Computer-aided reaction design. Development of a new facile procedure to synthesize 2-mercapto-3-alkoxycarboxylate on the basis of ab initio molecular orbital calculations

Article abstract of DOI:10.1021/jo990055n

This paper describes a new facial procedure to substitute a tosyloxy group in 2-(tosyloxy)alkanoate with SH^- to yield 2-mercaptoalkanoate on the basis of ab initio MO calculations. Combination of substrate and solvent effects can control both reactivity and selectivity of reaction for 2- (tosyloxy)3-alkoxycarboxylic acid which gave 2-mercapto-3-alkoxycarboxylic acid in good yield while its ethyl ester gave α,.β-unsaturated carboxylate ester as a main product. The difference of carboxylate moiety in the substrate causes remarkable change in reactivity and selectivity. To clarify origin of the difference, ab initio MO calculations in the gas phase and in DMF have been carried out. The solvent effect was considered at RHF/6-31+G* with the IPCM-SCRF model. It was confirmed that the substrate with an ester fragment prefers the E1cB to the S(N)2 mechanism. In the transition state of the S(N)2 mechanism with a carboxylate ion fragment, the nucleophile SH^- locates far from the reaction center due to the electrostatic repulsion between the COO^- fragment and SH^-. This repulsion causes high activation barrier in the gas phase while polar solvent can reduce the barrier height. Therefore, reaction conditions can control reactivity of carboxylic acid. On the basis of analysis of the MO calculations, subsequent experiments were designed for a new dianion system to synthesize 2-pyrimidinylthio carboxylic acid from 2-tosyloxy carboxylate. We succeeded in developing a new facile method that the two reactions for thioether carried out in a one-pot procedure in excellent yield.

Full text of DOI:10.1021/jo990055n

4768
J . Org. Chem. 1999, 64, 4768-4774
Com p u ter -Aid ed Rea ction Design . Develop m en t of a New F a cile
P r oced u r e to Syn th esize 2-Mer ca p to-3-a lk oxyca r boxyla te on th e
Ba sis of a b In itio Molecu la r Or bita l Ca lcu la tion s
Shohei Fukuda,1a,b Yuji Akiyoshi,1b and Kenzi Hori*,1a
Department of Chemistry and Chemical Engineering, Faculty of Engineering, Yamaguchi University,
Tokiwadai, Ube 755, J apan and Agrochemical Research Department, Ube Laboratory, Ube Industries,
Ltd.1978-5 Kogushi, Ube 755, J apan
Received J anuary 12, 1999
This paper describes a new facial procedure to substitute a tosyloxy group in 2-(tosyloxy)alkanoate
with SH- to yield 2-mercaptoalkanoate on the basis of ab initio MO calculations. Combination of
substrate and solvent effects can control both reactivity and selectivity of reaction for 2-(tosyloxy)-
3-alkoxycarboxylic acid which gave 2-mercapto-3-alkoxycarboxylic acid in good yield while its ethyl
ester gave R,â-unsaturated carboxylate ester as a main product. The difference of carboxylate moiety
in the substrate causes remarkable change in reactivity and selectivity. To clarify origin of the
difference, ab initio MO calculations in the gas phase and in DMF have been carried out. The
solvent effect was considered at RHF/6-31+G* with the IPCM-SCRF model. It was confirmed that
the substrate with an ester fragment prefers the E1cB to the SN2 mechanism. In the transition
state of the SN2 mechanism with a carboxylate ion fragment, the nucleophile SH- locates far from
the reaction center due to the electrostatic repulsion between the COO- fragment and SH-. This
repulsion causes high activation barrier in the gas phase while polar solvent can reduce the barrier
height. Therefore, reaction conditions can control reactivity of carboxylic acid. On the basis of
analysis of the MO calculations, subsequent experiments were designed for a new dianion system
to synthesize 2-pyrimidinylthio carboxylic acid from 2-tosyloxy carboxylate. We succeeded in
developing a new facile method that the two reactions for thioether carried out in a one-pot procedure
in excellent yield.
Sch em e 1
In tr od u ction
2-Mercapto carboxylate derivatives are frequently used
as intermediates for synthesizing many pesticides, medi-
cines and heterocyclic compounds.2 For example, Akiyoshi
et al. reported synthesis of 2-(4,6-dimethoxypyrimidin-
2-ylthio)-3-methoxy-3-methylbutanoic acid derivatives
4a -c3 via ester intermediate shown in Scheme 1. These
derivatives inhibit acetolactate synthase (ALS)4 in
branched chain amino acids biosynthesis as herbicides
and also have high herbicidal activities to various weeds.
In synthesizing the 2-mercapto carboxylate derivatives,
2-tosyloxy carboxylate ester 1 has been frequently used
as a precursor of 2-mercapto carboxylate 2.5 Although
several reagents such as sodium hydrosulfide hydrate,
thiourea, and dithiocarboxylate have been applied to
substitute the tosyloxy group of 1a with the SH one, no
corresponding ethyl 3-alkoxy-2-mercapto carboxylate 2a
was obtained. In many cases, R,â-unsaturated carboxyl-
ate ester 5 was obtained as one of the major products
shown in Scheme 2. That is, â-elimination mainly pro-
ceeds to form 5 when 1a was used as the substrate.
Fortunately, we found an alternative new and more
general route that used 2-tosyloxy carboxylic acid 1b for
substitution reaction with SH-.6 This reaction produced
2-mercapto carboxylic acid 2b in excellent yield more
than 85%. As the reaction of 1b with SH- yielded the
* To whom corresponding should be addressed.
(1) (a) Yamaguchi University. (b) Ube Laboratory. Ube Industries,
Ltd.
(2) (a) Hauptmann, S. Tetrahedron Lett. 1968, 9, 1317. (b) Kay, T.
T.; Punja, N. J . Chem. Soc. C 1970, 2409. (c) Santilli, A. A.; Kim, O.
H.; Wanser, S. V. J . Heterocycl. 1971, 8, 445. (d) Liebscher, J .;
Hartmann, H. Synthesis, 1979, 241. (e) Ishizaki, M.; Osada, S.; Kato,
S. J apan Kokai Koho 90172986, 1990.
(3) (a) Harada, K.; Abe, K.; Akiyoshi, Y.; Shiraishi, H.; Hayama, T.;
Yamomoto, K. J apan Kokai Koho 93125058, 1993. (b) Harada, K.; Abe,
K., Akiyoshi, Y.; Matsushita A.; Shiraishi I. J apan Kokai Koho
93148242, 1993. (c) Abe, K.; Akiyoshi, Y.; Shiraishi, H.; Shiraishi, I.;
Hayama, T. J apan Kokai Koho 93306274, 1993. (d) Abe, K.; Akiyoshi,
Y.; Shiraishi, H.; Shiraishi, I.; Hayama, T. Japan Kokai Koho 94009620,
1994. (e) Abe, K.; Akiyoshi, Y.; Shiraishi, H.; Koima, M.; Hayama, T.
J apan Kokai Koho 94016660, 1994. (f) Abe, K.; Akiyoshi, Y.; Shiraishi,
H.; Shiraishi, I.; Hayama, T.; Kuwata, T. J apan Kokai Koho 94025189,
1994. (g) Abe, K.; Akiyoshi, Y.; Shiraishi, H.; Shiraishi, I.; Hayama,
T. J apan Kokai Koho 94041091, 1994.
(5) (a) Oehrlein, H.; J eschke, R.; Ernst, B.; Bellus, D. Tetrahedron
Lett. 1989, 30, 3517. (b) Burkard U.; Effenberger F. Chem. Ber. 1986,
119, 1594.
(6) Abe, K.; Akiyoshi, Y.; Kojima, M.; Harada, K.; Matsushita, A.
J apan Kokai Koho 93148245.
(4) (a) LaRossa, R. A.; Falco S. C. Trends Biotechnol. 1984, 2, 158.
(b) Ray T. B. Plant Physiol. 1984, 75, 827. (c) Chaleff, R. S., Mauvis,
C. J . Science 1984, 224, 1443.
10.1021/jo990055n CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/06/1999

Computer-Aided Reaction Design
J . Org. Chem., Vol. 64, No. 13, 1999 4769
Sch em e 2
model molecules 6 as shown below for simplicity of
calculations. In the model molecules, a tosyl group in 1
was replaced with a mesyl group and the C(CH3)2(OCH3)
fragment with a methyl group. Methyl ester in 6a was
adopted instead of ethyl ester used in the experiments.
As experiments used large excess of sodium hydrosulfide
hydrate, the carboxyl group in 1b should release a proton
in the reaction conditions so that the reactant exists in
a carboxylate ion form.
required products, the key of the present method is to
use the carboxylic acid instead of the ester as a reactant.
The reactant 1 has a bulky substituents near the
reaction center. It is expected that steric hindrance
caused by the bulky group largely retarded substitution
reactions such as those of neopentyl bromide with sodium
alkoxides in protic solvents. In this case, the alkyl
neopentyl ethers are not major products because side
reactions such as neopentyl rearrangement, elimination,
and its decomposition proceed.7 However, there were
several successful examples of SN2 reaction of neopentyl
tosylate with the large steric hindrance. Stephenson
reported8 that the reaction of (S)-neopentyl-1-d tosylate
and nucleophiles proceeded through SN2 mechanism.
They proved that the inversion of the asymmetric envi-
ronment occurs in the experiment. Masada also obtained
results supporting the SN2 mechanism in the reaction of
neopentyl methanesulfonate and potassium 2,4,6-tri-tert-
butylphenoxide in 1,3-dimethyl-2-imidazolinone.9 There-
fore, the SN2 mechanism is applicable to substrates with
a bulky group in aprotic solvent, for example, DMSO,
DMF, DMI, and HMPA, when they have a leaving group
such as iodine, tosylate, mesylate, etc. Combination of
the good leaving group and aprotic solvent make the
reaction possible under the SN2 mechanism even though
the reactant has bulky substituents.
In the present experiment, we used 2-(tosyloxy)-3-
methoxy-3-methylbutyric acid and DMF. This is the
condition that the SN2 mechanism is applicable although
we used a different nucleophile, SH-. The kinetics data
of our preliminary experiments showed that the rate is
proportional to the product of 1b and SH-.10 Therefore,
we regarded the substitution in Scheme 1 as a reaction
in which SH- ion acts as nucleophile and replaces the
tosyloxy group of 1 through the SN2 mechanism.
In the present study, we performed ab initio molecular
orbital (MO) calculations at the MP2/6-31+G*//RHF/6-
31+G* level of theory in order to answer why the
carboxylic acid is suitable for controlling the reaction. 1
is too large to perform ab initio MO calculations with
large basis sets such as 6-31+G*.11 Therefore, we used
The reaction for 6b contains two negative charges, one
on carboxylate anion of 6b and the other on sulfide anion.
Hereafter, the SN2 reactions for the ester and the
carboxylate ion are called path A and B, respectively.
According to the results of MO calculations, we examined
a new type of the reaction which directly produces 4 with
R1 ) OH from 1. The reaction worked well in excellent
yield more than 90%.
Meth od of Ca lcu la tion
Ab initio MO calculations were performed by using the
Gaussian94 program.12 All the geometries were fully
optimized at the RHF/6-31+G* levels of theory, and the
energy relation among reactants, transition states (TSs),
and products were refined at the MP2/6-31+G*//RHF/6-
31+G* levels of theory. The vibration frequency calcula-
tion showed the obtained TS structures to have only one
imaginary frequency. Solvent effects were taken into
account for the optimized geometries by means of the
isodensity surface polarized continuum model (IPCM)
SCRF calculation13 with a dielectric constant of 36.7,
whose value corresponds to that of dimethylformamide
(DMF). The intrinsic reaction coordinates (IRCs)14 were
also calculated to analyze the mechanism in detail at the
RHF/6-31+G* level of theory. Table 1 summarizes total
and relative energies of the structures obtained.
Resu lts a n d Discu ssion
Exp er im en ta l Resu lts. The thiolation used sodium
hydrosulfide hydrate 3 by 5-hold equiv mol of reactants
(7) (a) Shaik, S. S.; Schlegel, H. B.; Wolfe, S. Theoretical aspects of
physical organic chemistry: the SN2 mechanism; Wiley: New York,
1992. (b) Fraser, G. M.; Hoffmann, H. M. R. J . Chem. Soc. (B) 1967,
425. (c) Davis, R.; Hudec, J . J . Chem. Soc., Perkin Trans. 2 1975, 1395.
(8) Stephenson, B.; Solladie, G.; S. Mosher, H. J . Am. Chem. Soc.
1972, 94, 4184.
(9) (a) Masada, H.; Tajima, K.; Taniguchi, K.; Yammamoto, T. Chem.
Lett. 1991, 753. (b) Masada, H.; Yammamoto, T.; Yamamoto, F. Nippon
Kagaku Kaishi 1995, 12, 1028.
(10) We have carried out kinetic experiments and found that the
present reaction proceeds in the second-order reaction which depends
on the first order with respect to the concentration of 1b and sodium
hydrosulfide hydrate.
(12) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T. A.; Petersson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Lantham, M. A.;
Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .;
Stefanov, B. B.; Nanayakara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andre’s, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; Defrees, D. J .; Baker, J .;
Stewart, J . J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian,
Inc.: Pittsburgh, PA, 1995.
(13) Foresman, J . B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J .;
Frisch, M. J . J . Phys. Chem. 1996, 100, 16098.
(11) Hehre, W. J .; Radom, L.; Schleyer, P.v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986; Chapter 4.
(14) (a) Fukui, K. Acc. Chem. Res. 1981, 14, 363. (b) Head-Gordon,
M.; Pople, J . A. J . Chem. Phys. 1988, 89, 5777.

4770 J . Org. Chem., Vol. 64, No. 13, 1999
Fukuda et al.
F igu r e 1. Optimized structures of transition states for paths A and B, and the intermediate 11a of the ElcB mechanism for 10.
Ta ble 1. Tota l En er gies (h a r tr ee) a n d Rela tive En er gies
(∆E, k ca l m ol-1) of Rea cta n ts, TSs, a n d P r od u cts of th e
SN2 Rea ction
linear in the TS, i.e., a typical TS geometry for SN2
reactions.
While the S(1)-C(2) and C(2)-O(3) lengths in 6b(TS)
turned out to be 3.110 and 2.276 Å, respectively. They
are much longer by 0.416 and 0.353 Å than the corre-
sponding lengths for 6a (TS). The two distances in 6b(TS)
are rather long in comparison with those of TSs for SN2
mechanisms concerning SH- ion as a nucleophile.15 The
S(1)-C(2)-O(3) angle of 6b(TS) was calculated to be
203.6°, which is larger by 30.9° than that of 6a (TS).
Therefore, the geometry of 6b(TS) largely deviates from
ordinal SN2 TS structures. The S(1)-O(6) distance was
calculated to be 4.255 Å which is longer by 0.949 Å than
the corresponding length of 6a (TS). The long S(1)-C(2)
distance and the wide S(1)-C(2)-O(3) angle are conve-
nient to avoid large electrostatic repulsion between the
two anion centers.
∆E
MP2/6-31+G*// (kcal
RHF/6-31-G* c mol-1
MP2/6-31+G*//
RHF/6-31+G*
(DMF)d
∆E
(kcal
mol-1
compound
)
)
6a + SH- a
6a (TS)
-1366.90011
-1366.87610
-1366.94865
-1327.18870
-1327.06955
-1327.23529
0
-1366.99506
-1366.96279
-1367.04208
-1327.38263
-1327.33362
-1327.43238
0
15.1
-30.5
0
74.8
-29.2
20.2
-29.5
0
30.8
-31.2
7a + 8a
6b + SH- b
6b(TS)
7b + 8b
a
Total energies for the geometries of the combined system
which were optimized by using the last geometry of IRC calcula-
b
tions. Relative energies (∆E) are estimated on the base of total
energies of reactants, 6a or 6b, and SH-. c Total energies of SH-,
6b, 7b, and 8 and SH- at MP2/6-31+G*//RHF/6-31+G* level of
theory are -398.22952, -928.95917, -664.69333, and -662.5419629
d
hartree, respectively. The MP2/6-31+G*//RHF/6-31+G* energies
of IPCM-SCRF calculations with ꢀ ) 36.7 in DMF for SH-, 6b,
7b,and 8 are -398.33268,-929.04995,-664.78855,and -662.64382
hartree, respectively.
It may not be true that 6b(TS) is a real TS for the SN2
reaction since its geometry largely deviates from the
ordinal TS structures for SN2 reactions. To confirm
whether 6(TS) connects 6 and 7, IRC calculations were
performed for both 6a (TS) and 6b(TS). Figure 2 displays
the geometry transformation together with potential
energy profiles along the IRCs. The C(2)-O(3) distance
at s ) -2.75 amu1/2 Bohr in path A is 1.500 Å, which is
similar to that in 6a (1.428 Å). The reaction almost
completes at s ) 5.15 amu1/2 Bohr because the S(1)-C(2)
distance is 1.883 Å, which is longer a little than that of
6a (1.834 Å). The potential energy goes up and down
sharply in both sides of the TS.
Ta ble 2. Geom er ica l P a r a m eter s of Rea cta n ts a n d TSs
for th e SN2 Rea ction
6a
6b
6a (TS) 6b(TS)
bond lengths S(1)-C(2)
-
-
2.694
1.428 1.464 1.923
3.110
2.276
3.839
4.255
3.334
3.778
203.6
(Å)
C(2)-O(3)
S(1)-C(4)
-
-
-
-
2.905
3.306
S(1)-O(6)
O(3)-O(6)
C(4)-C(5)
S(1)-C(2)-O(3)
2.638 2.710 2.902
4.218 3.713 4.054
bond angles
(deg)
-
-
172.7
In the geometry at s ) -7.90 amu1/2 Bohr in path B,
6b weakly interacts with SH- ion. Although the S(1)-
C(2) distance at s ) 7.90 amu1/2 Bohr is 1.993 Å, longer
slightly by 0.154 Å than that in 7b (1.839 Å), the product
thiol forms at this point. The potential energy goes up
and down gradually in both sides of the TS. Therefore,
6b(TS) is the real TS for the SN2 mechanism of 6b and
SH-. It should be noted that the IRC length for path B
is much longer than that for path A. This propensity well
correlates with the height of activation energies for each
path and the Hammond postulate,16 which expects that
and was performed in DMF, the polar aprotic solvent.
In path A, carboxylate ester 3 was added to 1a in DMF
and stirred for 30 min at 5-10 °C. TLC check indicated
that 1a disappeared and many spots appeared in the
reactant mixture. The reaction gave 5 in yield ca. 20%
as the main product, and no target compound 2a was
obtained. The amount of other compounds was too small
to isolate. For carboxylic acid in path B, 3 was added to
1b in DMF at room temperature and stirred for 3 h at
50-55 °C. The reaction gave 2b in excellent yield more
than 85% as the main product and no R,â-unsaturated
carboxylate ester was obtained.
Differ en ce of th e SN2 Mech a n ism betw een 6a a n d
6b. There are large difference between the carboxylate
ester and the carboxylic acid in the calculated TS
structures shown in Figure 1. Their geometrical pa-
rameters are summarized in Table 2. S(1)-C(2) and
C(2)-O(3) distances in 6a (TS) were calculated to be 2.694
and 1.923 Å. As the S(1)-C(2)-O(3) angle was calculated
to be 172.7°, S(1), C(2), and O(3) atoms align almost
(15) Several optimizations have been reported for TS geometries of
SN2 reactions using SH- ion. The distances between
S atom of
nucleophile and C atom of reaction center (dS-C) are shown below. (a)
dS-C; 2.537 Å in HS- - -CH2(CHdCH2)- - -Cl at the mp2/6-31++G**
level. Lee I.; Kim C. K.; Lee B-S. J . Comput. Chem. 1995, 16, 1045.
(b) dS-C; 2.515 Å in HS- - -CH2(CH2CN)- - -Cl at the mp2/6-31+G* level.
Chung D. S.; Kim C. K.; Lee B-S.; Lee I. J . Phys. Chem. A 1997, 101,
9097. (c) dS-C; 2.378 Å in HS- - -CH2(CH3)- - -SCH3 at the mp2/6-31+G*
level. Gronert, S.; Lee, J . M. J . Org. Chem. 1995, 60, 6731.
(16) (a) Hammond, G. S. J . Am. Chem. Soc. 1955, 77, 334. (b)
Streitwieser, A., J r. Chem. Rev. 1956, 56, 639.

Computer-Aided Reaction Design
J . Org. Chem., Vol. 64, No. 13, 1999 4771
F igu r e 2. Potential energy profiles and the geometry transformation along the IRC for 6a and 6b.
Ta ble 3. Ch a r ges Distr ibu tion for Molecu les Rela ted to th e SN2 Rea ction s. Va lu es in P a r en th eses Ar e Ch a n ges Rela tive
to Th ose of 6a or 6b
study. The Ea of 6a in DMF was calculated to be 20.2
kcal mol-1 which is larger by 5.1 kcal mol-1 than that in
a vacuum shown in Figure 3. On the other hand, that of
6b in DMF was calculated to be 30.8 kcal mol-1. The
solvent effect as dielectric field decreases the barrier by
as large as 44.0 kcal mol-1. Therefore, the polarity of
DMF reduces the electrostatic repulsion between the
COO- fragment and the SH-. Although this value is still
higher by 10.6 kcal mol-1 than that of 6a , the SN2
reaction with the two anions system is expected to
proceed in our experimental conditions.
The existence of two negative charges is the key for
understanding the reaction so that the solvent effects for
F igu r e 3. Energy-level diagrams (in kcal mol-1) of the
the charge distribution is analyzed. Table 3 summarizes
charge distribution of three fragments, i.e., SH-, the
reaction center, and the leaving group, of 6 and 6(TS) in
a vacuum and DMF. They are designated as Nu, RC, and
LG, respectively.
thiolation for 6a and 6b. At (a) the MP2/6-31+G*//RHF/6-
31+G* level of theory and (b) the MP2/6-31G*//RHF/6-31G*
level of theory including solvent effects of the IPCM-SCRF
method with DMF dielectric constant.
The charges of the MsO fragment (LG) of 6a in a
vacuum and in DMF were estimated to be -0.193 and
-0.206, respectively. As those for 6a (TS) are -0.564 and
-0.550, the difference of charges of between 6a and
6a (TS) are 0.371 and 0.344. The SH- fragment (Nu) loses
its charge by 0.355 in 6a (TS) and the MsO moiety gets
almost of this density. Although the Nu in 6a (TS, DMF)
loses its electron density by 0.472 which is larger than
that of 6a (TS), the MsO fragment in DMF accepts
electron density by 0.344 less than that of 6a (TS), and
a TS with a low activation energy locates an early point
on reaction coordinate.
Figure 3 displays the energy diagram for 6a and 6b
at the MP2/6-31+G*//RHF/6-31+G* level of theory. The
activation energies (Ea) in a vacuum were estimated to
be 15.1 and 74.8 kcal mol-1, respectively. 6b is higher
by 59.7 kcal mol-1 than 6a in term of the activation
energy. The electrostatic repulsion between the two anion
centers causes such a high activation barrier for 6b(TS).
Solven t Effects. The activation barrier in a vacuum
for path B is too high to expect that the SN2 reaction
proceeds under our experimental conditions. It is common
knowledge17 that polarity of solvents largely affects rates
of SN2 reactions. The IPCM-SCRF calculations were
adopted to estimate the solvent effect in the present
(17) (a) March, J . Advanced Organic Chemistry. Reactions, Mech-
anism and Structure, 3rd ed.; J ohn Wiley & Sons: New York, 1985;
pp 316. (b) Smith, S. G.; Fainberg, A. H.; Winstein, S. J . Am. Chem.
Soc. 1961, 83, 618. (c) J iali, G.; Xinfu, X. J . Am. Chem. Soc. 1993, 115,
9667.

4772 J . Org. Chem., Vol. 64, No. 13, 1999
Fukuda et al.
Sch em e 3
the residual charge stays in the RC. The electron density
smoothly flows from the nucleophile to the MsO fragment
in a vacuum while DMF disturbs its smooth redistribu-
tion. This is partially responsible for the increase in the
height of the activation barrier of path A in DMF.
As 6b has two negative charges, the electron density
in the SH- ion hardly moves to the reaction center in a
vacuum. For example, the SH- anion losses a small
amount of the charge in 6b(TS) (-0.092) while the LG
gains much larger charge (0.514) than that for 6a (TS).
The difficulty in redistributing charge of the two charges
system causes the activation barrier as high as 74.8 kcal
mol-1 in a vacuum. It is important to note that DMF
facilitates this charge redistribution very much. The SH-
anion in 6b(TS, DMF) loses the electron density by 0.227
although this value is still smaller than that of 6a (TS).
The increase of charge in the LG of 6b(TS, DMF) is the
largest in all the TSs obtained (0.578), and then the
electron density of the LG becomes -0.833. This easy
redistribution of electron density in DMF is effective to
reduce the activation energy of path B more than 40 kcal
Sch em e 4
mol-1
.
In the two negative charges SN2 reaction, the electro-
static repulsion prevents the SH- anion from attacking
the RC at the beginning of the reaction. However, the
charge of the RC flows by more than half of its initial
value to LG in the TS. This redistribution decreases the
electrostatic repulsion between the SH- and the COO-
fragment and makes MsO group leave easily from the
RC. The SH- anion can attack the RC after adequate
redistribution of electron density is occurred in the RC.
â-Elim in a tion of 1a . The main product for the reac-
tion of 1a and SH- is not the thiolated one but the R,â-
unsaturated carboxylate ester 5a as discussed above. The
first step for this product formation is extraction of the
R-proton at the C(2) of 1a to form an anionic intermediate
10a . Therefore, the stability of 10a is the key for
formation of 5.
The extraction of the R-proton is exothermic by 24.7
kcal mol-1; that is, the anionic intermediate is more
stable than 11a as expected. These results suggest that
E1cB mechanism21 is applicable to formation of R,â-
unsaturated carboxylate ester derivatives. Therefore,
forming the anionic intermediate proceeds instead of
replacing the TsO- fragment with the SH- ion. 12b is
expected to be very unstable because it possesses an
anionic carbon adjacent to the carboxylate ion. We
obtained no stable geometry for 12b. The calculated
results are consistent with the experimental one whereby
1b gave no R,â-unsaturated product.
New Syn th etic P r oced u r e. The ab initio MO calcu-
lations indicate that 2-tosyloxy carboxylate ester in path
A is too reactive to control its thiolation by using SH-
ion. In this path, it was observed that the reaction
produces undesired R,â-unsaturated product because the
ester reactant prefers E1cB to SN2 mechanism as dis-
cussed above. On the other hand, the electrostatic repul-
sion between the SH- ion and the carboxylic anion
fragment in path B causes not only lowering reactivity
of the reactant but also avoiding the undesired â-elimi-
nation. The SCRF calculations show that the appropriate
polarity of solvent reduces the activation energy of the
SN2 reaction very much. Therefore, we can control the
SN2 reaction with a strong SH- nucleophile by proper
combination of a changed substrate and a polar sol-
vent. According to this concept, we tried subse-
quent experiments and developed a new synthetic route
shown in Scheme 4. This route makes it possible to
synthesize 2-pyrimidinylthio carboxylic acid in a one-pot
procedure.
Two electron-withdrawing groups at the R-carbon,
COOCH3 and MsO, enhance acidity of the R-proton and
stabilize the anion formed.18 Furthermore, 1a contains
a leaving group (OMe) and two Me groups at the
â-position. These groups also stabilize the anion inter-
mediate very much.19 Although we used 3 by 5-fold equiv
mol of 1a in our experiments, the SH- ion mainly attacks
not the R-proton but the R-carbon atom because of its
high nucleophilic polarity.16 The OH- ion, which is
generated from H2O contained in sodium hydrosulfide
hydrate, acts as base extracting a proton at the R-car-
bon.20
To confirm stability of this anionic intermediate, 12a
as a model for 10a , shown in Scheme 3, was optimized,
and its optimized structure was depicted in Figure 1.
In a new procedure, 1b′ with a COO- fragment was
prepared with 1.0 mol equiv of sodium methoxide with
1b, followed by adding 1.1 mol equiv of sodium hydro-
sulfide hydrate. Stirring for 3 h at 50 °C gave 2b′. After
cooling the reaction mixture to 10 °C, adding 1.1 mol
equiv of 10% aqueous NaOH yielded 2b ′′ with two
(18) (a) Schleyer, P. v. R.; Kos A. Tetrahedron 1983, 39, 1141. (b)
Hoffmann, R.; Random, L.; Pople, J . A.; Schleyer, P. v. R.; Hehre, W.
J .; Salem, L. J . Am. Chem. Soc. 1972, 94, 6221.
(19) (a) Cann, P. F.; Stirling, J . M. J . Chem. Soc., Perkin Trans. 2
1974, 820. (b) Redman, R. P.; Stirling, J . M. Chem. Commun. 1970,
633.
(20) Chung, D. S.; Kim, C. K.; Lee, B. S.; Lee, I. J . Phys. Chem. A
1997, 101, 9097.
(21) Minato, T.; Yamabe S. J . Am. Chem. Soc. 1985, 107, 4621.

Computer-Aided Reaction Design
J . Org. Chem., Vol. 64, No. 13, 1999 4773
Calcd for C14H18O5S: C, 56.36; H, 6.08. Found: C, 55.83; H,
6.15.
negatively charged fragments. Subsequently, methane-
sulfonylpyrimidine were added and stirred for 10 min,
and finally, the reaction mixture was neutralized with
10% HCl. Thiolation and subsequent sulfidation pro-
ceeded in one pot and produced 4b in excellent yield more
than 92%.
3-Met h oxy-3-m et h yl-2-(4-t olu en esu lfon yloxy)b u t a n -
oic Acid (1b). To a solution of 1a (16.5 g, 0.05 mol) in ethanol
(60 mL) was added aqueous NaOH (5%, 30 mL), and the
mixture was allowed to stir at room temperature for 1 h.
Ethanol was removed under reduced pressure, and aqueous
HCl (5%, 40 mL) was added to the residue. The resulting
mixture was extracted with chloroform (50 mL). The organic
phase was washed with brine and dried over anhydrous
Na2SO4. The solvent was removed under reduced pressure to
afford solid, which was washed with 30 mL of hexane to give
1b as a white solid in 87.7% yield (13.2 g). 1H NMR (CDCl3) δ
1.26 (s, 3 H), 1.31 (s, 3 H), 2.45 (s, 3 H), 3.26 (s, 3 H), 4.77 (s,
2 H), 7.33-7.36 (m, 2 H), 7.82-7.85 (m, 2 H), 8.20-8.50 (br,
1 H). MS m/z: 303 (MH+). Anal. Calcd for C13H18O6S: C, 51.64;
H, 6.00. Found: C, 51.68; H, 5.98.
2-Mer ca p to-3-m eth oxy-3-m eth ylbu ta n oic Acid (2b). To
a solution of 1b (6.0 g, 0.02 mol) in DMF (30 mL) was added
sodium hydrosulfide hydrate (70% purity, 6.4 g, 0.08 mol) in
DMF (10 mL), and the mixture was allowed to stir at 60 °C
for 3 h. The resulting mixture was cooled and poured into
acidic ice-water (200 mL). The resulting mixture was ex-
tracted with ethyl acetate (20 mL × 2), and the organic phase
was washed with brine twice and dried over anhydrous
Na2SO4. The solvent was removed under reduced pressure, and
the residue was purified through chromatography (on silica
gel eluting with hexane-ethylacetate-acetic acid 1:1:0.1 v/v)
to give 2b as colorless oil in 92.2% yield (3.54 g). 1H NMR
(CDCl3) δ 1.39 (s, 6 H), 2.32 (d, J ) 9.2 Hz, 1 H), 3.30 (s, 3 H),
3.52 (d, J ) 9.2 Hz, 1 H), 9.45 (br, 1 H). MS m/z: 165 (MH+).
Anal. Calcd for C6H12O3S: C, 43.88; H, 7.37. Found: C, 43.43;
H, 7.26.
Con clu d in g Rem a r k s
An SN2 reaction cannot apply for thiolation of 2-tosyl-
oxy carboxylate ester 1a because of the stability of the
anionic intermediate extracted its R-proton. On the other
hand, thiolation of 2-tosyloxy carboxylic acid proceeds
through the SN2 reaction. Electrostatic repulsion between
the COO- fragment and the anionic nucleophile SH-
reduces the reactivity of the substrate for the â-elimina-
tion and gives selectivity for the SN2 reaction. The
combination of the two-charges technique and proper
solvent makes it possible to control the simple SN2
reaction. On the basis of above analysis of MO calcula-
tions, we succeeded in controlling the SN2 reaction
between sodium hydrosulfide hydrate with high nucleo-
philicity and 2-tosylate with high leaving ability. This
procedure can be performed in one pot and gives 4b in
excellent yield. The development of a facile procedure for
2-pyrimidinylthio carboxylic acid from 3-alkoxy-2-hy-
droxy carboxylate made it possible to synthesize many
3-alkoxy-2-(pyrimidinylthio)alkanoate derivatives.
2-(4,6-Dim e t h oxyp yr im id in -2-ylt h io)-3-m e t h oxy-3-
m eth ylbu ta n oic Acid (4b). To a solution of 2b (0.86 g, 5
mmol) in 4% aqueous NaOH (12 mL, 12 mmol) was added a
solution of 4,6-dimethoxy-2-methylsulfonylpyrimidine (1.09 g,
5 mmol) in acetone (5 mL) at 5 °C, and the mixture was
allowed to stir at room temperature for 1 h. The resulting
solution was added to 10 mL of 5% aqueous HCl solution and
extracted with ethyl acetate (20 mL × 2). The organic phase
was washed with brine twice and dried over anhydrous
Na2SO4. Ethyl acetate was removed under reduced pressure,
and the obtained pale brown solid was purified through
chromatography (hexanes-ethyl acetate-acetic acid 1:1:0.1
v/v) gave 4b as white solid in 81.2% yield (1.34 g). 1H NMR
(CDCl3) δ 1.45 (s, 3 H), 1.46 (s, 3 H), 3.34 (s, 3 H), 3.91 (s, 3
H), 3.92 (s, 3 H), 4.72 (s, 1 H), 5.75 (s, 1 H), 7.27 (s, 1 H). MS
m/z: 331 (MH+). Anal. Calcd for C12H18N2O5S: C, 47.67; H,
6.00. Found: C, 47.61; H, 5.92.
Syn th esis of 4b by th e On e-P ot Meth od . To a solution
of 1b (3.0 g, 0.01 mol) in DMF (30 mL) was added sodium
methoxide in methanol solution (28% solution, 1.93 g, 0.01 mol)
at 5 °C, and the reaction mixture was allowed to stir for 0.5 h.
Sodium hydrosulfide hydrate (70% purity, 0.8 g, 0.01 mol) was
added to the mixture, which was allowed to heat at 50 °C for
2 h. After cooling to room temperature, and then aqueous
NaOH (8% solution, 5.5 mL, 0.011 mol) was added to the
solution. Ten minutes later, to a solution of the mixture was
added 4,6-dimethoxy-2-(methylsulfonyl)pyrimidine (2.18 g,
0.01 mol) in DMF (5 mL), and the resulting mixture was
allowed to stir for 0.5 h. The resulting solution was poured
into aqueous HCl (5%, 30 mL) and extracted with ethyl acetate
(30 mL x 2). The organic phase was washed with brine twice
and dried over anhydrous Na2SO4. Ethyl acetate was removed
under reduced pressure. The obtained pale brown solid was
purified through chromatography (hexanes-ethyl acetate-
acetic acid 1:1:0.1 v/v) to give 4b as white solid in 92.4% yield
(3.05 g).
Exp er im en ta l Section
Gen er a l. Compounds 1a , 1b , 2b , and 4b were prepared
according to the methods described perviously.3a,6 All experi-
mental works were performed under nitrogen atomosphere.
1H NMR spectra were recorded at 300 MHz. All NMR spectra
are reported in ppm. Mass spectra were recorded in m/z.
Preparative chromatography was performed on Wako gel
C-200.
Eth yl 3-Meth oxy-3-m eth yl-2-(4-tolu en esu lfon yloxy)bu -
ta n oa te (1a ). To a solution of ethyl 3-methoxy-3-methyl-2-
hydroxybutanoate (17.6 g, 0.10 mol), p-toluenesulfonyl chloride
(20.9 g, 0.11 mol), and triethylamine (11.1 g, 0.11 mol) in
CH2Cl2 (150 mL) was added N,N-(dimethylamino)pyridine (0.2
g, 0.11 mol), and the mixture was allowed to stir at room
temperature for 12 h. The reaction mixture was poured into
water, and the organic phase was separated, washed with a
brine solution twice, and dried over anhydrous Na2SO4. The
solvent was removed under reduced pressure, and crude
product was obtained as oil. Chromatography (silica gel/
hexanes-ethyl acetate 3:1 v/v) gave 1a as colorless oil in 89.4%
yield (29.5 g). 1H NMR (CDCl3) δ 1.19 (t, J ) 7.2 Hz, 3 H),
1.26 (s, 3 H), 1.34 (s, 3 H), 2.43 (s, 3 H), 3.18 (s, 3 H), 4.07 (q,
J ) 7.2 Hz, 2 H), 4.71 (s, 1 H), 7.32-7.35 (d, 2 H), 7.80-7.83
(d, 2 H). MS m/z: 331 (MH+). Anal. Calcd for C15H22O6S: C,
54.53; H, 6.71. Found: C, 54.36; H, 6.76.
Eth yl 3-Meth yl-2-(4-tolu en esu lfon yloxy)-2-bu ten oa te
(5a ). To a solution of 1a (3.3 g, 0.01 mol) in DMF (30 mL) was
added sodium hydrosulfide hydrate (70% purity, 3.2 g, 0.03
mol) in DMF (10 mL), and the resulting mixture was allowed
to stir for 1 h at room temperature. The reaction mixture was
poured into 200 mL of acidic ice-water (5% HCl aqueous
solution) and extracted with ethyl acetate (20 mL × 2). The
organic phase was washed with brine twice and dried over
anhydrous Na2SO4. Removal of solvent in vacuo afforded oil,
which was subjected into column chromatography (hexanes-
ethyl acetate 1:6 v/v) to give 5 as colorless oil (20.1%, 0.62 g).
1H NMR (CDCl3) δ 1.16-1.20 (t, J ) 7.2 Hz, 3 H), 1.76 (s, 3
H), 2.15 (s, 3 H), 2.45 (s, 3 H), 4.07 (q, J ) 7.2 Hz, 2 H), 7.33-
7.36 (d, 2 H), 7.82-7.85 (d, 2 H). MS m/z: 299 (MH+). Anal.
Ack n ow led gm en t. The authors thank Computer
Center, Institute for Molecular Science at the Okazaki
National Research Institutes, for the use of the NEC
HSP computer and Library Program GAUSSIAN94.

4774 J . Org. Chem., Vol. 64, No. 13, 1999
Fukuda et al.
S.F. thanks Ube Industries Ltd. for the use of the Origin
computer including the GAUSSIAN94 program. The
authors also thank Mr. Yasuhiro Nagoshi in the Infor-
mation System Group, Corporate Planning and Admin-
istration Department at Ube Industries Ltd., for tech-
nical assistance on the MO calculations. This work was
supported in part by the Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture and
Sports, J apan.
Su p p or tin g In for m a tion Ava ila ble: Cartesian coordi-
nates and calculated total energies of all the structures in the
present study. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O990055N