Anomalous Z-isomer content in Wittig reaction products from keto-stabilised ylides with ortho-heteroatom substituted benzaldehydes

Article abstract of DOI:10.1016/j.tetlet.2012.09.123

Wittig reaction products of keto-stabilised ylides with ortho-substituted benzaldehydes are found to show significantly higher than expected Z-alkene content (up to 50%) compared to analogous reactions of the same ylides with benzaldehyde itself. A cooperative effect is seen whereby the unusual Z-content is further augmented if the ylide bears greater steric bulk in the α′-position. These results are consistent with our previous observations on reactions of all ylide types with aldehydes bearing a β-heteroatom. Significantly, the cooperative effect, previously seen only with semi-stabilised ylides, has now been extended to stabilised ylides. Both the anomalous increase in Z-content and the cooperative effect can be rationalised within the [2+2] cycloaddition mechanism of the Wittig reaction.

Full text of DOI:10.1016/j.tetlet.2012.09.123

Tetrahedron Letters 53 (2012) 6701–6704
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Anomalous Z-isomer content in Wittig reaction products from
keto-stabilised ylides with ortho-heteroatom substituted benzaldehydes
Peter A. Byrne a, Lee J. Higham b, Pádraic McGovern a, Declan G. Gilheany a,
⇑
a Centre for Synthesis & Chemical Biology, Conway Institute of Biomolecular and Biomedical Research, School of Chemistry and Chemical Biology, University College Dublin,
Belfield, Dublin 4, Ireland
b School of Chemistry, Bedson Building, Newcastle University, Newcastle upon Tyne NE1 7RU, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 June 2012
Revised 10 September 2012
Accepted 25 September 2012
Available online 2 October 2012
Wittig reaction products of keto-stabilised ylides with ortho-substituted benzaldehydes are found to
show significantly higher than expected Z-alkene content (up to 50%) compared to analogous reactions
of the same ylides with benzaldehyde itself. A cooperative effect is seen whereby the unusual Z-content
is further augmented if the ylide bears greater steric bulk in the a0-position. These results are consistent
with our previous observations on reactions of all ylide types with aldehydes bearing a b-heteroatom.
Significantly, the cooperative effect, previously seen only with semi-stabilised ylides, has now been
extended to stabilised ylides. Both the anomalous increase in Z-content and the cooperative effect can
be rationalised within the [2+2] cycloaddition mechanism of the Wittig reaction.
Keywords:
Wittig reaction mechanism
Keto-stabilised phosphonium ylide
ortho-Substituted benzaldehydes
Kinetic control in Wittig reactions
Z-Alkene
Ó 2012 Elsevier Ltd. All rights reserved.
[2+2] Cycloaddition
Introduction
Our explanation for the heteroatom-induced increase in
Z-selectivity involved an additional bonding interaction in the
The Wittig reaction1 is perhaps the most commonly used
method for the synthesis of alkenes. The many and varied applica-
tions of it and the extensive work that has been done on its
mechanism can be found in various excellent reviews.2–6 A list of
the very latest developments in the Wittig reaction can be found
in our recent publication on the mechanism.7 In the latter work,
we believe that we have definitively shown the unity of the mech-
anism of the Li-salt free Wittig reaction.7 By this we mean that all
ylide types (non-stabilised, semi-stabilised and stabilised) react by
the same mechanism. Our conclusions were based on the results of
Wittig reactions of b-heteroatom substituted aldehydes, which
showed unexpectedly high selectivity for Z-alkenes for all ylide
types (see Scheme 1). Such selectivity is not seen in reactions of
the same ylides with analogous aldehydes lacking the suitably
placed b-heteroatom.
We showed that this change of selectivity, which was rather
dramatic in many cases, could be easily accommodated within
the emerging consensus4,8–18 on the mechanism of the Li-salt free
Wittig reaction: [2+2] cycloaddition of ylide and aldehyde to give
an oxaphosphetane (OPA) followed by cycloreversion to phosphine
oxide and an alkene (see Scheme 2). Therefore our results also pro-
vided strong experimental support for that mechanism.19,20
cis-selective cycloaddition transition state (TS), between phospho-
rus and the b-heteroatom, as shown in Figure 1a.7
A subsidiary, but important, part of our study concerned an
unexpected co-operative effect seen in the case of semi-stabilised
ylides. This was also explicable on the proposed TS model, and is
illustrated by the data in Scheme 3: the placement of an ortho-
substituent on the ylide gives an even greater shift towards high
Z-selectivity than the corresponding reaction of the unsubstituted
benzylide. For example (Scheme 3), in reaction with ortho-
iodobenzaldehyde, the Z/E ratio changes from 96:4 to 99:1 on
placement of an iodo-substituent on the benzylide (ylide 1a vs
1b) even though in reaction with benzaldehyde the same change
has little effect (15:85 vs 28:72). This indicates a greater decrease
(by ca. 1 kJ molꢀ1) in the energy of the cis-selective TS in the reac-
tions of ortho-iodo ylide 1b than in the reactions of unsubstituted
benzylide 1a. According to our TS model, the increased 2–3 interac-
tions (see Fig. 1b for numbering) that would be engendered by the
larger aryl group on the ylide a-carbon of ortho-substituted benzy-
lides should result in a greater bias towards the cis-selective TS,
since it is better able to accommodate steric bulk in this position.
We refer to this as a ‘cooperative effect’ between the substituents
on the ylide and aldehyde which results in a greater shift in Z-selec-
tivity than is possible if either of the substituents is not present.
We were interested to see if this cooperative effect could be
extended to stabilised ylides. In our previous publication,7 we
⇑
Corresponding author. Tel.: +353 1 716 2308.
E-mail address: declan.gilheany@ucd.ie (D.G. Gilheany).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.09.123

6702
P. A. Byrne et al. / Tetrahedron Letters 53 (2012) 6701–6704
sterically bulky groups can be directly attached to the carbonyl
(i) NaHMDS
or KHMDS
n = 3 o r 4
R3
P
R3
alkyl
R2 = aryl
ester
group, mimicking the situation in the semi-stabilised cases.
Our objectives were to determine (i) whether there would be an
increase in Z-content on substitution of a heteroatom b to the
aldehyde, further confirming the generality of the main effect
and (ii) whether a cooperative effect (further increase in Z-content)
would be observed in reactions of keto-stabilised ylides with a
more heavily substituted a0-carbon (carbon-3 of the oxoalkylidene
moiety). The proposed cis-selective TS for these reactions, whose
existence we sought to confirm, is analogous to the one proposed
previously for reactions of non-stabilised, semi-stabilised and es-
ter-stabilised ylides with b-heteroatom substituted aldehydes
(see Fig. 1b).7 As in the previous study,7 we chose to focus on ylides
derived from methyldiphenylphosphine, since kinetic control has
previously been established in reactions of such ylides,7,16 but a
small number of triphenylphosphine-derived cases are also in-
cluded for comparison, as are some ester-stabilised ylides.
We found that there was indeed evidence for a cooperative ef-
fect with these ylides, but at a reduced level. The various new
substituted enones generated by these studies proved difficult to
handle and it was with great difficulty that many of the Z/E ratios
were obtained; some consistently underwent decomposition and,
in all cases the Z-isomers were extremely prone to isomerisation
to the E-isomer. Therefore, despite the precautions taken to
measure accurately the ‘kinetic Z/E ratio’ of each reaction,21 the
ratios quoted below represent only a lower boundary to the true
Z-content of these reactions.
Br
R3
Y
THF
R2
R2
(ii) -78 °C
n
Y
O
Y = F, Cl, Br, I, OMe,
OEt, SMe
H
n
Scheme 1. Highly Z-selective Wittig reactions of b-heteroatom-substituted alde-
hydes under conditions promoting kinetic control.
R3
P
R3
R3
R3
R3
R3
non-Wittig
processes
O
P
O
Betaine
OPA
R2
R1
R2
R1
R3
R3
R3
P
R2
R1
R2
R1
R3
R3
R3
O
P
P
R2
R3
R3
R1
O
cis
O
H
R3
trans
- (R3)3P
O
- (R3)3P
O
R1
R2
R1
R2
Alkene
Z
E
Scheme 2. [2+2] Cycloaddition mechanism for the Li salt-free Wittig reaction.
Reactions of ortho-heteroatom substituted benzaldehydes with
keto-stabilised and ester-stabilised ylides
(b)
(a)
All reactions were carried out using the standard set of condi-
tions established in our previous study,7 with the exception that,
for solubility reasons, certain reactions were carried out at higher
temperatures. The ylide was generated in situ from the parent
phosphonium bromide or chloride salt using the non-nucleophilic
NaHMDS base in THF (giving insoluble NaCl or NaBr) under an
atmosphere of dry nitrogen, and then cooled to ꢀ78 °C. The alde-
hyde was then added dropwise to the cold ylide solution, and
the mixture was stirred at low temperature for 20 min. Some of
the reactions were quenched at low temperature by addition of
acid to ensure that the reaction had occurred at that temperature.
The Z/E ratios of the alkenes produced in reactions of 2-oxoalky-
lidenemethyldiphenylphosphoranes with selected benzaldehydes
are shown in Table 1.
O
R
R3
R3
Y
P
R2
Y
P
R3
2
R3
3
R3
R3
H
1
H
H
H
O
O
4
Figure 1. (a) Favoured cis-selective TS in reactions of b-heteroatom substituted
aldehydes (e.g. ortho-substituted benzaldehyde) with non-stabilised (R2 = alkyl),
semi-stabilised (R2 = aryl) and ester-stabilised (R2 = CO2R) ylides; (b) proposed cis-
selective TS for similar reactions of keto-stabilised ylides, showing the numbering
of the forming ring.
Acetonylidenemethyldiphenylphosphorane (6, R = CH3 in the
scheme accompanying Table 1) was found to be insoluble in THF
below ꢀ50 °C and attempted Wittig reactions with it at tempera-
tures below that (with low temperature quenching) yielded no al-
kene product. Therefore its reactions were carried out at ꢀ45 °C or
20 °C. The reactions of this ylide with benzaldehyde (2a) showed
high E-selectivity (Table 1, entries 1 and 2), while its reactions with
THF
Ph
P
Me
Y
-78 °C
Ph
CHO
X
X
X = H, Y = H Z/E = 15:85
X = H, Y = I Z/E = 96:4
X = I, Y = H Z/E = 28:72
X = I, Y = I Z/E > 99:1
Y
1a X = H
1b X = I
2a X = H
2b X = I
ortho-heteroatom substituted benzaldehydes 2c–d showed
a
moderate increase in Z-content (up to 33% or 40% Z-isomer pro-
duced, see Table 1, entries 3 and 4). Thus the aldehyde ortho-
heteroatom-induced increase in Z-selectivity previously observed
in reactions of non-stabilised, semi-stabilised and ester-stabilised
ylides is also observed in reactions of keto-stabilised ylides. The
magnitude of the shift in the energy of the cis-selective TS is not
as great as was generally observed in our previously disclosed
results.7
Ylides with one or more substituents on carbon-3 of the
2-oxoalkylidene group (7 and 8) gave rise to increased E-selectivity
in their reactions with benzaldehyde (Table 1, entries 5 and 6), but
with ortho-heteroatom substituted benzaldehydes 2b–d (Table 1
entries 7–10) they gave considerably increased Z-content by
Scheme 3. Wittig reactions of benzaldehydes 2a and 2b with benzylides 1a and 1b
(produced in situ by the treatment of the parent phosphonium salts with NaHMDS
in THF).
focused on reactions of various alkyl ester stabilised ylides. These
showed the main effect of increased Z-selectivity, but there was
no analogue of the cooperative effect seen in reactions of benzy-
lides. We attributed this to the larger distance of the variable alkyl
group from the ylide
a-carbon in these compounds. In seeking a
series of stabilised ylides that would be more suitable candidates
for the demonstration of the cooperative effect, we turned our
attention to Wittig reactions of keto-stabilised ylides, in which

P. A. Byrne et al. / Tetrahedron Letters 53 (2012) 6701–6704
6703
Table 1
Table 2
Z/E ratioa for reactions of selected 2-oxoalkylidenemethyldiphenylphosphoranes 6–8
(generated in situ from the corresponding phosphonium salts 3–5)b with selected
benzaldehydes 2a–dc
Z/E ratioa for alkenes produced in the reactions of selected acetonylidenetriphenyl-
phosphoranes 18 and 19,b and (tert-butoxycarbonylmethylidene)triphenylphospho-
rane (20)c with selected benzaldehydes 2a and 2d–fd
CHO
-78 °C
CHO
2
Y
(i)
O
O
R
Ph
P
Z
O
Me
P
Me
P
Y
O
O
2
Y
NaHMDS
THF
Y
R
Ph
Ph
Ph
Ph
Ph
R
R
R
-78 or 20 °C
(ii) HCl, H2O
Ph
18 R = CH3
R = CH2OCH3
20 R = O(t-Bu)
2a
3 R = CH3
R = CH2Cl
5 R = t-Bu
6 R = CH3
7
R = CH2Cl
8 R = t-Bu
Y = H
2a
Y = H
9, 11, 21-26
9-17
19
2d Y = Br
4
2b Y = I
2e Y = OMe
2c Y = Cl
2f
Y = Me
2d
Y = Br
Entry
Ylide R
Aldehyde Y
Enone Z/E ratio
Entry
Ylide R
Aldehyde Y
Temp (°C)
Enone Z/E ratio
1
2
3
4
CH3
CH3
CH3
CH2OMe
O(t-Bu)
O(t-Bu)
O(t-Bu)
O(t-Bu)
H
Br
OMe
Br
H
Me
Br
OMe
3:97
1d
CH3
CH3
CH3
CH3
CH2Cl
t-Bu
CH2Cl
CH2Cl
t-Bu
t-Bu
H
H
Cl
Br
H
H
Cl
Br
Br
I
ꢀ45
20
19:81
20:80
33:67
40:60
12:88
17:83
50:50
50:50
48:52
50:50
11:89
10:90
17:83
3:97
3:97
20:80
19:81
2e
3d
ꢀ45
4e
20
5c,e
6c,e
7c
8c
5f
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
ꢀ78
6d,g
7d
8f
9d,g
10d,g
Z/E ratio determined by 1H NMR analysis of the crude product obtained after
aqueous work-up of the reaction mixture. See Supplementary data for full details.
Pre-formed ylides were used unless otherwise indicated.
Ylide 20 was generated in situ from the (commercially available) parent
phosphonium salt (tert-butoxycarbonylmethyl)triphenylphosphonium chloride.
Reactions of this ylide were carried out at 20 °C.
a
b
Z/E ratio determined by 1H NMR analysis of the crude product obtained after
a
c
aqueous work-up of the reaction mixture. See Supplementary data for full details.
Phosphonium salt counterion Z = Clꢀ except where otherwise noted.
b
c
Ylides 6–8 were generated at 20 °C from the parent phosphonium salts 3–5 in
d
Unless otherwise indicated, reactions were stirred at ꢀ78 °C for 20 min and
dry THF by treatment with NaHMDS. The ylide solution was then brought to the
reaction temperature indicated, and the appropriately substituted benzaldehyde
was added dropwise. After stirring for 20 min, the reaction was quenched as indi-
cated in footnotes d–f below.
then allowed to warm to 20 °C.
e
This result was reported in Ref. 22.
d
Quenched with 5% aqueous HCl after 20 min at the reaction temperature
(2a) and 2-methylbenzaldehyde (2f) in THF at 20 °C with, again,
very high E-selectivity (see Table 2, entries 5 and 6) while its reac-
tions with ortho-heteroatom substituted benzaldehydes, show
somewhat more Z-content (Table 2, entries 7 and 8), again clearly
demonstrating the operation of the ortho-heteroatom effect. The
absence of the effect from the reaction of 2-methylbenzaldehyde
(Table 2, entry 6) indicates that the operation of the effect, as in
our previous study,7 is dependent on the ortho-substituent of the
aldehyde bearing a lone pair of electrons.
indicated.
e
Reactions at 20 °C stirred for 4 h, then quenched with 5% aqueous HCl.
Removed from cooling bath after 20 min, and stirred for 12 h at room tem-
f
perature before being quenched with 5% aqueous HCl.
g
Counterion Z = Brꢀ.
comparison, indicating the operation of the same cooperative
effect previously observed in reactions of semi-stabilised ylides
with benzaldehydes.7 In summary, the unsubstituted acetonylide
6 reacts with benzaldehyde (2a) with reasonably high E-selectivity,
and reacts with ortho-heteroatom substituted benzaldehydes 2c–d
with somewhat increased Z-content. By comparison, the shift in
Z-content from the more E-selective reactions of 7 and 8 (which
have greater steric bulk at carbon-3 of the oxoalkylidene moiety)
with benzaldehyde to the reactions of the same ylides with ortho-
heteroatom substituted benzaldehydes 2b–d was much greater.
The Z/E ratios for alkenes produced in reactions of keto-
stabilised ylides derived from triphenylphosphine are shown in
Table 2. These were carried out using pre-formed ylide (i.e. not
generated in situ) at ꢀ78 °C, but were not quenched until they
had been allowed to warm to room temperature. Thus the actual
reaction temperature was not certain for these reactions.
The reaction of acetonylidenetriphenylphosphorane (18) with
benzaldehyde (2a) yielded the expected very high E-selectivity
(Table 2, entry 1), while its reactions with ortho-heteroatom
substituted benzaldehydes 2d–e showed somewhat more
Z-content (Table 2, entries 2 and 3). The reaction of 3-methoxy-2-
oxopropylidenetriphenylphosphorane (19) with 2-bromobenzalde-
hyde (2d) shows a further increase in Z-content (Table 2, entry 4).
Thus the trends observed for the reactions of 2-oxoalkylidene-
methyldiphenylphosphoranes detailed in Table 1 are reproduced
here—including the operation of the cooperative effect—albeit with
far less dramatic shifts in Z-content.
Discussion
It has previously been established that reactions of ester-
stabilised methyldiphenylphosphonium ylides are under kinetic
control.7,16 Although irreversibility of OPA formation has not been
explicitly demonstrated in reactions of keto-stabilised ylides, it
seems reasonable by analogy to assume that the reactions detailed
here are irreversible, especially in light of the increased amounts of
Z-isomer observed here, which occurred in THF where high
E-selectivity is the norm.2,4 Since the ortho-heteroatom effect
observed in our previous study is also present in reactions of
2-oxoalkylidenemethyldiphenylphosphoranes, and since the
ortho-heteroatom-induced increase in Z-content is augmented by
greater steric bulk at carbon-3 of the ylide oxoalkylidene moiety
(cooperative effect), we conclude that the anomalous amounts of
Z-alkene in these reactions is a consequence of the existence of a
cis-selecting TS (shown in Fig. 1b above) similar to that proposed
by us previously.
In the proposed TS, the new factor is a bonding interaction be-
tween the phosphorus and the ortho-heteroatom, causing the TS to
be of sufficiently low energy that it can become competitive with
the normally favoured trans-selective cycloaddition TS. The
crowded environment about phosphorus in the proposed TS re-
sults in a significant reorganisation of the phosphorus-substituents
in comparison with other Wittig TSs. The position of the aldehyde
aryl group is dictated by the phosphorus-heteroatom bond, and, as
Also included in Table 2 are some results from reactions of an
ester-stabilised ylide 20, two of which have been reported in a
previous publication.22 This ylide reacts with each of benzaldehyde

6704
P. A. Byrne et al. / Tetrahedron Letters 53 (2012) 6701–6704
a result, 1–3 steric interactions (between the aryl group and the
ylide phosphorus substituents, Fig. 1b) are relatively unimportant
in this TS. However, there is a concomitant increase in the potential
for 2–3 steric interactions, which are minimised if the ylide
conclude that these reactions are irreversible, and all occur through
a similar cis-selecting [2+2] cycloaddition transition state in which
a bond exists between the phosphorus and the b-heteroatom
substituent on the aldehyde. The consistency of the results further
strengthens the argument that all Li salt-free Wittig reactions
occur by the [2+2] cycloaddition mechanism proposed by Vedejs
and co-workers,8,9,16 and modified for reactions of stabilised ylides
by Aggarwal, Harvey and co-workers.10
a-carbon substituent occupies a pseudo-apical position in the
forming ring. Furthermore, such a TS would benefit from the
favourable anti-parallel orientation of the dipoles along the alde-
hyde C–O and ylide C–C(O) bonds, respectively, as is proposed to
occur in E-selective reactions of stabilised ylides.10 As a result of
the above factors, this TS is cis-selective.
Acknowledgements
The increased Z-content in reactions of ortho-heteroatom
substituted benzaldehydes with ylides that have greater steric bulk
at carbon-3 of the oxoalkylidene moiety (cooperative effect) can be
rationalised using the same TS arguments. Thus the increased 2–3
interactions that would be engendered by the larger substituent on
We thank sincerely the Irish Research Council for Science,
Engineering and Technology (IRCSET) for an EMBARK Scholarship
to P.A.B. We are also very grateful to UCD Centre for Synthesis
and Chemical Biology (CSCB) and the UCD School of Chemistry
and Chemical Biology for access to their extensive analysis facili-
ties, and especially to Dr. Jimmy Muldoon and Dr. Yannick Ortin
for advice and assistance in running NMR spectra.
the ylide a-carbon result in a greater bias towards the cis-selecting
TS since it is better able to accommodate steric bulk in this posi-
tion. That the cooperative effect exists in reactions of keto-
stabilised ylides but not in reactions of ester-stabilised ylides is
consistent with our initial working hypothesis: that the alkyl group
of the ester moiety in ester-stabilised ylides is too far removed
Supplementary data
from the ylide
the shape of the TS, while for keto-stabilised ylides the closer
proximity of the variable alkyl group to the ylide -carbon means
that differences in this group result in different stereoselectivity in
the cycloaddition step.
The existence of the proposed TS is also consistent with selectiv-
ity being dependent on the through-space phosphorus-heteroatom
bond, and thus essentially independent of the electrophilicity of the
carbonyl group, that is the effect is not a result of through-bond
electronic effects exerted by the ortho-heteroatom.
a-carbon for its steric bulk to have a bearing on
Supplementary data (containing details of experimental proce-
dures, assignments of alkene Z/E ratios, and full characterization of
all new phosphonium salts and alkenes) associated with this
article can be found, in the online version, at http://dx.doi.org/10.
1016/j.tetlet.2012.09.123.
a
References and notes
1. Wittig, G.; Geissler, G. Justus Liebigs Ann. Chem. 1953, 580, 44.
2. Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89, 863.
In the reactions of b-heteroatom substituted aldehydes that we
have studied, the steric effects of the phosphorus substituents, and
the influence of these effects on the observed diastereoselectivity,
are consistent across all ylide types, including keto-stabilised
ylides. We had previously noted7 that the magnitude of the shift
in selectivity for Z-alkene or cis-OPA in reactions of b-heteroatom
substituted aldehydes (compared with analogous unsubstituted
aldehydes) decreases in line with the steric bulk of the phospho-
nium moiety. In the context of our proposed TS model, it is easy
to envisage how sterically bulky substituents on phosphorus
would reduce the efficacy of phosphorus–heteroatom bonding. In
reactions of non-stabilised ylides, the magnitude of this shift de-
creases in the order P-phenyldibenzophosphole > Ph3P > EtPh2P;
for semi-stabilised ylides the trend is P-phenyldibenzophosp-
hole ꢁ MePh2P > Ph3P; and for both ester-stabilised and keto-
stabilised ylides we are now in a position to conclude that the
trend is MePh2P > Ph3P. The consistency and generality of these
results supports the existence of the proposed TS.
3. Johnson, A. W. Ylides and Imines of Phosphorus; Wiley: New York, 1993. pp 221–
305, Chapters 8 and 9.
4. Vedejs, E.; Peterson, M. J. Topics in Stereochemistry; In Eliel, E. L., Wilen, S. H.,
Eds.; Wiley: New York, 1994. Vol. 21.
5. Vedejs, E.; Peterson, M. J. Advances in Carbanion Chemistry In Snieckus, V., Ed.;
JAI: Greenwich, CN, 1996. Vol. 2.
6. Gosney, I.; Rowley, A. G. Organophosphorus Reagents in Organic Synthesis In
Cadogan, J. I. G., Ed.; Academic Press, 1979. pp 17–153, Chapter 2.
7. Byrne, P. A.; Gilheany, D. G. J. Am. Chem. Soc. 2012, 134, 9225.
8. Vedejs, E.; Marth, C. F.; Ruggeri, R. J. Am. Chem. Soc. 1988, 110, 3940.
9. Vedejs, E.; Marth, C. F. J. Am. Chem. Soc. 1988, 110, 3948.
10. (a) Robiette, R.; Richardson, J.; Aggarwal, V. K.; Harvey, J. N. J. Am. Chem. Soc.
2006, 128, 2394; (b) Robiette, R.; Richardson, J.; Aggarwal, V. K.; Harvey, J. N. J.
Am. Chem. Soc. 2005, 127, 13468; (c) Harvey, J. N. Faraday Discuss. 2010, 145,
487.
11. Bangerter, F.; Karpf, M.; Meier, L. A.; Rys, P.; Skrabal, P. J. Am. Chem. Soc. 1998,
120, 10653.
12. Maryanoff, B. E.; Reitz, A. B.; Mutter, M. S.; Inners, R. R.; Almond, H. R., Jr. J. Am.
Chem. Soc. 1985, 107, 1068.
13. Reitz, A. B.; Mutter, M. S.; Maryanoff, B. E. J. Am. Chem. Soc. 1984, 106, 1873.
14. Maryanoff, B. E.; Reitz, A. B.; Mutter, M. S.; Inners, R. R.; Almond, H. R., Jr.;
Whittle, R. R.; Olofson, R. A. J. Am. Chem. Soc. 1986, 108, 7664.
15. Appel, M.; Blaurock, S.; Berger, S. Eur. J. Org. Chem. 2002, 1143.
16. Vedejs, E.; Fleck, T. J. Am. Chem. Soc. 1989, 111, 5861.
17. Vedejs, E.; Marth, C. F. J. Am. Chem. Soc. 1990, 112, 3905.
18. For further details of the experimental evidence that has been amassed in
support of the [2+2] cycloaddition mechanism (including our own contribution
to this body of evidence), as well as details of historical significance to its
development, we refer the interested reader to Ref. 7.
Conclusion
Reactions of keto-stabilised ylides and triphenylphosphine-
derived ester-stabilised ylides with ortho-heteroatom substituted
benzaldehydes show an anomalous increase in the amount of
Z-isomer produced compared with the reactions of the same ylides
with benzaldehyde itself or with 2-methylbenzaldehyde. These re-
sults are entirely consistent with those previously observed in
reactions of non-stabilised, semi-stabilised and ester-stabilised
ylides with b-heteroatom substituted ylides,7 although the magni-
tudes of the shifts towards the Z-alkene are not as great in the pres-
ent study. Most significantly, we have demonstrated that the
cooperative effect that operates in reactions of benzylides with
ortho-heteroatom substituted benzaldehydes7 is replicated in
reactions of keto-stabilised ylides with the same aldehydes. We
19. As an aside, we note that betaines are not involved in the mechanism of the Li-
salt free Wittig but, if generated through a non-Wittig route, will collapse to an
OPA and hence give phosphine oxide and an alkene as shown as Scheme 2.
20. As a further aside, we also note that almost all Li-salt free Wittig reactions are
under kinetic control. A very small number of reactions of non-stabilised ylides
are known in which the formation of the cis-OPA isomer only is reversible,
resulting in enhanced production of trans-OPA (and hence E-alkene) relative to
the initial amount formed of this intermediate. Most of these involve ethylides
and benzaldehydes – see Ref. 7 for a full discussion of the phenomenon and an
exhaustive list of the examples for which it is known to operate.
21. Reaction vessels and NMR tubes were protected from light where possible by
being wrapped in aluminium foil. NMR samples of the crude reaction mixtures
were obtained as soon as the work-up of the reaction was complete.
22. Byrne, P. A.; Rajendran, K. V.; Muldoon, J.; Gilheany, D. G. Org. Biomol. Chem.
2012, 10, 3531.