Diastereoselectivity of enolate anion protonation. H/D exchange of β-substituted ethyl butanoates in ethanol-d

Article abstract of DOI:10.1021/ja962631s

The stereochemistry of base-catalyzed H/D exchange on 13 β-substituted ethyl butanoates in ethanol-d has been studied in order to analyze the steric and electronic factors which control the diastereoselectivity of electrophilic attack on enolate anions. E

Full text of DOI:10.1021/ja962631s

J. Am. Chem. Soc. 1997, 119, 479-486
479
Diastereoselectivity of Enolate Anion Protonation. H/D
Exchange of â-Substituted Ethyl Butanoates in Ethanol-d
Jerry R. Mohrig,* Robert E. Rosenberg,1 John W. Apostol, Mark Bastienaansen,
Jordan W. Evans, Sonya J. Franklin, C. Daniel Frisbie, Sabrina S. Fu,
Michelle L. Hamm, Christopher B. Hirose, David A. Hunstad, Thomas L. James,
Randall W. King, Christopher J. Larson, Hallie A. Latham, David A. Owen,
Karin A. Stein, and Ronald Warnet2
Contribution from the Department of Chemistry, Carleton College, Northfield, Minnesota 55057
ReceiVed July 29, 1996X
Abstract: The stereochemistry of base-catalyzed H/D exchange on 13 â-substituted ethyl butanoates in ethanol-d
has been studied in order to analyze the steric and electronic factors which control the diastereoselectivity of
electrophilic attack on enolate anions. Electrophilic deuteration of the enolate anion also determines the stereoselectivity
of 1,4-conjugate addition of ethanol-d to R,â-unsaturated esters. Experimental conditions were selected which
rigorously exclude the effects of ion pairing and aggregation. The research showed that stereoelectronic factors
generally produce higher stereoselection than steric effects do. Electronegative heteroatom substituents at C-3 produced
a 10:1 ratio of the 2R*,3R*/2R*,3S* 2-deuteriobutanoates. In the most stable transition states for electrophilic attack,
these electronegative substituents occupy an antiperiplanar position to the forming C-D bond. Only with a â-tert-
butyl substituent did steric effects produce high stereoselection, and it fell off rapidly with a decrease in carbon
branching. Protonation of acyclic â-ethoxy aldehyde and ketone enolates follows the same diastereoselectivity pattern
as the â-ethoxy ester enolate, but protonation of the cyanocarbanion from a â-ethoxy nitrile gives much lower
stereoselection.
Enols and enolate anions are intermediates in many important
reactions of carbonyl compounds. There is substantial interest
in how enzymes activate protons R to the carbonyl groups of
esters and carboxylates and how they control the chemistry of
enolic species.3,4 The key position played by enolates in organic
synthesis, especially in the Michael and aldol reactions, has led
to an enormous wealth of data in recent years. Stereocontrol
has been of particular importance, especially with conforma-
tionally-mobile acyclic substrates.5 However, in spite of its
fundamental importance, the diastereoselectivity of electrophilic
attack on enolate anions is much less well studied than the
stereochemistry of nucleophilic attack at the carbonyl group.
Computational chemists have discussed the importance of
both steric and stereoelectronic effects in electrophilic reactions
of carbon-carbon double bonds.6-8 Much of the theoretical
work on the stereochemistry of electrophilic attack on enolates
derives from the wider understanding of nucleophilic attack at
the carbonyl group. From an experimental viewpoint, Lodge
and Heathcock state the problem nicely in their treatment of
steric and stereoelectronic effects in the diastereofacial dif-
ferentiation of nucleophilic additions to chiral aldehydes.9
The most frequent interpretation used for stereoselection in
the reactions of enolate anions with electrophiles is basically
steric in nature. This has been described by Fleming and his
co-workers in terms of 1.10 Transition state 1 has the approach
of the electrophile on the face of the π-bond opposite to the
largest group, with the smallest group at the stereogenic center
gauche to the π-system. This avoids destabilizing allylic 1,3-
interactions between the “inside” gauche group and the groups
syn to it on the enolate double bond. When the medium group
does not occupy the “outside” position, greater A strain is
present.
Studying enolate anion protonation offers a potentially
valuable way to analyze the various steric and electronic factors
that control the diastereoselectivity of electrophilic attack.
Surprisingly few systematic investigations of the stereochemistry
of enolate protonation have been reported. Although islands
of understanding are available for cyclic11 and acyclic10,11
systems, the variable effects of aggregation have been a
stumbling block with regard to a fundamental and comprehen-
sive grasp of the steric and electronic factors controlling the
stereochemistry.12-14 Current efforts to sort out equilibria and
kinetics when aggregation is a factor should make this less
X Abstract published in AdVance ACS Abstracts, January 1, 1997.
(1) Present address: Department of Chemistry, SUNY, College at
Geneseo, Geneseo, NY 14454.
(2) Simpson College, Indianola, IA 50125.
(3) Gerlt, J. A.; Gassman, P. G. Biochemistry 1993, 32, 11943-11952.
(4) Mohrig, J. R.; Moerke, K. A.; Cloutier, D. M.; Lane, B. D.; Person,
E. C.; Onasch, T. B. Science 1995, 269, 527-529.
(5) Eliel, E. L.; Wilen, S. H. Stereochemistry of Carbon Compounds;
Wiley-Interscience, 1994; Chapter 12.
(6) (a) Paddon-Row, M. N.; Rondan, N. G.; Houk, K. N. J. Am. Chem.
Soc. 1982, 104, 7162-7166. (b) Houk, K. N.; Paddon-Row, M. N.; Rondan,
N. G.; Wu, Y.-D.; Brown, F. K.; Spellmeyer, D. C.; Metz, J. T.; Li, Y.;
Loncharich, R. J. Science 1986, 231, 1108-1117.
(7) Bach, R. D.; Badger, R. C.; Lang, T. J. J. Am. Chem. Soc. 1979,
101, 2845-2848.
(8) McGarvey, G. L.; Williams, J. M. J. Am. Chem. Soc. 1985, 107,
1435-1437.
(9) Lodge, E. P.; Heathcock, C. H. J. Am. Chem. Soc. 1987, 109, 3353-
3361.
(10) Crump, R. A. N. C.; Fleming, I.; Hill, J. H. M.; Parker, D.; Reddy,
N. L.; Waterson, D. J. Chem. Soc., Perkin Trans. 1 1992, 3277-3294.
(11) (a) Zimmerman, H. E. Acc. Chem. Res. 1987, 20, 263-268. (b)
Zimmerman, H. E.; Chang, W.-H. J. Am. Chem. Soc. 1959, 81, 3634-
3643.
S0002-7863(96)02631-5 CCC: $14.00 © 1997 American Chemical Society

480 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Mohrig et al.
Figure 1. Formation of the 2R*,3R* and 2R*,3S* diastereomers in
base-catalyzed H/D exchange of â-substituted ethyl butanoate deriva-
tives.
difficult in the future.15 With the current state of knowledge,
however, it seems wise to avoid the complexities of the
aggregates ubiquitous in nonpolar organic solvents when
attempting to study the structural factors that control the
diastereoselectivity of the reactions of enolate anions with
electrophiles.
We have studied the hydrogen-deuterium exchange of
carbonyl compounds in ethanol-d under conditions that rigor-
ously avoid the formation of aggregated species. In addition,
to minimize the complex conformational effects that so often
dominate the stereochemistry of cyclic compounds, we chose
to focus our study on a series of ethyl butanoates with substi-
tuents at the 3-position (2) and their aldehyde and ketone analogs
(Figure 1). This has allowed investigation of the intrinsic effect
of a neighboring stereogenic center upon the diastereoselectivity
of the electrophilic deuteration of enolate anions.
We have presented our initial results in preliminary form and
now describe our work in full.16 To gain further insight into
this process, we have also carried out ab initio MO calculations
on the transition states for gas-phase protonation of the enolate
anion of 3-fluorobutanoic acid by HCN.17
Figure 2. The enolate pathway for H/D exchange.
Table 1. Stereoselection of H/D Exchange on Ethyl Butanoates
Substituted at C-3
chemical shift of
diastereotopic
C-2 deuteronsa
% 2R*,3R*
2R*,3R*
(3)
2R*,3S*
(4)
diastereomer
compd
substituent
(3/4 ratio)c
2a
2b
2c
2d
2e
2f
2g
2h
2i
OEt
OPh
CMe3
OCMe3
SCMe3
OMe
CF3
CH(CO2Et)2
Ph
2.13
2.21
1.92
2.22
2.42
2.11
2.35
2.18
2.35
1.84
1.91
1.95
1.90
2.45
2.61
2.32
2.44
2.56b
2.43
2.55
2.45
2.43
2.07
2.16
2.12
2.09
91 (10.1)
91 (10.1)
90 (9.0)
89 (8.1)d
88 (7.4)
87 (6.7)
83 (4.9)
79 (3.8)
75 (3.0)
75 (3.0)
70 (2.3)
68 (2.1)
59 (1.4)
Diastereoselectivity of H/D Exchange
2j
2k
2l
CN
CHMe2
CH2CMe3
CH2Me
Since the prime stereochemical determinant in enolate anion
protonation is the influence of the neighboring stereogenic
center, we have examined a large number of â-substituents in
order to discern the relative importance of their steric and
electronic consequences. At the outset, however, it is important
to ask if we are indeed looking at the chemistry of solvated
enolate anions, without chelation or aggregation interactions.
We used ethyl 3-ethoxybutanoate (2a) to help answer this
question.
First of all, we chose to use ethanol-d (EtOD), a relatively
polar solvent, and very dilute base concentrations (0.01-0.06
M NaOEt) in our experiments. The use of 0.01 M KOEt with
0.031 M 18-crown-6 produced no change in the stereochemistry.
With 0.04 M (CH3)4NOH as the basic catalyst, over a range of
16-24% H/D exchange, we obtained the same result (93% 3a)
as with NaOEt and KOEt.
2m
a δ in ppm using benzene (δ 7.15) as solvent. b In CHCl3 (δ 7.24).
c Determined by repetitive 30-76 MHz 2H NMR integrations at a range
of conversions. d Same result for the ethyl and tert-butyl esters.
diastereoselectivities (80% and 72% 3a, respectively); however,
when Me4NOH was used as the catalyst, and chelation became
impossible, the amount of 3a was 88% and 92%, respectively.
Thus, when the effects of ion pairing are reduced, even in
nonpolar solvents, the stereoselectivity becomes virtually the
same as when EtOD is the solvent.
Although unlikely, it is possible that H/D exchange of 2a
could proceed through an elimination-addition mechanism. In
order to test this hypothesis, we carried out a double isotope
experiment. Dideuterated 2a was used as the substrate in an
ethoxide-catalyzed H/D exchange experiment using 13C-labeled
ethanol as the reaction solvent. After over 95% H/D exchange
had occurred, along with complete exchange of 13C into the
ethyl ester, the 13C NMR peak for the ethoxy group at C-3 was
less than 1% above natural isotopic abundance. Thus, the H/D
exchange proceeds by base-catalyzed proton removal and
electrophilic deuteration by EtOD, with no competing elimina-
tion-addition pathway (Figure 2).
Since the Me4N+ cation is unable to coordinate with the
enolate anion, it is difficult to believe that ion pairing plays
any major role in the diastereoselectivity that we observe, unless
the stereochemistry is markedly resistant to aggregation phe-
nomena. That this is not the case was shown by solvent
studies.13 The use of NaOEt as the base and benzene or
tetrahydrofuran as the solvent produced lower H/D exchange
(12) Gerlach, U.; Haubenreich, T.; Hu¨nig, S.; Keita, Y. Chem. Ber. 1993,
126, 1205-1215.
(13) Mohrig, J. R.; Lee, P. K.; Stein, K. A.; Mitton, M. J.; Rosenberg,
R. E. J. Org. Chem. 1995, 60, 3529-3532.
(14) Seebach, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 1624-1654.
(15) (a) Collum, D. B. Acc. Chem. Res. 1992, 25, 448-454. (b) Krom,
J. A.; Streitwieser, A. J. Am.Chem. Soc. 1992, 114, 8747-8748. (c) Abbotto,
A.; Streitwieser, A. J. Am. Chem. Soc. 1995, 117, 6358-6359.
(16) (a) Mohrig, J. R.; Fu, S. S.; King, R. W.; Warnet, R.; Gustafson,
G. J. Am. Chem. Soc. 1990, 112, 3665-3667. (b) Mohrig, J. R.; Franklin,
S. J.; King, R. W.; Larson, C. J.; Hirose, C. B.; Frisbie, C. D.; Warnet, R.
Abstracts of the 199th National Meeting of the American Chemical Society,
Boston, MA, April 1990; paper ORGN 404.
Table 1 presents stereochemical data for H/D exchange with
EtOD/NaOEt on 13 different butanoate esters representing a
wide range of substituents at C-3.
The first four entries in Table 1 (2a-d), where H/D exchange
gives the highest diastereoselectivity, are within experimental
error of one another, and 2e,f produce stereoselection nearly as
high. All but 2c have electronegative substituents at the
â-position. With compounds having â-oxygen substituents,
steric effects of the group attached to oxygen play a minimal
role and electronic effects dominate. Comparison of the
stereoselection for 2d and 2l is especially revealing. Whereas
(17) Rosenberg, R. E.; Mohrig, J. R. J. Am. Chem. Soc. 1997, 119, xxxx-
xxxx.

DiastereoselectiVity of Enolate Anion Protonation
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 481
Table 2. Stereoselection of H/D Exchange on â-Ethoxy Carbonyl
carbonyl compounds as well. The H/D exchange of the tert-
butyl ketone 6 is consistent with the results obtained for
â-ethoxy esters. The aldehyde 5 shows somewhat less but not
substantially different stereoselectivity. However, the nitrile 7,
not having a trigonal center at C-1, provides much lower
diastereoselection. It is possible that our usual NMR-structure
correlation does not extend to the nitrile, whose configurational
assignment was not determined independently, so we are
uncertain if the stereoselection is actually reversed from the
general pattern.
Lower stereoselection for electrophilic attack of the nitrile’s
conjugate base was not unexpected and is consistent with
Fleming’s results on the alkylation of carbanions R to a CtN
group.10 Presumably, this low diastereoselection results from
the lack of destabilizing 1,3-interactions with the gauche group
in the cyanocarbon anion.19 The consequence is that the two
diastereomeric transition states are much more equal in energy
than is the case with enolate anions. Fishbein and Jencks have
pointed out that it is not clear whether a negative charge R to
a cyano group produces a planar, resonance-stabilized carbanion
or an sp3-hybridized carbanion.20 Boche has shown that a
cyclopropyl R-cyanocarbanion has a tetrahedral configuration,
in contrast to the comparable enolate anion; in addition, his
calculations on CH2CN- indicate a nonplanar R-carbon atom.21
Our computations, which included electron correlation at the
MP2 level, are also consistent with a largely tetrahedral
carbanion, quite different from the planar enolate anion.22
Consistent with their pKa’s, the rate of H/D exchange varied
substantially for different substrates. For example, the aldehyde
5 reacted about 10 times faster than the nitrile 7 under
comparable conditions. Among the â-substituted butanoate
esters, the alkyl-substituted substrates reacted significantly
slower than those with â-heteroatom substituents.
Compounds
chemical shift of
diastereotopic
C-2 deuteronsa
% 2R*,3R* diastereomerc
compd
2R*,3R* 2R*,3S*
(2R*,3R*/2R*,3S*)
OEt
O
5
6
7
1.91
2.42
2.52
2.18
2.84
2.63b
84 (5.3)
H
OEt
OEt
O
88 (7.3)
44 (0.8)
CMe3
CN
a δ in ppm using benzene (δ 7.15) as solvent. b In acetone (δ 2.04).
c Determined by repetitive 76 MHz 2H NMR integrations at a range of
conversions.
a â-tert-butoxy substituent gives an 8:1 ratio of the two product
diastereomers, the ratio for the larger neopentyl substituent is
only 2:1; the importance of an electronic factor seems inescap-
able. However, compound 2c, which has a â-tert-butyl sub-
stituent, clearly shows that steric effects can also be important
in stereoselection. The same steric factors are likely involved
with a â-phenyldimethylsilyl group, which we have studied
using the methyl â-phenylpropionate system; here 85% of the
2R*,3R* diastereomer was produced.18
In order to further probe the importance of stereoelectronic
effects on H/D exchange, we attempted to see if there were a
Hammett Fσ-relationship for analogs of 2b. We compared the
effect of the â-phenoxy group with the â-m-nitrophenoxy and
â-p-aminophenoxy substituents. We hoped that this range of
Hammett σ values of 1.3 might be sufficient to discern a non-
zero F value. However, the diastereoselectivities of H/D
exchange for all three compounds were essentially within
experimental error of one another; deuteration of the m-
nitrophenoxy compound gave 92% of its 2R*,3R* diastereomer,
while the p-aminophenoxy compound gave 89%.
The last three entries in Table 1, all with â-alkyl substituents,
show that steric effects fall off rapidly with a decrease in carbon
branching. Among the alkyl substituents, even the isopropyl
group (2k) shows only modest stereoselection, forming 70%
3k. A neopentyl substituent (2l) produces only slightly more
stereoselection than an ethyl group (2m). Clearly, branching
more than two atoms removed from the π-system of the enolate
anion affects the stereochemistry of electrophilic attack only
slightly. Entry 2g, with a â-trifluoromethyl group, is interesting
in that it gives reasonably high stereoselection even though it
is a carbon substituent of modest size. Entries 2h,i show
intermediate diastereoselection and probably involve a combina-
tion of steric and electronic factors. Although the â-cyano group
of 2j produces lower stereoselection than other electron-
withdrawing substituents, its small size argues that electronic
effects must be operating.
While the data in Table 1 show that different â-substituted
ethyl esters can produce substantially different diastereoselec-
tivities, the effect of different carbonyl groups remained to be
shown. For entry 2d of Table 1, data are reported for two
different â-tert-butoxy esters having different alkyl groups; the
ethyl and tert-butyl esters gave exactly the same stereoselectivity
in H/D exchange. Table 2 presents the stereoselection of three
additional compounds: 3-ethoxybutanal (5), 2,2-dimethyl-5-
ethoxy-3-hexanone (6), and 3-ethoxybutanenitrile (7).
Stereochemistry and Mechanism of 1,4-Conjugate
Addition
The original motivation for our enolate work was to under-
stand the stereochemistry of conjugate nucleophilic addition to
R,â-unsaturated esters. We have shown that the conjugate
addition of ethanol-d (EtOD) to ethyl (E)-2-butenoate (8)
produces 92 ( 1% of the 2R*,3R* diastereomer (3a), the same
amount within experimental error as that produced in the H/D
exchange of 2a in EtOD (Figure 3). This suggests that the same
enolate intermediate is responsible for the stereochemistry that
we have observed in conjugate addition and in H/D exchange
experiments.
The stereoconvergence that we observed in addition reactions
of 2-methyl-2-propanethiol-d to 8 and its (Z)-isomer is further
evidence that electrophilic deuteration of a common enolate
anion intermediate produces the diastereoselectivity. Reaction
of 8, ranging over 15-90% completion, gave 89 ( 0.4% 3e,
whereas reaction of ethyl (Z)-2-butenoate gave 83 ( 1.4% 3e.
The H/D exchange of 2e in EtOD and 0.01 M NaOEt gave
88% 3e at 11% conversion, consistent with enolate protonation
as the stereochemical determinant in the conjugate additions.
This kind of stereoselectivity has also been observed by
Fleming,10 although stereospecificity in the conjugate addition
of sulfur nucleophiles has been reported using THF and lithium
thiophenoxides.23
(19) Bordwell, F. G.; Van Der Puy, M.; Vanier, N. R. J. Org. Chem.
1976, 41, 1883-1885.
(20) Fishbein, J. C.; Jencks, W. P. J. Am. Chem. Soc. 1988, 110, 5087-
5095.
(21) Boche, G. Angew. Chem., Int. Ed. Engl. 1989, 28, 277-297.
(22) The dihedral angle at C-2 of the R-cyanocarbanion conjugate base
of acetonitrile was calculated to be 130.3° (tetrahedral ) 120°, trigonal )
180°) using ab initio MP2 (fc)/6-31G* calculations as described in ref 17.
The exchange data in Table 2 show that the high diastereo-
selection which we found for ethyl esters extends to other
(18) We thank Prof. Ian Fleming for providing this compound.

482 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Mohrig et al.
Apparently, the rate of conjugate addition slows appreciably
when the nucleophile is the bulky tert-butoxide, and Michael
addition competes effectively, thereby removing the desired
product from the reaction mixture. However, the synthesis of
2d and its tert-butyl analog could be carried out by the acid-
catalyzed reaction of 2-methylpropene with the appropriate
3-hydroxybutanoate ester.
The three substrates shown in Table 2 are all known
compounds. The â-ethoxy aldehyde 5 and nitrile 7 were
synthesized by ethoxide-catalyzed conjugate addition of ethanol
to crotonaldehyde and crotononitrile, respectively. Synthesis
of ketone 6 was more complex. Although Powell and Wasser-
man reported that it could not be made from acetaldehyde and
pinacolone,24 this methodology was satisfactory in our hands.
Following the aldol condensation and dehydration to give largely
the (E)-alkene, conjugate addition of ethanol provided 6.
The results from stereochemical investigations such as this
one rest on a foundation of unambiguous configurational
assignments. In our case this means a firm NMR-structure
correlation for the diastereotopic protons R to the carbonyl group.
During our research on the stereochemistry of addition-
elimination reactions of butanoate esters and thioesters substi-
tuted at C-3, we synthesized over 10 different compounds
stereospecifically deuterated at C-2 and provided unambiguous
configurational assignments for each of them. In every case,
the NMR spectrum of the 2R*,3R* diastereomer has the C-2
Figure 3. The common stereochemistry of 1,4-conjugate addition and
H/D exchange.
In every case where we have compared the diastereoselec-
tivities of nucleophilic conjugate addition and H/D exchange,
the percentage of 3 has been the same within experimental error
of our 2H NMR integration methodology (e(2%). In addition
to the same diastereomeric ratios of 3a/4a and 3e/4e in conjugate
addition and H/D exchange, the two reactions involving
3-ethoxybutanal (5) gave 86.5 ( 0.5% and 83.7 ( 1% of the
2R*,3R* diastereomer, respectively. Also, ethoxide-catalyzed
conjugate addition of diethyl malonate to 8, as well as H/D
exchange on diethyl 2-(ethoxycarbonyl)-3-methylpentanedioate
(2h), gave 79% of 3h. All of our experimental evidence points
to the fact that electrophilic deuteration of the enolate anion
determines the stereochemistry of base-catalyzed 1,4-conjugate
addition as well as of H/D exchange.
Ethoxide-catalyzed nucleophilic additions of EtOD to the R,â-
unsaturated alkenes were generally slower than H/D exchange
on the â-ethoxy adducts under comparable conditions. For this
reason and because 8 isomerizes to the (Z)-isomer in the
presence of EtO-/EtOD, the stereochemistry of conjugate
addition was determined from reactions in which the conversion
to product was less than 10%; at 5% addition of EtOD to 8,
4% ethyl (Z)-2-butenoate had formed. In the case of adduct 7,
the rate of H/D exchange was so much faster than nucleophilic
addition of EtOD that we were unable to determine the
stereochemistry of the addition reaction.
1
proton in the downfield portion of the AB pattern of the H
spectrum and the C-2 deuteron upfield in the 2H spectrum.13,25
This has proved to be a useful empirical pattern for making
relative configurational assignments. In the case of 2b, 2f, 2h,
and 2j-m, as well as 6 and 7, this NMR-structure correlation
was assumed to hold. The relative configurations at C-2 and
C-3 were determined unambiguously for all other products of
H/D exchange reported herein; in every case the configurational
assignments for these compounds also matched the general
NMR-structure correlation.
In most cases CDCl3 was not a useful solvent for separating
the diastereotopic protons, but benzene-d6 had a dramatic effect
on their resolution, usually separating them by over 0.2 ppm.
The use of benzene as the 2H NMR solvent proved to be
invaluable for obtaining good baselines between the diaste-
reotopic deuteron peaks and thus good integrations.
We have used two strategies for our configurational assign-
ments. The first is based upon 3-hydroxy[2-2H1]butanoate (9),
synthesized by NaBD4 reduction of the epoxide of (E)-2-
butenoate.25a This SN2 reaction proceeds with inversion of
configuration at C-2 and produces the 2R*,3R* diastereomer
of 9. The NMR-structure correlation of 2a was assigned by
its synthesis from the alkoxide salt of 9 with ethyl iodide.16 In
the case of 2d, a sample of the tert-butyl ester which had
undergone H/D exchange was transformed into 9 by reaction
with sulfuric acid in CH2Cl2, giving the same relative integra-
tions for the upfield and downfield 2H NMR peaks. The NMR-
structure correlation for the aldehyde 5 was carried out by O2
oxidation to the carboxylic acid, followed by acid-catalyzed
esterification to 2a.
Synthesis and Proof of Configuration
Ten of the substrates shown in Table 1 are known compounds.
The majority are easily synthesized by esterification of the
commercially-available â-substituted carboxylic acids or by
base-catalyzed conjugate addition of the appropriate nucleophile
to alkene 8. In the case of 2b, conditions could not be found
where phenol would add to 8, so conjugate addition to ethyl
2-butynoate was followed by hydrogenation using Wilkinson’s
catalyst. Syntheses of 2c, 2g, and 2k were built on Horner-
Emmons-Wadsworth methodology, followed again by hydro-
genation.
Potassium tert-butoxide-catalyzed nucleophilic addition of
tert-butyl alcohol to tert-butyl (E)-2-butenoate was not an
effective route to 2d because the ester enolate of the conjugate
addition product reacted further with the R,â-unsaturated ester
to give a dimeric product, tert-butyl 3-methyl-4-(tert-butoxy-
carbonyl)-4-hexenoate. At 49% substrate conversion, only 12%
tert-butyl 3-tert-butoxybutanoate had formed, and longer reac-
tion times gave lower conversion to the desired product.
The second strategy is based upon the syn addition of D2
across the CdC of 2-butenoate esters substituted at C-3.25b When
both diastereomers of the alkenes have been available, their
configurations were proved by NMR chemical shifts and NOE
effects. Thus, the NMR-structure correlations for 2c and 2e
(24) Powell, S. G.; Wasserman, W. J. J. Am. Chem. Soc. 1957, 79, 1934-
1938.
(25) (a) Mohrig, J. R.; Vreede, P. J.; Schultz, S. C.; Fierke, C. A. J.
Org. Chem. 1981, 46, 4655-4658. (b) Mohrig, J. R.; Dabora, S. L.; Foster,
T. F.; Schultz, S. C. J. Org. Chem. 1984, 49, 5179-5182.
(23) Miyata, O.; Shinada, T.; Ninomiya, I.; Naito, T.; Date, T.; Okamura,
K.; Inagaki, S. J. Org. Chem. 1991, 56, 6556-6564.

DiastereoselectiVity of Enolate Anion Protonation
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 483
were accomplished by homogeneous catalytic deuterogenation
of alkenes whose configurations were known. The syntheses
of both 2c and 2g used Horner-Emmons-Wadsworth meth-
odology in the C-C bond forming steps. In the former, using
pinacolone and the conjugate base of triethyl phosphonoacetate,
the reaction produced a 97:3 E/Z mixture of the known ethyl
3,4,4-trimethyl-2-pentenoate.26 The (E)-isomer was deutero-
genated, giving the (2R*,3S*)-[2,3-2H2] derivative of 2c; in this
case our NMR-structure relationship also correlates with the
1H NMR assignments of the C-2 diastereotopic protons of 2c
made by Hart and Krishnamurthy, assuming likely conforma-
tional populations.27 Our relative configurations at C-2 and C-3
is also made more complex by the fact that steric and electronic
effects can reinforce one another. Therefore, experimentally,
it is difficult to separate them. Whether electronic stabilization
of the transition state plays a role is made ambiguous because
the outcome is the same as with the steric considerations
commonly cited. The suggestion that stereoelectronic effects
can be important in reactions of enols was made 40 years ago
to account for the axial preference in reactions at the R-position
of cyclic ketones.31 However, the axial preference for removal
or introduction of a proton in decalones and tert-butylcyclo-
hexanone (∼5.5-fold) has been considered too small to conclude
whether it results from stereoelectronic or steric effects.32
Fleming and his co-workers have suggested that there is
considerable doubt about how much electrophilic attack on a
double bond adjacent to a stereogenic center is controlled by
steric and how much by electronic factors.33 They found that
carbon substituents and electropositive heteroatoms give a
relatively orderly pattern of stereoselection that can be under-
stood using 1, but electronegative substituents were more
difficult to explain, and the detection of an electronic component
to the diastereoselectivity was considered tentative. Contrary
to our conclusions, they suggested that there is a reasonable
probability that oxygen substituents might be more or less
orthogonal rather than antiperiplanar to the developing bond.10
Many computations have pointed to the importance of
stereoelectronic effects in electrophilic reactions of enols and
enolate anions. The calculations of Houk and his colleagues
generally support the view that an allylic electron-donating
substituent will take a perpendicular relationship to a π-system,
so that it can participate in a hyperconjugative interaction.6 The
less sterically demanding remaining substituent is gauche to the
enolate in the favored transition state. This idea is comple-
mentary to the explanations that have been offered for the
corresponding reactions involving nucleophilic attack on a
carbonyl group.
In addition, it has been suggested that the σ* molecular orbital
of an electronegative substituent may also have substantial
amplitude for bonding to a π-system.34 Hehre and his col-
leagues have suggested that these interactions may be particu-
larly important with electron-rich allylic alcohols and ethers.35
In our reactions, this would imply that the stabilizing interaction
has the antiperiplanar electronegative group at C-3 accepting
electron density from the electron-rich π-system of the enolate.
The electron delocalization, made possible by the antiperiplanar,
electronegative â-substituent, can stabilize the transition state
for protonation. An alternate view of the stereoelectronic effect
is that it is really an electrostatic effect, where the best transition-
state rotamers are those that minimize dipole-dipole repulsions
while also minimizing gauche interactions. When dipole-
dipole repulsions are lowered in the transition state for proton
transfer, as they are when a â-electronegative substituent is
antiperiplanar to the forming C-H bond, that transition state is
stabilized.
1
of 2i are also consistent with Hart’s H NMR assignments. In
the case of 2g, reaction of 1,1,1-trifluoroacetone with the
conjugate base of triethyl phosphonoacetate produced largely
ethyl (E)-3-methyl-4,4,4-trifluoro-2-butenoate,28 which was deu-
terogenated to give the 2R*,3S* diastereomer.
Discussion
A great deal has been learned about the fundamental kinetic
and thermodynamic properties of enols and enolates in recent
years, particularly through the work of Kresge and his associ-
ates.29 Much less is understood about the stereochemistry of
these important reaction intermediates. The synthesis of
preformed enolate anions, using strong bases such as lithium
diisopropylamide (LDA) in tetrahydrofuran (THF) or hexa-
methylphosphoric triamide (HMPA), has been tremendously
effective in organic synthesis; however, general structure-
reactivity correlations for the supramolecular aggregates pro-
duced under these reaction conditions are still poorly understood.
The dramatic dissociative effects of lithium and tetraalkylam-
monium salts, as well as crown ethers, are a case in point.30
Comparison of our results, under conditions where ion pairing
and aggregation phenomena play no role, with those of Hart
and Krishnamurthy, where they do, is useful in this regard.27
With LDA in THF, and acetic acid-d as the electrophile, they
found that 75% of the 2R*,3R* diastereomer was produced from
the â-tert-butyl ester 2c; our research showed 90% of the
2R*,3R* diastereomer. With the â-phenyl compound 2i they
found 61% of the 2R*,3R* diastereomer, compared to 75%
under our conditions. Though the trends are in the same
direction, the numbers themselves differ substantially. Perhaps
even more relevant is the research of Hu¨nig and his colleagues
on the stereochemistry of protonation and alkylation of enolates
produced with LDA in the 4-tert-butylcyclohexyl system,
including ester, ketone, nitrile, and sulfone enolates.12 After a
particularly careful investigation, they found that no general rule
for the diastereoselectivity of protonation was possible because
of incomplete knowledge of the aggregates present. Our work
with ethyl 3-hydroxy- and 3-alkoxybutanoates is another case
in point. Here we showed that the diastereoselectivity of enolate
protonation can even be reversed under aggregation conditions.13
Understanding the fundamental causes that produce stereo-
selection in the reactions of enolate anions with electrophiles
In any case, our results can only be explained by recognizing
the important role of electronic factors in the stereoselection of
(26) Weedon, A. C. Can. J. Chem. 1984, 62, 1933-1939.
(27) Hart, D. J.; Krishnamurthy, R. J. Org. Chem. 1992, 57, 4457-4470.
(28) (a) Poulter, C. D.; Satterwhite, D. M. Biochemistry 1977, 16, 5470-
5478. (b) Abele, H.; Haas, A.; Lieb, M. Chem. Ber. 1986, 119, 3502-
3506.
(29) (a) Chiang, Y.; Kresge, A. J. Science 1991, 253, 395-400. (b)
Keeffe, J. R.; Kresge, A. J.; Schepp, N. P. J. Am. Chem. Soc. 1990, 112,
4862-4868.
(30) (a) Gerlach, U.; Haubenreich, T.; Hu¨nig, S. Chem. Ber. 1994, 127,
1981-1988. (b) Seebach, D.; Beck, A. K.; Studer, A. In Modern Synthetic
Methods 1995; Ernst, B., Leumann, C., Eds.; Verlag Helvetica Chimica
Acta: Basel, VCH: Weinheim, 1995; Vol. 7, pp 1-178. (c) Nakamura,
E.; Tanaka, K.; Fujimura, T.; Aoki, S.; Williard, P. G. J. Am. Chem. Soc.
1993, 115, 9015-9020.
(31) Corey, E. J.; Sneen, R. A. J. Am. Chem. Soc. 1956, 78, 6269-
6275.
(32) Toullec, J. In AdVances in Physical Organic Chemistry; Gold, V.,
Bethell, D., Eds.; Academic Press: London, 1982; Vol. 18, pp 1-77.
(33) Fleming, I.; Lewis, J. J. J. Chem. Soc., Chem. Commun. 1985, 149-
151.
(34) (a) Kozikowski, A. P.; Ghosh, A. K. J. Am. Chem. Soc. 1982, 104,
5788-5789. (b) Hori, K.; Higuchi, S.; Kamimura, A. J. Org. Chem. 1990,
55, 5900-5905.
(35) (a) Kahn, S. D.; Pau, C. F.; Chamberlin, A. R.; Hehre, W. J. J. Am.
Chem. Soc. 1987, 109, 650-663. (b) Kahn, S. D.; Hehre, W. J. J. Am.
Chem. Soc. 1987, 109, 666-671.

484 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Mohrig et al.
particular, if the enolate configuration were an important factor
in the stereoselection, it is difficult to envision how virtually
the same diastereomeric mixture of products would result from
the H/D exchange of aldehyde 5 and tert-butyl ketone 6, which
are predicted to produce different enolate geometries under
kinetic control.5
Figure 4. The trans enolate, the likely H/D exchange intermediate.
electrophilic attack on the enolate anions. We see no other way
to understand the pattern for H/D exchange on the â-substituted
ethyl butanoates shown in Table 1. All but one of the
substituents which produce the highest diastereoselectivities have
an electronegative oxygen or sulfur atom attached to the
stereocenter at C-3. The OCMe3 substituent gives a much
higher diastereoselectivity than the larger CH2CMe3 substituent.
The size of the group attached to the oxygen plays virtually no
role in the stereoselection; OEt, OPh, OCMe3, and OMe produce
almost the same result. So, the effect of the OR group is
independent of R.
With substituents that have no electronegative â-heteroatoms,
the stereoselection drops dramatically. Only tert-butyl, trifluo-
romethyl, and cyano are exceptions to this general pattern. The
tert-butyl case shows that steric factors also influence the
stereoselection of H/D exchange, but the effect is small unless
substantial γ-branching is present. The pattern of stereoselection
is CMe3 . CHMe2 ∼ CH2CMe3 > CH2Me. It is not unlikely
that our results using a â-phenyldimethylsilyl group (85%
2R*,3R*) are also due partly to steric factors, as has previously
been suggested in the case of a Ph(MeS)2C substituent.36
With 83% of the 2R*,3R* diastereomer produced by H/D
exchange of 2g, the effect of the â-trifluoromethyl group is also
substantial. It is likely that minimization of dipole-dipole
repulsions plays an important role. The situation is somewhat
muddied by disagreements over the effective size of a trifluo-
romethyl group, however. Trifluoromethyl has been estimated
to be the size of an isopropyl group,37 which causes the
production of only 70% of the 2R*,3R* diastereomer, although
it has also been estimated to be somewhat larger.38 Nonetheless,
it is difficult to ascribe the entire stereochemical influence of
the â-CF3 group to steric effects. Even though only 75% of
the 2R*,3R* diastereomer is produced by H/D exchange of 2j,
the effect of a â-cyano group, which is very small (roughly the
size of a fluoro substituent), must be ascribed to an electronic
effect.
Rotation about the C-2/C-3 σ-bond of the enolate anion
should be fast on the reaction time scale so that ready
interconversion of the conformers at the stereogenic center can
take place. Chiang et al. have determined that the first-order
rate constant for protonation of the enolate anion of acetaldehyde
with water is 8.8 × 102 s-1, and it is 6.6 s-1 for the analogous
reaction of isobutyraldehyde.40 In neither case is protonation
of the enolate fast enough to compete with single-bond rotation.
The low rotation barriers obtained from our ab initio calculations
bear this out. Apeloig et al. have also reported ab initio
calculations that support low rotation barriers for the adjacent
σ-bonds in carbanions stabilized by the CtN and NO2 groups.41
Having the conformation at C-3 relative to the π-system as the
important stereochemical determinant in the electrophilic attack
is quite consistent with the Curtin-Hammett principle, since
we are not talking about the enolate anion itself but instead the
transition states for its protonation.
Conclusion
Our experimental results clearly indicate that both stereo-
electronic and steric effects are important factors in determining
the diastereoselectivity of enolate anion protonation. However,
stereoelectronic effects are more dramatic under our conditions
where supramolecular aggregates play no role. The stereo-
chemistry of conjugate nucleophilic addition is also controlled
by the enolate protonation step. With our substituted ethyl
butanoate substrates, the most stable transition states have the
electronegative group or very bulky group at C-3 occupying an
antiperiplanar position to the C-D bond that is forming, with
the proton gauche to the π-system (10).
The final point that must be considered is the geometry of
the enolate anion. Although we have no direct experimental
evidence on the point, with esters 2a-m it is not unlikely that
we are dealing with kinetically-formed ester enolates, the trans
isomers (Figure 4), which have been shown in a number of
studies to form from esters using LDA/THF.5 If the trans isomer
is also formed faster under our conditions, one would expect
that its interconversion to the cis isomer will not compete with
the fast deuteration step.
Fortunately, ab initio calculations on the protonation of the
enolate anion of 3-fluorobutanoic acid indicate that the dia-
stereoselectivities which we observe could be obtained from
either cis or trans enolate isomers.17 In addition, Fleming and
McGarvey have argued that the stereoselection of enolate
protonation should be relatively unaffected by whether the anion
has the cis or trans geometry.8,10,39 This is also consistent with
the pattern of diastereoselectivities found in Table 2. In
Experimental Section
General Procedures. C6D6 and C6H6 were used for NMR spectra;
2
multiple H NMR integrations were performed for determination of
reaction diastereoselectivities. Preparatory GC used an 8 ft × 3/8 in.
Carbowax 20 M column. Melting points were calibrated against a
benzoic acid standard. For H/D exchanges, glassware was soaked in
NaHCO3 solution and then rinsed with H2O; all glassware was oven-
dried. Sodium ethoxide (NaOEt) and potassium ethoxide (KOEt)
solutions in ethanol-d (EtOD) were prepared from Na and K metal,
respectively, and stored under N2. Tetramethylammonium deuteroxide
(Me4NOD, e0.2 atom % H) was made by two exchanges in D2O of
Me4NOH‚5H2O (Aldrich, 99%). Ethyl (E)-2-butenoate (8) was sepa-
rated from the (Z)-isomer by careful fractional distillation; 8 of 99.9+%
purity was used in the conjugate addition experiments. Benzene was
distilled from Na and THF and dimethyl-2-oxohexahydropyrimidine
(DMPU) from CaH2 before use. The silica gel used in chromatography
was Merck Kieselgel 60, 70-230 mesh. EtOD (99+ atom % D),
[1-13C]ethanol (98 atom % 13C), and all other reagents except where
noted were purchased from Aldrich and used without further purifica-
(36) Kawasaki, H.; Tomioka, K.; Koga, K. Tetrahedron Lett. 1985, 26,
3031-3034.
(37) Bott, G.; Field, L. D.; Sternhell, S. J. Am. Chem. Soc. 1980, 102,
5618-5626.
(38) (a) Charton, M. J. Am. Chem. Soc. 1969, 91, 615-618. (b) Della,
E. W. J. Am. Chem. Soc. 1967, 89, 5221-5224.
(39) Fleming, I.; Lewis, J. J. J. Chem. Soc., Perkin Trans. 1 1992, 3257-
3266.
(40) (a) Chiang, Y.; Hojatti, M.; Keeffe, J. R.; Kresge, A. J.; Schepp, N.
P.; Wirz, J. J. Am. Chem. Soc. 1987, 109, 4000-4009. (b) Chiang, Y.;
Kresge, A. J.; Walsh, P. A. J. Am. Chem. Soc. 1986, 108, 6314-6320.
(41) Apeloig, Y.; Karni, M.; Rappoport, Z. J. Am. Chem. Soc. 1983,
105, 2784-2793.

DiastereoselectiVity of Enolate Anion Protonation
J. Am. Chem. Soc., Vol. 119, No. 3, 1997 485
tion. Ethyl 2-butynoate (Farchan, 99%), 3-nitrophenol (Eastman
Kodak), and isobutylene (Matheson) were also used without purification.
Elemental analyses were by Galbraith Labs, Inc., Knoxville, TN.
Esters 2a and 2f were prepared using literature procedures.13 The
triester 2h42 (bp 125 °C at 0.9 Torr) was synthesized in ethanol by
ethoxide-catalyzed conjugate addition of diethyl malonate to 8, and
ethyl 3-cyanobutanoate (2j)43 by the addition of HCN. Ethyl 3-phen-
ylbutanoate (2i)44 was prepared by the reaction of 3-phenylbutanoic
acid, thionyl chloride and ethanol, and ethyl 3-methylpentanoate (2m)45
by H2SO4-catalyzed azeotropic esterification of 3-methylpentanoic acid
in ethanol and toluene. Ethyl 3,5,5-trimethylhexanoate (2l)46 was made
by oxidation of 3,5,5-trimethylhexanal with CrO3/H2SO4/AcOH at 0
°C, followed by esterification with thionyl chloride and ethanol. 2,2-
Dimethyl-5-ethoxy-3-hexanone (6)24 was synthesized by the aldol
condensation of pinacolone and ethanal (LDA/THF at -78 °C),
followed by dehydration with p-toluenesulfonic acid (50 °C, 3 h) and
ethoxide-catalyzed addition of EtOH.
and a 20:1 hexane/Et2O eluent. After solvent removal, 3.9 g (86%) of
2b was recovered (99% purity): 1H NMR (CDCl3) δ 1.23 (t, 3H), 1.35
(d, 3H), 2.46-2.85 (2 dd, 2H), 4.13 (q, 2H), 4.82 (m, 1H), 6.90-7.31
(m, 5H); MS m/z 208 (M+), 163, 94.
Ethyl 3-(3-Nitrophenoxy)butanoate. A 9.05 g sample of 3-nitro-
phenol (64.4 mmol) was dissolved in a solution of 34 mL of DMPU
and 16 mL of THF and the solution slowly added to a stirred slurry of
0.80 g of NaH (32 mmol) in 20 mL of DMPU at 0 °C. After 1 h, 3.6
g of ethyl 2-butynoate (31.5 mmol) was added. About 70% reaction
had occurred after 23 h at room temperature, and it reached completion
after further heating at 40 °C for 22 h. Addition of 300 mL of H2O
produced a solid, which was filtered, dissolved in 80 mL of Et2O, and
washed with 5 × 20 mL of 4% Na2CO3 solution. After evaporation
of the solvent, the product was chromatographed on 75 g of silica gel
(Et2O) to produce a mixture containing 3% of the (Z)- and 97% of the
(E)-alkene. Recrystallization from hexane gave 4.95 g (63%) of fluffy
white crystals of ethyl (E)-3-(3-nitrophenoxy)-2-butenoate: mp 69.5-
1
General Method for H/D Exchanges. Using syringe or glovebag
techniques, exchanges were done under N2 with 0.05 M NaOEt (0.01
M for 2e, 2h, and 2i) and 0.5 M ester substrate. To 4.0 mL of stirred
solvent was added 2 mmol of the ester and the reaction initiated with
0.3 mL of 0.8 M base. When the reaction had proceeded to 2-15%
71 °C; H NMR (CDCl3) δ 1.20 (t, 3H), 2.49 (s, 3H), 4.09 (q, 2H),
4.85 (s, 1H), 7.36 (d, 1H), 7.56 (t, 1H), 7.89 (m, 1H), 8.08 (d, 1H);
MS m/z 251 (M+), 206, 178, 160.
Hydrogenation of ethyl (E)-3-(3-nitrophenoxy)-2-butenoate with Rh-
(PPh3)3Cl in benzene was carried out under the usual conditions for 10
h. After workup, GC analysis indicated two products, in addition to
unreacted starting material. After acid extraction to separate the
3-aminophenoxy products, chromatography on 100 g of silica gel using
a 20:1 hexane/Et2O eluent gave 1.35 g (27%) of a colorless liquid,
ethyl 3-(3-nitrophenoxy)butanoate (99+% purity): 1H (CDCl3) δ 1.24
(t, 3H), 1.32 (d, 3H), 2.56-2.90 (2 dd, 2H), 4.18 (q, 2H), 4.96 (m,
1H), 7.2-7.8 (m, 4H); MS m/z 253 (M+), 208, 115.
Ethyl 3-(4-Aminophenoxy)butanoate. Using 62 mL of a 3:1
mixture of DMPU/THF with 0.60 M ethyl 2-butynoate, 0.62 M sodium
4-nitrophenoxide, and 0.61 M 4-nitrophenol, the conjugate addition was
run at 65 °C for 7 d. Addition of 300 mL of H2O gave a crystalline
product. After being washed with 4 × 10 mL of 5% Na2CO3, it was
chromatographed on 40 g of silica gel using a 3:1 Et2O/hexane eluent
and recrystallized from hexane to yield 7.32 g (73%) of ethyl (E)-3-
(4-nitrophenoxy)-2-butenoate as yellowish needles: mp 86-87 °C; 1H
NMR (CDCl3) δ 1.20 (t, 3H), 2.47 (s, 3H), 4.10 (q, 2H), 4.99 (s, 1H),
7.15 (d, 2H), 8.25 (d, 2H); NOE’s between the vinyl and allyl protons
and the vinyl and C-2 aromatic protons were consistent with those
calculated for the (E)-isomer; MS m/z 251 (M+), 206, 178, 160.
Hydrogenation under the usual conditions for 7 d and workup using
pentane to precipitate the Rh(PPh3)3Cl catalyst was followed by three
extractions with 1.0 M HCl. Neutralization of the acidic water layer
with 1.0 M NaOH produced 1.9 g (47%) of a red liquid, ethyl 3-(4-
aminophenoxy)butanoate: 1H NMR (CDCl3) δ 1.20 (t, 3H), 1.27 (d,
3H), 2.38-2.77 (2 dd, 2H), 3.4 (broad s, 2H), 4.10 (q, 2H), 4.58 (m,
1H), 6.53-6.76 (dd, 4H); MS m/z 223 (M+), 178, 109.
1
exchange of one proton as determined by H NMR, it was quenched
with 0.06 mL of 2 M D2SO4 which brought the pH to 5-6. After the
reaction was quenched, the reaction mixture was either extracted with
hexane/brine or chromatographed on 30 g of silica gel (Et2O). Upon
drying and evaporation of the solvent, the typical recovery of the ester
was over 80%. Pure samples of products for NMR analyses were
recovered by preparatory GC. Exchange reactions were carried out
for 1-15 min at room temperature except for 2c (30-90 m), 2d (∼1
h), 2i (2-5 h), 2j (1-3 s), 2k and 2m (10-24 h), 2l (24-120 h), 5 (0
°C, 20-80 s), and 6 (10-60 s). With 2b and its 3-nitro and 4-amino
analogs, 0.16-0.25 M substrate was used, and elimination competed
with H/D exchange. With 2h the proton at C-2 was pre-exchanged
with EtOD before the actual H/D exchange was carried out.
Synthesis and D/H Exchange of Ethyl 3-Ethoxy[2,2-2H2]butanoate
(2a-d2) in [1-13C]EtOH. Deuteration of 2a was carried out using two
exchanges of 2 d each at room temperate in a 0.5 M solution of KOEt
1
in EtOD under N2; H NMR integrations showed only 0.5% proton
content at C-2 and 13% deuterium (1H) at C-4. The D/H exchange of
1.0 mmol 2a-d2 was done for 60 h with 0.04 M KOEt in 1.8 mL of
ethanol, one-third of which was [13C]EtOH. After the usual workup,
1H NMR showed about 97% loss of deuterium at C-2. The 13C NMR
peak for C-1 of the 3-ethoxy group was enhanced only 0.3% above
natural isotopic abundance, whereas the ester ethyl group was com-
pletely exchanged.
Ethyl 3-Phenoxybutanoate (2b). Ethyl 2-butynoate (3.37 g, 0.94
M in 25 mL of DMPU/7 mL of THF) was reacted in the presence of
1.0 M NaOPh and 0.94 M phenol for 24 h at room temperature.
Addition of 25 mL of H2O, extraction with 4 × 20 mL of hexane,
followed by back-extraction with 35 mL of H2O, 30 mL of 1.3 M
NaOH, and 25 mL of H2O, drying over K2CO3, evaporation of the
solvent, and vacuum distillation (bp 106-108 °C at 0.04 Torr) produced
5.0 g (80%) of ethyl (E)-3-phenoxy-2-butenoate.47
A 4.5 g sample of ethyl (E)-3-phenoxy-2-butenoate (22 mmol), along
with 0.86 g of tris(triphenylphosphine)rhodium(I) chloride (0.97 mmol),
was dissolved in 70 mL of benzene which had been deoxygenated with
N2 in a pressure reactor. After the reactor was sealed and flushed twice
with 15 atm of H2, it was charged with 30 atm of H2. After 4 d at 46
°C, the reaction had proceeded to 99% completion. The mixture was
evaporated to remove as much benzene as possible, leaving a brown
oily residue. The catalyst was removed by addition of 50 mL of hexane,
vacuum filtration, and flash chromatography using 20 g of silica gel
Ethyl 3,4,4-Trimethylpentanoate (2c).27 Reaction of 3,3-dimethyl-
2-butanone and the conjugate base of triethyl phosphonoacetate in 1,2-
dimethoxyethane gave a 97:3 E/Z ratio of ethyl 3,4,4-trimethyl-2-
pentenoate.26 The alkene was hydrogenated over PtO2 in EtOH at 3.4
atm, and after solvent evaporation 2c was recovered in 81% overall
yield (bp 80 °C at 5 Torr). Configurational assignments were made
1
2
by H and H NMR analysis of the reduction product of ethyl (E)-
3,4,4-trimethyl-2-pentenoate with D2/Rh(PPh3)3Cl in benzene (27 atm,
55 °C, 10 d), a known syn addition,25b followed by exhaustive D/H
exchange of 2c-d2 (0.5 M) in EtOH (0.27 M NaOEt, 90 °C, 5 h) to
assign the R-H chemical shifts. The 2H NMR spectrum of the
(2R*,3S*)-[2H2] diastereomer of 2c-d2 had peaks at δ 2.32 and 1.82,
1
and the H spectrum was missing the usual HC-2 peak at δ 2.32 and
HC-3 peak at δ 1.82 but not the HC-2 peak at δ 1.92. Thus, the 2R*,3S*
diastereomer has the C-2 deuteron downfield.
tert-Butyl 3-tert-Butoxybutanoate. Reduction of tert-butyl aceto-
acetate with a 20% excess of NaBH4 in 95% EtOH (2.0 M) for 21 h,
followed by neutralization with AcOH, solvent evaporation, brine/Et2O
extractions, and vacuum distillation (bp 55 °C, 0.5 Torr), gave a 67%
yield of tert-butyl 3-hydroxybutanoate. Anal. Calcd for C8H16O3: C,
59.97; H, 10.07. Found: C, 59.78; H, 10.04.
Isobutylene gas was bubbled for 4 h through a frit into a solution of
16.0 g of tert-butyl 3-hydroxybutanoate in 0.2 L of CH2Cl2 containing
0.9 mL of H2SO4 and the solution allowed to sit overnight; GC analysis
(42) Michael, A. Ber. Dtsch. Chem. Ges. 1900, 33, 3748.
(43) Elderfield, R. C.; Claflin, E. F.; Mertel, H. E.; McCurdy, O. L.;
Mitch, R. T.; VerNooy, C. D.; Wark, B. H.; Wempen, I. M. J. Am. Chem.
Soc. 1955, 77, 4819-4822.
(44) Rupe, H.; van Walraven, F. HelV. Chim. Acta 1930, 13, 361-377.
(45) Croxall, W. J.; Van Hook, J. O. J. Am. Chem. Soc. 1950, 72, 803-
808.
(46) Turner, E. E.; Turner, L. J. Chem. Soc. 1951, 2543-2545.
(47) Chalchat, J.-C.; Theron, F.; Vessiere, R. Ann. Chim. 1971, 269-
274.

486 J. Am. Chem. Soc., Vol. 119, No. 3, 1997
Mohrig et al.
showed 87% conversion to product. Extraction with 3 × 100 mL of
10% NaHCO3 solution, drying, solvent evaporation, and flash chro-
matography on silica gel using a 4:1 hexane/Et2O eluent gave 5.1 g
(24%) of tert-butyl 3-tert-butoxybutanoate (bp 42 °C at 0.01 Torr):
1H NMR δ 1.11 (s, 9H), 1.17 (d, 3H), 1.39 (s, 9H), 2.21-2.31 (dd,
1H), 2.46-2.56 (dd, 1H), 4.07 (m, 1H). Anal. Calcd for C12H24O3:
C, 66.63; H, 11.18. Found: C, 66.59; H, 11.30.
Ethyl 3-Methyl-4,4,4-trifluorobutanoate (2g) and Ethyl 3,4-
Dimethylpentanoate (2k). Reaction of triethyl phosphonoacetate with
NaH in 1,2-dimethoxyethane and subsequent addition of 1,1,1-
trifluoroacetone at 0 °C, refluxing overnight, neutralization with AcOH,
Et2O/H2O extractions, drying, and distillation produced 35% of ethyl
3-methyl-4,4,4-trifluoro-2-butenoate, largely the (E)-isomer.28 Hydro-
genation in Et2O at room temperature over PtO2 using 3.4 atm of H2,
followed by removal of the catalyst, evaporation of the solvent and
preparatory GC, gave pure 2g.51 Configurational assignments for the
deuterated 2g which was produced by H/D exchange were made by
reducing ethyl (E)-3-methyl-4,4,4-trifluoro-2-butenoate with D2 and
PtO2/C6H12; the major 1H peak was at δ 2.2, the expected position for
the 2R*,3S* diastereomer.
The same synthetic procedure, but with 3-methyl-2-butanone instead
of 1,1,1-trifluoroacetone, gave 2k52 (99% purity, 40% yield).
Addition of Diethyl [2,2-2H2]propanedioate to 8. The reaction
was done in EtOD to 3-22% completion with 0.15 M 8, 0.18 M diethyl
malonate-d2, and 0.03 M NaOEt at room temperature, giving over the
entire range 79 ( 1% of the 2R*,3R* diastereomer of diethyl
2-(ethoxycarbonyl)-3-methyl[2,4-2H2]pentanedioate.
3-Ethoxybutanal (5). Reaction of 2-butenal and NaOEt (0.05 M)
in ethanol at 0 °C for 5 min gave 7.5 g of 5 (53%). Purification by
vacuum distillation gave 553 of 96% purity: 1H NMR δ 0.85 (d, 3H),
1.00 (t, 3H), 1.91 (2 dd, 1H), 2.18 (2 dd, 1H), 3.0-3.3 (m, 2H), 3.5
(m, 1H), 9.4 (dd, 1H).
Oxidation of 5, which had been exchanged in EtOD (2H NMR
showed an 85/15 ratio of peaks at δ 1.9 and 2.2, respectively), was
done in C6H6 over a 10 d period by bubbling O2 into the solution every
2 d. Azeotropic esterification of the 3-ethoxybutanoic acid product
with EtOH/H2SO4/toluene gave ethyl 3-ethoxy[2H1]butanoate, with 2H
NMR peaks at δ 2.13 and 2.44 in an 85/15 ratio. Thus, the 2R*,3R*
diastereomer of 5 has the C-2 deuteron upfield.
The tert-butyl groups were removed in 4.0 mL of CH2Cl2 by
bubbling N2 for 6 h through a solution of 18 µL of H2SO4 and 0.32 g
of tert-butyl 2-deuterio-3-tert-butoxybutanoate, which had been pro-
2
duced by H/D exchange in EtOD/NaOEt (15% D at C-2; H peaks at
δ 2.20 and 2.42 were in a 90:10 ratio, respectively). The product was
extracted into 1 mL of H2O. Its 2H NMR spectrum gave a 90:10 ratio
of two peaks, the larger at δ 2.45 and the smaller at 2.56, corresponding
to a 90:10 mixture of (2R*,3R*/2R*,3S*)-3-hydroxy[2-2H1]butanoic
acids.25a
Ethyl 3-tert-Butoxybutanoate (2d). Reaction of isobutylene/H2-
SO4 with ethyl 3-hydroxybutanoate in CH2Cl2 and the usual workup
gave 2d in 32% yield (99.9% purity, bp 32 °C at 0.01 Torr): 1H NMR
δ 0.95 (t, 3H), 1.07 (s, 9H), 1.13 (d, 3H), 2.22-2.33 (dd, 1H), 2.47-
2.57 (dd, 1H), 3.95 (t, 2H), 4.04 (m, 1H).
Addition of 2-Methyl-2-propanethiol-d to Ethyl (E)- and (Z)-2-
Butenoate. To 150 mL of EtOH was added 5.0 g (0.22 mol) of Na
metal, followed by 21.4 g (0.24 mol) of 1,1-dimethylethanethiol, and
the solution was stirred for 30 min. The white sodium thiolate salt
was vacuum filtered under N2, washed with anhydrous Et2O, and dried
in a vacuum oven over P2O5 at 100 °C and 47 Torr for 12 h. A 2.4 g
(21.5 mmol) sample of the thiolate salt was dissolved in 48 mL of
EtOD, and 3.77 mL of D2SO4 in EtOD (1.43 M, 5.38 mmol) was added.
The Na2SO4 precipitate was removed by centrifugation. The conjugate
addition reactions of Me3CSD to 8 and to ethyl (Z)-2-butenoate (99%
(Z), made by addition of H2 to ethyl 2-butynoate on Pd/BaSO4)48 were
carried out with 0.19 M Me3CSD and Me3CSNa and 0.186 M alkene
at room temperature for 1-20 min; no differences in stereochemistry
were observed over 6-89% completion. At 20 min the (Z)-2-butenoate
had rearranged to a Z/E ratio of 95/5. The reactions were neutralized
to pH 7 with D2SO4/EtOD, the Na2SO4 was removed by centrifugation,
hexane and Et2O were added, and the solution was extracted twice with
H2O, dried, and evaporated. Pure products were recovered by prepara-
tory GC.
Conjugate additions of EtOD to 2-butenal were done with 0.5 M
substrate and 0.05 M NaOEt at -15 °C for 30-60 s (3-10%
conversion). 2H NMR samples were purified by preparatory GC.
3-Ethoxybutanenitrile (7). Reaction of 15 g of crotononitrile in
95 mL of 0.36 M NaOEt in EtOH for 1 h at room temperature gave
15.5 g (62%) of 7 (bp 72 °C at 40 Torr):54 1H NMR (DMSO-d6) δ
1.09 (t, 3H), 1.16 (d, 3H), 2.63 (dd, 1H), 2.73 (dd, 1H), 3.45 (q, 2H),
3.67 (m, 1H).
Ethyl 3-(tert-Butylthio)butanoate (2e).16 To 78 mL of 1 M NaOEt
in EtOH were added 14.2 g (0.16 mol) of 1,1-dimethylethanethiol and
8.9 g (78 mmol) of 8. After being stirred for 7 h at room temperature,
the reaction mixture was chromatographed on silica gel using hexane
and then Et2O as eluents. Vacuum distillation at 65 °C and 0.25 Torr
gave 10.2 g of 2e (59%), of 99+% purity: 1H NMR (CDCl3) δ 1.28
(t, 3H), 1.35 (s, 9H), 1.37 (d, 3H), 2.45 (dd, 1H), 2.60 (dd, 1H), 3.19
(m, 1H), 4.15 (q, 2H).
Ethyl 3-(tert-butylthio)-2-butenoate was synthesized by reacting 1,1-
dimethylethanethiol (49 mmol) with ethyl 2-butynoate (45 mmol) in
77 mL of ethanol in the presence of NaOEt (2.2 mmol) for 2 h.49,50
The (E)- and (Z)-isomers were separated on a silica gel column using
2-5% Et2O in hexane. The (Z)-isomer gave a significant NOE
enhancement between the allyl and vinyl protons, whereas the (E)-
isomer gave no enhancement.
Acknowledgment. We are grateful for support from the
National Science Foundation (NSF Grant CHE-8505408), the
National Institutes of Health (NIH Grant GM40018), the Camille
and Henry Dreyfus Foundation, Inc. for a fellowship to R.E.R.
and for a student stipend, the NSF (Grant CHE-8409822) and
the 3M Foundation for funding the purchase of a 200 MHz
NMR spectrometer, and the Hewlett-Packard Co. for the gift
of a capillary GC. Acknowledgment is made to the donors of
the Petroleum Research Fund, administered by the American
Chemical Society, for partial support of this research. Student
stipends were also provided by grants from the Howard Hughes
Medical Institute and the 3M Foundation. We thank M. Mitton
for the NMR-structure correlation for aldehyde 5 and Dr.
Stephen B. Philson and the University of Minnesota for 46 and
Ethyl (E)-3-(tert-butylthio)-2-butenoate was deuterogenated in deoxy-
genated benzene with D2 (11 atm)/Rh(PPh3)3Cl at 40 °C for 4 d.
Workup in the usual way gave a product that was 52% reduced.
Preparatory GC gave pure ethyl (2R*,3S*)-3-(tert-butylthio)[2,3-2H2]-
butanoate; 2H NMR (CHCl3) δ 2.58 and 3.16. Thus, the 2R*,3S*
diastereomer has the C-2 deuteron downfield.
2
76 MHz H NMR spectra.
JA962631S
(48) Taschner, M. J.; Rosen, T.; Heathcock, C. H. In Organic Syntheses;
Kende, A. S., Ed.; Wiley, 1986; Vol. 64, pp 108-113.
(49) Truce, W. E.; Tichenor, G. J. J. Org. Chem. 1972, 37, 2391-2396.
(50) Chalchat, J.-C.; Theron, F. Bull. Soc. Chim. France 1973, 3361-
3366.
(51) Weinges, K.; Kromm, E. Liebigs Ann. Chem. 1985, 90-102.
(52) Levene, P. A.; Marker, R. E. J. Biol. Chem. 1935, 111, 299-312.
(53) Boekelheide, V.; Fritz, H.; Ross, J. M.; Kaempfen, H. X. Tetrahe-
dron 1964, 20, 33-41.
(54) Bruylants, P. Bull. Soc. Chim. Belg. 1922, 31, 225-230.