interatomic distances and the NPA charges, is likely to
play a particular role. The same pattern holds for the three
other isomers with the 2,5-, 2,4- and 2,3-substitution
patterns, each giving mainly the equatorial-axial products
under basic kinetic conditions.
In contrast, for the (Z)-Michael acceptor, a clear 1,3-
interaction strongly disfavours the transition state leading to
the 2,6-trans isomer, leaving the 2,6-cis isomer as the
preferred product.
In this study, a comprehensive analysis of the literature
precedents was carried out, which led us to recognize
distinct selectivity patterns. However, in the absence of a
focused study comparing the stereoselectivities under
different conditions, a clear picture had not emerged
pertaining to the origin of the selectivity. We hope that our
contribution brings together all these stereo-determining
The case of the (Z)-alkene
The remaining issue to be addressed is why the
corresponding (Z)-isomer 1b of the Michael acceptor
delivers the cis-tetrahydropyran 2a, even under kinetic
conditions.2k This question appears to have been largely
avoided up until now. With a verified model for the TS for
the (E)-alkenes, we were able to examine this issue by
calculating the corresponding transition states for the
cyclisation of the (Z)-alkene (Figure 6). Again, early transition
states were found, as judged by the length of the forming C-
O bond (2.22 and 2.25 Å). The transition state 6.1-trans
leading to the trans isomer 3a was found to be very
significantly disfavoured compared to the transition state
6.1-cis that leads to the cis isomer 2a. This can be easily
ascribed to a severe 1,3-interaction between the ester group
and the pseudo-axial hydrogen atom at position 4 (THP
numbering;  relative to the carbonyl). Indeed, the distance
between them is within the sum of their Van der Waals radii.
The severity of this clash is in accordance with the very high
diastereoselectivity (97:3) reported by Banwell.2k
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 11 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Chemical Science
ARTICLE
9
B. W. Gung, M. B. Francis, J. Org. Chem. 19V9ie3w, A5r8tic,le6O1n7li7ne-
factors. Equally importantly, we hope that the present study
will provide guidance for THP synthesis as well.
6179.
DOI: 10.1039/C9OB00750D
10 S. L. Schreiber, Science 2000, 287, 1964-1969.
11 For an example of the C2 opening of styrene oxide, see G.
M. Whitesides, J. D. Roberts, J. Am. Chem. Soc. 1965, 87,
4878-4888.
12 We have previously shown that TBS deprotection and oxa-
Michael addition can occur in tandem fashion under these
conditions, see refs 3a, c and f.
Conflicts of interest
There are no conflicts to declare.
13 N. Greeves, W.-M. Lee, Tetrahedron Lett. 1997, 38, 6445-
6448.
Acknowledgements
14 E. K. Starosin, D. B. Furman, A. V. Ignatenko, A. P. Barkova,
G. I. Nikishin, Russ. Chem. Bull. 2006, 55, 2016-2019.
15 Details have been deposited with the Cambridge
Crystallographic Data Centre and may be obtained at
http://www.ccdc.cam.ac.uk; CCDC deposition numbers:
trans-19 1895903; cis-20 1895904; trans-21 1895902.
16 We also examined hexamethyldisilazide bases, but we
obtained erratic results. This is likely to be due to
aggregation of these bases: R. E. Mulvey, S. D. Robertson,
Angew. Chem. Int. Ed. 2013, 52, 11470-11487.
RR thanks IISER Tirupati, and SERB, DST, Govt. of India for the
Early Career Research Award (ECR/2016/000041); DC, AHXY
and RWB thank Pfizer and Nanyang Technological University
for generous support of this work; AHXY thanks the
Economic Development Board and Pfizer for an EDB IPP
Scholarship. We thank Dr. Li Yongxin and Dr. Rakesh Ganguli
(NTU) for determination of the X-Ray structures.
17 It is possible that the basicity of the alkoxides is enhanced
by the sequestration of the counter cation by the crown
ether, so the increased proportion of cis isomer may be
due to equilibration.
18 We also found the behaviour of this base to be capricious.
Commercial material was inactive, while material
prepared in our own laboratory had to be used promptly.
19 We have previously observed this effect in a different
context: R. W. Bates, C. J. Gabel, J. Ji, T. Rama-Devi,
Tetrahedron, 1995, 51, 8199-8212.
Notes and references
1
For reviews of intramolecular oxa-Michael addition, see (a)
J. Hu, M. Bian, H. Ding, Tetrahedron Lett. 2016, 57, 5519-
5539; (b) N. M. Nasir, K. Ermanis, P. A. Clarke, Org. Biomol.
Chem. 2012, 12, 3323-3335 (c) H. Fuwa, Heterocycles 2012,
86, 1255-1298; (d) C. F. Nising, S. Bräse, Chem. Soc. Rev.
2012, 41, 988-999; (e) I. Larossa, P. Romea, F. Urpí,
Tetrahedron 2008, 64, 2683-2723; (f) P. A. Clarke, S.
Santos, Eur. J. Org. Chem. 2006, 2045-2053.
(a) A. Gómez-Palomino, M. Pellicena, K. Krämer, P. Romea,
F. Urpí, G. Aullón, J. M. Padrón, Org. Biomol. Chem. 2017,
15, 1842-1862; (b) L. Yang, Z. Lin, S.-H. Huang, R. Hong,
Angew. Chem., Int. Ed. 2016, 55, 6280-6284; (c) S. J.
Gharpure, J. V. K. Prasad, K. Bera, Eur. J. Org. Chem. 2014,
3570-3574; (d) I. Paterson, G. W. Haslett, Org. Lett. 2013,
15, 1338-1341; (e) S. Takahashi, Y. Akita, T. Nakamura, H.
Koshino, Tet. Asymmetry 2012, 23, 952-958; (f) H. Fuwa, N.
Ichinokawa, K. Noto, M. Sasaki, J. Org. Chem. 2012, 77,
2588-2607; (g) C. Bressy, F. Allais, J. Cossy, Synlett 2006,
3455-3456; (h) A. Bhattacharjee, O. Soltani, J. K. De
Brabander, Org. Lett. 2002, 4, 481-484; (i) C. Schneider, A.
Schuffenhauer, Eur. J. Org. Chem. 2000, 73-82; (j) S. V. Ley,
D. S. Brown, J. A. Clase, A. J. Fairbanks, I. C. Lennon, H. M.
I. Osborn, E. S. E. Stokes, D. J. Wadsworth, J. Chem. Soc.,
Perkin Trans. 1 1998, 2259-2276; (k) M. G. Banwell, C. T.
Bui, H. T. T. Pham, G. W. Simpson, J. Chem. Soc., Perkin
Trans. 1 1996, 967-969. For a comprehensive listing and
discussion, see the S.I.
20 Note: where two conformers are likely to be of
significance, only one is shown for clarity.
2
21 MacroModel; Schrödinger, LLC: New York, NY, 2016.
22 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.
Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J.
Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V.
N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N.
Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V.
Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J.
W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D.
Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski,
D. J. Fox, Gaussian 09, Revision E.01; Gaussian, Inc.:
Wallingford CT, 2009.
23 (a) J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys.
2008, 10, 6615−6620. (b) J.-D. Chai, M. Head-Gordon, J.
Chem. Phys. 2008, 128, 084106.
24 A. V. Marenich, C. J. Cramer, D. G. Truhlar, J. Phys. Chem. B
2009, 113, 6378−6396.
25 C. Y. Legault, CYLview, 1.0b; Université de Sherbrooke,
2009 (http://www.cylview.org).
26 As all acids appear to give the same stereochemical
outcome, the role of the counter ion was disregarded.
27 See the S.I. for the definition of pyramidalisation.
28 The importance of the Bürgi-Dunitz angle in intramolecular
cyclisations has been questioned: C. I. Bayly, F. Grein, Can.
J. Chem. 1989, 67, 2173-2177.
29 We thank a reviewer for the suggestion of repulsion
between the oxygen atoms.
30 D. H. R. Barton Experientia 1950, VI/8, 316.
3
(a) R. W. Bates, P. Song, Tetrahedron, 2007, 63, 4497-4499;
(b) R. W. Bates, K. Palani, Tetrahedron Lett., 2008, 49,
2832-2834; (c) R. W. Bates, P. Song, Synthesis, 2010, 2935-
2942; (d) R. W. Bates, T. G. Lek, Synthesis 2014, 1731-1738;
(e) R. W. Bates, L. Li; K. Palani; W. Phetsang, J. K. Loh, Asian
J. Org. Chem. 2014, 7, 792-796; (f) R. W. Bates, K. Wang, G.
Zhou, D. Z. Kang, Synlett 2015, 751-754; (g) D. Csókás, R.
W. Bates, Synlett 2019, 30, 178-180.
4
5
For an example, see ref 2h.
M. A. Ramírez, J. M. Padrón, J. M. Palazón, V. S. Martín, J.
Org. Chem. 1997, 62, 4584-4590.
6
7
8
O. Yonemitsu, T. Yamazaki, J.-i. Uenishi, Heterocycles 1998,
49, 89-92.
R. Takagi, H. Nishitani, S. Takenami, K. Okada, S. Kojima, K.
Ohkata, Bull. Chem. Soc. Jpn. 2001, 74, 1901-1907.
K. Ermanis, Y.-T. Hsiao, U. Kaya, A. Jeuken, P. A. Clarke,
Chem. Sci. 2017, 8, 482-490.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins