SN2' versus SN2 reactivity: Control of regioselectivity in conversions of Baylis-Hillman adducts

Article abstract of DOI:10.1002/chem.200902487

TiCl₄-induced Baylis-Hillman reactions of α,β- unsaturated carbonyl compounds with aldehydes yield the (Z)-2-(chloromethyl) vinyl carbonyl compounds 5, which react with 1,4diazabicyclo[2.2.2]octane (DABCO), quinuclidine, and pyridines to give the allylammonium ions 6. Their combination with less than one equivalent of the potassium salts of stabilized carbanions (e.g. malonate) yields methylene derivatives 8 under kinetically controlled conditions (SN2' reactions). When more than one equivalent of the carbanions is used, a second SN2' reaction converts 8 into their thermodynamically more stable allyl isomers 9. The second-order rate constants for the reactions of 6 with carbanions have been determined photometrically inDMSO. With these rate constants and the previously reported nucleophilespecific parameters N and s for the stabilized carbanions, the correlation log k (20°C) = s(N + E) allowed us to calculate the electrophilicity parameters E for the allylammonium ions 6 (-19 <￡ <-18). The kinetic data indicate the SN2' reactions to proceed via an addition-elimination mechanism with a rate-determining addition step.

Full text of DOI:10.1002/chem.200902487

FULL PAPER
DOI: 10.1002/chem.200902487
SN2ꢀ versus SN2 Reactivity: Control of Regioselectivity in Conversions of
Baylis–Hillman Adducts
Mahiuddin Baidya, Grygoriy Y. Remennikov, Peter Mayer, and Herbert Mayr*[a]
Dedicated to Dr. Klaus Rçmer on the occasion of his 70th birthday
Abstract: TiCl4-induced Baylis–Hill-
man reactions of a,b-unsaturated car-
bonyl compounds with aldehydes yield
the (Z)-2-(chloromethyl)vinyl carbonyl
compounds 5, which react with 1,4-
When more than one equivalent of the
carbanions is used, a second SN2ꢀ reac-
tion converts 8 into their thermody-
namically more stable allyl isomers 9.
The second-order rate constants for the
reactions of 6 with carbanions have
been determined photometrically in
DMSO. With these rate constants and
the previously reported nucleophile-
specific parameters N and s for the sta-
bilized carbanions, the correlation log
k (208C)=sACTHNUGRTENUNG(N + E) allowed us to cal-
diazabicyclo
G
(DABCO),
culate the electrophilicity parameters E
for the allylammonium ions 6 (ꢀ19<E
<ꢀ18). The kinetic data indicate the
SN2ꢀ reactions to proceed via an addi-
quinuclidine, and pyridines to give the
allylammonium ions 6. Their combina-
tion with less than one equivalent of
the potassium salts of stabilized car-
banions (e.g. malonate) yields methyl-
ene derivatives 8 under kinetically con-
trolled conditions (SN2ꢀ reactions).
Keywords: allylic substitution
·
·
tion–elimination mechanism with
rate-determining addition step.
a
A
E
relationships
·
regioselectivity
AHCTUNGTRENNUNG
Introduction
reactions may either proceed via allyl cations, through addi-
tion–elimination mechanisms, or via concerted SN2ꢀ process-
es.[1–3] Whereas allylic substitutions through intermediate
allyl cations have intensively been explored,[2] only few ex-
amples involving the addition–elimination mechanism
(bottom of Scheme 1) have been reported.[3a,b] Bordwell and
co-workers concluded that the concerted SN2ꢀ mechanism
may well be a myth.[3c,3d]
Nucleophilic substitutions of allyl derivatives often proceed
with allylic rearrangements. As depicted in Scheme 1, these
Since 1968, a large variety of allyl derivatives have
become available through Baylis–Hillman reactions.[4] Be-
cause of their high functionality, these products are valuable
building blocks in organic synthesis, and the regioselective
conversion of these products is an important challenge for
synthetic chemists. The acetoxy groups in Baylis–Hillman
acetates 1 have been substituted by nitrogen, oxygen, sulfur,
and carbon nucleophiles (Scheme 2).[5–9] The catalysis of
Scheme 1. Nucleophilic substitutions with allylic rearrangement.
these reactions by 1,4-diazabicycloACTHNUTRGNEUNG[2.2.2]octane (DABCO)
has been rationalized through the intermediacy of allylam-
monium ions 2, the formation of which has been proven in
few cases.[5c,10]
While amine-catalyzed Baylis–Hillman reactions yield b-
hydroxy-a-methylene carbonyl compounds 4, the corre-
sponding TiCl4-induced reactions give rise to the stereose-
lective formation of (Z)-2-(chloromethyl)vinyl carbonyl
[a] M. Baidya, Dr. G. Y. Remennikov, Dr. P. Mayer, Prof. Dr. H. Mayr
Department Chemie und Biochemie
Ludwig-Maximilians-Universitꢁt Mꢂnchen
Butenandtstrasse 5–13, 81377 Mꢂnchen (Germany)
Fax : (+49)89-2180-77717
E-mail: Herbert.Mayr@cup.uni-muenchen.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902487.
Chem. Eur. J. 2010, 16, 1365 – 1371
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1365

Table 1. Reactions of tertiary amines with the Baylis–Hillman adducts 5.
NR3
R1
R2
X
Product
Yield[a]
[%]
Z:E[b]
OMe
H
Me
Cl
Cl
Cl
Br
Br
Cl
Cl
Br
Cl
Cl
Cl
Cl
6a
6b
6c
6d
6e[c]
6 f
6g
6h
6i
86
82
82
83
87
87
95
93
73
40
43
46
93:7
Me
92:8
Scheme 2. Nucleophilic substitution reaction of Baylis–Hillman acetate in
the presence of DABCO.
NO2
OMe
H
Me
90:10
30:70
40:60[c]
92:8
OMe
OMe
Me
compounds 5 (Scheme 3).[11] We now report that the reac-
tions of compounds 5 with tertiary amines and pyridines
lead to the formation of isolable allylammonium salts, the
OMe
NO2
H
Me
94:6
OMe
Me
17:83
7:93
OMe
OMe
OMe
OMe
Me
6j
9:91
Scheme 3. Tertiary amine and TiCl4-induced Baylis–Hillman reactions.
Me
6k
6l
7:93
alleged intermediates of the amine-catalyzed nucleophilic
substitutions of Baylis–Hillman acetates (Scheme 2). By in-
vestigating products and kinetics of the reactions of these
ammonium ions with various carbanions, we were able to
elucidate the involved mechanisms and derive conditions for
the regioselective conversions of Baylis–Hillman products.
Me
8:92
1
[a] Isolated yield. [b] Determined by H NMR spectroscopy. [c] The same
reaction in a 1:1 H2O/THF solvent mixture produced 6e with a Z/E ratio
of 18:82; ref. [10b].
Results
Isolation of the ammonium salts 6: Treatment of the readily
accessible 2-(halomethyl)vinyl ketones or esters 5[12] with
DABCO or quinuclidine in absolute THF resulted in the
formation of the allylammonium halides 6a–g, which pre-
cipitated as colorless powders (Scheme 4, Table 1). The sub-
Figure 1. Configurational assignment of the ammonium salts 6a–l by
NOESY experiments (arrows indicate observed NOE effects).
duced with a high excess of the E isomers, and the E/Z ratio
was even greater than 90:10 for compounds 6i–l where X=
Cl (Table 1).
Scheme 4. Preparation of ammonium salts 6 from Baylis–Hillman ad-
ducts 5.
Reactions of the Baylis–Hillman ammonium ions 6 with
carbACTHUNRTGNEUNGanions 7: Treatment of 6a with 1.2 equivalents of potas-
sium malonate 7a in DMSO for 10 s or 1 min at ambient
temperature gave mixtures of 8a and 9a in ratios of 96:4
and 92:8, respectively (Scheme 5). When the reaction time
was extended to 24 h, only the styrene derivative 9a was ob-
tained. Because the isolated methylene derivative 8a was
also transformed into 9a by treatment with 7a, one can con-
clude that 8a is the kinetically controlled product, whereas
9a is the product of thermodynamic control.
When strictly kinetically controlled conditions were estab-
lished by treating the ammonium salts 6 with only 0.9 equiv-
alents of the carbanions 7a–g, the methylene derivatives
8a–m were produced selectively (Table 2). Only the reac-
stituted pyridinium salts 6h–l were prepared analogously,
but the isolated yields were lower (except for 6h) because
of their partial solubility in THF. The pyridinium salts 6h–l
are more sensitive to moisture than the corresponding
DABCO and quinuclidine salts 6a–g.
NOESY experiments (Figure 1) showed that the ammoni-
um chlorides 6a–c and 6 f,g obtained with DABCO and qui-
nuclidine were predominantly formed as Z isomers (>
90%), whereas the ammonium bromides 6d,e (R2 =OMe)
were predominantly formed as E isomers (> 60%).[10b] In
contrast, all pyridinium salts 6h–l (X=Cl, Br) were pro-
1366
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1365 – 1371

Control of Regioselectivity in Conversions of Baylis–Hillman Adducts
FULL PAPER
Table 2. Reactions of the ammonium salts 6a–l (1.0 equiv) with substoi-
chiometric amounts (0.90 equiv) of the potassium salts of the carbanions
7a–g (in DMSO, at ambient temperature, 20 h).
Ammonium salt
Carbanion
Product
Yield
[%]
6a
6a
6a
6a
7a
7b
7c
7d
8a
8b
8c
94
81[c]
[b,c]
Scheme 5. Plausible reaction mechanism for the formation of 9a.
tions of the malononitrile anion 7e with the ammonium
salts 6a, 6 f, and 6i gave complex product mixtures which
have not been identified.
8d
79[c]
[a]
6a
6b
7e
7a
8e
8 f
94
6b
7 f
87
6c
6c
6d
7a
7b
7a
8g
8h
8i
85
91
90
The expected SN2ꢀ product 8n from the reaction of the di-
medone carbanion 7h with 6a was not detectable; instead
the tetrahydrochromone 10 was isolated as a 1:1 mixture of
diastereomers (Scheme 6). One of the diastereomers crystal-
lized and was characterized by X-ray analysis (Figure 2).
Reactions of the allyl chloride 5a with nucleophiles: Com-
pounds 8a and 9a, which have been obtained from the am-
monium salt 6a (Scheme 5) can also be directly synthesized
from the Baylis–Hillman chloride 5a (Scheme 7). Combina-
tion of equimolar amounts of 5a and K-7a in DMSO gave a
26:74 mixture of 8a and 9a (Table 3, entry 1).[13] When ther-
modynamic product control was realized by treating 5a with
1.2 equivalents of K-7a in the presence or absence of
DABCO, the arylidene derivative 9a was formed exclusively
(Table 3, entries 2, 3). The resulting 88:12 mixture of (E)-9a
and (Z)-9a was separated by column chromatography. Only
8a, the kinetically controlled product, was observed and iso-
lated with 63% yield, when these reagents were combined
in THF/H2O solution (Table 3, entry 4). It can be expected
that compounds 8b–m as well as the thermodynamically fa-
vored isomers 9b–m can be synthesized analogously from
the allyl chlorides 5 and the carbanions 7 without generating
6d
7 f
8j
85
6e
6e
7a
7g
8k
8l
95
87
6e
7 f
8m
88
71
the corresponding ammonium ions
(Scheme 5, Table 2).
6 as intermediates
6 f
6 f
6i
6i
6i
6j
6k
7a
7e
7a
7d
7e
7a
7a
8a
[a]
8a
78
75
Quantification of the electrophilicities of the ammonium
salts 6: In numerous publications, we have demonstrated
that the rates of the reactions of diarylcarbenium ions with
nucleophiles follow Equation (1), in which E is the electro-
8d
[a]
8a
8a
85
82
[a] Complex mixture. [b] Cyclized product was also observed (see text).
[c] Small amount of SN2 product was also observed.
Chem. Eur. J. 2010, 16, 1365 – 1371
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1367

H. Mayr et al.
bands of the ammonium salts 6
partially overlap with those of
the substitution products 8,
sufficient changes of the UV
spectra allowed us to follow
the reaction kinetics photomet-
rically by using a diode array
spectrometer coupled to an in-
Scheme 6. Reaction of the dimedone anion 7h with the DABCO salt 6a in DMSO.
sertion probe described previ-
ously (see also the Supporting
Information).[14k] Pseudo-first-order rate constants kobs were
obtained by fitting the decays of the absorbances of the sty-
rene chromophor in the ammonium ions 6 to monoexponen-
tial functions. Because the employed allylammonium ions 6
only had a configurational purity of 90–93%, the observed
monoexponential decays indicate that the electrophilicities
of the Z and E isomers do not differ significantly. Plots of
kobs versus the concentrations of the carbanions 7 were
linear (Figure 3), and the second-order rate constants, which
correspond to the slopes of these correlation lines, are listed
in Table 4.
Figure 2. Crystal structure of cis-10 obtained from the reaction of 6a and
K-7h.
Scheme 7. Reaction of 5a with the carbanion of diethyl malonate (7a).
Figure 3. Exponential decay of the absorbance A of 6i at 340 nm and
linear correlation of the pseudo-first-order rate constants kobs versus [7d]
for the reaction of 6i with 7d in DMSO at 208C.
Table 3. Reaction of 5a with 7a under various conditions (2 h, room
temperature).
Entry
5a:7a:DABCO ratio
Solvent
Yield of
Yield of
8a [%]
9a [%]
1
2
3
4
1.0:1.0:0
1.0:1.2:0
1.0:1.2:1.0
1.0:1.2:1.0
DMSO
DMSO
DMSO
THF/H2O (1:1)
24
0
0
67
93
65
0
Only for the reactions of the potassium salt of diethyl
malonate (7) with several ammonium ions 6 the kobs versus
[7] plots showed slight deviations from linearity. As speci-
fied in the Supporting Information, these deviations were
small and have been neglected in the evaluations.
63
philicity parameter, N is the nucleophilicity parameter, and
s is the nucleophile-specific slope parameter.[14] This rela-
tionship also holds for nucleophilic additions of carbanions
to quinone methides,[15] which can be considered as stabi-
lized benzhydryl cations, and to several Michael acceptors.[16]
Because of strongly overlapping UV bands, the photomet-
ric monitoring of the reactions of the coumarin-derived am-
monium ion 6m[10a] with the carbanions 7a–f is problematic.
Therefore, the reaction of 6m with the anion of Meldrumꢀs
acid (7 f) was followed by 1H NMR spectroscopy in
[D6]DMSO. The reactivity of 6m was furthermore character-
ized through its reaction with the blue carbanion 7i (potas-
sium salt of bis(4-nitrophenyl)methyl anion),[18] which has
been followed photometrically under pseudo-first-order con-
ditions using the electrophile 6m in high excess (Scheme 8).
Details are given in the Supporting Information.
log kð20 ꢁCÞ ¼ sðN þ EÞ
ð1Þ
To characterize the electrophilicities of the ammonium
ions 6 by Equation (1), we have studied the kinetics of their
reactions with some of the carbanions 7. Although the UV
1368
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1365 – 1371

Control of Regioselectivity in Conversions of Baylis–Hillman Adducts
FULL PAPER
Table 4. Second-order rate constants (k) for the reactions of ammonium
salts 6 with the carbanions 7 in DMSO at 208C and averaged E parame-
ters for the ammonium ions 6.
groups (6a,f) and pyridine leaving groups (6i–l) shown in
Table 4 allow us to differentiate between these alternatives.
In previous work we have demonstrated that DABCO is a
106-fold better nucleofuge than DMAP.[19] If breaking of the
[b]
Ammonium
ions (E)
Carbanions[a]
k
A
kcalcd
[mꢀ1 sꢀ1
]
[mꢀ1 sꢀ1
ACHTUNGTERNNUNG ]
ꢀ
C N bond were involved in the rate-determining step, 6a
6a (ꢀ18.6)
6a (ꢀ18.6)
6a (ꢀ18.6)
6a (ꢀ18.6)
6 f (ꢀ18.9)
6 f (ꢀ18.9)
6 f (ꢀ18.9)
6 f (ꢀ18.9)
6i (ꢀ19.1)
6i (ꢀ19.1)
6i (ꢀ19.1)
6i (ꢀ19.1)
6j (ꢀ17.9)
6j (ꢀ17.9)
6j (ꢀ17.9)
6j (ꢀ17.9)
6k (ꢀ18.2)
6l (ꢀ18.9)
6m (ꢀ17.3)
6m (ꢀ17.3)
7a
7c
7d
7e
7a
7c
7d
7e
7a
7c
7d
7e
7a
7c
7d
7e
7a
7a
7 f
7i
4.26
11.6
should be considerably more reactive than 6i, because both
compounds differ only in the nature of the leaving groups
(DABCO vs. DMAP). The observation that both com-
pounds show almost identical electrophilic reactivities indi-
cates the occurrence of an addition–elimination mechanism
with the addition step rate-determining.
[c]
6.18
7.64
5.0
3.3
7.7
2.49
[c]
4.21
5.22
5.59
0.42
2.62
2.24
22.1
3.06
17.1
13.1
21.0
7.35
3.2
2.2
5.6
0.64
2.3
1.6
30.8
4.1
The fact that the reactions of 5 and 6 with carbanions give
the methylene compounds 8 under conditions of kinetic con-
trol, and not the thermodynamically favored isomer 9, indi-
cates that these compounds prefer SN2ꢀ over SN2 reactions
which might produce compounds 9 directly. It can, there-
fore, be assumed that the allylammonium ions 6a–l, formally
SN2 products of the reactions of 5 with amines, are analo-
gously formed via two consecutive SN2ꢀ reactions
(Scheme 9).
13.6
9.1
[d]
[d]
8.78ꢄ10ꢀ4[e]
77.4
1.3ꢄ10ꢀ3
58.7
[a] N and s parameters in DMSO are from refs. [15a,17,18]: 7a, N=
20.22, s=0.65; 7c, N=18.82, s=0.69; 7d, N=19.62, s=0.67; 7e, N=
19.36, s=0.67; 7 f, N=13.91, s=0.86; 7i, N=19.92, s=0.67. [b] kcalcd from
E, N, and s using Equation (1). [c] Change of the UV spectrum was not
mono-exponential. [d] kcalcd =k, because only one rate constant was used
1
for the calculation of E. [e] Determined by H NMR kinetics.
Scheme 9. SN2 reaction via two consecutive SN2ꢀ reactions.
Conclusions
Scheme 8. Reaction of 6m with 7i in DMSO.
The mechanistic alternatives for SN2ꢀ mechanisms depicted
in Scheme 1 can be summarized in a More OꢀFerrall–Jencks
diagram as depicted in Figure 4.[20] Whereas the pathways
via allyl cations (path a through the bottom-right corner)
and carbanions (path e through the top-left corner) are well
Substitution of the second-order rate constants k listed in
Table 4 and the previously published N and s parameters for
the carbanions 7[15a,17,18] into Equation (1) yielded the elec-
trophilicity parameters E for 6 which were averaged and
listed in Table 4. Comparison of the experimental rate con-
stants k and those calculated from N, s, and the averaged E
values derived in this work shows agreement within a factor
of 4. It has thus been demonstrated that Equation (1), which
presently covers 40 orders of magnitude, can also be applied
to the SN2ꢀ substitutions studied in this work.
Discussion
The observed second-order kinetics for the reactions of the
allylammonium ions 6 with the carbanions 7 exclude the SN1
mechanism and are in line with a concerted SN2ꢀ reaction
(Scheme 1, middle) as well as with an addition–elimination
mechanism (Scheme 1, bottom).
The similar reactivities of the allylammonium ions (6,
R1 =OMe, R2 =Me) with DABCO and quinuclidine leaving
Figure 4. More OꢀFerrall–Jencks diagram.
Chem. Eur. J. 2010, 16, 1365 – 1371
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1369

H. Mayr et al.
defined, many variants of the concerted SN2ꢀ pathways are
conceivable. Apart from the synchronous SN2ꢀ mechanism
along the diagonal (c), one can imagine a concerted mecha-
nism in which the transition state carries partial carbocation
(b) or partial carbanion character (d). In line with Bord-
wellꢀs skepticism about the feasibility of bimolecular con-
certed mechanisms involving four or more bonds,[3] our ex-
periments clearly show that the SN2ꢀ reactions with the allyl-
stable allylic isomers with terminal double bonds (e.g. 8)
under conditions of kinetic control (e.g. by using less than 1
equivalent of the nucleophiles). If stabilized carbanions are
used as nucleophiles, which have a moderate nucleofugality,
they can be replaced by the same or another nucleophile to
yield the thermodynamically favored allylic isomers 9 with
an internal double bond. Full control of regioselectivity is
thus achieved.
ACHTUNGTRENNUNGammonium ions studied in this work proceed via rate-deter-
mining Michael additions followed by the elimination of the
amines (route (e) in Figure 4).
The rates of the Michael additions can be described by
Equation (1), and Scheme 10 shows that the electrophilicity
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (Ma 673/21-2) and the
Fonds der Chemischen Industrie for support of this work. Valuable sug-
gestions by Dr. Armin R. Ofial are gratefully acknowledged.
[1] General reviews: a) R. H. DeWolfe, W. G. Young, Chem. Rev. 1956,
56, 753–901; b) R. M. Magid, Tetrahedron 1980, 36, 1901–1930;
c) L. A. Paquette, C. J. M. Stirling, Tetrahedron 1992, 48, 7383–7423;
d) J. A. Marshall, Chem. Rev. 1989, 89, 1503–1511; e) A. Alexakis,
J. E. Bꢁckvall, O. Krause, O. Pamies, M. Dieguez, Chem. Rev. 2008,
108, 2796–2823.
[2] Reactions through allyl cations: a) W. G. Young, S. Winstein, H. L.
Goering, J. Am. Chem. Soc. 1951, 73, 1958–1963; b) H. L. Goering,
J. P. Blanchard, E. F. Silversmith, J. Am. Chem. Soc. 1954, 76, 5409–
5418; c) H. L. Goering, E. F. Silversmith, J. Am. Chem. Soc. 1955,
77, 1129–1133; d) H. L. Goering, J. T. Doi, J. Am. Chem. Soc. 1960,
82, 5850–5854; e) H. L. Goering, G. S. Koermer, E. C. Linsay, J.
Am. Chem. Soc. 1971, 93, 1230–1234; f) R. A. Sneen, Acc. Chem.
Res. 1973, 6, 46–53; g) M. Speranza, A. Troiani, J. Org. Chem. 1998,
63, 1020–1026; h) A. Streitwieser, E. G. Jayasree, F. Hasanayn, S. S.-
H. Leung, J. Org. Chem. 2008, 73, 9426–9434.
[3] Addition–elimination mechanisms: a) F. G. Bordwell, D. A. Schex-
ayder, J. Org. Chem. 1968, 33, 3240–3246; b) F. G. Bordwell, T. G.
Mecca, J. Am. Chem. Soc. 1972, 94, 5825–5829; c) F. G. Bordwell,
Acc. Chem. Res. 1970, 3, 281–290; d) F. G. Bordwell, T. G. Mecca, J.
Am. Chem. Soc. 1972, 94, 5829–5837; e) F. G. Bordwell, G. A.
Pagani, J. Am. Chem. Soc. 1975, 97, 118–123.
[4] a) P. I. Dalko, Enantioselective Organocatalysis, Wiley-VCH, Wein-
heim, 2007; b) A. Berkessel, H. Grçger, Asymmetric Organocataly-
sis, Wiley-VCH, Weinheim, 2005; c) D. Basavaiah, A. J. Rao, T. Sa-
tyanarayana, Chem. Rev. 2003, 103, 811–891; d) V. Declerck, J. Mar-
tinez, F. Lamaty, Chem. Rev. 2009, 109, 1–48; e) P. Radha Krishna,
R. Sachwani, P. S. Reddy, Synlett 2008, 2897–2912; f) K. Morita, Z.
Suzuki, H. Hirose, Bull. Chem. Soc. Jpn. 1968, 41, 2815; g) A. B.
Baylis, M. E. D. Hillman, DE-B 215513, 1972; [Chem. Abstr. 1972,
77, 34174q]; h) S. E. Drewes, G. H. P. Roos, Tetrahedron 1988, 44,
4653–4670; i) D. Basavaiah, K. V. Rao, R. J. Reddy, Chem. Soc. Rev.
2007, 36, 1581–1588; j) V. K. Aggarwal, S. Y. Fulford, G. C. Lloyd-
Jones, Angew. Chem. 2005, 117, 1734–1736; Angew. Chem. Int. Ed.
2005, 44, 1706–1708.
[5] For nitrogen nucleophiles: a) H.-L. Cui, X. Feng, J. Peng, J. Lei, K.
Jiang, Y.-C. Chen, Angew. Chem. 2009, 121, 5847–5850; Angew.
Chem. Int. Ed. 2009, 48, 5737–5740; b) J. Li, X. Wang, Y. Zhang,
Tetrahedron Lett. 2005, 46, 5233–5237; c) J. N. Kim, H. J. Lee, K. Y.
Lee, J. H. Gong, Synlett 2002, I173–I175; d) S.-H. Kwon, C.-W. Cho,
Bull. Korean Chem. Soc. 2008, 29, 1835–1838; e) R. Galeazzi, G.
Martelli, M. Orena, S. Rinaldi, Synthesis 2004, 2560–2566; f) W.
Zhong, Y. Zhao, B. Guo, P. Wu, W. Su, Synth. Commun. 2008, 38,
3291–3302; g) J. H. Gong, H. R. Kim, E. K. Ryu, J. N. Kim, Bull.
Korean Chem. Soc. 2002, 23, 789–790; h) M. Ciclosi, C. Fava, R. Ga-
leazzi, M. Orena, J. Sepulveda-Arques, Tetrahedron Lett. 2002, 43,
2199–2202; i) A. Patra, A. K. Roy, S. Batra, A. P. Badhuri, Synlett
2002, 1819–1822.
Scheme 10. Comparison of electrophilicity parameters E (from this work
and refs. [15] and [16]; see also ref. [21]).
parameters of compounds 6a,f,i–l range between ꢀ19<E<
ꢀ18, comparable to quinone methides and benzylidene mal-
onates,[15,16] which enables us to calculate rate constants for
the reactions with various nucleophiles and to predict poten-
tial reaction partners. Reviews on scope and limitations of
the use of the reactivity parameters E, N, and s have been
published.[14]
The fact, that the stepwise SN2ꢀ reactions of the Baylis–
Hillman products 5 and 6 are considerably faster than the
corresponding SN2 reactions, can systematically be employed
in synthesis design. Because the Baylis–Hillman halides 5 as
well as the Baylis–Hillman ammonium ions 6 are the ther-
modynamically more stable allylic isomers (internal double
bonds), they are easily accessible under conditions of ther-
modynamic control. Since both 5 and 6 carry good leaving
groups, their SN2ꢀ reactions with strong nucleophiles (e.g.
carbanions) selectively yield the thermodynamically less
[6] For oxygen nucleophiles: a) J. N. Kim, H. J. Lee, J. H. Gong, Tetra-
hedron Lett. 2002, 43, 9141–9146; b) Y. Du, X. Han, X. Lu, Tetrahe-
1370
www.chemeurj.org
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 1365 – 1371

Control of Regioselectivity in Conversions of Baylis–Hillman Adducts
FULL PAPER
dron Lett. 2004, 45, 4967–4971; c) A. Arfaoui, F. Beji, T. B. Ayed,
H. Amri, Synth. Commun. 2008, 38, 3717–3725.
[7] For hydride and sulfur nucleophiles: a) Y. J. Im, J. M. Kim, J. H.
Mun, J. N. Kim, Bull. Korean Chem. Soc. 2001, 22, 349–350;
b) S. R. S. S. Kotti, X. Xu, G. Li, A. D. Headley, Tetrahedron Lett.
2004, 45, 1427–1431.
[8] For carbon nucleophiles: a) D. J. V. C. van Steenis, T. Marcelli, M.
Lutz, A. L. Spek, J. H. V. Maarseveen, H. Hiemstra, Adv. Synth.
Catal. 2007, 349, 281–286; b) L. D. S. Yadav, V. K. Rai, Tetrahedron
Lett. 2009, 50, 2414–2419; c) Y. M. Chung, J. H. Gong, T. H. Kim,
J. N. Kim, Tetrahedron Lett. 2001, 42, 9023–9026; d) V. Singh, G. P.
Yadav, P. R. Maulik, S. Batra, Tetrahedron 2008, 64, 2979–2991.
[9] For P-based catalysts: a) C.-W. Cho, M. J. Krische, Angew. Chem.
2004, 116, 6857–6859; Angew. Chem. Int. Ed. 2004, 43, 6689–6691;
b) C.-W. Cho, J.-R. Gong, M. J. Krische, Org. Lett. 2004, 6, 1337–
1339; c) H. Park, C.-W. Cho, M. J. Krische, J. Org. Chem. 2006, 71,
7892–7894; d) T.-Z. Zhang, L.-X. Dai, L.-X. Hou, Tetrahedron:
Asymmetry 2007, 18, 1990–1994; e) Y.-Q. Jiang, Y.-L. Shi, M. Shi, J.
Am. Chem. Soc. 2008, 130, 7201–7203.
nisms, Synthesis and Applications (Ed.: K. Matyjaszewski), Marcel
Dekker, New York, 1996, pp. 51–136; d) H. Mayr, O. Kuhn, M. F.
Gotta, M. Patz, J. Phys. Org. Chem. 1998, 11, 642–654; e) H. Mayr,
T. Bug, M. F. Gotta, N. Hering, B. Irrgang, B. Janker, B. Kempf, R.
Loos, A. R. Ofial, G. Remennikov, H. Schimmel, J. Am. Chem. Soc.
2001, 123, 9500–9512; f) H. Mayr, B. Kempf, A. R. Ofial, Acc.
Chem. Res. 2003, 36, 66–77; g) H. Mayr, A. R. Ofial in Carbocation
Chemistry (Eds.: G. A. Olah, G. K. S. Prakash), Wiley, Hoboken
(N.J.), 2004, Chapter 13, pp. 331–358; h) A. R. Ofial, H. Mayr, Mac-
romol. Symp. 2004, 215, 353–367; i) H. Mayr, A. R. Ofial, Pure
Appl. Chem. 2005, 77, 1807–1821; j) H. Mayr, A. R. Ofial, Nachr.
Chem. 2008, 56, 871–877; k) H. Mayr, A. R. Ofial, J. Phys. Org.
Chem. 2008, 21, 584–595.
[15] a) R. Lucius, R. Loos, H. Mayr, Angew. Chem. 2002, 114, 97–102;
Angew. Chem. Int. Ed. 2002, 41, 91–95; b) D. Richter, N. Hampel,
T. Singer, A. R. Ofial, H. Mayr, Eur. J. Org. Chem. 2009, 3203–
3211.
[16] a) T. Lemek, H. Mayr, J. Org. Chem. 2003, 68, 6880–6886; b) F. See-
liger, S. T. A. Berger, G. Y. Remennikov, K. Polborn, H. Mayr, J.
Org. Chem. 2007, 72, 9170–9180; c) S. T. A. Berger, F. H. Seeliger,
F. Hofbauer, H. Mayr, Org. Biomol. Chem. 2007, 5, 3020–3026;
d) O. Kaumanns, R. Lucius, H. Mayr, Chem. Eur. J. 2008, 14, 9675–
9682; e) O. Kaumanns, H. Mayr, J. Org. Chem. 2008, 73, 2738–2745.
[17] T. Bug, H. Mayr, J. Am. Chem. Soc. 2003, 125, 12980–12986.
[18] S. T. A. Berger, T. Lemek, H. Mayr, ARKIVOC 2008, (x), 37–53.
[19] a) M. Baidya, S. Kobayashi, F. Brotzel, U. Schmidhammer, E.
Riedle, H. Mayr, Angew. Chem. 2007, 119, 6288–6292; Angew.
Chem. Int. Ed. 2007, 46, 6176–6179; b) F. Brotzel, B. Kempf, T.
Singer, H. Zipse, H. Mayr, Chem. Eur. J. 2007, 13, 336–345.
[20] a) R. A. More OꢀFerrall, J. Chem. Soc. B 1970, 274–277; b) W. P.
Jencks, Chem. Rev. 1972, 72, 705–718; c) W. P. Jencks, Acc. Chem.
Res. 1980, 13, 161–169.
[10] a) S. E. Drewes, O. L. Njamela, N. D. Emslie, N. Ramesar, J. S. Field,
Synth. Commun. 1993, 23, 2807–2815; b) D. Basavaiah, N. Kumara-
gurubaran, Tetrahedron Lett. 2001, 42, 477–479.
[11] a) M. Shi, J.-K. Jiang, S.-C. Cui, Molecules 2001, 6, 852–868; b) M.
Shi, Y.-S. Feng, J. Org. Chem. 2001, 66, 406–411; c) M. Shi, J.-K.
Jiang, Tetrahedron 2000, 56, 4793–4797; d) G. Li, J. Gao, H.-X. Wei,
M. Enright, Org. Lett. 2000, 2, 617–620; e) Z. Han, S. Uehira, H.
Shinokubo, K. Oshima, J. Org. Chem. 2001, 66, 7854–7857.
[12] a) R. Buchholz, H. M. R. Hoffmann, Helv. Chim. Acta 1991, 74,
1213–1220; b) F. Ameer, S. E. Drewes, N. D. Emslie, P. T. Kaya,
R. L. Mann, J. Chem. Soc. Perkin Trans. 1 1983, 2293–2295.
[13] A similar transformation was previously observed by Drewes and
Emslie when synthesizing precursors for sceleranecic and integerri-
necic acids: S. E. Drewes, N. D. Emslie, J. Chem. Soc. Perkin Trans.
1 1982, 2079–2083.
[21] For nucleophilicity and electrophilicity parameters see the database
on the WWW: http://www.cup.lmu.de/oc/mayr.
[14] a) H. Mayr, M. Patz, Angew. Chem. 1994, 106, 990–1010; Angew.
Chem. Int. Ed. Engl. 1994, 33, 938–957; b) H. Mayr, M. Patz in
Frontiers of Studies on Organic Reaction Mechanisms (Eds.: T.
Okuyama, S. Tomoda, H. Yamataka), Tokyo Kagaku Dojin, Tokyo,
1995, pp. 33–43; c) H. Mayr in Cationic Polymerization: Mecha-
Received: September 9, 2009
Revised: October 7, 2009
Published online: December 2, 2009
Chem. Eur. J. 2010, 16, 1365 – 1371
ꢃ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1371