Divergent pathways and competitive mechanisms of metathesis reactions between 3-arylprop-2-ynyl esters and aldehydes: An experimental and theoretical study

Article abstract of DOI:10.1002/chem.201402551

Mechanistic studies of the reaction between 3-arylprop-2-ynyl esters and aldehydes catalyzed by BF₃Et₂O were performed by isotopic labeling experiments and quantum chemical calculations. The reactions are shown to proceed by either a classical alkyne-carbonyl metathesis route or an unprecedented addition-rearrangement cascade. Depending on the structure of the starting materials and the reaction conditions, the products of these reactions can be Morita-Baylis-Hillman (MBH) adducts that are unavailable by traditional MBH reactions or E- and Z-α,β-unsaturated ketones. ¹⁸O-Labeling studies suggested the existence of two different reaction pathways to the products. These pathways were further examined by quantum chemical calculations that employed the DFT(wB97XD)/6-311+G(2d,p) method, together with the conductor-like screening model for realistic solvation (COSMO-RS). By using the wB97XD functional, the accuracy of the computed data is estimated to be 1-2kcalmol^-1, shown by the careful benchmarking of various DFT functionals against coupled cluster calculations at the CCSD(T)/aug-cc-pVTZ level of theory. Indeed, most of the experimental data were reproduced and explained by theory and it was convincingly shown that the branching point between the two distinct mechanisms is the formation of the first intermediate on the reaction pathway: either the four-membered oxete or the six-membered zwitterion. The deep mechanistic understanding of these reactions opens new synthetic avenues to chemically and biologically important α,β-unsaturated ketones.

Full text of DOI:10.1002/chem.201402551

DOI: 10.1002/chem.201402551
Full Paper
&
Reaction Mechanisms
Divergent Pathways and Competitive Mechanisms of Metathesis
Reactions between 3-Arylprop-2-ynyl Esters and Aldehydes: An
Experimental and Theoretical Study
[c]
[a]
Cristina Trujillo,[a] Goar Sꢀnchez-Sanz,[a, b] Ieva Karpaviciene, Ullrich Jahn,
ˇ
˙
[c]
[a]
ˇ
˙
ˇ
Inga Cikotiene,* and Lubomꢁr Rulꢁsek*
Abstract: Mechanistic studies of the reaction between 3-ar-
ylprop-2-ynyl esters and aldehydes catalyzed by BF3·Et2O
were performed by isotopic labeling experiments and quan-
tum chemical calculations. The reactions are shown to pro-
ceed by either a classical alkyne–carbonyl metathesis route
or an unprecedented addition–rearrangement cascade. De-
pending on the structure of the starting materials and the
reaction conditions, the products of these reactions can be
Morita–Baylis–Hillman (MBH) adducts that are unavailable by
traditional MBH reactions or E- and Z-a,b-unsaturated ke-
tones. 18O-Labeling studies suggested the existence of two
different reaction pathways to the products. These pathways
were further examined by quantum chemical calculations
that employed the DFT(wB97XD)/6-311+G(2d,p) method, to-
gether with the conductor-like screening model for realistic
solvation (COSMO-RS). By using the wB97XD functional, the
accuracy of the computed data is estimated to be 1–2 kcal
molꢀ1, shown by the careful benchmarking of various DFT
functionals against coupled cluster calculations at the
CCSD(T)/aug-cc-pVTZ level of theory. Indeed, most of the ex-
perimental data were reproduced and explained by theory
and it was convincingly shown that the branching point be-
tween the two distinct mechanisms is the formation of the
first intermediate on the reaction pathway: either the four-
membered oxete or the six-membered zwitterion. The deep
mechanistic understanding of these reactions opens new
synthetic avenues to chemically and biologically important
a,b-unsaturated ketones.
Introduction
because this structural motif is found in many pharmaceutical
and bioactive natural compounds.[3] The most common classi-
cal synthetic approaches toward the synthesis of a,b-unsatu-
rated ketones are aldol-type condensations or Wittig reac-
tions.[4] More recently, synthetic protocols such as carbonyla-
tive Heck coupling of aryl halides and styrenes,[5] Meyer–Schus-
ter rearrangements of propargylic alcohols,[6] and catalytic
alkene metathesis[7] have been reported. These methods have
their own scope but are sometimes problematic because of
limited selectivity (aldol-type condensations), formation of
waste (Wittig reaction), harsh conditions and requirement for
specialized equipment (carbonylation reactions, which use
toxic carbon monoxide), or preparation of vinyl ketones as
starting materials (alkene metathesis).
The formation of carbon–carbon bonds is at the heart of syn-
thetic organic chemistry.[1] Among the vast number of methods
available, those that are atom economic and environmentally
benign are the most attractive for application in organic syn-
thesis.[2] Access to a,b-unsaturated ketones has a high priority
ˇ
[a] Dr. C. Trujillo, Dr. G. Sꢀnchez-Sanz, Dr. U. Jahn, Dr. L. Rulꢁsek
Institute of Organic Chemistry and Biochemistry
Gilead Sciences Research Center & IOCB
Academy of Sciences of the Czech Republic
Flemingovo nꢀm. 2, 166 10 Praha 6 (Czech Republic)
E-mail: rulisek@uochb.cas.cz
[b] Dr. G. Sꢀnchez-Sanz
The alkyne–carbonyl metathesis reaction in the presence of
a Lewis acid represents an atom-economic strategy for the
preparation of a,b-disubstituted a,b-unsaturated ketones. It
can be considered a useful alternative to the above-mentioned
methods; many of them do not work well for this substitution
pattern. The carbonyl–alkyne metathesis is especially attractive
because the carbonyl group and double bond form simultane-
ously, thus selectivity issues do not arise. This reaction was first
reported by Vieregge and co-workers in 1959.[8] The authors
proposed a mechanism that involved three steps: 1) nucleo-
philic attack of the alkyne to the Lewis acid (LA) activated alde-
hyde with formation of ion-pair intermediate I, 2) electrocyclic
School of Physics & Complex and Adaptive Systems Laboratory
University College Dublin, Belfield, Dublin 4 (Ireland)
ˇ
ˇ
˙
˙
[c] I. Karpaviciene, Dr. I. Cikotiene
Department of Organic Chemistry
Faculty of Chemistry, Vilnius University
Naugarduko 24, 03225, Vilnius (Lithuania)
E-mail: inga.cikotiene@chf.vu.lt
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402551. It contains product characteri-
1
zation, copies of H and 13C NMR spectra, the equilibrium geometries and
molecular energies (primary computational data) of all the molecules stud-
ied, the B3LYP+D3 reaction pathways for all three reactions studied theo-
retically, and a comparison between various functionals and different sol-
vent energies.
Chem. Eur. J. 2014, 20, 10360 – 10370
10360
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Alkyne–carbonyl metathesis reaction.
ring closure to give an oxete intermediate II, 3) cycloreversion
to the corresponding a,b-unsaturated ketone E (Scheme 1).
This mechanism was further supported by Middleton in
1965 by isolation of the corresponding oxete intermediate
from the reaction between ethoxyacetylene and hexafluoroa-
cetone.[9] Moreover, it was shown that the isolated oxete un-
derwent cycloreversion to form the final unsaturated ester. The
plausibility of the suggested three-step mechanism was also
confirmed by computational methods,[10] albeit only for forma-
tion of methyl acrylate from formaldehyde and methoxyacety-
lene at the HF/6-31G* and B3LYP/6-31G* level of theory. The
authors provided arguments in favor of a pathway that in-
volved the formation of the CꢀC bond in the oxete intermedi-
ate.
Scheme 3. Scope and limitations of the classical Morita–Baylis–Hillman
reaction.
ducts 5, which are unavailable by traditional MBH reactions
from aryl vinyl ketones because of the high reactivity of the
starting materials in fast follow-up processes (Scheme 3).[15]
Propargylic esters are a specific class of alkynes with inter-
esting chemical properties. It should be noted that usually
gold-initiated carbophilic activation of the alkyne unit is the
dominant synthetic application which leads to 1,2- or 1,3-acyl-
oxy shifts and formation of metal vinyl carbenoid species
or metal-complexed allenic intermediates, respectively
(Scheme 4).[16]
It should be noted that the usual outcome of the intermo-
lecular alkyne–carbonyl metathesis reaction is the formation of
thermodynamically stable E enones and that a number of
oxo-[11] and carbophilic[12] Lewis or Brønsted acids have been
employed in inter- and intramolecular alkyne–carbonyl meta-
thesis reactions.
This method has not been frequently used in the past, but
recently we found that alkyne–carbonyl metathesis between
3-arylprop-2-ynyl esters 1 and aldehydes 2 has a very broad
scope and other product classes can be accessed
(Scheme 2).[13] It was found that the structure of both starting
materials influences the outcome of the reaction significantly.
When esters 1 reacted with aliphatic aldehydes, E enones 3
were always formed. In contrast, reactions with aromatic alde-
hydes gave mixtures of E- and Z-enones 3 and 4.[14] Surprising-
ly, the combination of an electron-donating group (EDG) in
1 with an electron-withdrawing group (EWG) on the benzalde-
hydes 2 led to the unprecedented and selective formation of
acetylated or benzoylated Morita–Baylis–Hillman (MBH) ad-
Scheme 4. Metal-mediated 1,2- and 1,3-acyloxy shifts of propargylic esters.
Previously described results[13] indicate a broad and diver-
gent applicability of the alkyne–carbonyl metathesis protocol.
Nothing is known about the mechanistic course of the oxophil-
ic Lewis acid mediated alkyne–carbonyl metathesis of esters 1.
The formation of the MBH adducts is also not understood.
In this work, we present an experimental and theoretical
study that aims to shed light on the divergent mechanistic
pathways that lead either to a,b-unsaturated ketones 3 and 4
or MBH adducts 5. Extensive mechanistic experiments with
18O-labeled aldehydes show the fate of the labeled oxygen
atom. Several mechanistic pathways can be proposed on this
basis, which are rationalized by a thorough and detailed theo-
retical study, which includes benchmark studies to validate the
applied DFT methods against the reference coupled cluster
(CCSD(T)/aug-cc-pVTZ) calculations. Our results bear many im-
Scheme 2. The diversity of products from reactions between 3-arylprop-2-
ynyl esters 1 and aldehydes 2.
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10361
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
plications for further development of these useful alkyne–car-
bonyl metathesis reactions.
the MBH adduct 5cd. Besides the main product 5cd, 12% of
hydrolyzed compound 5.1cd with a free-hydroxy group was
also isolated. The aldehyde oxygen atom was incorporated
into the CH-18O(C=O)Ph and CH-18OH positions.
Results and Discussion
The incorporation of the 18O atom into the products was
proven by 13C NMR spectroscopy.[17] Inverse-gated 13C NMR de-
coupling experiments[18] were used to allow integration of the
areas of the carbonyl signals. Upfield chemical shifts of approx-
imately 0.04 ppm (4 Hz) were found for the 18O-carbonyl
carbon atoms in both 3aa and 3ab. Similarly, 4 Hz upfield
shifts were observed in the 13C NMR spectra for the ester car-
bonyl groups of 3ac, 3bd, and 4ac, whereas no change in
chemical shift occurred at the ketone carbon atoms. The
13C NMR spectra of 5bd and 5cd showed upfield-shifted
values for both carbon atoms of the 18O-labeled CH-O(C=O)R
functionalities. Moreover, the IR spectra of 3aa and 3ab clearly
showed a shift of the carbonyl bands to lower wavenumbers,
specifically from n˜ =1738 to 1708 cmꢀ1 for the ester carbonyl
bands and from n˜ =1652 to 1646 cmꢀ1 for the ketone carbonyl
bands. Similar shifts of the ester carbonyl bands to lower
wavenumbers (n˜ =1738 and 1732 cmꢀ1 to n˜ =1714 and
1711 cmꢀ1, respectively) were observed in the IR spectra of
3ac, 3bd, and 4ac.
Experimental data for the reaction between 3-arylprop-2-
ynyl esters 1a–d and aldehydes 2a–d
Both the aldehydes 2 and starting alkynes 1 were carefully
chosen for the current study (Table 1) to cover the full reactivi-
ty spectrum indicated above (Scheme 2).[13]
The reaction between 3-phenylprop-2-ynylcarboxylate (1a)
and 18O-acetaldehyde (2a) provided 18O-labeled E enone 3aa
as the exclusive product. The 18O atom was incorporated in
both the ester carbonyl and keto groups in a 1.8:1 ratio
(Table 1, entry 1). The reaction between 1a and 18O-2-fluoro-
benzaldehyde (2b) led to formation of both enones 3ab and
4ab in a 2:1 ratio. After isolation and purification of the prod-
ucts it was found that the 18O atom was incorporated into
both carbonyl groups of 3ab. The ratio of labeled ketone and
labeled ester groups in 3ab was 1:3. In contrast, the 18O label
was found only in the ester carbonyl group of 4ab (Table 1,
entry 2).
The experiments with the more electron-poor aldehyde sub-
strates revealed exclusive incorporation of the 18O label in the
ester groups of products 3–5. Specifically, the reaction be-
tween 1a and 2,4-dichlorobenzaldehyde (2c) provided com-
pounds 3ac and 4ac with labeled ester carbonyl groups
(Table 1, entry 3). Reactions between 1b or 1c and 18O-4-nitro-
benzaldehyde (2d) displayed a divergent outcome (Table 1, en-
tries 4 and 5). Alkyne 1b and aldehyde 2d were stirred in
CH2Cl2 in the presence of BF3·Et2O (1 equiv) at room tempera-
ture to provide 5bd and a small amount of 3bd (Table 1,
entry 4). It was proven that the 18O atom was incorporated
into the ester carbonyl group of 3bd and the sp3 hybridized
oxygen atom of the ester functionality in 5bd. In contrast, the
reaction between 1c and 2d led to the exclusive formation of
The data presented in Table 1 indicate that at least two com-
peting reaction pathways operate to lead to the formation of
the products. Path 1 explains the formation of E enones 3 by
a classical alkyne–carbonyl metathesis route, which consists of
a [2+2] reaction via intermediates I and II (Scheme 5). In con-
trast, the location of the 18O label in the ester groups of 3–5
cannot be explained by this pathway, but suggests a mecha-
nism that includes intramolecular nucleophilic attack of the
ester group onto the initial ion-pair intermediate I to form
a six-membered zwitterion V, which stabilizes by an acyl group
transfer via VI and IV to the products (Scheme 5, path 2).
To gain support for path 2, additional control experiments
were performed (Table 2). As shown earlier,[13] the reactions be-
tween 1a and aldehydes 2 under the optimal conditions
Table 1. Reactions of selected 3-arylprop-2-ynyl esters (1) with 18O-labeled aldehydes (2).
Entry
Alkyne
Aldehyde
R1
Reaction time
[h]
Products
Isolated overall
yield [%]
3
4
5
1
Ar
R
2
Labeled
ketone
Labeled
ester
carbonyl
1
2
3
4
5
1a
1a
1a
1b
1c
Ph
Ph
Ph
Me
Me
Me
Me
Ph
2a
2b
2c
2d
2d
Me
2-FC6H4
2,4-Cl2C6H4
4-NO2C6H4
4-NO2C6H4
24
24
24
5 min
5 min
3aa
27
65
71
89
78
1
0.5
–
–
–
1.8
1.5
2
0.15
–
–
1
1
–
–
–
–
–
1
1
3ab, 4ab
3ac, 4ac
3bd, 5bd
5cd[a]
4-MeOC6H4
4-MeOC6H4
[a] Hydrolyzed 5.1cd with the 18O-label in the hydroxy group was also isolated in 12% yield.
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10362
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ity of 3bd (Table 2, entry 7).
Lowering the reaction tempera-
ture to 108C resulted in slower
conversion of the starting mate-
rial, but formation of mixture of
3bd and 5bd in 1:1 ratio
(Table 2, entry 8). It is interesting
to note that a mixture of 3bd
and 4bd formed during pro-
longed stirring of 1b and 2d in
the
presence
of
BF3·Et2O
(1 equiv) at room temperature
for 24 h (Table 2, entry 9). The
same result was obtained when
pure 5bd was stirred in CH2Cl2
in the presence of BF3·Et2O
(1 equiv). These results support
path 2 (Scheme 5).
To prove the role of the prop-
argylic ester group in the formal
alkyne–carbonyl metathesis reac-
tion, 4-(4-methoxyphenyl)but-3-
ynyl acetate (1d) was applied in
the BF3·Et2O catalyzed reaction
with 18O-labeled aldehyde 2c
(Scheme 6). Introduction of the
additional methylene group
should effectively shut down
Scheme 5. Two plausible mechanistic pathways. Path 1 (red arrows): Classical alkyne–carbonyl metathesis route,
followed by 1,3 carboxylate migration (as proposed originally).[13] Path 2 (blue arrows): A novel nucleophilic addi-
tion/rearrangement cascade reaction. LA=BF3·Et2O. Depicted structures represent either energy minima or transi-
tion states.
(BF3·Et2O (1 equiv), CH2Cl2, rt) were slow and took 24 h for full
conversion (Table 2, entries 1, 3, and 5). The major products of
these three reactions were a,b-unsaturated ketones 3 and 4.
However, when the BF3·Et2O/TMSOTf couple was used at lower
temperature (ꢀ108C), known to generate the more powerful
Lewis acid BF2OTf in situ,[19] the previously unobserved MBH
adducts 5 formed competitively to the major E- and Z-enones
3 and 4 (Table 2, entries 2, 4, and 6). 3-(4-Methoxyphenyl)prop-
2-ynyl acetate (1b) reacted very smoothly with 2d in the pres-
ence of BF3·Et2O (1 equiv) at room temperature[13] and 5bd
was isolated as the main product, together with a small impur-
path 2, because formation of V and the subsequent fragmenta-
tion to VI will not be possible. In the event, (E)-4-(2,4-dichloro-
phenyl)-3-(4-methoxybenzoyl)but-3-enyl acetate (3dc) was
formed as the exclusive product of the reaction. Moreover, the
13C NMR spectrum of isolated 3dc revealed a 0.05 ppm (5 Hz)
upfield shift of the ketone carbonyl resonance, whereas the
ester carbonyl carbon atom remained unshifted. This indicated
that the 18O atom was incorporated in the keto group and the
classical alkyne–carbonyl metathesis via the oxete intermediate
proceeded.
Table 2. Reactions of selected 3-arylprop-2-ynyl esters 1 and aldehydes 2 with BF3·OEt2 in CH2Cl2.
Entry
Alkyne
Aldehyde
Additive
T [8C]
Reaction time [h][a]
Product ratio
Overall yield [%]
1
2
3
4
5
6
7
8
9
1a
1a
1a
1a
1a
1a
1b
1b
1b
2a
2a
2b
2b
2c
2c
2d
2d
18O-2d
–
20
ꢀ10
20
ꢀ10
20
ꢀ10
20
10
24
0.5
24
0.5
24
0.5
5 min
25 min
24
3aa/4aa/5aa=1:0:0
3aa/4aa/5aa=1:0:0.3
3ab/4ab/5ab=2:1:0
3ab/4ab/5ab=5.3:1.1:1
3ac/4ac/5ac=2:1:0
3ac/4ac/5ac=1:0:1.5
3bd/4bd/5bd=0.15:0:1[d]
3bd/4bd/5bd=1:0:1
3bd/4bd/5bd=1.3:1:0
26[b]
16[b]
77
69
69
55
82
51
64
TMSOTf[c]
–
TMSOTf[c]
–
TMSOTf[c]
–
–
–
20
[a] Isolation of products was performed after full conversion of the starting alkyne 1. [b] The low overall yields can be explained by possible self-condensa-
tion side reaction of the aliphatic aldehyde under the reaction conditions. [c] To solve the problem of slow reactivity of the starting materials at lower tem-
peratures we used the synergistic couple of BF3·Et2O and TMSOTf. [d] Compounds 3bd and 5bd were isolated as a mixture due to their similiar Rf values.
1
The product ratio was determined from the H NMR spectrum.
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10363
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the CCSD(T)/aug-cc-pVTZ values
for the small system, both
wB97XD and B3LYP+D3 have
low standard deviations of s
ꢂ1.6 kcalmolꢀ1 (s is almost
equivalent to the commonly
used mean-unsigned or mean-
absolute deviation, MUD/MAD—
one of the most important quan-
tities to characterize the accura-
cy of a given set of data with re-
spect to the reference set); the
B3LYP+D3 value also has a re-
markably low mean-signed devi-
ation (a measure of “systematic
shift” in the calculated values) of
ꢀ0.14 kcalmolꢀ1 (cf. ꢀ1.95 kcal
Scheme 6. Isotopic labeling experiment for the reaction between 4-(4-methoxyphenyl)but-3-ynyl acetate (1d) and
18O-labeled 2,4-dichlorobenzaldehyde (2c).
Theoretical calculations of plausible reaction pathways
molꢀ1; wB97XD). The M06-2X functional is almost equally good
whereas other functionals tested perform considerably worse.
These trends are more-or-less repeated in two other compari-
sons: with the CCSD(T)/aug-cc-pVDZ references values for the
larger model system, as well as with the CCSD(T)/aug-cc-pVDZ
calculations for smaller systems (see Table S1 in the Supporting
Information). Again, wB97XD, B3LYP+D3, and M06-2X are the
best from the DFT functionals studied, whereas MP2 and SCS-
MP2 have a very good performance for the larger model
system. From the data presented in Tables 3 and S1 (see the
Supporting Information), another very important parameter
can be highlighted: the energy difference (DEP1/P2) between
Comments on the accuracy of the computed activation
energies
The experimental data, notably the observed ratios of the
products formed, is suggestive of fairly small free-energy differ-
ences between the transition-state barriers for various path-
ways. With this in mind, reference CCSD(T)/aug-cc-pVTZ and
CCSD(T)/aug-cc-pVDZ calculations were carried out for small
model systems, namely HCHO·BF3 + H-CꢁC-CH2ꢀOꢀCHO and
CF3CHO·BF3 + CH3OꢀCꢁCꢀCH2ꢀOꢀCOCH3. All structures on
paths 1 and 2 were considered and constructed from the reac-
tion between 1b and 2d (see
Figure 4, below) by truncation of
the system and optimization of
Table 3. The performance of the selected DFT functionals and SCS-MP2 method with respect to the reference
CCSD(T) values.[a] Relative energies of various intermediates and transition states along the HCHO·BF3
H-CꢁC-CH2ꢀOꢀCHO reaction pathway (with respect to the sum of energies of isolated reactant).
+
hydrogen atoms only. The model
reaction pathways perfectly cap-
ture all of the important features
of the target systems and can
thus be considered as excellent
benchmark systems. Notably, the
size of the larger system (28
atoms) was impractical for the
Structure
Method
CCSD(T) SCS-MP2
30.0 40.8
wB97XD
B3LYP+D3
M06-2X
I
II
V
VI
30.2
ꢀ26.8
ꢀ9.6
4.8
29.5
ꢀ21.1
ꢀ5.1
5.3
29.2
ꢀ24.7
ꢀ11.9
6.9
ꢀ23.8 ꢀ20.7
ꢀ8.2
ꢀ1.4
5.4
10.1
3bd
4bd
IV
5bd
5bd’
TSR!I
TSI!II
TSI!V
TSII!3bd
TSV!VI
TS3bd!IV
TSVI!5bd
TSIV!5bd
TS5bd’!3bd
s(MSD)[b] max–min-
(dev)[c]
ꢀ37.5 ꢀ33.5
ꢀ34.2 ꢀ30.2
ꢀ17.7 ꢀ12.3
ꢀ36.1 ꢀ32.2
ꢀ39.4 ꢀ35.1
ꢀ40.4
ꢀ36.6
ꢀ21.9
ꢀ38.3
ꢀ40.8
23.4
34.4
6.5
ꢀ2.8
18.9
ꢀ20.5
3.5
ꢀ37.9
ꢀ34.5
ꢀ18.3
ꢀ36.2
ꢀ37.9
22.5
35.9
8.7
ꢀ1.2
17.7
ꢀ18.1
3.5
ꢀ37.0
ꢀ33.4
ꢀ20.8
ꢀ36.1
ꢀ39.4
22.1
33.5
5.4
0.0
20.5
ꢀ17.9
5.7
ꢀ18.9
ꢀ16.3
CCSD(T)/aug-cc-pVTZ
calcula-
tions, therefore the smaller aug-
cc-pVDZ basis set was used in-
stead. For the former (small)
system and selected functionals,
the computed data are summar-
ized in Table 3; the full table for
both systems and all the tested
functionals is available in the
22.3
36.2
7.8
29.9
43.7
15.8
5.0
ꢀ2.0
19.9
27.1
ꢀ15.8 ꢀ10.2
3.9 8.2
ꢀ16.2 ꢀ10.6
ꢀ15.0 ꢀ9.5
1.92 (5.86) 7.62
Supporting
Table S1.
Information,
ꢀ20.5
ꢀ19.1
ꢀ18.1
ꢀ17.5
1.65 (ꢀ1.95) 5.77
1.58 (ꢀ0.14) 5.68
1.76 (ꢀ0.70) 5.65
Based on the data presented
in Tables 3 and S1 (see the Sup-
porting Information), we decided
to use wB97XD and B3LYP+D3
functionals for the rest of this
study. When referenced against
[a] aug-cc-pVTZ basis set; all values are in kcalmolꢀ1
.
[b] s is the standard deviation (similar
deviation); MSD=mean-signed deviation=
(Ei,XꢀEi,CCSD(T)). [c] max/min(dev)=max/min(Ei,XꢀEi,CCSD(T)).
to
MUD/MAD=mean-unsigned/mean-absolute
P
1/n
i
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10364
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the key (branching) points to divert the reaction either to
path 1 or 2 (i.e. TSI!II versus TSI!V). In all three cases, the
values are almost quantitatively predicted by wB97XD and the
DDEP1/P2(CCSD(T)-wB97XD) amounts to 1.1, 0.4, and 0.1 kcal
molꢀ1 for the two basis sets and two systems studied. The
B3LYP+D3 and M06-2X calculations are only slightly worse.
The presented benchmark values also provided estimates of
the error bars that must be taken into account in analysis of
the computed activation barriers and reaction energies for
target reactions. It can be readily observed that at least 1–
2 kcalmolꢀ1 uncertainty in the relative values of computed ac-
tivation energies and stabilities of reaction intermediates must
be allowed for.
than the reactants (see Figure S4 in the Supporting Informa-
tion). This stabilization even outweighs the generic unfavorable
entropy contribution of approximately 10–15 kcalmolꢀ1 associ-
ated with an A+B!C chemical reaction and results in a favora-
ble free-energy change of DG(1a!II)=ꢀ3.2 kcalmolꢀ1. For
the TSII!3aa transition state, which corresponds to opening of
the four-membered ring of intermediate II, a barrier of DG°-
(TSII!3aa)=14.7 kcalmolꢀ1 (DE°(TSII!3aa)=16.3 kcalmolꢀ1 in
vacuo; see Figure S4 in the Supporting Information) has been
calculated. Enone 3aa is the most stable compound on the re-
action pathway (DG(3aa)=ꢀ28.0 kcalmolꢀ1). The possibility of
interconversion of 3aa into 5aa (DG(5aa)=ꢀ23.6 kcalmolꢀ1
)
via six-membered intermediate IV was also investigated. The
two barriers along this 3aa!5aa pathway have been calculat-
ed as DG°(TS3aa!IV)=19.5 kcalmolꢀ1 (ꢀ8.5 kcalmolꢀ1 with re-
spect to the reactants, R) and DG°(TSIV!5aa)=5.7 kcalmolꢀ1
(ꢀ5.5 kcalmolꢀ1 with respect to R).
The reaction between 3-phenylprop-2-ynyl acetate (1a) and
acetaldehyde (2a)
Path 2 (Figure 1, red) has a similar profile for TSR!I to six-
membered intermediate V to that found for path 1, but TSI!V
lies DDG=0.7 kcalmolꢀ1 above transition state TSI!II (see
Figure 2 for equilibrium structures). From V, intermediate VI
can be obtained through the cleavage of the CꢀO bond and
Both mechanistic pathways for the reaction between 1a and
acetaldehyde 2a were examined computationally (Figure 1
and Figures S1–S7 in the Supporting Information) at the
wB97XD/6-311+G(2d,p) and B3LYP+D3/def2-TZVP level of
theory by employing the COSMO-RS solvation model. It is
worth mentioning that in the initial step, both paths 1 and 2
proceed via the same intermediate I and transition state TSR!I
.
For the initial step, an activation energy of DG°(TSR!I)=
23.1 kcalmolꢀ1 was obtained at the wB97XD/6-311+G(2d,p)
level, which corresponds to addition of the alkyne unit to the
aldehyde. Transition state TSR!I is characterized by formation
of the CꢀC bond between the aldehyde carbon atom and the
carbon atom of the triple bond of the starting alkyne to form
intermediate I (Scheme 5). In the next step of path 1, the alkox-
ide that results from the aldehyde undergoes nucleophilic
attack onto the positively charged carbon atom, which leads
to the formation of a four-membered oxete intermediate II.
Despite the anticipated strain in the four-membered ring, II is
energetically more stable by DE(1a!II)=ꢀ25.5 kcalmolꢀ1
Figure 2. The equilibrium geometries of two key transition states that divert
the reaction to path 1 or 2.
formation of a C=C bond via
TSV!VI with
a
barrier of
DG°(TSV!VI)=16.6 kcalmolꢀ1.
The subsequent step VI!5aa
involves acyl transfer of the ter-
minal acetyl group to the alkox-
ide ion via TSVI!5aa. The 5aa
adduct reorients to 5aa’, which
is the optimal conformer to lead
to 3aa. This reorientation occurs
in each of the reactions studied.
Finally, in
a subsequent step
5aa’ converts into 3aa via the
transition state TS5aa’!3aa. This
last step is associated with
a small barrier of 4.8 kcalmolꢀ1
.
The experimental data sug-
gest that 3aa is obtained by
both paths 1 (four-membered in-
Figure 1. Gibbs energy levels for paths 1 (black) and 2 (red) calculated at the DFT(wB97XD)/6-311+G(2d,p)//RI-
PBE+D3/def2-SVP level of theory and COSMO-RS solvation method for the reaction between 1a and 2a.
termediate) and
2 (six-mem-
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10365
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
directly connects R and II. The computed activation energy for
the R!II reaction step is DG°(TSR!II)=22.6 kcalmolꢀ1
(DE°(TSR!II)=12.4 kcalmolꢀ1 in vacuo). The transformation
from II to 3ab was found to be similar to that for 3aa, with
a slightly lower barrier of DG°(TSII!3ab)=7.4 kcalmolꢀ1 (cf.
14.7 kcalmolꢀ1 for II!3aa; Figure 1). The same holds true for
the final steps that lead to 5ab. The first barrier for the 3ab!
IV reaction step is DG°(TS3ab!IV)=22.6 kcalmolꢀ1 higher than
that for the analogous step 3aa!IV (19.5 kcalmolꢀ1), whereas
the subsequent barrier to give 5ab has a slightly lower
DG°(TSIV!5ab)=5.3 kcalmolꢀ1 (cf. 5.7 kcalmolꢀ1 on the 5aa
pathway; Figure 1).
Table 4. Relative free energies (DG) and relative gas-phase energies (DE)
[kcalmolꢀ1] and transition-state barriers of products 3–5 obtained by
using the DFT(wB97XD)/6-311+G(2d,p)//RI-PBE+D3/def2-SVP method
(DE) and COSMO-RS solvation method (DG).
Compound
DG[a]
DE[a]
DDG°[b]
DDE°[b]
3aa
4aa
5aa’(3aa/4aa)
3ab
4ab
5ab’ (3ab/4ab)
3bd
4bd
ꢀ28.0
ꢀ43.9
4.8
15.0
–
22.4
18.7
–
24.7
21.4
–
17.7
20.7
–
29.9
17.3
–
29.0
21.1
–
ꢀ25.4
ꢀ44.7
ꢀ17.0/ꢀ21.9
ꢀ25.8
ꢀ40.5/ꢀ43.0
ꢀ44.3
ꢀ22.5
ꢀ42.3
ꢀ21.4/ꢀ17.2
ꢀ27.1
ꢀ44.2/ꢀ37.6
ꢀ46.0
ꢀ26.3
ꢀ45.9
On path 2 (Figure 3, red), transition state TSR!V directly con-
nects R and V, which involves addition of the aldehyde and
formation of the CꢀO bond between the carboxylate and ben-
zylic position. Transition state TSR!V lies DG°(TSR!V)=21.4 kcal
molꢀ1 above the reactants. From intermediate V, which is
almost isoergonic with the reactants R (DG(V)=0.7 kcalmolꢀ1),
intermediate VI is formed. The activation barrier was calculated
as DG°(TSV!VI)=16.5 kcalmolꢀ1. In the next step, VI transforms
to 5ab’ through a free-energy barrier of DG°(TSVI!5ab)=
13.8 kcalmolꢀ1, which is higher than the equivalent barrier
found for 5aa’ (6.5 kcalmolꢀ1). Finally, enone 3ab is obtained
from 5ab’ and this step is associated with an activation free
energy of 22.4 kcalmolꢀ1.
It can be concluded that 3ab is DG=4.4 kcalmolꢀ1 more
stable than 5ab’. Therefore, the conversion from 5ab’ to 3ab
is thermodynamically favourable. The corresponding barrier
has been calculated as DG°(TS5ab’!3ab)=22.4 kcalmolꢀ1. Inter-
estingly, the transition-state barriers associated with the first
steps in both reaction mechanisms (DG°(TSR!II)=22.6 kcal
molꢀ1 and DG°(TSR!V)=21.4 kcalmolꢀ1) also correlate qualita-
tively with the experimental ratio (Table 1, entry 2).
Product 4ab is DG(4ab)=ꢀ22.5 kcalmolꢀ1 more stable than
reactants R and 3.3 kcalmolꢀ1 less stable than 3ab (similar to
the relationship between 3aa/4aa; Figure 1). The analysis of
the theoretical energy profiles for both pathways (Figure 3 and
Figure S1–S7 in the Supporting Information) show that product
3ab is 3.3 kcalmolꢀ1 more stable
5bd’(3bd/4bd)
ꢀ26.3/ꢀ19.6
ꢀ49.2/ꢀ40.1
[a] DG and DE are referenced with respect to reactants. [b] DDG°, and
DDE°, correspond to the transition barrier (TS5’!3/4) for the elementary
5!3/4 reaction step (i.e. DG(5)=0);
bered intermediate). Indeed, it is consistent with the calcula-
tions that both mechanisms might operate competitively. Fur-
thermore, the low barrier of 4.8 kcalmolꢀ1 to reach 3aa from
5aa’ (Table 4) by path 2 and the lower thermodynamic stability
of 5aa’ with respect to 3aa by 11.0 kcalmolꢀ1 explain the for-
mation of 3aa and the absence of 5aa among the products.
Product 4aa is 2.6 kcalmolꢀ1 less stable than 3aa. Additionally,
the transition-state barrier for conversion of 5aa to 4aa is, at
15 kcalmolꢀ1, three times larger than for 5aa’ to 3aa, which
may explain why 4aa is not observed experimentally.
The reaction between 3-phenylprop-2-ynyl acetate (1a) and
2-fluorobenzaldehyde (2b)
Despite huge efforts in studying computationally the reaction
between 1a and 2b (Figure 3 and Figures S1–S7 in the Sup-
porting Information), neither transition state TSR!I nor a stable
intermediate I was found. All attempts to optimize the struc-
ture led either to the reactants or transition state TSR!II, which
than 4ab, which qualitatively
supports the 2:1 3ab/4ab ex-
perimental ratio. However, the
energy barrier TS5ab’!4ab, which
leads to 4ab, is 3.7 kcalmolꢀ1
lower than TS5ab’!3ab. The 3ab/
4ab ratio can be explained in
terms of the first transition-state
barrier, TSR!V
.
This barrier
amounts to 23.3 kcalmolꢀ1 en
route to 3ab, whereas TSR!V
reaches 31.4 kcalmolꢀ1 in the
case of 4ab. This fact allows E
enone 3ab to be obtained at
a higher rate than the Z isomer
4ab. Moreover, the experimental
data, together with TLC and
NMR spectroscopic analysis,
Figure 3. Gibbs energy levels for paths 1 (black) and 2 (red) calculated at the DFT(wB97XD)/6-311+G(2d,p)//RI-
PBE+D3/def2-SVP level of theory and COSMO-RS solvation method for the reaction between 1a and 2b.
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10366
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
showed that E enones 3 formed first and isomerization to Z
enones 4 occurred only after some time.
ergonic for the decisive transition-state barriers, TSI!II versus
TSI!V. Apparently, our computational approach is not able to
discriminate between the two pathways. The estimated error
bar of at least 1–2 kcalmolꢀ1 obtained from the benchmark cal-
culations (see above) has to be recalled and may prohibit
a final and unambiguous conclusion for this case. Still, it can
be seen in Figure 4 that a transition-state barrier for the trans-
formation 5bd’!3bd of DG°(TS5bd’!3bd)=24.7 kcalmolꢀ1 has
been found on path 2, which comprises the highest activation
energy found in all of the theo-
The reaction between 3-arylprop-2-ynyl acetate (1b) and
4-nitrobenzaldehyde (2d)
For path 1 in the reaction between 1b and 2d the lowest bar-
rier for the R!I step of all three studied reactions was found:
DG°(TSR!I)=16.2 kcalmolꢀ1 (Figure 4). Once I has formed, the
retically studied reactions for
this step. This may explain why
5bd was isolated as a product
experimentally. However, at ele-
vated temperature or during
prolonged stirring of the reac-
tion mixture, products 3bd and
4bd become thermodynamically
accessible
(Table 2,
entry 9).
Moreover, it is very important to
note that the relatively low acti-
vation energies of the reactions
of 1b or 1c with 2d result in
smooth conversion of the start-
ing materials in 5 min compared
with 24 h for the 1a+2a–d reac-
tions). Therefore, it was possible
to isolate kinetic products 5bd
and 5cd prior to their conver-
sion to thermodynamically more
stable products 3 and 4.
Figure 4. Gibbs energy levels for paths 1 (black) and 2 (red) calculated at the DFT(wB97XD)/6-311+G(2d,p)//RI-
PBE+D3/def2-SVP level of theory and COSMO-RS solvation method for the reaction between 1b and 2d.
Finally, we attempt to provide
next reaction step is almost barrier-less (DG°(TSI!II)=0.1 kcal
molꢀ1) and, thus, almost immediate formation of intermediate
II can be expected. Subsequently, 3bd is formed via transition
state TSII!3bd with DG°(TSII!3bd)=9.6 kcalmolꢀ1 barrier, which
is lower than the II!3aa barrier (14.7 kcalmolꢀ1; Figure 1) but
slightly higher than that of II!3ab (7.4 kcalmolꢀ1; Figure 2).
Like in the other reactions, the formation of 5bd occurs via in-
termediate IV, with a large activation free energy (18.1 kcal
molꢀ1) for the transformation 3bd!IV and a small barrier
(4.9 kcalmolꢀ1) for the IV!5bd step. In summary, the reaction
between 1b and 2d has the lowest activation energy of all
three computationally investigated R!3n processes. This is
consistent with the experimental findings, in which the reac-
tion 1b+2d yielded 82% of 3bd and 5bd’ in the shortest re-
action time of 5 min (Table 2, entry 7).
one qualitative argument that may provide an intuitive ratio-
nale for the two divergent pathways described above. To this
end, we calculated the wB97XD/6-311+(G(2d,p) Mulliken
charges on the oxygen atom that originated from the alde-
hyde that interacts with BF3. When these charges for the TSI!II
versus TSI!V structures were compared we noticed that for the
reaction 1a+2a (Figure 1), the Mulliken charge on this oxygen
atom is ꢀ0.12e more negative for the former transition state
(TSI!II). The trend is reversed for the reaction 1b+2d
(Figure 4), for which the oxygen atom in TSI!V is ꢀ0.08e more
negative. Therefore, it might be an evolving electron density
on the oxygen atom that governs the course of the reaction at
the branching point.
Conclusion
Path 2 for 1b+2d has a similar energy and free-energy pro-
file to the 1a+2a reaction (compare Figures 1 and 4) and is
characterized by a small initial barrier of DG°(TSI!V)=1.9 kcal
molꢀ1. The subsequent activation free energies found for steps
V!VI, VI!5bd, and 5bd’!3bd are DG° =19.7, 19.9, and
24.7 kcalmolꢀ1, respectively.
Reactions between 3-arylprop-2-ynyl esters and aldehydes that
lead to the formation of various a,b-unsaturated ketones have
been studied by using 18O-labeling experiments and computa-
tional methods. The obtained computational results suggest
that both mechanisms, via either a four- or six-membered in-
termediate, are plausible and energetically feasible, which ex-
plains the observed mechanistic dichotomy. It was also proved
that the formation of the MBH adducts always proceeds by
a new addition–rearrangement cascade, which includes nucle-
It should be noted that experimental evidence suggests that
5bd is formed exclusively via six-membered intermediate V by
path 2, whereas the theoretical calculations do not fully sup-
port this finding. Paths 1 and 2 are calculated to be almost iso-
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10367
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
organic solution was dried over sodium sulfate, concentrated
under reduced pressure, and the residue was used in the next step
without purification.
ophilic attack of the alkyne to the Lewis acid activated alde-
hyde, followed by an intramolecular nucleophilic stabilization
of the vinylic carbocation by the ester carbonyl group and con-
comitant formation of a six-membered zwitterion. An acyl
group transfer completes this cascade by formation of the ki-
netic MBH carboxylates. Uniquely, this new 1,3-acyl shift in
propargylic esters is induced by addition of electrophilic alde-
hydes and does not require alkyne activation by transition-
metal catalysis. Thus acceptor-substituted benzaldehydes were
shown to dramatically switch from the classical alkyne–carbon-
yl metathesis pathway via four-membered intermediates to the
newly discovered addition–rearrangement cascade via six-
membered zwitterions. Therefore, on one hand, the present
synthetic method provides a useful approach to MBH deriva-
tives that have been difficult to access by classical MBH reac-
tions. On the other hand, prolonged reaction times allow the
synthesis of thermodynamically more stable 2-aroyl-3-arylallyl
acetates. Efforts for the selective preparation of a variety of
a,b-unsaturated ketones is currently underway in our laborato-
ry and the results will be reported in due course.
General procedure for the synthesis of compounds 3–5
Boron trifluoride etherate (0.142 g, 0.13 mL, 1 mmol) was added to
a solution of alkyne 1 (1 mmol) and aldehyde 2 (1 mmol) in dry
CH2Cl2 (5 mL). The resultant solution was stirred at rt. After comple-
tion of the reaction (observed by TLC), the solution was quenched
with a saturated aqueous solution of sodium bicarbonate. The or-
ganic layer was separated, washed with water (2ꢃ20 mL), and
dried over anhydrous Na2SO4. The solvent was evaporated under
reduced pressure and the residue was purified by flash column
chromatography (hexane/ethyl acetate).
Computational Details
The quantum chemical calculations reported in this work were per-
formed with the TURBOMOLE 6.4[21] and Gaussian 09[22] programs.
Geometry optimizations (minima and transition states) were carried
out at the DFT level and employed the density-fitted (see below)
Perdew–Burke–Ernzerhof (PBE) functional[23] in conjunction with
Grimme’s D3 empirical dispersion correction[24] (RI-DFT+D3
method) and the def2-SVP basis set on all of the atoms.[25,26]
Experimental Section
Reference values for small model systems (truncated from approxi-
mately 20 key structures along the studied reaction pathways)
were obtained by the coupled cluster method with singles, dou-
bles, and non-iterative triples (CCSD(T)) that employed the aug-cc-
pVTZ and aug-cc-pVDZ basis sets.[27] A variety of DFT functionals
were benchmarked against the reference values (most of the data
are shown in the Supporting Information), as well as the popular
Moller–Plesset perturbation method (MP2) and the spin-compo-
nent scaling second-order MP2 (SCS-MP2) method (with recom-
mended cOS =6/5 and cSS =1/3 parameters).[28] The wB97XD[29] func-
tional was selected as probably the most accurate and consistent
to obtain final single-point energies for the key intermediates and
transition states along the studied pathways (Path 1 and Path 2).
The reported single-point energies were obtained by using the 6–
311+G(2d,p) basis set. Almost equally good performance in bench-
mark calculations was found for the B3LYP+D3 functional;[30] the
reaction profiles calculated with the B3LYP+D3/def2-TZVP[26] values
can be found in the Supporting Information and are referred to in
the discussion. Wherever possible, calculations were expedited by
expansion of the Coulomb integrals to an auxiliary basis set, by
using the resolution-of-identity (RI) approximation (density fit-
ting).[31]
General information
IR spectra were recorded with a PerkinElmer FT spectrophotometer
1
Spectrum BX II as KBr discs. H and 13C NMR spectra were recorded
with a Varian Unity INOVA spectrometer (300 MHz) or Bruker
(400 MHz) in [D]chloroform; the residual solvent signal was used as
an internal standard. HRMS was performed with a Dual-ESI Q-TOF
6520 (Agilent Technologies) spectrometer. All reactions and the
purity of the synthesized compounds were monitored by TLC
(silica gel 60 F254 aluminum plates; Merck). Visualization was ac-
complished by UV light. Final purification of synthesized com-
pounds was performed with a CombiFlash (Teledyne Isco) flash
chromatography system.
Compound 1a was prepared according to a literature proce-
dure.[20]
Characterization data of the compounds synthesized can be found
in the Supporting Information.
General procedure for the synthesis of alkynes 1b–d
Under argon atmosphere, the appropriate terminal alkyne
(2.1 mmol) was added to a mixture of 1-iodo-4-methoxybenzene
(0.468 g, 2 mmol), [PdCl2(PPh3)2] (0.28 g, 0.4 mmol) and triethyla-
mine (6 mmol) in THF (5 mL). After stirring the resultant mixture at
rt for 5 min, copper(I) iodide (38 mg, 0.2 mmol) was added. The
mixture was stirred under argon at rt for 1–4 h. After completion
of the reaction (observed by TLC), the solvent was evaporated
under reduced pressure and the crude residue was purified by
flash column chromatography (hexane/ethyl acetate).
Solvation (free) energies of all studied species were calculated by
using the COSMO-RS method[32,33] (conductor-like screening model
for realistic solvation) as implemented in the COSMOtherm pro-
gram,[34] by using the “BP TZVP C30 1201.ctd” parametrization
file. The geometries were first optimized in CH2Cl2 (e=8.93) by
using the Becke–Perdew (B-P86) functional[30a,35] and the COSMO
implicit solvation model.[36] The COSMO-RS calculations were per-
formed by the recommended protocol, which included RI-BP86/
def-TZVP calculations with e=1 (ideal conductor) and e=
1 (vacuum) as a prerequisite for the final calculations in CH2Cl2. The
Gibbs free energy was subsequently calculated as the sum of the
following contributions, shown in [Eq. (1)]:
General procedure for the synthesis of 18O-labeled
aldehydes 2
Para-Toluene sulfonic acid (p-TSA; 0.034 g, 0.2 mmol) was added to
a solution of the appropriate aldehyde (2 mmol) and 18O-water
(0.12 mL, 6 mmol) in absolute THF (5 mL). The resultant solution
was stirred at 508C for 12 h. An aqueous saturated solution of
K2CO3 was added and the mixture was extracted with CH2Cl2. The
G ¼ EelþGsolvþEZPVEþpVꢀRT lnðqtransqrotqvib
Þ
ð1Þ
Eel is the in vacuo energy of the system (at the DFT(X)/TZP//RI-
PBE+D3/def2-SVP level of theory; X=wB97XD or B3LYP+D3 and
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10368
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
114, 2509; Angew. Chem. Int. Ed. 2002, 41, 2403; e) H. Yang, R. G. Carter
J. Org. Chem. 2010, 75, 4929; f) R. Miyaji, K. Asano, S. Matsubara, Org
Lett. 2013, 15, 3658; g) J. Louie, C. W. Bielawski, R. H. Grubbs, J. Am.
Chem. Soc. 2001, 123, 11312; h) A. K. Chatterjee, T.-L. Choi, R. H. Grubbs,
F. D. Toste, Adv. Synth. Catal. 2002, 344, 634.
TZP=6–311+G(2d,p) and def2-TZVP, respectively). Gsolv is the solva-
tion free energy, calculated by using the RI-BP86/def-TZVP(COSMO-
RS, e=1, e=1) method described above. The zero-point vibra-
tional energy (DEZPVE), thermal contributions to enthalpy, and en-
tropic contributions were calculated by using the analytical har-
monic vibrational frequencies at the RI-PBE+D3/def2-SVP level.
The standard formulas of statistical thermodynamics that corre-
spond to the ideal gas, rigid rotor, and harmonic oscillator approxi-
mations[37] were used to obtain thermodynamic functions at 298 K
and 1 atm.
The correction of (1.9ꢃDn) kcalmolꢀ1, which corresponds to the
difference between the concentration of the ideal gas at 298 K and
1 atm and its 1 mollꢀ1 concentration, was applied for the reactions
(processes) in which the number of moles (Dn) changed.
[8] a) H. Vieregge, H. J. T. Bos, J. F. Arens, Rec. Trav. Chim. 1959, 78, 664;
b) H. Vieregge, H. M. Schmidt, J. Renema, H. J. T. Bos, J. F. Arens, Rec.
Trav. Chim. 1966, 85, 929.
[9] W. J. Middleton, J. Org. Chem. 1965, 30, 1307.
[10] M. Oblin, M. Rajzmann, J.-M. Pons, Tetrahedron 2001, 57, 3099.
[11] a) S. G. Viswanathan, C.-J. Li, Tetrahedron Lett. 2002, 43, 1613; b) L. W.
Xu, L. Li, C. G. Xia, P. Q. Zhao, Helv. Chim. Acta 2004, 87, 3080; c) M.
Curini, F. Epifano, F. Maltese, O. Rosati, Synlett 2003, 0552; d) R. P.
Hsung, C. A. Zificsak, L. L. Wei, C. J. Douglas, H. Xiong, J. A. Mudler, Org.
Lett. 1999, 1, 1237; e) C. Gonzꢀlez-Rodrꢁguez, L. Escalante, J. A. Varela, L.
Castedo, C. Saꢀ, Org. Lett. 2009, 11, 1531; f) T. Jin, F. Yang, C. Liu, Y. Ya-
mamoto, Chem. Commun. 2009, 3533.
[12] a) T. Jin, Y. Yamamoto, Org. Lett. 2007, 9, 5259; b) T. Jin, Y. Yamamoto,
Org. Lett. 2008, 10, 3137; c) E. M. L. Sze, W. Rao, M. J. Koh, P. W. H. Chan,
Chem. Eur. J. 2011, 17, 1437; d) J. U. Rhee, M. J. Krische, Org. Lett. 2005,
7, 2493.
Acknowledgements
The project was supported by the European Social Fund under
[13] I. Karpaviciene, I. Cikotiene, Org. Lett. 2013, 15, 224.
ˇ
the Global Grant measure (Grant No. VP1–3.1-SMM-07 K-01–
[14] It was also found that the presence of an electron-donating group on
benzaldehydes diminishes the reaction rates and stimulates formation
of a 2:1 adduct (cf. ref. [13]).
[15] a) M. Shi, C.-Q. Li, J.-K. Jiang, Helv. Chim. Acta 2002, 85, 1051; b) D. Basa-
vaiah, V. V. L. Gowriswari, T. K. Bharathi, Tetrahedron Lett. 1987, 28, 4591;
c) M. Shi, C.-Q. Li, J.-K. Jiang, Molecules 2002, 7, 721.
[16] For mechanistically different gold-catalyzed reactions of propargylic
esters, see: a) N. Marion, S. P. Nolan, Angew. Chem. 2007, 119, 2806;
Angew. Chem. Int. Ed. 2007, 46, 2750; b) Y. Peng, L. Cui, G. Zhang, L.
Zhang, J. Am. Chem. Soc. 2009, 131, 5062; c) J. W. Cran, M. E. Krafft,
Angew. Chem. 2012, 124, 9532; Angew. Chem. Int. Ed. 2012, 51, 9398;
d) L. Zhang, J. Am. Chem. Soc. 2005, 127, 1684; e) Y. Yu, W. Yang, F. Ro-
minger, A. S. K. Hashmi, Angew. Chem. 2013, 125, 7735; Angew. Chem.
Int. Ed. 2013, 52, 7586; f) R. K. Shiroodi, V. Gevorgyan, Chem. Soc. Rev.
2013, 42, 4991.
002), Grant Agency of the Czech Republic (Grant No. 14–
31419S), and by the Institute of Organic Chemistry and Bio-
chemistry, Academy of Sciences of the Czech Republic (project
RVO: 61388963). G.S.-S. thanks the Human Frontier Science
Program (project reference: LT001022/2013-C) for support.
G.S.-S, and C.T thank Centro de Computaciꢄn Cientꢁfica of
UAM for the allocation of computing time. We thank Prof.
Martin Kotora for fruitful discussions of the reactions studied.
Keywords: 3-arylprop-2-ynyl esters
calculations · isotopic labeling · ketones · Morita–Baylis–
·
density functional
Hillman adducts
[17] J. C. Vederas, J. Am. Chem. Soc. 1980, 102, 374.
[18] P. Giraudeau, J. L. Wang, ꢆ. Baguet, C. R. Chim. 2006, 9, 525.
[19] E. L. Myers, C. P. Butts, V. K. Aggarwal, Chem. Commun. 2006, 4434.
[20] Y. Nishimoto, A. Okita, M. Yasuda, A. Baba, Org. Lett. 2012, 14, 1846.
[21] TURBOMOLE V6.4 2012, a development of University of Karlsruhe and
Forschungszentrum Karlsruhe GmbH, 1989–2007, TURBOMOLE GmbH,
since 2007; available from www.turbomole.com.
[1] a) Comprehensive Organic Synthesis Vol. 3 (Eds.: B. M. Trost, T. Fleming),
Pergamon, Oxford, 1991; b) C. M. R. Volla, I. Atodiresei, M. Rueping,
Chem. Rev. 2014, 114, 2390; c) A. Bhunia, S. R. Yetraa, A. T. Biju, Chem.
Soc. Rev. 2012, 41, 3140; d) S. H. Cho, J. Y. Kim, J. Kwaka, S. Chang,
Chem. Soc. Rev. 2011, 40, 5068.
[22] Gaussian 09, Revision A.2, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢇ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[2] a) B. M. Trost, Angew. Chem. 1995, 107, 285; Angew. Chem. Int. Ed. Engl.
1995, 34, 259; b) B. M. Trost, Science 1991, 254, 1471; c) R. A. Sheldon,
Pure Appl. Chem. 2000, 72, 1233; d) C. M. Marson, Chem. Soc. Rev. 2012,
41, 7712; e) M. B. Gawande, V. D. B. Bonifꢀcio, R. Luque, P. S. Brancoa,
R. S. Varma, Chem. Soc. Rev. 2013, 42, 5522.
[3] a) B. L. Feringa, Acc. Chem. Res. 2000, 33, 346; b) S. Woodward, Chem.
Soc. Rev. 2000, 29, 393; c) A. Alexakis, C. Benhaim, Eur. J. Org. Chem.
2002, 3221; d) F. Lꢄpez, A. J. Minnaard, B. L. Feringa, Acc. Chem. Res.
2007, 40, 179; e) S. Sulzer-Mossꢅ, A. Alexakis, Chem. Commun. 2007,
3123.
[4] a) Modern Aldol Reactions, Vols. 1 and 2 (Ed.: R. Mahrwald), Wiley-VCH,
Weinheim, 2004; b) B. M. Trost, C. S. Brindle, Chem. Soc. Rev. 2010, 39,
1600; c) J. Mlynarski, J. Paradowska, Chem. Soc. Rev. 2008, 37, 1502;
d) A. Ramazani, A. R. Kazemizadeh, E. Ahmadi, N. Noshiranzadeh, A.
Souldozi, Curr. Org. Chem. 2008, 12, 59; e) K. M. Pietrusiewicz, M. Za-
blocka, Chem. Rev. 1994, 94, 1375; f) B. E. Maryanoff, A. B. Reitz, Chem.
Rev. 1989, 89, 863.
[23] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865.
[24] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
[25] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297.
[26] A. Schꢈfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829.
[27] D. E. Woon, T. H. Dunning, Jr., J. Chem. Phys. 1993, 98, 1358.
[28] S. Grimme, J. Chem. Phys. 2003, 118, 9095.
[29] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008, 10, 6615.
[30] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098; b) C. Lee, W. Yang, R. G. Parr,
Phys. Rev. B 1988, 37, 785; c) A. D. Becke, J. Chem. Phys. 1993, 98, 5648;
d) P. J. Stephens, F. J. Devlin, M. J. Frisch, C. F. Chabalowski, J. Phys.
Chem. 1994, 98, 11623.
[5] X. F. Wu, H. Neumann, A. Spannenberg, T. Schulz, H. J. Jiao, M. Beller, J.
Am. Chem. Soc. 2010, 132, 14596.
[6] a) K. H. Meyer, K. Schuster, Chem. Ber. 1922, 55, 819; b) S. Swaminathan,
K. V. Narayan, Chem. Rev. 1971, 71, 429; c) Y. Sugawara, W. Yamada, S.
Yoshida, T. Ikeno, T. Yamada, J. Am. Chem. Soc. 2007, 129, 12902.
[7] a) K. Matsuo, K. Ohtsuki, T. Yoshikawa, K. Shishido, K. Yokotani-Tomita,
M. Shindo, Tetrahedron 2010, 66, 8407; b) M. A. Evans, J. P. Morken, Org
Lett. 2005, 7, 3371; c) S. Ghorai, B. H. Lipshutz, K. Voigtritter, J. Org.
Chem. 2011, 76, 4697; d) H. Wakamatsu, S. Blechert Angew. Chem. 2002,
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10369
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[31] K. Eichkorn, O. Treutler, H. ꢇhm, M. Hꢈser, R. Ahlrichs, Chem. Phys. Lett.
1995, 240, 283.
[32] A. Klamt, J. Phys. Chem. 1995, 99, 2224.
[33] A. Klamt, V. Jonas, T. Buerger, J. C. W. Lohrenz, J. Phys. Chem. 1998, 102,
5074.
[36] A. Klamt, G. Schuurmann, J. Chem. Soc. Perkin Trans. 2 1993, 799.
[37] F. Jensen, Introduction to Computational Chemistry, John Wiley & Sons,
New York, 1999.
[34] F. Eckert, A. Klamt, COSMOtherm, version C3.0, release 12.01; COSMO-
logic GmbH & Co. KG.
Received: March 11, 2014
[35] a) S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200; b) J. P.
Perdew, Phys. Rev. B 1986, 33, 8822.
Published online on July 10, 2014
Chem. Eur. J. 2014, 20, 10360 – 10370
www.chemeurj.org
10370
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim