On the Origin of E/Z Selectivity in the Modified Julia Olefination - Importance of the Elimination Step

Article abstract of DOI:10.1002/ejoc.201201634

The mechanism and origin of high E selectivity in the modified Julia olefination of aromatic aldehydes have been explored by computational and experimental means. Reversibility of addition and hence selectivity of the formation of sulfinate 5 is very variable and depends on the nature of the sulfone substrate. However, in all cases, elimination occurs through a concerted antiperiplanar and synperiplanar mechanism for sulfinates anti-5 and syn-5, respectively. Both syn and anti diastereomeric pathways thus lead preferentially to the (E)-alkene. Copyright

Full text of DOI:10.1002/ejoc.201201634

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201201634
On the Origin of E/Z Selectivity in the Modified Julia Olefination –
Importance of the Elimination Step
[a]
[a]
ˇ
ˇ
Raphaël Robiette* and Jirí Pospísil*
Dedicated to Professor Alois Fürstner on the occasion of his 50th birthday
Keywords: Density functional calculations / Elimination / Stereoselectivity / Reaction mechanisms / Olefination
The mechanism and origin of high E selectivity in the modi-
fied Julia olefination of aromatic aldehydes have been ex-
plored by computational and experimental means. Revers-
ibility of addition and hence selectivity of the formation of
sulfinate 5 is very variable and depends on the nature of the
sulfone substrate. However, in all cases, elimination occurs
through concerted antiperiplanar and synperiplanar
mechanism for sulfinates anti-5 and syn-5, respectively. Both
syn and anti diastereomeric pathways thus lead preferen-
tially to the (E)-alkene.
a
In the reaction of lithiated alkyl-BT-sulfones with alkyl
aldehydes (R1, R2 = alkyl), S. Julia et al. showed that the
initial addition is nonreversible and that subsequent steps
are stereospecific, elimination occurring exclusively by a
concerted antiperiplanar elimination (i.e. from transoid 5).[5]
The low selectivity observed in these cases is thus a direct
consequence of the poor diastereocontrol in the initial ad-
dition step of sulfone anion 2 onto aldehyde 1.
Introduction
The modified Julia olefination allows the synthesis of
alkenes in a single step from metallated benzothiazol-2-yl
(BT) sulfones and aldehydes (Figure 1).[1,2] Over the past
two decades, this olefination reaction has emerged as a
powerful tool for carbon–carbon double bond formation,
in particular when two complex molecular fragments
should be connected.[3] The attractiveness of this connective
olefination reaction arises from its versatility, its wide func-
tional group tolerance, and the mild reaction conditions un-
der which the reaction proceeds. The only shortcoming of
the reaction is the difficulty of predicting and controlling
the stereoselectivity of the newly formed double bond,
which is a limitation to an even broader use of this reaction.
The experimental results showed that the stereochemical
outcome of this process depends highly on the nature of the
reactants: reaction of BT-sulfones 2 with aromatic alde-
hydes (compound 1, R2 = aryl) generally gives high E selec-
tivity, whereas their reaction with aliphatic aldehydes (com-
pound 1, R2 = alkyl) furnishes alkenes with little or no ste-
reochemical bias.[1,4]
For aromatic aldehydes (R2 = Ar), the situation is dif-
ferent: elimination from 3 is not stereospecific.[5,6] In order
to account for this nonstereospecificity of the elimination
and the observed high E selectivity, Julia et al. suggested
that in these cases elimination occurs by a direct loss of
benzothiazolonate from intermediates 5 to yield a zwitter-
ion such as 6.[5c] Stabilization of 6 by the aryl group should
allow conformational equilibration and favor formation of
the (E)-olefin upon loss of SO2 from the more stable anti
zwitterion 6.
Results and Discussion
In our calculations[7–9] on the reaction of benzaldehyde
with lithiated ethyl- and benzyl-BT-sulfones (2a and 2b), we
find a slightly exothermic addition step to form alkoxides 3
(Figure 2). For the syn alkoxide (syn-3), the free energy bar-
rier to Smiles rearrangement[10,11] is computed to be lower
than that for addition, thus predicting a nonreversible ad-
dition in both cases (R1 = Me or Ph).[12] In the case of the
anti isomer, unfavorable 1,2 steric interactions in TS-Smiles
(Smiles transition state) lead to an increase of the barrier
to Smiles rearrangement.[13] This has the result that ad-
dition of benzyl-BT-sulfone 2b (R1 = Ph) becomes revers-
ible. These predictions could be confirmed by crossover ex-
Although the global mechanistic sequence depicted in
Figure 1 is known and widely accepted since the pioneering
work of S. Julia, a detailed atomistic account of the mecha-
nism and selectivity of this reaction is still lacking.[5]
[a] Institute of Condensed Matter and Nanosciences, Université
catholique de Louvain
Place Louis Pasteur 1,
box L4.01.02, 1348 Louvain-la-Neuve, Belgium
E-mail: raphael.robiette@uclouvain.be
jiri.pospisil@uclouvain.be
Homepage: http://www.uclouvain.be/raphael.robiette
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201201634.
836
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 836–840

On the Origin of E/Z Selectivity in the Modified Julia Olefination
Figure 2. Computed pathways for the formation of sulfinate 5 (free
energies in kcalmol–1 relative to reactants).
Scheme 1. Crossover experiments. Reagents and conditions: (a)
TBAF (2 equiv.), LiCl (5 equiv.), p-NO2C6H4CHO (1.1 equiv.),
THF, –78 °C.
preference for (Z)-alkene formation for the reaction of
benzyl-BT-sulfone (2b). Experimentally, a high E selectivity
(Ͼ 96:4) is observed in both cases.[1] Indeed, as postulated
by Julia and shown by our crossover experiments (see
E/Z selectivity for stylbene formation in Scheme 1),[6] both
Figure 1. Generally accepted mechanism and selectivity in the
modified Julia reaction.
periments in which the two diastereomeric β-alkoxy-BT- diastereomeric pathways lead, in fact, to (E)-alkene.
sulfones 3b (R1 = R2 = Ph) were independently generated
In order to identify the mechanism of formation of (E)-
by another route in the presence of a more reactive alde- alkene from sulfinate 5, we thus investigated the formation
hyde (Scheme 1).[6] Indeed, deprotection of anti-7 and syn- of zwitterion 6 as proposed by Julia et al.[5c] Every attempt
7 in the presence of p-NO2C6H4CHO gave 48% and Ͻ5% to optimize such an intermediate, or a carbanion resulting
of the olefin, respectively, in which the more reactive alde- from the loss of SO2, failed, our results favoring a concerted
hyde had been incorporated, proving the (at least partial) elimination instead.[15]
reversibility of the alkoxide formation in the former case
and the nonreversible character of the syn addition.
However, while our calculations indicate that anti iso-
mers should undergo antiperiplanar elimination to give (E)-
As a consequence, sulfinate adducts 5 should be formed olefins, syn isomers are actually predicted to preferentially
with a low anti selectivity in the case of ethyl-BT-sulfone eliminate from cisoid syn-5 by a concerted synperiplanar
(2a), addition being nonreversible and poorly selective.[6] elimination, thus leading also to (E)-alkenes (Figures 3 and
On the other hand, for benzyl-BT-sulfone (2b) the observed 4).[12]
reversibility of anti addition should yield predominantly syn
In order to provide experimental evidence for this con-
certed elimination from cisoid syn-5 (synperiplanar elimi-
sulfinate 5.[14]
Accordingly, if the elimination step were stereospecific nation) in the reaction of aromatic aldehydes, we performed
via an antiperiplanar arrangement as in the reaction of ali- deuterium-labeled experiments. We reasoned that if we
phatic aldehydes,[1,5] our calculations would predict an E could form stereoselectively monodeuterated 5-d, in which
selectivity for the reaction of ethyl-BT-sulfone (2a) and a 1,2 steric strain is absent, we would get information on the
Eur. J. Org. Chem. 2013, 836–840
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
837

R. Robiette, J. Pospísˇil
SHORT COMMUNICATION
Table 1. Deuteration experiments.[a]
Entry Ar
Olefin
Yield [%][b]
E/Z[c,d]
1
2
3
p-ClC6H4
C6H5
p-MeOC4H6
9a
9b
9c
89
92
87
81:19
90:10
94:6
[a] Reaction conditions: EtSH (1.2 equiv.), LiHMDS (1.1 equiv.)
THF, –78 °C (2 h) to room temp. (6 h). [b] Refers to the pure iso-
lated compound. [c] Based on the crude 1H NMR spectra. [d] Mean
of three independent experiments.
Figure 3. Computed pathways for alkene formation from 5 (free
energies in kcalmol–1 relative to reactants).
action also explains the fact that, in the case of anti-5 iso-
mer, it is the antiperiplanar elimination which is favored
(see Figure 3). Indeed, in this case it is the synperiplanar
TS which involves this unfavorable 1,2 steric interaction.
Another difference between the two transition state struc-
tures is the degree of coordination to the lithium cation: the
synperiplanar TS involves coordination of lithium by both
the sulfinate and the BT groups, whereas the antiperiplanar
arrangement allows coordination only by the BT group (see
Figure 4).[18]
This leaves, however, the question of why syn-5 un-
dergoes synperiplanar elimination in the reaction of aro-
matic aldehydes while exclusive antiperiplanar elimination
is observed in the reaction of alkyl aldehydes.[5] Calcula-
tions on model systems reveal that the presence of the
phenyl group also plays an important role in the electronic
stabilization of the synperiplanar TS: Substitution by a
phenyl group leads to
a
higher stabilization (by
Figure 4. Transition state structures for the elimination step from
syn-5a (top) and syn-5b (bottom) (free energies in kcalmol–1 rela-
tive to reactants).
2.8 kcalmol–1) of the synperiplanar TS than the antiper-
iplanar one (Table 2). A natural bond orbital (NBO) analy-
sis shows that the key interaction responsible for this acti-
vation is an electronic donation from the π system of the
phenyl to the positively charged aldehyde carbon atom of
the TS. Corresponding E(2) values are 58.4 and
52.8 kcalmol–1 for syn- and antiperiplanar TSs, respectively,
mechanism of the elimination (Table 1). Indeed, if elimi-
nation from 5-d involves formation of a carbocation such
as 6, an approximately 50:50 mixture of (E)- and (Z)-styr-
ene-(β)-d would be obtained, whereas if elimination occurs
selectively by a concerted mechanism, only one isomer
would be formed: the (E)-alkene if elimination is synperi- Table 2. Influence of the nature of the aldehyde (R2) on the stereo-
selectivity of the elimination (free activation energies in kcalmol–1
planar and the (Z)-alkene if it is antiperiplanar. Addition
from transoid sulfinates).
of EtSLi onto 8-d[6,16] at –78 °C allowed clean formation
of sulfinates 5-d, which immediately undergo elimination to
provide the corresponding deuterated styrenes.[17] In each
case, selective (E)-alkene formation was observed. These re-
sults thus give support to the concerted synperiplanar elimi-
TS synperiplanar
ΔGrel dC–O
TS antiperiplanar
[a]
[a]
R2
Charge on C(1)[b] ΔGrel sC–O
Charge on C(1)[b]
nation mechanism.
Analysis of elimination TSs for the reaction of 2a and 2b
shows that an important factor contributing to this unex-
pected preference for synperiplanar elimination from syn-5
is the destabilizing 1,2 steric strain present in the antiper-
H
Me
Ph
30.5
32.1
25.9
2.07
2.16
1.98
0.329
0.347
0.354
30.9
33.8
29.1
2.00
2.00
1.96
0.241
0.302
0.323
[a] C(1)–O distances (in Å) in the TS. [b] NPA (natural population
iplanar TS (but absent in the synperiplanar one). This inter- analysis) charges (in a. u.) in TS.
838
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 836–840

On the Origin of E/Z Selectivity in the Modified Julia Olefination
Figure 5. Rationale for observed high E selectivity in modified Julia olefination of aromatic aldehydes.
che Scientifique – FNRS (Project FRFC No. 2.4502.05). R. R. and
consistent with a higher stabilization in the former case,
probably because of the slightly more positively charged
C(1) carbon atom in the synperiplanar TSs (see Table 2).
Interestingly, the higher stabilization of the synperiplanar
TS by electronic donation from the aryl group suggests an
explanation for the fact that electron-poor aldehydes give
lower E selectivities than electron-rich ones:[5c] Electron-
poor aryl groups stabilize elimination TSs to a lesser extent,
thus decreasing the preference for synperiplanar elimi-
nation, and hence they lower the E/Z selectivity. This is sup-
ported by the decrease of the E/Z ratio upon going from
p-methoxy- to p-chloro-substituted aryl derivatives in our
crossover experiments (see Table 1).
J. P. are, Chercheur Qualifié and Chargé de recherches F. R. S.-
FNRS, respectively. R. R. thanks Prof. Jeremy N. Harvey for fruit-
ful discussions.
[1]
For reviews, see: a) I. E. Markó, J. Pospísˇil in Science of Synthe-
sis, vol. 47b (Ed.: A. de Meijere), Thieme, Stuttgart, New York,
2010, pp. 105–160; b) C. Aïssa, Eur. J. Org. Chem. 2009, 1831–
1844; c) K. Plesniak, A. Zarecki, J. Wicha, Top. Curr. Chem.
2007, 275, 163–250; d) R. Dumeunier, I. E. Markó in Modern
Carbonyl Olefination (Ed.: T. Takeda), Wiley-VCH, Weinheim,
2004, pp. 104–150; e) P. R. Blakemore, J. Chem. Soc. Perkin
Trans. 1 2002, 2563–2585.
[2]
[3]
Several heteroaryl substrates have been developed for this pro-
cess and the nature of the heteroaryl group has shown to have,
in some cases, an influence on the E/Z selectivity (see ref.[1]).
BT-sulfones being the original, and still most popular, sub-
strates in modified Julia reaction, the present study will focus
on such systems.
The modified Julia olefination is commonly used in total syn-
theses of natural and unnatural biologically active molecules.
For selected examples, see ref.[1d] and a) R. Cribiú, C. Jäger, C.
Nevado, Angew. Chem. 2009, 121, 8938–8941; Angew. Chem.
Int. Ed. 2009, 48, 8780–8783; b) K. Sastraruji, T. Sastraruji,
S. G. Pyne, A. T. Ung, A. Jatisatienr, W. Lie, J. Nat. Prod. 2010,
73, 935–941; c) K. Micoine, A. Fürstner, J. Am. Chem. Soc.
2010, 132, 14064–14066; d) J. Santos, B. M. Illescas, N. Martín,
J. Adrio, J. C. Carretero, R. Viruela, E. Ortí, F. Spänig, D. M.
Guldi, Chem. Eur. J. 2011, 17, 2957–2964; e) C. Rink, F. Sasse,
A. Zubriene, D. Matulis, M. E. Maier, Chem. Eur. J. 2010, 16,
14469–14478; f) P. Gowrisankar, S. A. Pujari, K. P. Kaliappan,
Chem. Eur. J. 2010, 16, 5858–5862.
Conclusions
In summary, we have clarified the mechanism of elimi-
nation in modified Julia reactions of aromatic aldehydes
and showed that elimination occurs through a concerted
antiperiplanar and synperiplanar mechanism in the case of
anti- and syn-sulfinate, respectively. The high experimental
E selectivity is thus explained by E-selective elimination,
from both the syn and the anti diastereomer (Figure 5).
Identification of synperiplanar elimination as the main
pathway for elimination from syn-sulfinate and understand-
ing of factors controlling the stereoselectivity of elimination
now allow us to rationalize some key experimental observa-
tions relating to modified Julia olefination of aromatic alde-
hydes and lithiated BT-sulfones. This analysis should assist
in the design of new reagents and reaction conditions and
allow further development of the modified Julia reaction
process for highly E/Z-selective synthesis of alkenes.
[4]
Despite numerous experimental studies, no general solution to
the control of this E/Z selectivity could be found thus far. See:
a) F. Billard, R. Robiette, J. Pospísˇil, J. Org. Chem. 2012, 77,
6358–6364; b) J. Pospísˇil, Tetrahedron Lett. 2011, 52, 2348–
2352; c) S. Surprenant, W. Y. Chan, C. Berthelette, Org. Lett.
2003, 5, 4851–4854; d) P. R. Blakemore, W. J. Cole, P. J. Kocien´-
ski, A. Morley, Synlett 1998, 26–28; e) A. B. Charette, H. Le-
bel, J. Am. Chem. Soc. 1996, 118, 10327–10328.
Supporting Information (see footnote on the first page of this arti-
cle): Full computational and experimental details including pro-
cedures and characterization data for all compounds and optimized
Cartesian coordinates of all optimized structures.
[5]
[6]
a) J. B. Baudin, G. Hareau, S. A. Julia, O. Ruel, Tetrahedron
Lett. 1991, 32, 1175–1178; b) J. B. Baudin, G. Hareau, S. A.
Julia, O. Ruel, Bull. Soc. Chim. Fr. 1993, 130, 336–357; c) J. B.
Baudin, G. Hareau, S. A. Julia, O. Ruel, Bull. Soc. Chim. Fr.
1993, 130, 856–878.
See Supporting Information for full details and complementary
experiments.
Acknowledgments
This work was supported by the Université catholique de Louvain
(UCL) (Fonds Spéciaux de Recherche) and the Fonds de la Recher-
Eur. J. Org. Chem. 2013, 836–840
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
839

R. Robiette, J. Pospísˇil
SHORT COMMUNICATION
[7]
[12] The barrier to rotation (between cisoid and transoid conform-
ers) in alkoxides and sulfinates indicates, in both cases, a rapid
equilibrium that has no important role in determining reactiv-
ity or selectivity. See the Supporting Information for full data.
[13] This is in good agreement with experiments, which showed a
higher stability of anti-β-alkoxy-BT-sulfones over their syn iso-
mers (see ref.[5b]).
[14] One may expect sterically hindered alkyl-BT-sulfones to involve
increased barriers to Smiles rearrangement for anti isomers,
which in turn could lead to a reversal of addition. In these
cases, a syn selectivity for sulfinate formation may thus well be
observed.
[15] Additionally, during our experimental study of this reaction,
no direct or indirect evidence for cation formation was ob-
served.
[16] See the Supporting Information for the synthesis of 8-d.
[17] Control experiments showed that compound 5-d does not re-
verse back to benzaldehyde and lithiated sulfone. We have also
checked that no isotopic scrambling was occurring. See the
Supporting Information for details.
Calculations were carried out at the SCS-MP2/aug-cc-
pVTZ(THF)//B3LYP/6–31+G*(THF) level of theory with the
Jaguar 6.5[8] and ORCA 2.7[9] programs. Unless mentioned
otherwise, reported energies are free energies relative to reac-
tants. Since the nature of the counterion has a small influence
on the E/Z selectivity (see ref.[4]), a lithium cation was included
in all our calculations, LDA being the most common base used
in these reactions. See Supporting Information for full compu-
tational details.
[8]
Jaguar, version 6.5, Schrödinger, LLC, New York, NY, 2005.
[9] F. Neese, ORCA, An Ab Initio, DFT, and Semiempirical Elec-
tronic Structure Package, version, 2.7.0, University of Bonn,
Germany, 2010.
[10] Interestingly, our results indicate that, conversely to what is
commonly admitted, transfer of the heterocycle occurs by a
concerted pathway, i.e., in a single elementary step (see the Sup-
porting Information for full details). For a study on single-step
and multistep mechanisms of aromatic nucleophilic substitu-
tion, see: M. N. Glukhovstev, R. D. Bach, S. Laiter, J. Org.
Chem. 1997, 62, 4036–4046.
[18] See Supporting Information for a detailed discussion of lithium
cation positioning.
[11] For a review on Smiles rearrangement, see W. E. Truce, E. M.
Kreiber, W. W. Brand in Organic Reactions, vol. 18, John
Wiley & Sons, New York, 1970, p. 99.
Received: December 4, 2012
Published Online: January 8, 2013
840
www.eurjoc.org
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2013, 836–840