Reversible generation of metastable enols in the 1,4-addition of thioacetic acid to α,β-unsaturated carbonyl compounds

Article abstract of DOI:10.1021/jo3021709

Addition of thioacetic acid to reactive α,β-unsaturated carbonyl compounds like acrolein or crotonaldehyde in acetone-d₆ generates metastable (E)-and (Z)-1-alkenols, which tautomerize slowly at ambient temperature. The 1,4-addition of thioacetic acid and crotonaldehyde to (Z)-3-(acetylsulfanyl)-1-propen-1-ol is reversible with K_eq = 5.5 ± 0.5 L/mol. A concerted, cyclic 1,4-addition mode is proposed to explain the preferred (Z)-stereoselectivity in lower polarity, nonprotic solvents.

Full text of DOI:10.1021/jo3021709

Note
pubs.acs.org/joc
Reversible Generation of Metastable Enols in the 1,4-Addition of
Thioacetic Acid to α,β-Unsaturated Carbonyl Compounds
Lukas Hintermann* and Aleksej Turockin
̌
Department Chemie, Technische Universitat Munchen, Lichtenbergstrasse 4, 85748 Garching b. Munchen, Germany
̈
̈
̈
S
* Supporting Information
ABSTRACT: Addition of thioacetic acid to reactive α,β-
unsaturated carbonyl compounds like acrolein or crotonalde-
hyde in acetone-d6 generates metastable (E)- and (Z)-1-
alkenols, which tautomerize slowly at ambient temperature.
The 1,4-addition of thioacetic acid and crotonaldehyde to (Z)-
3-(acetylsulfanyl)-1-propen-1-ol is reversible with Keq = 5.5
0.5 L/mol. A concerted, cyclic 1,4-addition mode is proposed to explain the preferred (Z)-stereoselectivity in lower polarity,
nonprotic solvents.
Simple enols1 lacking steric or electronic stabilization are only a
minute fraction in the tautomeric equilibrium with their parent
aldehydes or ketones. The enol content of ethanal, for example,
amounts to merely 0.00006%.2 Nevertheless, excess concen-
trations of metastable enol solutions have been obtained by
photoelimination3,4 or careful hydrolyses of enol orthoest-
ers.1,5,6 Chin7 and Bosnich8 described elegant transition-metal-
catalyzed isomerizations of allylic alcohols to enols,9 which are
stable for minutes to hours at ambient temperature and a stable
enol, 2-methyl-1-propen-1-ol, has even been obtained at low
temperatures (−20 °C).7b Still, simple enols remain unusual,
and few NMR or reactivity studies are available. We now report
that metastable enols are generated in the conjugate hydro-
thiolation of α,β-unsaturated aldehydes or vinyl ketones with
thioacetic acid.10 Solutions with 0.3−0.5 mol/L of simple enols
are obtained straightforwardly by mixing commercially available
starting materials in several solvents with no precautions. The
enols have been studied by NMR spectroscopy for hours at
ambient temperature. The mechanism of addition of the
reaction involves a reversible 1,4-addition to an enol, followed
by rate-limiting tautomerization.
d6.13 They correspond closely to those previously observed in
(Z)-1-propen-1-ol (δOH = 8.07 ppm,12 JHOCH = 5.9 Hz at −80
°C).6b The values are consistent with an s-trans arrangement
around the O−C-1 single bond in (Z)-3a,6b excluding an
intramolecular hydrogen bond as a potential stabilizing factor.
Similarly, the NMR data of the minor isomer (E)-3a (3JHCOH
=
9.1 Hz, δOH = 7.54 ppm) are consistent with the s-cis
conformation, as also seen in (E)-1-propen-1-ol (δOH = 7.99
ppm, JHCOH = 9.5 Hz at −80 °C).6b It follows that the
3
acetylsulfanyl group has no specific effect on structure and
bonding of (E)-3a14 and (Z)-3a, which are truly simple enols.1
The 1H,13C-HSQC NMR experiment recorded at ambient
temperature correlates the characteristic 13C resonances of the
enol carbons C-1 (δ = 144.2/146.7 ppm) and C-2 (δ = 99.3/
99.9 ppm) with H-1 (δ = 6.44/6.57 ppm) and H-2 (δ = 4.35/
4.79 ppm) in (Z)- and (E)-3a, respectively (Figure 1).
An analogous experiment with crotonaldehyde (2b) as
substrate in acetone-d6 gave similar results, with two notable
changes: first, the metastable enol 3b was almost exclusively
formed as the (Z)-3b isomer, whose spectral characteristics are
3
comparable to (Z)-3a (δOH = 7.70 ppm, JHCOH = 6.6 Hz; s-
The addition of thioacetic acid (1; 2 equiv) to a solution of
acrolein (2a) in acetone-d6 at ambient temperature (Scheme 1)
generated a solution consisting mainly of enols (Z)-3a (69 mol
%) and (E)-3a (15 mol %).11 Compound 2a was largely
consumed (≤2 mol %) according to NMR spectra recorded
after 15 min. For the major product isomer, the chemical shift
of the enolic OH-hydrogen (δOH = 7.67 ppm)12 and its
coupling to H-1 (3JHOCH = 6.9 Hz) are well resolved in acetone-
trans); second, 2b was not completely consumed by an excess
of thioacetic acid. The composition of the reaction mixture
relative to the initial concentration of 2b (0.38 mol·L−1 = 100
mol %) over the course of several days is shown in Table 1.
Preferential generation of the less stable8 (Z)-enol in this
uncatalyzed 1,4-addition could hint at a hydrogen-bond assisted
mechanism (Scheme 2). It appears that thioacid 1 is not
sufficiently strong to fully protonate the substrate carbonyl
group, but instead undergoes a concerted, cyclic 1,4-addition to
the s-cis isomer of 2 (Scheme 2). In the case of acrolein (2a) as
a particularly strong electrophile, competitive addition to the s-
trans conformer of 2 by a neutral or ionic mechanism could
account for generation of some (E)-3a. The proposal of
Scheme 1. Simple Enols from the 1,4-Addition of Thioacetic
Acid to Acrolein
Received: October 3, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo3021709 | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
Values for Keq have been obtained by converting the mol %
composition of species 1, 2b, and (Z)-3b from Table 1 to
concentrations and insertion in eq 1. In spite of large changes
of the individual concentrations, the value of Keq is constant
over most of the observation range, as expected for a true
equilibrium constant;18 only the value obtained after 20 min is
low, indicating incomplete equilibration.
Generation of metastable enols and the existence of
equilibria analogous to eq 1 has also been established with
other unsaturated carbonyl substrates (Table 2). The extent to
which enol is formed depends on the reactivity of the carbonyl
acceptor, which in turn, reflects its ground-state stabilization.19
The concentration of the most reactive acceptor acrolein (2a)
was too small to allow reliable calculation of Keq. Similar to
crotonaldehyde (2b), (E)-2-hexenal (2c) underwent fast initial
addition to (Z)-3c, accompanied by a small amount of (E)-3c.
After 3 h, the product ratio for 3c had changed to (Z/E)-3c ≈
2:1, which means that (E)-3c is generated but also tautomerizes
more slowly than (Z)-3c. The Keq value for 1,4-addition to give
(Z)-3c calculated after 20 min was low (1.2 L·mol−1) but
further rose and remained constant at 3.3 L·mol−1 after 1 day. A
(Z)-enol 3d was also detected in the conjugate addition to
methyl vinyl ketone (2d) after 20 min, but it had largely
disappeared within 1 h, when close to 99% of the tautomerized
1,4-addition product 4 was present.
Figure 1. 1H,13C-HSQC of an acrolein/thioacetic acid mixture.
1H,13C-HSQC (500 MHz, acetone-d6, 295 K) of a freshly prepared
mixture of acrolein and thioacetic acid (2 equiv). Main signals are due
to (E)-3a and (Z)-3a; minor signals are due to a hemithioacetal
follow-up product, vide infra.
Table 1. Time-Dependent Composition of a Thioacetic Acid
a
(1)/Crotonaldehyde (2b) Mixture in Acetone-d6
With the least reactive ketone, 3-penten-2-one (2e), a small
amount (Z)-enol 3d was detected after 20 min, but none
remained after 3 h, when the usual thia-Michael addition had
gone to completion. The Keq values listed in Table 2 for
generation of (Z)-3c and (Z)-3d are probably low relative to
the true equilibrium constants because tautomerization
competes with equilibration of the initial 1,4-addition. Similarly,
1,4-addition of substrates 1 + 2 to form the stereoisomeric
enols (E)-3 is slow relative to the tautomerization (E)-3→4,
and equilibration with (E)-enols is not established. The
generation of enols from acrolein (2a) was also realized in
other solvents including CDCl3, C6D6, DMSO-d6, and MeOH-
d4. In DMSO-d6, the major enol observed after 20 min was (E)-
3a (70%, besides 12% (Z)-3a), as was the case in MeOH-d4
(47% (E)-3a, 27% (Z)-3a). The results point to a disruption of
the concerted transition state (Scheme 2) by strong donor or
protic solvents. Enols were even obtained in an aqueous
medium consisting of D2O/MeOH-d4 (1:2):20 after 15 min,
46% of (E)-3a and 3% of (Z)-3a were detected. Subsequent
tautomerization was completed after 4 h.
Other Michael acceptors have not been included in Table 2,
since enols were not observed in their 1,4-additions. These
include cinnamaldehyde (3-phenyl-2-propenal) and mesityl
oxide (4-methyl-3-penten-2-one), which slowly gave only the
usual thia-Michael addition products (mesityl oxide: t50% = 20 h
and cinnamaldehyde: t50% = 96 h, where t50% denotes time to
50% conversion of starting materials). Neither were enol
intermediates observed in conjugate additions of the cyclic
enones 2-cyclopenten-1-one and 2-cyclohexen-1-one, though
they reacted markedly faster (t50% = 1 h).21
time (h)
b
0
0.3
1
5
35
15
60
2b
100
208
36
105
51
26
112
61
22.5
5.1
b
1
105
50
0.8
1.0
21
5.6
64
20
33
(Z)-3b
(E)-3b
4b
3.4
0.4
0.4
4.3
3.6
1.3
0.1
3.9
55
5.5
0.4
7.2
5.5
10.6
75.3
5.3
5b
c
Keq
a
Reaction in acetone-d6 at 295 K with 2.08 equiv of 1. Molar
composition in mol % relative to initially added 2b (100 mol % = 0.38
mol/L) from H NMR peak integrations relative to internal standard
(see the Supporting Information). Values from a separate experiment
1
b
c
with only 2.0 equiv of 1. Equilibrium constant for generation of (Z)-
3b; see eq 1. Units are L·mol−1.
Scheme 2. Concerted 1,4-Addition of Thioacetic Acid as a
Possible Explanation of the Selective Generation of (Z)-
Enols 3
Scheme 2 recommends itself for evaluation by in silico
methods.
Even though 1 and 2b react to 60−70% conversion within
the first 15 min of mixing, the reaction levels off and remaining
2b is only slowly consumed over several hours (Table 1).15,16
This is due to the existence of a fast private equilibrium in the
1,4-addition reaction of 1 and 2b to (Z)-3b.17 The equilibrium
constant for this reaction is defined by eq 1:
The fate of the reaction mixtures deserves some comment: in
conjugate additions with a 2-fold excess of thioacetic acid (1) to
aldehydes 2a−c in either acetone-d6, C6D6, or CDCl3, the final
reaction products (>80%) are hemithioacetals 5a−c in
equilibrium with regular thia-Michael addition products 4a−c
(Scheme 3a). Whiting and co-workers have previously observed
these anti Erlenmeyer compounds22 as intermediates in thia-
Keq = [(Z)‐3b]/[1]·[2b] = 5.5 0.5 L·mol−1
Michael additions of 1,15 and Bohme has isolated distillable
(1)
̈
B
dx.doi.org/10.1021/jo3021709 | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
a
Table 2. Time-Dependent Generation of Metastable Enols from Thioacetic Acid and Various Acceptors in Acetone-d6
a
b
Reactions were performed in acetone-d6 at 295 K with 2 equiv of 1. Mol% of enol relative to initially added 2 (100 mol % = 0.4−0.6 mol/L) after
c
d
ca. 20 min. Mol % of enol after the time indicated in brackets. Keq was calculated from species concentrations after 20 min reaction time or later as
defined in eq 1. Keq after 1 h. Keq after 3 h. See the Supporting Information for additional data.
e
f
5b is the sum of two individual constants (23 and 18 L·mol−1)
Scheme 3. Hemithioacetals of Aldehydes and Thioacetic
Acid
for the two diastereomers of 5b. Interestingly, no 1,2-addition
of thioacetic acid to the carbonyl groups of substrates 2 has
been observed. This process may be kinetically and
thermodynamically disfavored relative to 1,4-addition, since it
would break off conjugation in the unsaturated aldehydes.
The generation of an enol as the initial product by ionic 1,4-
addition of a heteronucleophile to an acceptor alkene (hetero-
Michael addition) can be expected on the basis of the
kinetically faster protonation of enolates at oxygen rather
than at carbon.2a,24,25 However, such enols are hardly ever
observed because of fast tautomerization under reaction
conditions. Nevertheless, from the conjugate addition of
thioacetic acid (1) to carbonyl acceptors 2 it is possible to
obtain and study solutions of metastable simple enols by mixing
readily available starting materials at ambient temperature. The
remarkable stability of the simple enols under the conditions of
this experiment is apparently due to the weakly acidic character
of the reaction medium, which suppresses tautomerization via
ionization of the enol. At the same time, the acidity level is too
hemithioacetals of lower aldehydes with 1.23 A quantification of
the representative equilibria 1 + 4b = 5b (Scheme 3a) and 1 +
6 = 7 (Scheme 3b) was readily achieved by 1H NMR
spectroscopy. The hemithioacetalization constant for 4b + 1 =
C
dx.doi.org/10.1021/jo3021709 | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
(10) While preparing the manuscript we became aware of a
precedent where thiobenzoic acid was added to α-benzylacrolein
over 7 days at −18 °C in CH2Cl2 to generate a (Z)-enol, which was
then tautomerized to aldehyde with asymmetric induction. No spectral
data or yields had been reported, however. A related observation (with
thioacetic acid) from a doctoral thesis was mentioned: Henze, R.;
Duhamel, L.; Lasne, M.-C. Tetrahedron: Asymmetry 1997, 8, 3363−
3365.
low to induce acid-catalyzed tautomerization. A notable aspect
of these reactions is the presence of an equilibrium 1 + 2 = (Z)-
3, which is established in spite of the metastable character of
product (Z)-3. Stereoselective generation of (Z)-enols in
weakly hydrogen bonding solvents can be explained by
assuming a concerted addition mechanism via hydrogen-
bonding activation (Scheme 2). The rate-limiting step in
these thia-Michael additions is tautomerization of enols 3 to
regular thia-Michael products 4.25
(11) The ratio of (E)-3a and (Z)-3b isomers varied to some extent
with the concentration and the quality of the thioacetic acid sample
used; this observation could point to a competing, minor radical
addition pathway.
EXPERIMENTAL SECTION
General Procedure for Generating Metastable Enols. A
■
(12) The chemical shift δ(1H) of OH depends on concentration and
temperature, causing minor deviations from Capon’s data.6b
solution of the Michael acceptor and CH2Cl2 (internal standard) in
3
1
(13) The JHOCH coupling is not resolved in C6D6 or CDCl3.
deuterated solvent was analyzed by H and 13C NMR spectroscopy.
(14) An intramolecular hydrogen bridge is impossible in (E)-3.
(15) Ilyashenko, G.; Whiting, A.; Wright, A. Adv. Synth. Catal. 2010,
352, 1818−1825.
(16) Similar behavior was also observed but differently interpreted by
Whiting and co-workers.15 Their reaction conditions were slightly
different; thus, their explanation might also be valid.
Thioacetic acid was then added by microsyringe and the reaction
mixture monitored by NMR spectroscopy.
Generation of a Solution of 3-Acetylsulfanyl-1-propen-1-ol
(3a). In an NMR tube, freshly distilled acrolein (2a; 17 μL, 0.25 mmol,
100%) and CH2Cl2 (internal standard, 16 μL) were dissolved in 400
μL of acetone-d6 at ambient temperature. Thioacetic acid (1; 37 μL,
0.52 mmol, 208 mol %) was added, the NMR tube closed, and the
tube thoroughly shaken. NMR spectra (1H, 13C) were recorded in
regular time intervals. The following components (mol % relative to
2a) were detected after 20 min: 1 (76%), 2a (2%), (E)-3a (15%), (Z)-
3a (69%), 4a (1%), 5a (9%), and acetic acid (23%). The analytical
recovery of compounds derived from 2a is 96% (2a + (E)-3a + (Z)-3a
+ 4a + 5a). The total enol concentration was 84% = 0.45 mol·L−1.
(17) The term private equilibrium specifies an equilibration between a
set of starting materials and one specific among several possible
reaction products. We thank Johannes Schluter for suggesting the
̈
presence of an equilibrium in the conjugate addition reaction.
(18) 2D EXSY and selective 1D irradiation experiments have been
performed with an intent to study this equilibrium. However, since t50%
for equilibration of 1 + 2 = (Z)-3 at rt is in the order of minutes for the
forward reaction, exchange kinetics are too slow for obtaining
meaningful results for the reverse reaction. We thank Dr. Wolfgang
ASSOCIATED CONTENT
* Supporting Information
■
Eisenreich, Technische Universitat Munchen, for assistance with these
̈
̈
S
NMR experiments.
Experimental procedures and spectral data for various reaction
mixtures and species. This material is available free of charge via
the Internet at http://pubs.acs.org.
(19) Krenske, E. H.; Petter, R. C.; Zhu, Z.; Houk, K. N. J. Org. Chem.
2011, 76, 5074−5081.
(20) Reactions in D2O were inhomogeneous.
(21) Since cyclic enones are fixed in the s-trans conformation; they
cannot undergo a concerted addition such as displayed in Scheme 2.
(22) Referring to the Erlenmeyer rule, which states that compounds
of the general form R2C(YH)(X) will decompose to R2CY and HX
(X, Y = heteroelements); a discussion of the origin and scope of this
rule is in preparation (L.H.).
AUTHOR INFORMATION
Corresponding Author
*E-mail: lukas.hintermann@tum.de.
Notes
■
(23) Bohme, H.; Bezzenberger, H.; Clement, M.; Dick, A.; Nurnberg,
̈
̈
The authors declare no competing financial interest.
E.; Schlephack, W. Chem. Ber. 1959, 623, 92−102.
(24) Kresge, A. J. Acc. Chem. Res. 1975, 8, 354−360.
(25) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic
Chemistry; University Science Books: Sausalito, CA, 2006; p 567.
REFERENCES
(1) Hart, H. Chem. Rev. 1979, 79, 515−528.
(2) Data for enol content of carbonyl compounds: (a) Wirz, J. Adv.
Phys. Org. Chem. 2010, 44, 325−356. (b) Keeffe, J. R.; Kresge, A. J.;
Schepp, N. P. J. Am. Chem. Soc. 1990, 112, 4862−4873. (c) Keeffe, J.
R.; Kresge, A. J.; Schepp, N. P. J. Am. Chem. Soc. 1988, 110, 1993−
1995.
■
(3) Chiang, Y.; Hojatti, M.; Keeffe, J. R.; Kresge, A. J.; Schepp, N. P.;
Wirz, J. J. Am. Chem. Soc. 1987, 109, 4000−4009 and references cited
therein.
(4) Henne, A.; Fischer, H. Angew. Chem. 1976, 88, 445−446; Angew.
Chem., Int. Ed. Engl. 1976, 15, 435−435.
(5) Hoffmann, H. M. R.; Schmidt, E. A. Angew. Chem. 1973, 85,
227−227; Angew. Chem., Int. Ed. Engl. 1973, 12, 239−240 and
references cited therein.
(6) (a) Capon, B.; Rycroft, D. S.; Watson, T. W.; Zucco, C. J. Am.
Chem. Soc. 1981, 103, 1761−1765. (b) Capon, B.; Siddhanta, A. K. J.
Org. Chem. 1984, 49, 255−257. (c) Chiang, Y.; Kresge, A. J.; Walsh, P.
A. J. Am. Chem. Soc. 1986, 108, 6314−6320.
(7) (a) Park, J.; Chin, C. S. J. Chem. Soc., Chem. Commun. 1987,
1213−1214. (b) Chin, C. S.; Lee, S. Y.; Park, J.; Kim, S. J. Am. Chem.
Soc. 1988, 110, 8244−8245.
(8) Bergens, S. H.; Bosnich, B. J. Am. Chem. Soc. 1991, 113, 958−967.
(9) For a highly efficient catalyst based on bifunctional ruthenium
complexes, see: Larsen, C. R.; Grotjahn, D. B. J. Am. Chem. Soc. 2012,
134, 10357−10360.
D
dx.doi.org/10.1021/jo3021709 | J. Org. Chem. XXXX, XXX, XXX−XXX