C–O cleavage via InIII alkoxide intermediates: In situ 13C NMR analysis of the mechanism of an enantioselective in-mediated cyclopropanation reaction

Article abstract of DOI:10.1016/j.tet.2020.131786

The mechanism of asymmetric cyclopropanation of dibenzylideneacetone and benzylideneacetone by in situ generated allyl indium reagents in the presence of methyl mandelate as a chiral modifier has been studied by in situ ¹³C{¹H} NMR in conjunction with ¹³C/²H labelling and mass spectrometry. Two indium alkoxides were identified, the first arising from indium mediated allylation of the ketone, the second arising from reaction of an in situ liberated homoallylic via a LiI mediated reaction with excess allyl indium reagent. On acidification, protonation at oxygen induces C–O rather than In–O cleavage and the incipient tertiary allylic cation is stereoselectivly allylated with approximately 90% si selectivity, via what is assumed to be a mandelate-chelated indium allyl reagent.

Full text of DOI:10.1016/j.tet.2020.131786

Tetrahedron xxx (xxxx) xxx
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
CeO cleavage via InIII alkoxide intermediates: In situ 13C NMR analysis
of the mechanism of an enantioselective in-mediated
cyclopropanation reaction
, b, *
Jennifer L. Slaughter a, Guy C. Lloyd-Jones a
a School of Chemistry, Cantock’s Close, Bristol University, Bristol, BS8 1TS, UK
b School of Chemistry, University of Edinburgh, Edinburgh, EH9 3FJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The mechanism of asymmetric cyclopropanation of dibenzylideneacetone and benzylideneacetone by in
situ generated allyl indium reagents in the presence of methyl mandelate as a chiral modifier has been
studied by in situ 13C{1H} NMR in conjunction with 13C/2H labelling and mass spectrometry. Two indium
alkoxides were identified, the first arising from indium mediated allylation of the ketone, the second
arising from reaction of an in situ liberated homoallylic via a LiI mediated reaction with excess allyl
indium reagent. On acidification, protonation at oxygen induces CeO rather than IneO cleavage and the
incipient tertiary allylic cation is stereoselectivly allylated with approximately 90% si selectivity, via what
is assumed to be a mandelate-chelated indium allyl reagent.
Received 14 October 2020
Received in revised form
12 November 2020
Accepted 16 November 2020
Available online xxx
Keywords:
Indium
Stereoselective reactions
Mechanism
Crown Copyright © 2020 Published by Elsevier Ltd. All rights reserved.
In situ NMR
13C-labelling
1. Introduction
forms, to be an effective modifier for the process [8]. After opti-
misation of the conditions, enone 2a could be converted into
Pioneering work by Butsugan and co-workers [1] around three
decades ago, led to allyl indium reagents (1X), prepared in situ from
allyl halides (allyl-X) and In metal [2], becoming popular reagents
for organic synthesis [3,4]. Although a rather ill-defined mixture of
cyclopropane 4a with up to 94/6 er. Three factors were found to be
essential for efficient and selective cyclopropanation of 2a: i) gen-
eration of the allyl indium mixture (1I) from allyl iodide, ii) use of
ꢀ2 equiv. of 5(H), and iii) addition of LiI prior to the aqueous HCl
work-up. By coupling the homoallyl vinylcyclopropanation with
Ru-catalysed ring closing metathesis [9], the methodology pro-
vided an enantioenriched bicyclo[4.1.0]hept-2-ene (6a) from a
simple enone subtrate [8].
species (1X) is generated, comprising InI and InIII
m-halide-bridged
dimers and aggregates [3], the reagents effect a range of mild and
chemoselective allylation processes. For example, they undergo
highly [1,2]-selective addition to
a,b-unsaturated ketones (2),
affording the corresponding allylic alcohols (3) in good yield after
acid work-up (aq. HCl) [1]. In 1998 we reported that if enone 2a
For further development of this intriguing asymmetric reaction
(2a/4a), mechanistic insight is pivotal. However, many aspects of
the process were unclear, including i) the sequence of generation of
bonds A, B and C (Scheme 1), ii) the roles of the modifier 5(H), the
LiI, and the acid, and the requirement for their addition in the
specified order, and iii) why two equivalents [10], of 5(H) and an
excess of 1x are required. Herein we report on a13C/2H-labelling/in
situ 13C NMR/MS study that allows rationalisation of these prior
observations [8], and reveals that homoallylic alcohols (3) are key
reaction intermediates. This in turn allows us to elucidate the ste-
reochemistry attending generation of bonds B and C; a process in
which modifier 5(H) is shown to act intermolecularly. These in-
sights will be of utility for design of new InIII-mediated, C-X bond-
(R ¼
b-styryl, Scheme 1) is added to an excess of 1Br, and the work-
up modified by addition of LiBr then aq. HCl, then instead of alcohol
3a being isolated, the reaction affords homoallyl vinylcyclopropane
4a [5]. In other words there is an overall deoxygenative addition of
two allyl units to the enone 2a to generate a cyclopropane [6,7]. By
screening a range of simple enantiopure chiral additives, we
identified methyl mandelate (5(H)), readily available in both R and S
* Corresponding author. School of Chemistry, University of Edinburgh, Edinburgh,
EH9 3FJ, UK.
E-mail address: guy.lloyd-jones@ed.ac.uk (G.C. Lloyd-Jones).
https://doi.org/10.1016/j.tet.2020.131786
0040-4020/Crown Copyright © 2020 Published by Elsevier Ltd. All rights reserved.
Please cite this article as: J.L. Slaughter and G.C. Lloyd-Jones, CeO cleavage via InIII alkoxide intermediates: In situ 13C NMR analysis of the
mechanism of an enantioselective in-mediated cyclopropanation reaction, Tetrahedron, https://doi.org/10.1016/j.tet.2020.131786

J.L. Slaughter and G.C. Lloyd-Jones
Tetrahedron xxx (xxxx) xxx
4a, Scheme 2. The three contiguous 13C labels also provide infor-
mation regarding changes in hybridisation and electronegativity of
1
the substituents at C(1) (via modulation of JCC) [11], as well as
identifying intermediates in which coordination to indium induces
the two 13C-labelled carbons adjacent to C(1) to become
diastereoisotopic.
13C{1H} NMR spectroscopic analysis (see SI) of a D8-THF solution
of allyl indium reagent 1I, immediately after addition of enone 13C3-
2a (C(1), t, 187 ppm, 1JCC ¼ 56 Hz) directly to the NMR tube under
N2, the showed that enone 2a is completely converted to a new
species (C(1), t, 80 ppm, 1JCC ¼ 49 Hz) in which C(1) is sp3 hybridised
and bound to an electronegative atom, i.e. oxygen. This first inter-
mediate (13C3-7a) is thus assigned as an indium alkoxide, arising
from a standard In-mediated [1,2]-allylation [1] of the enone [12].
Addition of LiI to indium alkoxide 13C3-7a led to a second indium
Scheme 1. Preparation of bicyclo[4.1.0]hept-2-ene (S,S)-6a via asymmetric indium-
mediated homoallyl cyclopropanation (2a/(S)-4a; R ¼
b-styryl) then ring-closing
metathesis [8]. Conditions: (i) In (4 equiv.), allyl iodide (6 equiv), (S)-5(H) (2.0
equiv), THF, RT; ii) Et2O, 1 M HCl, RT; iii) LiI (1 equiv.), 40 min, RT; (iv) 5 mol %
RuCl2(PCy3)2]CHPh, CH2Cl2, RT. Overall yield 44%.
1
alkoxide intermediate 13C3-8a (C(1), t, 78 ppm, JCC ¼ 48 Hz). In
both 13C3-7a and 13C3-8a, the carbons adjacent to C(1) are inequi-
valent (DdC ¼ 1 and 2 ppm, respectively), indicative of diaster-
eotopicity at these sites, induced by the other ligands (allyl, I, THF)
present on indium [13]. On addition of 1 M HCl (aq), intermediate
13C3-8a immediately generated the cyclopropane 13C3-4a (C(1), t,
30 ppm, 1JCC ¼ 56 Hz), by acid-mediated reaction with the excess
indium allyl reagent 1I.
cleaving, and CeC bond-forming reactions.
2. Results and discussion
In situ 1H NMR spectroscopic analysis of the mechanism of the
cyclopropanation reaction (2a/4a) was not productive, due to
very broad and uninformative signals arising from the dynamic
aggregation of indium intermediates. We thus switched to 13C{1H}
NMR spectroscopy and employed triply-13C-labelled enone [13C3]-
2a, readily prepared from commercially-available [13C3]-acetone, to
allow selective analysis of the enone-derived component
throughout the reaction. Using this technique we were able to
unambiguously track the carbonyl carbon, C(1), in 2a as a triplet,
that migrates a remarkable 157 ppm upfield on conversion of 2a to
13C{1H} NMR spectroscopic analysis of the same sequence, but
conducted in the presence of modifier (S)-5(H) revealed that a
different species (13C3-3a) is generated immediately after addition
of the allyl indium reagent 1I. In this new species, C(1) is 5 ppm
upfield (t, 75 ppm, 1JCC ¼ 49 Hz) of indium alkoxide 13C3-7a and the
adjacent carbons are not diastereotopic, suggesting that the oxygen
at C(1) is not bound to indium. In a subsequent step, that was
markedly accelerated by the LiI, 13C3-3a reacted with excess allyl
Scheme 2. In situ 13C{1H} NMR spectroscopic study of the reaction of dienone [13C3]- 2a with allyl indium species (1I) to generate cyclopropane 4a and tetraene 9a, with addition of
two equivalents of enantiopure methyl mandelate (S)-5(H) as a chiral modifier. The allyl indium reagent, 1I, is generated in situ from In (4 equiv.) þ allyl iodide (6 equiv) and is a
mixture of InI/InIII species; L ¼ unspecified ligands: iodide/
m
-iodide, allyl, alkoxide, solvent (D8-THF).
2

J.L. Slaughter and G.C. Lloyd-Jones
Tetrahedron xxx (xxxx) xxx
Scheme 3. Comparison of the asymmetric cyclopropanation of enones 2a,b and homoallylic alcohols 3a and ( )-3b. Conditions (i) In (4 equiv.), allyl iodide (6 equiv), (S)-5(H) (2.0
equiv), THF, RT; ii) LiI (1 equiv.), 40 min, RT; iii) Et2O, 1 M HCl, RT.
Scheme 4. Resolution of ( )-3b and assignment of configurations by convergent synthesis. The relative configuration of intermediate 14 established by X-ray crystallography [17].
Conditions: (i) preparative chiral HPLC, Chiralcel OD; (ii); 3b/15, EtOH, Pd/C, H2; 85% yield. (iii) allyl-TMS, DCM, ꢂ78 ꢁC, TfOH, 48h, then NEt3; 50% yield; (iv) 14, THF, ꢂ78 ꢁC, NH3, N2,
Na (2.5 equiv.); then MeOH; 94% yield.
indium reagent 1 to generate the same alkoxide (13C3-8a) as is
observed in the absence of modifier (S)-5(H), after reaction with LiI/
1I. On addition of HCl (1 M), cyclopropane (S)e13C3-4a was again
immediately generated, and chiral HPLC analysis showed it to be of
94/6 er. The intermediate species 13C3-3a, was subsequently iden-
tified as the homoallylic alcohol. This arises from rapid protonation
of the incipient and unobserved indium alkoxide (13C3-7a) by the
alcohol moiety of the modifier (S)-5(H). Addition of (S)-5(H) to an
in-situ generated sample of indium alkoxide 13C3-7a confirmed
that protonation to generate alcohol 3a is rapid, and thus the
equilibrium between the alkoxides (13C3-7a and (S)-5-In) strongly
favours the latter, presumably due to chelation of indium by the a-
carbonyl. Both the OH and the ester were previously found to be
essential components for selectivity using modifier (S)-5(H) [8].
13C{1H} NMR spectroscopic analysis of the reaction of inde-
pendently prepared alcohol 13C3-2a with LiI/1I confirmed that this
also generates indium alkoxide 13C3-8a (1 h at 60 ꢁC) and then
racemic cyclopropane ( )-13C3-4a, on addition of 1 M HCl. The
same reaction conducted with stoichiometric modifier (S)-5(H)
gave (S)e13C3-4a in 93/7 er, together with a trace of a tetraene, 13C3-
9a. Acidification before conversion of the alcohol 3C3-3a to indium
alkoxide 3C3-8a, resulted solely in elimination to generate the tet-
raene 13C3-9a and no cyclopropane 13C3-4a; confirming that it is the
indium alkoxide 8a, not the alcohol 3a, that is the active interme-
diate for cyclopropane generation.
The above experiments establish that the first stage in the
overall cyclopropanation process is generation of bond A (Scheme
1), through a conventional indium-mediated [1,1,2]-allylation of
the enone (2a). Indium(I)-allyl species are generally accepted as
being the most active for carbonyl allylation [3,14], implying that
intermediate 7a is an indium(I) alkoxide, as well as the modifier-
derived alkoxide (S)-5(InI).
Scheme 5. Match/mismatch selectivity with the modifier (S)-5(H) in the cyclo-
propanation of homoallylic alcohol 3a and chiral modifier effect of (S)-(ꢂ)-3b on the
reaction of ( )-2H4-3b. Conditions (i) In (4 equiv.), allyl iodide (6 equiv), THF, RT with
(S)-5(H) (2.0 equiv) if indicated; ii) LiI (1 equiv.), 40 min, RT; iii) Et2O, 1 M HCl, RT.; iv) In
(8 equiv.), allyl iodide (12 equiv), (S)-(ꢂ)-3b (1 equiv.) THF, RT, with selective analysis of
2H4-4b in presence of 4b by chiral-HPLC MS.
Both the indium (I) alkoxide 7a and alcohol 3a undergo LiI-
accelerated reaction with allyl indium reagents 1I to generate an
3

J.L. Slaughter and G.C. Lloyd-Jones
Tetrahedron xxx (xxxx) xxx
Scheme 6. Generic scheme to account for the stereochemical outcomes on indium mediated homoallyl-cyclopropanation of 1 ab and 3 ab, proceeding via indium(III) alkoxide 8 ab,
via intermolecular allyation by (S)-5-modified indium allyl reagent.
intermediate (8a) that, based on 13C{1H} NMR is structurally similar
to 7a, but shows profoundly different behaviour on acidification:
indium alkoxide 8a gives cyclopropane 4a, whereas indium
alkoxide 7a gives tetraene 9a, presumably via alcohol 3a [12]. The
difference can be ascribed to In(I) versus In(III) oxidation state in
the alkoxide, in which the greater oxophilicity of In(III) results in
fission of the CeO rather than OeIn bond upon protonation at
oxygen. This conclusion is supported by the observation that
sequential reaction of homoallylic alcohol 3a with n-BuLi (1.0
equiv.); InI3, (1.0 equiv.); allyl indium reagent 1I; modifier (S)-5(H),
and then acidification (1 M HCl) efficiently generated cyclopropane
4a (80%, 93/7 er), whereas the analogous process with InI gave
mostly tetraene 9a, and ꢃ15% 4a [15]. 13C{1H} NMR spectroscopic
analysis of the same sequence, but using alcohol 13C3-3a, confirmed
that the stable [16] lithium alkoxide 13C3-10a (C(1), t, 76 ppm,
1JCC ¼ 47 Hz) reacted cleanly with the InI3 to generate an InIII
underwent enantioselective cyclopropanation using (S)e5H (see
Scheme 2). Although 4b is generated as a mixture of syn/anti di-
astereomers, there is a similar level of stereocontrol at C(2) (ca.
90:10) to that in 4a generated from 2a/3a, Scheme 3, suggesting
that the same prochiral face of the homoallyl unit reacts with C(1)
on generation of bond C. On this basis, the C(2) stereocentres in syn/
anti 4b are assigned by analogy to (S)-4a; the latter having previ-
ously been definitively assigned via a convergent asymmetric
synthesis from hex-1-ene [7].
The stereochemistry at C(1) during the pathway(s) leading to
generation of bond B was probed by cyclopropanation of the en-
antiomers of 3b, resolved from ( )-3b by preparative chiral HPLC.
The configurations of (R)-3b and (S)-3b were assigned by conver-
gent synthesis involving alcohols (S)-12 and (R)-15, Scheme 4. The
diastereoselective allylation procedure of Tietze [18] employing
pseudoephedrine-derived reagent 13, was used to prepare (R)-15,
via intermediate 14, for which the relative configurations have been
unambiguously established by X-ray crystallography [18].
Cyclopropanation of alcohol 3b under the conditions of Scheme
3, results in syn/anti selectivity that depends on match/mismatch
with the modifier (S)-5(H), Scheme 5. For example, reaction of (R)-
3b with 1I/(S)-5(H) gave 72% syn-1R,2S-4b (96/4 er) whereas the
enantiomer (S)-3b gave 71% anti-1S,2S-4b (99/1 er). Moreover,
cyclopropanation of (S)-3b and (R)-3b, in the absence of modifier
(S)-5(H), gave 4b in non-racemic form, with the CeC unit being
predominantly formed via inversion at C(1), ruling out fully-
developed carbocation intermediates [19,20]. Curiously, the reac-
tion of ( )-3b in the absence of modifier (S)-5(H) gave a different
ratio of diastereomers (40/60 syn/anti) to the net product from
individual reactions of (S)-3b and (R)-3b (48.5/52.5). This suggested
that alcohol 3b acts as a chiral modifier for its own cyclo-
propanation reaction. This was tested by co-reaction of (S)-3b with
racemic ( )-2H4-3b, in the absence of modifier (5(H)). MS analysis
of the isotope distribution in each of the four stereoisomers after
physical separation by iterative analytical chiral HPLC confirmed
that cyclopropanation product 2H4-4b is indeed generated in non-
racemic form due to the presence of co-reacting (S)-3b.
1
alkoxide 13C3-11a (C(1), t, 77 ppm, JCC ¼ 48 Hz) in which the
adjacent carbons were equivalent, consistent with solely iodide
ligands (L ¼ I) at indium. Overall this suggests that the function of
the LiI in the reaction is to accelerate InI/InIII exchange of the
alkoxide (7a) with the allylating reagent (1I) or other indium spe-
cies generated in situ, so that the key CeOeIn(III) intermediate (8a)
is generated. An alternative interpretation is that LiI changes the
aggregation state [17] of 7a versus 8a. However the 13C NMR
chemical shift for C(1) in series 3a (ROH, 75 ppm); 10a (ROLi,
76 ppm); 11a (ROInIII, 77 ppm); 8a (ROIn(L), 78 ppm); and 7a
(ROIn(L0), 80 ppm) suggests a progressive decrease in covalency at
oxygen, with the difference between 7a and 8a (DdC ¼ 2 ppm)
larger than expected for a change in aggregation state. Moreover, in
situ 1H NMR analysis (see SI) of the chemical shifts[1bc,3c] of the
methylene unit in the mixture of allylindium reagents remaining
after generation of 7a from 2a indicate they are predominantly of
the form [(allyl)3-nInIIIIn]m (n,m ¼ 1,2; dH CH2eIn ¼ 2.12 and
2.06 ppm, d, 3JHH ¼ 8 Hz).[1bc] After addition of LiI and conversion of
7a to 8a, the allylindium reagents are predominantly of the form
3
allyl-InI (dH CH2eIn ¼ 1.71 ppm, d, JHH ¼ 8 Hz).[3bc,14].
Enone 2b, and the corresponding homoallylic alcohol 3b, also
4

J.L. Slaughter and G.C. Lloyd-Jones
Tetrahedron xxx (xxxx) xxx
Despite extensive efforts, we were not able to isolate or identify
References
the reagent generated from 1I and 5(H) (see Scheme 5). Reactions
[1] (a) S. Araki, Y. Butsugan, N. Katsumura, K.-I. Kawasaki, J. Chem. Soc. Perkin
Trans. 1 (1991) 499e501;
with analogous modifiers in which the
a-hydroxy group is pro-
tected ee.g. (S)-5(Ac) and (S)-5(Me) give no enantioselectivity (50/
50 er) [7], suggesting that the allylindium is chelated [21], e.g. as an
indate, Li[5((S)eIn-allyl)]. The consistent level of stereoselectivity
(b) S. Araki, Y. Butsugan, H. Ito, N. Katsumura, J. Organomet. Chem. 369 (1989)
291e296;
(c) S. Araki, Y. Butsugan, H. Ito, J. Org. Chem. 53 (1988) 1831e1833;
(d) S. Araki, Y. Butsugan, H. Ito, Synth. Commun. 18 (1988) 453e458.
[2] (For discussion and leading references, see:) (a) K. Jess, E.M. Hanada,
H. Peacock, S. A Blum, Organometallics 39 (2020) 2575e2579;
(b) V. Dorn, A. Chopa, G. Radivoy, RSC Adv. 6 (2016) 23798e23803;
(c) P. P Fedorov, A.I. Popov, R.L. Simoneaux, Russ. Chem. Rev. 86 (2017)
240e268.
[3] (For leading references, see:) (a) K. Kozinowski, J. Org. Chem. 83 (2018)
14314e14322;
(b) W.J. Bowyer, B. Singaram, A.M. Sessler, Tetrahedron 67 (2011) 7449e7460;
(c) T.H. Chan, Y. Yang, J. Am. Chem. Soc. 121 (1999) 3228e3229.
[4] (For selected reviews and examples, see: ) (a) V. Nair, S. Ros, C.N. Jayan,
B.S. Pillai, Tetrahedron 60 (2004) 1959e1982;
(approximately 90/10 er) induced by (S)-5(H) and the identical 13
C
chemical shifts for all intermediates generated from 4a in the
presence/absence of 5(H), suggest that the allyl is delivered inter-
molecularly by an alkoxide-modified allyl indium reagent to a
carbocationoid (16); a generic scheme that accounts for the ste-
reochemical outcomes at C(1) and C(2) is shown in Scheme 6.
3. Conclusions
(b) T.-P. Loh, G.-L. Chua, Chem. Commun. (2006) 2739e2749;
(c) J. Auge, N. Lubin-Germain, J. Uziel, Synthesis (2007) 1739e1764;
We have used in situ 13C{1H} NMR in conjunction with 13C/2H
labelling and mass spectrometry to study the asymmetric
homoallyl-cyclopropanation of enone 2a (Scheme 1) using a
mixture of in situ generated allylindium reagents (1I) and chiral
modifier (S)-5(H) [7]. A key outcome is the identification of acid-
labile In(I) alkoxide 7a as a primary intermediate from In(I)-
mediated allylation of the ketone 2a. In the absence of the modi-
fier 5(H), the In(I) alkoxide 7a undergoes LiI mediated In(I/III) ex-
change with excess reagent (1I), leading to the key In(III) alkoxide
intermediate 8a. In the presence of (S)-5(H), the In(I) alkoxide 7a is
rapidly protonated, leading to homoallylic alcohol 3a and
consuming 1 equivalent of the modifier, to generate (S)-5-In(L);
where L are unspecified ligands. Alcohol 3a also undergoes LiI
mediated reaction with excess reagent (1I) to generate indium (III)
alkoxide 8a. On acidification of the solution of 8a, protonation at
oxygen induces CeO rather than In(III)eO cleavage [20]. This is
undoubtedly assisted by stabilisation of the incipient tertiary allylic
cation by the homoallyl chain to give a highly charge-delocalised
system (16). Allylation of the homoallyl terminus in transient
cation 16 by (S)-5(In-allyl) is stereoselective, proceeding with
approximately 90% si selectivity, and presumably involving a
mandelate-chelated indium reagent [21]. In the case of a cation
derived from a chiral precursor (e.g. 3b), racemization in the
homoallyl-cation is slower than its trapping by the allylating spe-
cies, leading to cyclopropane generation with substantial net
inversion (76%) of the CeO bond. Overall the results suggest that
In(III) reagents may provide considerable utility for the selective
cleavage of other CeO or C-X units in the presence of nucleophilic
reagents.
ꢀ
(d) P.H. Lee, Bull. Kor. Chem. Soc. 28 (2007) 17e28;
(e) U.K. Roy, S. Roy, Chem. Rev. 110 (2010) 2742, 2535. A. K. Singh, Synlett,
2013, 24, 1457e1458;
(f) M. Bigovic, S. Skaro, V. Maslak, R.N. Saicic, Tetrahedron Lett. 54 (2013)
6624e6626;
(g) Z.-L. Shen, S.-Y. Wang, Y.-K. Chok, Y.-H. Xu, T.-P. Loh, Chem. Rev. 113
(2013) 271e401;
(h) G. Lemonnier, N. Van Hijfte, M. Sebban, T. Poisson, S. Couve-Bonnaire,
Pannecoucke, Tetrahedron 70 (2014) 3123e3133;
(i) X. Ma, B. Bolte, M.G. Banwell, A.C. Willis, Org. Lett. 18 (2016) 4226e4229;
(j) D. Kumar, S.R. Vemula, N. Balasubramanian, G.R. Cook, Acc. Chem. Res. 49
(2016) 2169e2178;
(k) Y.-S. Yang, S.-W. Wang, Y. Long, Curr. Org. Chem. 20 (2016) 2865e2880;
€
€
(l) M. Gintner, C. Denner, C. Schmolzer, M. Fischer, P. Frühauf, H. Kahlig,
W. Schmid, Monatsh. Chem. 150 (2019) 849e860;
(m) A.J. Sudha, N.A. Aslam, A. Sandhu, M. Yasuda, A. Baba, S.A. Babu, Tetra-
hedron 76 (2020) 131217.
€
[5] (a) H.A.F. Hoppe, G.C. Lloyd-Jones, M. Murray, T.P. Peakman, K.E. Walsh,
Angew. Chem., Int. Ed. Engl. 37 (1998) 1545e1547;
(b) S.M. Capps, G.C. Lloyd-Jones, M. Murray, T.P. Peakman, K.E. Walsh, Tetra-
hedron Lett. 39 (1998) 2853e2856;
€
(c) S.M. Capps, T.P. Clarke, J.P.H. Charmant, H.A.F. Hoppe, G.C. Lloyd-Jones,
M. Murray, T.P. Peakman, R.A. Stentiford, K.E. Walsh, P.A. Worthington, Eur. J.
Org Chem. (2000) 963e974.
[6] (Selected Reviews on cyclopropane structure and reactivity: ) (a) K. Mizuno,
N. Ichinose, Y. Yoshimi, J. Photochem, C. Photobiol, Photochem. Rev. (2000)
167e193;
(b) F. Gnad, O. Reiser, Chem. Rev. 103 (2003) 1603e1624;
(c) L.K. Sydnes, Chem. Rev. 103 (2003) 931, 164;
(d) H.-U. Reissig, R. Zimmer, Chem. Rev. 103 (2003) 1151e1196;
(e) M. Yu, B.L. Pagenkopf, Tetrahedron 61 (2005) 321e347;
(f) H. Rinderhagen, P.A. Waske, J. Mattay, Tetrahedron 62 (2006) 6589e6593;
(g) H. Pellissier, Tetrahedron 64 (2008) 7041e7095.
[7] (Reviews on Cyclopropane synthesis: ) (a) W.A. Donaldson, Tetrahedron 57
(2001) 8589e8627;
(b) S. Ma, Acc. Chem. Res. 36 (2003) 701e712;
(c) A.B. Charette, H. Lebel, J.-F. Marcoux, C. Molinaro, Chem. Rev. 103 (2003)
977e1050;
(d) W. Brandt, T. Thiemann, L.A. Wessjohn, Chem. Rev. 103 (2003)
1625e1648;
Declaration of competing interest
(e) F.C. Engelhardt, M.J. Schmitt, R.E. Taylor, Tetrahedron 59 (2003)
5623e5634;
(f) G. Maas, Chem. Soc. Rev. 33 (2004) 183e190;
(g) J.L. García Ruano, A.M. Martín Castro, E. Torrente, Phosphorus, Sulfur and
Silicon 180 (2005) 1445e1446;
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
(h) J. Szymoniak, P. Bertus, Top. Organomet. Chem. 10 (2005) 107e132;
(i) G. Bartoli, G. Bencivenni, R. Dalpozzo, Synthesis 46 (2014) 979e1029.
[8] D.D.P. Laffan, G.C. Lloyd-Jones, A.J. Parker, P.D. Wall, J.L. Slaughter, Tetrahedron
62 (2006) 11402e11412.
Acknowledgements
[9] (a) G.C. Lloyd-Jones, M. Murray, R.A. Stentiford, P.A. Worthington, Eur. J. Org
Chem. (2000) 975e985;
€
We thank Prof. Dr Lutz Tietze and Dr Tom Kinzel (Gottingen) for
the provision of reagent 13 and detailed advice for the preparation
of intermediates crucial in assignment of the configuration of the
resolved enantiomers of 3b. The Universities of Bristol and Edin-
burgh provided generous support. Dr David Laffan (AstraZeneca)
provided assistance with the synthesis and resolution of alcohol 3b.
(b) P.E. Standen, D. Dodia, M.R.J. Elsegood, S.J. Teat, M.C. Kimber, Org. Biomol.
Chem. 10 (2012) 8669e8676.
[10] There is no evidence for a non-linear relationship between the modifier ee
and the selectivity (see reference [8]), suggesting that only a single mandelate
unit is present in the modified reagent, e.g. (S)-5-In-(L)-allyl.
[11] Typical ranges for one bond carbonꢂcarbon coupling constants are: sp2ꢂsp2
1JCC 75ꢂ80 Hz, sp2ꢂsp3 JCC 35ꢂ60 Hz, and sp3ꢂsp3 JCC 25ꢂ55 Hz, see
Hansen, P. E.; Wray, V. Org. Magn. Reson 15 (1981) 102e103.
[12] Careful Aqueous Work-Up of an Independent Identically Prepared Solution of
7a Afforded the Corresponding Alcohol 3a. Acidification of 3a (HCl) Generates
Tetraene 9a.
1
1
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tet.2020.131786.
[13] Diastereotopicity Can Be Induced by the Presence of a Stereocentre at the
5

J.L. Slaughter and G.C. Lloyd-Jones
Tetrahedron xxx (xxxx) xxx
Indium Bound to Oxygen in the Alkoxide, or to Presence of a Stereocentre at
43 (1972) 257e264;
an Indium that Is
m-halide Bridge to the Indium Alkoxide.
(b) J.S. Gynane, L.G. Waterworth, I.J. Worrall, J. Organomet. Chem. 40 (1972)
[14] There Has Been Considerable Debate Regarding the Active Species in the
Mixture of Reagents Generated - See Reference [3]. A Small Amount of Re-
sidual Indium Metal Present after Generation of (1I) Was Found to Improve
Yields of the Cyclopropane Products (4). LiI May Possibly Induce Compro-
portionation with In(III), or Further Reaction with Excess Allyliodide.
[15] InBr3 Functioned Equally as Well as InI3; ꢃ 3% 4a Was Obtained with InCl3.
[16] There was no anionic oxy-Cope rearrangement of the lithium alkoxide 13C3-
10a at these temperatures, even when the analogous potassium alkoxide was
generated using KH. The reaction requires heating to reflux - see: S.K. Mandal,
Sk R. Amin, W.E. Crowe J. Am. Chem. Soc. 123 (2001) 6457e6458.
C9eC10.
[18] L.F. Tietze, K. Schiemann, C. Wegner, C. Wulff, Chem. Eur J.
4 (1998)
1862e1869.
[19] For an example of a homoallyl cation to vinylcyclopropane interconversion,
proceeding with inversion of configuration, see B.J. Melancon, N.R. Perl,
R.E. Taylor, Org. Lett. 9 (2007) 1425e1428.
[20] For indium-mediated C-O cleavage to generate cations, see: C. Behloul,
ꢀ
A. Chouti, D. Guijarro, F. Foubelo, C. Najera, M. Yus Tetrahedron 72 (2016)
7937e7941.
[21] For
a review on indium chelation see: B.K. Zambron Synthesis (2020)
[17] (We thank an anonymous reviewer for noting this and the following refer-
ences: ) (a) M.J.S. Gynane, L.G. Waterworth, I.J. Worrall, J. Organomet. Chem.
1147e1180.
6