Asymmetric catalysis, 125: Synthesis of the stereoisomers of methohexital by palladium-catalyzed allylation

Article abstract of DOI:10.1002/(sici)1099-0682(199901)1999:1<51::aid-ejic51>3.0.co;2-y

Allylation of 1-methyl-5-(1'-methylpent-2'-ynyl)barbituric acid (MBS) with allyl acetate using in situ catalysts of palladium(II) acetylacetonate and chiral phosphane imine ligands resulted in the enantioselective formation of 5-allyl-1-methyl-5-(1'-methylpent-2'-ynyl)barbituric acid (Methohexital), an important anesthetic drug. Both, MBS and Methohexital contain two stereogenic carbon atoms. In MBS, the asymmetric centre in the barbiturate system is labile due to enolization. The asymmetric centre in the hexyne side chain is stable and racemic. The two asymmetric centres of Methohexital are stable and give rise to four stereoisomers, two diastereomeric racemates. An analysis of the isomers of MBS and Methohexital was established on the basis of ¹H NMR and, in particular, GC including a base-line separation of the four stereoisomers of Methohexital. The stereoselectivity of the allylation is difficult to control, because the new quaternary asymmetric centre in the barbiturate ring of Methohexital is formed within the nucleophile, attacking the η³-allyl ligand of the catalyst from the side opposite to the palladium atom. Classical optically active ligands, such as diop or norphos, give only 2-6 % ee. Chiral phosphane imine ligands are a successful class of compounds, synthesized by Schiff base condensation of (2-formylphenyl)diphenylphosphane with optically active primary amines. The most efficient ligands have a hydroxymethyl and a bulky alkyl substituent at the asymmetric centre in the imine part, e.g. the L-iso-leucinol and the L-tert-leucinol derivatives 5 and 7. In the Pd-catalyzed allylation of MBS a kinetic resolution and the effect of the enantioselective catalyst interplay, the contributions of which are separated. For MBS the best stereoselectivity factor of the kinetic resolution s = k(R)/k(S) was 2.6 and 83 % 'ee' were achieved. The corresponding values for Methohexital were s = 3.5 and 80 % ee in the α-dl pair. For 10 mixtures of Methohexital stereoisomers the anesthetic doses for rats were determined. With 9.1 mg/kg body weight of the animal the sample obtained from the catalysis with the D-α-phenylglycinol derivative 8 gave a much lower anesthetic dose than the widely used narcotic BrevimytalNatrium, the sodium salt of the α-dl racemate of Methohexital, with 13.0 mg/kg body weight.

Full text of DOI:10.1002/(sici)1099-0682(199901)1999:1<51::aid-ejic51>3.0.co;2-y

FULL PAPER
Asymmetric Catalysis, 125[°]
Synthesis of the Stereoisomers of Methohexital
by Palladium-Catalyzed Allylation
Henri Brunner,*[a] Irmgard Deml,[a] Wolfgang Dirnberger,[a] Karl-Peter Ittner,[b]
Walter Reißer,[a] and Markus Zimmermann[b]
Dedicated to Professor Bernt Krebs on the occasion of his 60th birthday
Keywords: Asymmetric catalysis / Allylic alkylation / 1-Methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric acid / Palladium
catalysts / Chiral phosphane imine ligands / Methohexital / Anesthetic dose
Allylation of 1-methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric
acid (MBS) with allyl acetate using in situ catalysts of
palladium(II) acetylacetonate and chiral phosphane imine
ligands resulted in the enantioselective formation of 5-allyl-
1-methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric acid (Metho-
hexital), an important anesthetic drug. Both, MBS and
Methohexital contain two stereogenic carbon atoms. In MBS,
the asymmetric centre in the barbiturate system is labile due
to enolization. The asymmetric centre in the hexyne side
chain is stable and racemic. The two asymmetric centres of
Methohexital are stable and give rise to four stereoisomers,
two diastereomeric racemates. An analysis of the isomers of
MBS and Methohexital was established on the basis of 1H
NMR and, in particular, GC including a base-line separation
Chiral phosphane imine ligands are a successful class of
compounds, synthesized by Schiff base condensation of (2-
formylphenyl)diphenylphosphane with optically active
primary amines. The most efficient ligands have
hydroxymethyl and bulky alkyl substituent at the
asymmetric centre in the imine part, e.g. the -iso-leucinol
and the -tert-leucinol derivatives 5 and 7. In the Pd-
a
a
L
L
catalyzed allylation of MBS a kinetic resolution and the effect
of the enantioselective catalyst interplay, the contributions of
which are separated. For MBS the best stereoselectivity
factor of the kinetic resolution s = kR/kS was 2.6 and 83 %
“ee” were achieved. The corresponding values for
Methohexital were s = 3.5 and 80 % ee in the α-dl pair. For
10 mixtures of Methohexital stereoisomers the anesthetic
doses for rats were determined. With 9.1 mg/kg body weight
of the animal the sample obtained from the catalysis with the
of the
four stereoisomers of Methohexital. The
stereoselectivity of the allylation is difficult to control,
because the new quaternary asymmetric centre in the
barbiturate ring of Methohexital is formed within the
nucleophile, attacking the η3-allyl ligand of the catalyst from
the side opposite to the palladium atom. Classical optically
active ligands, such as diop or norphos, give only 2–6 % ee.
D
-α-phenylglycinol derivative
8
gave
a
much lower
anesthetic dose than the
widely
used narcotic
Brevimytal Natrium, the sodium salt of the α-dl racemate of
Methohexital, with 13.0 mg/kg body weight.
the synthesis of 5-allyl-1-methyl-5-(1Ј-methylpent-2Ј-ynyl)-
barbituric acid (Methohexital). Methohexital is a pharma-
ceutically important short-time anesthetic (Eli Lilly).[5Ϫ7]
For a narcosis on an average 150 mg of substance are
needed per person.[8,9] The sodium salt of the α-dl racemate
of Methohexital is commercially available (trade name Brev-
imytal Natrium).[5Ϫ11] Besides the allylic group there is a
1Ј-methylpent-2Ј-ynyl substituent in 5-position of the barbi-
turic acid skeleton. Therefore, two asymmetric centres re-
sult, one in 5-position of the barbiturate system and one in
2Ј-position of the hexyne side chain. Consequently, there
are four stereoisomers, two diastereomeric racemates (R,R)/
(S,S) and (R,S)/(S,R). Gibson and Doran found different
narcotic effects for the four stereoisomers.[12,13] The individ-
ual enantiomers can be obtained by regiospecific conden-
sation of the enantiomers with respect to the 1Ј-methylpent-
2Ј-ynyl side chain and subsequent diastereomer separ-
ation.[13] The advantage of a synthesis of Methohexital by
Pd-catalyzed allylic alkylation is the accessibility of differ-
Introduction
Recently, we reported on the application of the enantiose-
lective palladium-catalyzed allylic alkylation in the chemis-
try of barbituric acid derivatives.[2] The starting material
1,5-dimethylbarbituric acid could be converted into the
product 5-allyl-1,5-dimethylbarbituric acid in an enantio-
meric excess of up to 34% ee by variation of the bases and
the optically active ligands.[2Ϫ4] In the present paper we
transferred the experiences gained in the enantioselective al-
lylation of the model system 1,5-dimethylbarbituric acid to
[°]
Part 124: Ref.[1]
[a]
Institut für Anorganische Chemie der Universität,
D-93040 Regensburg, Germany
Fax: (internat.) ϩ 49(0)941/9434439
E-mail: henri.brunner@chemie.uni-regensburg.de
[b]
Institut für Pharmakologie der Universität,
D-93040 Regensburg, Germany
Fax: (internat.) ϩ 49(0)941/9434771
E-mail: karl-peter.ittner@klinik.uni-regensburg.de
Eur. J. Inorg. Chem. 1999, 51Ϫ59
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ1948/99/0101Ϫ0051 $ 17.50ϩ.50/0
51

H. Brunner, I. Deml, W. Dirnberger, K.-P. Ittner, W. Reißer, M.. Zimmermann
FULL PAPER
Scheme 1. Keto-enol tautomerism of the labile asymmetric centre of MBS in the barbiturate system
ently enriched diastereomeric and enantiomeric mixtures in lute configuration is not known. The difference in area be-
one reaction step.
tween the two peaks is called “combined enantiomeric ex-
We describe the preparation of the starting material 1- cess” and abbreviated by “ee”. For a racemic MBS sample
methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric acid (MBS), the a “combined enantiomeric excess” of 0% “ee” was meas-
analysis of its stereoisomers and its Pd-catalyzed allylation ured on the CP-Chirasil-DEX-CB column used for the
as well as the analysis of the diastereomers and enantiomers analysis.
of the product Methohexital and the results of the catalytic
runs (variation of base and optically active ligand). Then we
report on the separation of the contributions of the kinetic 5-Allyl-1-methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric
resolution and the optically active catalyst, which superpose Acid (Methohexital) and Its Stereoisomer Analysis
in the enantioselective Pd-catalyzed allylation of MBS. Fi-
In the Pd-catalyzed allylation of MBS a configurationally
stable second asymmetric centre in 5-position of the bar-
biturate system of the product Methohexital is formed
(Scheme 2).
nally, the Methohexital mixtures synthesized are tested with
regard to their narcotic effect in rats.[3,4]
1-Methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric Acid
(MBS) and Analysis of Its Stereoisomers
1-Methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric acid (MBS)
is synthesized in five steps.[3,4,10Ϫ15] MBS contains two
asymmetric centres. Therefore, it forms two diastereomeric
pairs of enantiomers. The signals of the four stereoisomers
of MBS appear in approximately equal intensities in the 1H-
NMR spectrum (N-methyl groups) after addition of the
shift reagent (S)-(ϩ)-1-(9Ј-anthryl)-2,2,2-trifluoroethanol.
To follow the accumulation of the stereoisomers of MBS
during the catalytic process a GC separation of the stereo-
isomers on a chiral CP-Chirasil-Dex-CB column (length 25
m), coated with permethylated β-cyclodextrin, was devel-
Scheme 2. Synthesis of Methohexital by Pd-catalyzed allylation of
MBS
The four stereoisomers of Methohexital were separated
oped. Two peaks of equal areas were obtained, each of with a GC system consisting of an achiral and a chiral col-
which contains the two stereoisomers of MBS, which in- umn (Figure 1a). The achiral column was a polar Dura-
terconvert by the keto-enol tautomerism of the labile asym- bond-DB-225 column (length 30 m) and the chiral column
metric centre of the barbiturate system (Scheme 1).
was a Chiraldex-B-DA column (length 30 m). The two en-
This keto-enol tautomerism is only possible for these two antiomeric pairs are designated α and β and the configura-
stereoisomers, which have the same configuration in the tion in the hexyne side chain 1 and 2. Thus, the four stereo-
hexyne side chain (bold configuration symbol), because isomers of Methohexital are: α1, α2, β1, β2 (Figure 1a).
only the asymmetric centre in the barbiturate system is in- The two α-peaks belong to one pair of enantiomers (α1 and
volved in the tautomerism. Thus, the first peak arises either α2) with the literature name α-dl[12][13] and the two β-peaks
from the combination (R,S)/(R,R) and the second peak to the other pair of enantiomers (β1 and β2) with the litera-
from the combination (S,R)/(S,S) or vice versa. The abso- ture name β-dl. These two pairs of enantiomers are either
52
Eur. J. Inorg. Chem. 1999, 51Ϫ59

Asymmetric Catalysis, 125
FULL PAPER
(R,R)/(S,S) and (R,S)/(S,R) or vice versa. The indices 1 and
2 are choosen such, that stereoisomers with the same index
have the same configuration in the hexyne side chain. Thus
α1 and β1 have either the configuration (R,R)/(R,S) or
(S,S)/(S,R) (analogously for α2 and β2).
Which of the stereoisomers of Methohexital have the
same configuration in the hexyne side chain, was estab-
lished from a chromatogram of a catalysis with a yield of
94%. As the starting material MBS was racemic with re-
spect to the configuration in the hexyne side chain, the
product stereoisomers can be combined to the dia-
stereomeric pairs (R,R)/(R,S) or (S,S)/(S,R) only in this
way, that the pairs do not have more than yields of 50%.
Thus, the first and the fourth as well as the second and the
third peak must have the same configuration in the hexyne
side chain (Figure 1a).
The four stereoisomers of Methohexital could also be
separated with a permethylated β-cyclodextrin column Chi-
raldex-B-PM (length 30 m). In this analysis the second and
the fourth peak belong to the α-dl pair of Methohexital
and the first and the third to the β-dl pair (Figure 1b). A
Methohexital sample resulting from an achiral catalysis
with the ligand triphenylphosphane (Table 1, entry 1) gave
an enantiomeric excess, as expected, of 0% ee in the α and
β series. The diastereomeric excess of α pair versus β pair
was measured to be 8% de with the B-PM column com-
pared to 7% de with the two-column system DB-225 ϩ
B-DA (Figure 1a). In the GC investigations of Methohexi-
tal samples care had to be taken to separate the starting
material MBS completely from the product Methohexital,
Figure 1. Separation of the four Methohexital isomers: a) on a DB-
225 ϩ B-DA column system (catalysis with ligand triphenylphos-
phane; Table 1, entry 1; 7% de of α pair); b) on a B-PM column
because both with the DB-225 ϩ B-DA column system as (catalysis with ligand 7; Table 2, entry 3; sample heavily enriched
in α2)
well as with the B-PM column there was overlap[3,4]
.
With a qualitative TLC test it is possible to monitor the
progress of the chromatography (allyl acetate, AAMBS,
Methohexital, and MBS show different yellow spots after
dipping into a dilute KMnO4 solution). By changing the
eluent to dichloromethane/acetonitrile (3:1) the unreacted
starting material MBS can be eluted. After removal of the
solvent Methohexital forms a colourless oil, which crys-
tallizes subsequently. The yield is determined by weighing.
The analysis of the Methohexital stereoisomers is achieved
by GC.
Standard Catalytic Allylation of MBS to
Methohexital
In the Pd-catalyzed allylation (Scheme 2) MBS is dis-
solved in dichloromethane/toluene (1:1). To the clear solu-
tion a base (NEt3 or DBU) is added in 5% excess. Then 1
mol-% of palladium(II) acetylacetonate and 4 mol-% of a
monodentate or 2 mol-% of a bidentate optically active
phosphane are dissolved in the reaction mixture. The reac-
tion is started by addition of allyl acetate in 11% excess.
The homogeneous solution containing the “in situ” Pd/li-
gand complex is stirred for 72 h at 38°C. If the ligands are
not soluble in dichloromethane/toluene (1:1), the dichloro-
methane content is increased.
Chiral Ligands in the Palladium-Catalyzed Allylation
of MBS
The reaction is stopped by addition of 0.2  hydrochloric
In the Pd-catalyzed allylation of MBS the achiral ligand
acid. From the organic layer an oily product is obtained, triphenylphosphane as well as a series of chiral ligands[2]
which is chromatographated on silica gel with dichloro- were used. The phosphane imine ligands 2a, 2b, 3Ϫ10 were
methane/acetonitrile (25:1). First, the yellow (2-formyl- obtained by Schiff base condensation of (2-formylphenyl)-
phenyl)diphenylphosphane, the hydrolysis product of the diphenylphosphane with the following primary amines[2]
:
phosphane imine ligands, is eluted, then the excess of allyl (R)-(Ϫ)-2-amino-1-butanol, (S)-(ϩ)-2-amino-1-butanol, -
acetate and small amounts of the doubly allylated by-prod- alaninol, -valinol, -leucinol, -isoleucinol, -tert-leucinol,
uct 3,5-diallyl-1-methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric -α-phenylglycinol, -phenylalaninol, -tert-butyl tert-leu-
acid (AAMBS), formed in a second allylation (N-al- cinate. Further chiral ligands were (Ϫ)-(4S)-2-[2Ј-(diphenyl-
lylation[2][4]). Then the product Methohexital is eluted. phosphanyl)phenyl]-4-isopropyl-1,3-oxazoline
(11),[16,17]
Eur. J. Inorg. Chem. 1999, 51Ϫ59
53

H. Brunner, I. Deml, W. Dirnberger, K.-P. Ittner, W. Reißer, M.. Zimmermann
FULL PAPER
(ϩ)-1,2-bis{bis[3Ј,5Ј-bis(N-methyliden-(R)-alaninol)phen- mate of 7% de was obtained (Table 1, entry 1). Phosphane
yl]phosphanyl}ethane (12)[2,18] and (Ϫ)-1,2-bis{bis[3Ј,5Ј- imine ligands induce a high diastereomeric and enantio-
bis(N-methylidene-(S)-phenylalaninol)phenyl]phosphanyl}- meric excess, if there is a hydroxymethyl substituent at the
ethane (13)[18] (Scheme 3). For a comparison the standard asymmetric centre near the imine group, which acts as an
ligands (Ϫ)-norphos[19,20] and (Ϫ)-diop[21,22] were included. anchor for the incoming barbiturate anion.[2] The variation
The results of the catalyses are summarized in Table 1.
of the other side chain at the asymmetric centre leads to
Scheme 3. The phosphane imines 2Ϫ10, the phosphane oxazoline 11 and the phosphane octaimines 12 an 13
With the achiral triphenylphosphane in the Pd-catalyzed distinct differences in the diastereomeric and enantiomeric
allylation of MBS a diastereomeric excess of the α-dl race- excess. With the (R) ligand 2a an enantiomeric excess in the
54
Eur. J. Inorg. Chem. 1999, 51Ϫ59

Asymmetric Catalysis, 125
FULL PAPER
Table 1. Synthesis of Methohexital by palladium-catalyzed allylation of MBS with allyl acetate; procatalyst 1 mol-% [Pd(acac)2], Pd/
ligand ratio 1:4, solvent 25 mL of CH2Cl2 ϩ 25 mL of toluene, 38°C, 72 h; GC analysis with the DB-225 ϩ B-DA two-column system
Entry
Ligand
Base
Yield
[%][c]
ee
(α1, α2)
[%]
ee
(β1, β2)
[%]
“ee”[a]
(α1ϩβ1,
α2ϩβ2) [%]
de
(α1, β1)
[%]
de
(α2, β2)
[%]
de
s[b]
(α1ϩα2,
β1ϩβ2) [%]
1
1
DBU
NEt3
NEt3
NEt3
DBU
NEt3
NEt3
NEt3
82
52
52
66
62
26
94
33
43
63
89
91
91
63
64
12
10
59
62
74
93
0
0
0
7 (α1)
62 (α1)
3 (β1)
6 (α1)
8 (α1)
3 (β1)
12 (β1)
7 (α2)
7 (α1ϩα2)
40 (α1ϩα2)
37 (α1ϩα2)
23 (α1ϩα2)
24 (α1ϩα2)
42 (α1ϩα2)
29 (α1ϩα2)
1.0
2.3
2.5
1.5
1.1
2.9
1.3
2
2a
2b
3
47 (α1)
50 (α2)
24 (α2)
15 (α2)
62 (α2)
33 (α2)
62[d] (α2)
62 (α2)
51 (α2)
40 (α2)
40 (α2)
26[d] (α1)
28[d] (α1)
19 (α2)
1 (α2)
21 (β2)
14 (β1)
8 (β1)
27 (α1ϩβ1)
30 (α2ϩβ2)
11 (α2ϩβ2)
4 (α2ϩβ2)
43 (α2ϩβ2)
2 (α2ϩβ2)
1 (α2)
3
60 (α2)
36 (α2)
38 (α2)
61 (α2)
68 (α2)
64[d] (α2)
65 (α2)
69 (α2)
71 (α2)
71 (α2)
3[d] (α2)
0[d]
23 (α2)
5 (α2)
4 (α2)
6 (α2)
15 (β2)
13 (β2)
4[d] (α2)
4
5
3
16 (β1)
3 (β1)
6
4
7
5
55 (β1)
8[d] (β1)
12 (β1)
38 (β1)
59 (β1)
61 (β1)
32[d] (β2)
28[d] (β2)
6 (β1)
8, 9
6
42[d] (α2ϩβ2) 6[d] (β1)
44[d] (α1ϩα2) 3.0
42 (α2ϩβ2)
24 (α2ϩβ2)
2 (α2ϩβ2)
3 (α2ϩβ2)
5[d] (α2ϩβ2)
8[d] (α1ϩβ1)
8 (α2ϩβ2)
2 (α1ϩβ1)
0
3 (α1ϩβ1)
3 (α2ϩβ2)
3 (α2ϩβ2)
0[d]
7 (β1)
44 (α1ϩα2)
39 (α1ϩα2)
26 (α1ϩα2)
25 (α1ϩα2)
3.3
2.3
1.2
1.3
10
11, 12
6
7
DBU
NEt3
11 (β1)
23 (β1)
22 (β1)
55[d] (α1)
52[d] (α1)
1 (β1)
13
14
15
16
17
18
19
20
21
8
NEt3
DBU
DBU
NEt3
DBU
NEt3
NEt3
NEt3
NEt3
30[d] (α1ϩα2) 1.6
8
28[d] (α1ϩα2) 1.3
9
12 (α1ϩα2)
3 (α1ϩα2)
3 (α1ϩα2)
5 (α1ϩα2)
6 (β1ϩβ2)
0
1.3
1.0
1.0
1.1
1.1
1.1
10
5 (β1)
0
10
1 (α2)
2 (β1)
1 (α1)
11
3 (α1)
4 (β1)
4 (α1)
12[e,f]
13[e]
(Ϫ)-
norphos[e]
8 (α1)
13 (β2)
19 (β2)
6[d] (β2)
5 (α1)
12 (α1)
18 (α1)
5[d] (α1)
15[d] (α1)
10[d] (α1ϩα2) 1.0
22
(Ϫ)-diop[e] NEt3
93
2[d] (α1)
2[d] (β2)
0[d]
11[d] (α1)
7[d] (α2)
9[d] (α1ϩα2)
1.0
[a] “ee” ϭ “combined enantiomeric excess” (see text). Ϫ [b] s ϭ stereoselectivity factor. Ϫ [c] After chromatography. Ϫ [d] GC analysis with
the B-PM column. Ϫ [e] Pd/ligand ratio 1:2. Ϫ [f] 150 mL of CH2Cl2 ϩ 25 mL of toluene was used to dissolve ligand 12.
α-dl pair of 47% ee (α1) and in the β-dl pair of 21% ee (β1) using NEt3 as a base. The diastereoselectivities are between
is achieved, the diastereomeric excess for the α-dl pair being 25 and 29% de for the α-dl pair (entries 7, 11, 12). If the
40% de (entry 2). The isomers added in parentheses after side chain at the asymmetric centre contains a phenyl group
the ee and de values are the favoured isomers. The anal- (ligands 8 and 9), the induced enantiomeric excess is higher
ogous (S) ligand 2b gives a similar enantiomeric excess of for ligand 8 than for ligand 3 and distinctly lower for ligand
the opposite stereoisomers, a similar diastereomeric excess 9 than for ligand 2b (entries 13Ϫ15). A substitution of the
for the α-dl pair and the same chemical yield (entry 3). A hydroxymethyl group at the asymmetric centre by a tert-
shortening of the ethyl group at the asymmetric centre in butyl ester group in the ligand 10 reduces the enantiomeric
the ligands 2a and 2b to a methyl group in the ligand 3 excess appreciably independent of the base used (entry 16).
diminishes the diastereomeric and enantiomeric excess con-
siderably, although the chemical yield increases (entries 4, antiomeric excess in the product Methohexital similar to
5). the standard ligands (Ϫ)-norphos and (Ϫ)-diop (entries 17,
The phosphane oxazoline ligand 11 gives only a low en-
Branching of the alkyl group as in the ligands 4Ϫ7 leads 20, 21). The ligands 12 and 13 belong to a different group
to the highest stereoselectivities in the Pd-catalyzed al- of phosphane imine ligands with two phosphorus atoms.
lylation of MBS observed so far similar to the Pd-catalyzed They afford an enantiomeric excess in the various stereoiso-
allylation of 1,5-dimethylbarbituric acid[2]. With the ligands mers of 8Ϫ19% ee using the base NEt3 (entries 18, 19). Up
4 and 6, which contain an isopropyl group and a sec-butyl to now ligand 12 is the only one, which induces a dia-
group, respectively, a high enantiomeric excess in the α-dl stereomeric excess in favour of the β-dl pair (6% de).
pair of 62% ee (α2) was attained, associated with a low
enantiomeric excess of 3Ϫ12% ee (β1) in the β-dl pair, the
diastereomeric excess for the α-dl pair being 42 and 43% de.
With NEt3 as a base, the yields are only 26Ϫ43% (entries 6,
8, 9). Using DBU leads to an increased yield of 63%. The
enantiomeric excess in the α-dl pair decreases to 51% ee
The Interplay of Kinetic Resolution and Catalyst
Contribution
In the allylation of 1,5-dimethylbarbituric acid the 1,5-
(α2), whereas the enantiomeric excesses in the β-dl pair rises dimethylbarbiturate anion is formed, which no longer con-
to 38% ee (β1) (entry 10). tains an asymmetric centre. Thus, there is no kinetic resolu-
The ligands 5 and 7, which carry an isobutyl group and tion in this reaction and the enantiomeric excess of the
a tert-butyl group at the asymmetric centre, respectively, product 5-allyl-1,5-dimethylbarbituric acid is only depend-
lead to a lower enantiomeric excess in the α-dl pair [33 and ent on the enantioselectivity of the chiral catalyst.[2] On the
40% ee (α2)] but to a higher enantioselectivity in the β-dl contrary, the starting material 1-methyl-5-(1Ј-methylpent-
pair [55, 59, and 61% ee (β1)] with yields of up to 94% 2Ј-ynyl)barbituric acid (MBS) for the synthesis of Me-
Eur. J. Inorg. Chem. 1999, 51Ϫ59
55

H. Brunner, I. Deml, W. Dirnberger, K.-P. Ittner, W. Reißer, M.. Zimmermann
FULL PAPER
Table 2. The interplay of kinetic resolution and catalyst contribution in the Pd-catalyzed allylation of MBS with ligand 7; procatalyst 1
mol-% of [Pd(acac)2], Pd/ligand ratio 1:4, solvent CH2Cl2/toluene (1:1), 38°C, base DBU; GC analysis of MBS with a CP-Chirasil-Dex-
CB column and of Methohexital with a B-PM column
A
B
C
D
E
F
G
H
I
J
K
Methohexital
MBS
Entry Time Yield
ee
ee
“ee”[a]
de
de
de
S[b]
(from A,
unreacted “ee”[a] s[b]
MBS
(α1, α2) (β1, β2) (α1ϩβ1,
(α1, β1) (α2, β2) (α1ϩα2,
(1., 2.) (from A,
[%]
[h]
[%][c]
[%]
[%]
α2ϩβ2) [%] [%]
[%]
β1ϩβ2) [%] D calcd.) [%][c]
J calcd.)
1
2
3
4
5
6
7
3
17
20
26
32
49
65
91
77 (α2) 7 (β1)
80 (α2) 6 (β1)
60 (α2ϩβ2) 11 (β1) 76 (α2) 59 (α1ϩα2) 4.5
62 (α2ϩβ2) 18 (β1) 76 (α2) 58 (α1ϩα2) 5.0
79
77
70
65
48
Ϫ
7
5 (1.)
9 (1.)
1.7
2.3
4.5
9
80 (α2) 13 (β1) 60 (α2ϩβ2) 20 (β1) 77 (α2) 58 (α1ϩα2) 4.9
78 (α2) 17 (β1) 57 (α2ϩβ2) 19 (β1) 77 (α2) 57 (α1ϩα2) 4.7
61 (α2) 30 (β1) 36 (α2ϩβ2) 12 (β1) 71 (α2) 45 (α1ϩα2) 2.9
14 (1.) 2.7
19 (1.) 2.8
40 (1.) 3.5
12
72
72
72
42 (α2) 67 (β1)
46 (α2) 72 (β1) 6 (α2ϩβ2)
0
27 (β1) 75 (α2) 24 (α1ϩα2) 1.0
24 (β1) 82 (α2) 32 (α1ϩα2) 1.7
Ϫ
Ϫ
83 (1.) 2.3
[a] “ee” ϭ “combined enantiomeric excess”. Ϫ [b] s ϭ stereoselectivity factor. Ϫ [c] After chromatography.
thohexital contains two asymmetric centres. On addition of of Methohexital can be used to represent the overall conver-
base the asymmetric centre in the barbiturate system is re- sion.
moved, but the asymmetric centre in the hexyne side chain
After 3 h, the first sample taken gave a yield of 17% of
is preserved. In the present study the asymmetric centre in Methohexital and 79% of starting material MBS (Table 2,
the hexyne side chain of the MBS used was racemic. There- entry 1). The yield of Methohexital increased from 20%
fore, two different effects can influence the distribution of after 4.5 h (77% MBS) via 26% after 9 h (70% MBS) and
the stereoisomers in the Pd-catalyzed allylation of MBS, the 32% after 12 h (65% MBS) to 49% after 72 h (48% MBS),
kinetic resolution and the enantioselective catalysis, which as shown in columns A and I of Table 2 (entries 2Ϫ5). In
are both concentration-dependent. The contributions of comparable reactions Methohexital yields of 65% (MBS
these two effects can be calculated provided the yields of not determined) and 91% (7% MBS) were isolated after 72
Methohexital and MBS as well as the diastereomeric and h (entries 6, 7).
enantiomeric excess of Methohexital and MBS are known.
Figure 2 shows the correlation between the yield of Me-
These data were obtained from a standard catalysis with thohexital (column A) and the increase of the “combined
ligand 7 by taking aliquots after 3, 4.5, 9, 12, and 72 h. The enantiomeric excess” of MBS (column J). With the decrease
yields of Methohexital and MBS were determined and their of the amount of unreacted MBS down to 7% after 72 h
stereoisomer composition was analysed by GC as described (entry 7) the enantioselectivities of the remaining MBS rises
above. In Table 2 the results and calculations are summar- to an enrichment of 83% “ee” (1.). The designation (1.)
ized together with those of two other catalytic runs.
means an excess of the first peak of MBS in the GC chro-
matogram.
The Kinetic Resolution in the Pd-Catalyzed Allylation
of MBS with Ligand 7
Due to the kinetic resolution in the reaction with the op-
tically active catalyst there is a chiral discrimination of the
MBS anions, which have opposite configurations in the
hexyne side chain. One is consumed faster to give the prod-
uct Methohexital, whereas the other is enriched in the unre-
acted starting material. This difference in reactivity shows
up in the intensity ratio of the first and the second MBS
GC peak (Table 2, column J). The “combined enantiomeric
excess”, calculated as [(Խpeak1 Ϫ peak2Խ)/(peak1 ϩ peak2)]
ϫ 100% reflects the reactivity difference of the MBS stereo-
isomers, which have opposite configurations in the hexyne
side chain.
Figure 2. Correlation between the “combined enantiomeric excess”
of MBS and the yield of Methohexital
Addition of the yield of Methohexital (column A) and
the amount of unreacted MBS (column I) leads to retrieval
rates of 96Ϫ98%. The missing 2Ϫ4% are lost by the treat-
With the rate constants kR and kS for the reaction of the
ment with 0.2  hydrochloric acid, by the chromatographic two MBS anion enantiomers, the stereoselektivity factor
purification or by the formation of the doubly allylated by- s ϭ kR/kS is a measure for the kinetic resolution.[23,24] It
product AAMBS. In a reasonable approximation the yield can be calculated from the amount and the enantiomeric
56
Eur. J. Inorg. Chem. 1999, 51Ϫ59

Asymmetric Catalysis, 125
FULL PAPER
excess of the unreacted starting material with Equation
Columns E and F show the diastereomeric excess α1 vs.
1.[23,24] The values of the stereoselectivity factor s of MBS β1 and α2 vs. β2, which involves the influence of the asym-
metric centre in the hexyne side chain and the optically ac-
tive catalyst on the formation of the new asymmetric centre
in the barbiturate ring. For the diastereomeric excess α1 vs.
C ϭ conversion: 0 < C < 1 (column A); assumption: retrieval rate
of 100%; ee ϭ “ee” ϭ “combined enantiomeric excess” in the unre-
acted starting material MBS: 0 < “ee” < 1 (column J)
β1 values between 11 and 27% de (β1) were achieved (col-
umn E) and for α2 vs. β2 values between 71 and 82% de
(α2) (column F). The diastereomeric excess α1 vs. β1 and
α2 vs. β2 is relatively independent of the turnover (columns
A, E, F). The overall diastereomeric excess α pair vs. β pair
decreases from 59 to 24% de (α pair) (column G) depending
on the yield of Methohexital (column A).
are between 1.7 and 3.5 with an average of 2.6 (column K).
The effect of the kinetic resolution with respect to the
product Methohexital can be calculated with the help of the
“combined enantiomeric excess” α1 ϩ β1 vs. α2 ϩ β2 (vs. ϭ
versus) in column D, obtained from the four GC peaks of
the Methohexital stereoisomers. The two components of
each sum have the same configuration in the hexyne side
chain. It decreases from 60Ϫ62% “ee” for a Methohexital
yield of 17Ϫ26% via 36% “ee” for a yield of 49% to 6%
“ee” for a yield of 91% (columns A and D) with increasing
conversion (entries 1Ϫ5, 7). The value of 0% “ee” for a
Methohexital yield of 65% does not fit into this series (en-
try 6). The stereoselectivity factor s (values between 1.7 and
5.0, column H) is calculated with Equation 2.[23,24]
Kinetic Resolution and Catalyst Contribution for the
Other Ligands
The influence of the kinetic resolution in the Pd-cata-
lyzed allylation of MBS is calculated for all the optically
active ligands with Equation 2 using enantiomeric excesses
α1 vs. α2 and β1 vs. β2 for Methohexital. The results are
summarized in Table 1. As expected s is 1.0 for the achiral
catalysis with triphenylphosphane, as both enantiomeric ex-
cesses α1 vs. α2 and β1 vs. β2 are 0% ee. The influence
of the kinetic resolution is small for most of the ligands
investigated (values for s between 1.0 and 1.6) except for
the catalyses with the ligands 2a, 2b, 4, and 6 (values for s
from 2.3 to 3.3).
C ϭ conversion: 0 < C < 1 (column A); assumption: retrieval rate
of 100%; ee ϭ “ee” ϭ “combined enantiomeric excess” in the prod-
uct Methohexital: 0 < “ee” < 1 (column D)
The contribution of the optically active catalyst in the
Pd-catalyzed allylation of MBS for all the ligands is obvious
from the enantiomeric excess in the α and β series, although
most of the reactions were stopped at higher conversions.
It was small for the standard ligands (Ϫ)-norphos, (Ϫ)-
diop, the phosphane oxazoline ligand 11 and the phosphane
imine ligand 10 containing a tert-butyl ester substituent
with an enantiomeric excess between 1 and 6% ee. For the
remaining phosphane imine ligands, which carry a hydroxy-
methyl substituent at the asymmetric centre, the catalyst
contribution was much higher with up to 62% ee (Table 1).
Although the absolute configurations are not known, the
starting material MBS and the product Methohexital can
be correlated with respect to the configuration in the
hexyne side chain. During the reaction the isomers of MBS
appearing in the first GC peak are enriched. Thus, they
react more slowly to the product Methohexital. The faster
reacting MBS isomers lead to a growth of the GC peaks of
the Methohexital stereoisomers α2 and β2. Therefore, they
must have the same configuration in the hexyne side chain
as the isomers of the second MBS peak.
The Contribution of the Optically Active Catalyst in
the Pd-Catalyzed Allylation of MBS with Ligand 7
Determination of the Anesthetic Doses of
Methohexital Stereoisomer Mixtures in Rats
In addition to the kinetic resolution the optically active
catalyst contributes to the stereoselectivity of the Pd-cata-
Methohexital is a common anesthetic widely used to in-
lyzed allylation of MBS. This is reflected in the enantiom- itiate narcoses.[5Ϫ11] The four stereoisomers of Methohexi-
eric excess in the α series which decreases from 80% ee (α2) tal exhibit different anesthetic effects in rats, dogs and
to 42% ee (α2) (Table 2, column B), whereas in the β series monkies.[12,13] Injected as a solution of the sodium salt into
it rises from 6% ee (β1) to 72% ee (β1) (column C). Figure rats, the β-l isomer of Methohexital is the most effective of
1b shows the highest enantiomeric excess of 80% ee (α2) in the stereoisomers with an anesthetic dose of 7 mg/kg body
the α-dl pair of Methohexital obtained up to now (entry 3). weight of the animal. The α-d stereoisomer with 11 mg/kg
The contribution of the optically active catalyst is con- and the β-d stereoisomer with 17 mg/kg follow. For a nar-
centration-dependent, because the composition of the MBS cosis with the α-l stereoisomer 35 mg/kg are necessary. With
stereoisomers changes during the reaction due to the kinetic 10 mg/kg the β-dl racemate is more effective than the α-dl
resolution. At the beginning of the reaction MBS is ra- racemate with 15 mg/kg. However, because of undesirable
cemic. Thus, the influence of the optically active catalyst stimulating post-anesthetic effects of the β-dl racemate
can be read off directly from the enantiomeric excess of caused by the β-l stereoisomer the extensively used narcotic
Methohexital in the beginning of the MBS allylation.
is the α-dl racemate (Eli Lilly).
Eur. J. Inorg. Chem. 1999, 51Ϫ59
57

H. Brunner, I. Deml, W. Dirnberger, K.-P. Ittner, W. Reißer, M.. Zimmermann
FULL PAPER
Table 3. 10 Methohexital samples and their isomer composition, the reference Brevimytal Natrium and the anesthetic doses (AD) in
Sprague Dawley rats of body weight between 560 and 720 g[25][26]
Sample
Ligand
α1 [%]
α2 [%]
β1 [%]
β2 [%]
AD [mg/kg]
Average AD [mg/kg]
1
1
27
27
23
23
10.8, 9.3, 11.1
14.6, 13.5, 11.1
15.7, 17.7, 14.8
17.4, 16.3, 16.4
16.1, 15.3, 15.7
8.9, 9.4, 8.9
17.9, 14.5, 15.6
14.7, 13.7, 14.3
14.1, 8.9, 9.5
22.0, 21.0, 20.1
16.3, 12.4, 10.3
10.4
13.0
16.1
16.7
15.7
9.1
16.0
14.2
10.9
21.0
13.0
2
2a
51.5
17
18.5
51.5
58.5
43
12
18
3
2b
18
13.5
13.5
8
4
4
13.5
22
14.5
27
5
5
6
8
41
23
13
23
7
8
41
24
12
23
8
3
26
36
22
16
9
6
13.5
19
58.5
44
15
13
10
11
7
29.5
Ϫ
7.5
Ϫ
reference[a]
50
50
[a] Protonated Brevimytal Natrium (see Experimental Section).
CC3), 151.2/151.1 (HNϪCOϪNϪCH3), 86.6/86.5 (HCϪCϵC),
76.8 (CϵCϪCH2), 54.3 (OCϪCϪCO), 31.8/31.6 (HCϪCϵC),
27.7/27.4 (NϪCH3), 20.0/19.7 (H3CϪCH), 13.8 (CH2ϪCH3), 12.1
(CϵCϪCH2). Ϫ The signals appear doubled because of the two
MBS diastereomers. Addition of (S)-(ϩ)-1-(9Ј-anthryl)-2,2,2-tri-
fluoroethanol (Super-Pirkle) to MBS in the molar rate 1:12 results
also in a separationof of the enantiomers: δ ϭ 3.059/3.054/3.043/
3.035 (4s, NϪCH3), 0.940/0.938/ 0.933/ 0.927 (4 t, 3J ϭ 7.5 Hz,
CϵCϪCH2ϪCH3). Ϫ MS (EI, 70 eV); m/z (%): 222.1 (7) [Mϩ],
207.1 (45) [Mϩ Ϫ CH3], 193.1 (46) [Mϩ Ϫ C2H5], 81.1 (35)
For 10 mixtures of Methohexital stereoisomers, obtained
by Pd-catalyzed enantioselective allylation of MBS, the
anesthetic dose (AD) in rats was determined using the cor-
neal stimulus technique[25] as described elsewhere.[26] The
results were compared with those of a sample of protonated
Brevimytal Natrium (sample 11).
Table 3 summarizes the anesthetic doses obtained by
standardising to 1 kg body weight of the rat. The average
anesthetic dose of the samples 1, 6, and 9 was lower than
that of Brevimytal Natrium with 13.0 mg/kg (Table 3). In [C6H9ϩ], 79.1 (100). Ϫ C11H14N2O3 (222.24): calcd. C 59.45, H
6.35, N 12.60; found C 59.52, H 6.34, N 12.63. Ϫ GC analysis of
MBS: 25-m CP-Chirasil-Dex-CB column (chemically connected
with permethylated β-cyclodextrin; 0.25 mm inner diameter, 0.25
µm film thickness; from Chrompack), column temp. 165°C, He
pressure 1.2 bar, injector temp. 250°C, detector temp. 230°C (flame
ionisation). Retention times: (1.) 23.29 min, (2.) 24.97 min; dead
time t0 (CH2Cl2) ϭ 1.21 min. Separation factor: α ϭ 1.07, resolu-
tion R ϭ 3.52.
particular, sample 6 with 9.1 mg/kg had a distinctly reduced
anesthetic dose. Deviations from the average value were
small with this sample. However, it is conspicuous, that
samples 6 and 7, which showed marked differences in the
anesthetic doses, had a closely related stereoisomer compo-
sition (Table 3).[26]
Synthesis of Methohexital by Enantioselective Palladium-Catalyzed
Allylation of MBS with Allyl Acetate: 1.07 g (4.81 mmol) of MBS
was dissolved in 25 mL of dichloromethane/toluene (1:1) at 38°C
under nitrogen. To the clear solution 5.05 mmol of the base was
added [DBU (752 µL, 769 mg) or NEt3 (700 µL, 511 mg), respec-
tively]. After 5 min of stirring, 14.6 mg (4.81и10Ϫ2 mmol) of Pd-
(acac)2 and 19.2и10Ϫ2 mmol of a monodentate or 9.61и10Ϫ2 mmol
of a bidentate ligand were rinsed into the reaction vessel with an-
Experimental Section
The instruments used were the same as described in the preceding
paper.[2] Ϫ All the syntheses and catalytic reactions were carried
out with dried solvents under nitrogen. Ϫ Palladium(II) acetylace-
tonate and allyl acetate (Merck) were used without further purifi-
cation. DBU (Fluka) was dried with CaCl2, NEt3 (Merck) with
CaH2, and distilled.
1-Methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric Acid (MBS):[3,4,13] other 25 mL of dichloromethane/toluene (1:1). After 2.0 min, 576
10.9 g (0.474 mol) of sodium was dissolved in 200 mL of refluxing
absolute ethanol. After addition of 17.6 g (0.237 mol) of N-meth-
µL (5.35 mmol, 536 mg) of allyl acetate was added to the clear
yellow solution. Usually, the solution was stirred for 72 h at 38°C.
ylurea and stirring for 15 min, 38.0 g (0.158 mol) of diethyl (1- For work-up the solution was diluted with 50 mL of dichlorometh-
methylpent-2-ynyl)malonate[4,13,14] was added to the clear boiling
solution within 10 min. A solid precipitated after 30 min. After 1.5
h of reflux, the solvent was removed and the product dissolved in
about 750 mL of water. With vigorous stirring 80 mL of conc.
hydrochloric acid was added to give an oil, which solidified. After
30 min of stirring, the product was filtered off and washed with
cold water. Recrystallization from 2-propanol/water (2:1) provided
colourless crystals. Yield 20.8 g (59%), m. p. 90Ϫ91.5°C. Ϫ IR
ane and the reaction was stopped with 50 mL of 0.2  hydrochloric
acid. The organic layer was dried with Na2SO4, the Na2SO4 was
filtered off and washed with 50 mL of dichloromethane. The sol-
vent was removed. The remaining oily product was chromatog-
raphed on silica gel with dichlormethane/acetonitrile (25:1). If
phosphane imine ligands were used in the catalysis, first the yellow
(2-formylphenyl)diphenylphosphane was eluted, followed by the
excess of allyl acetate and small amounts of doubly allylated MBS
(AAMBS). Then the product Methohexital was eluted. The sepa-
ration of the products was monitored by a qualitative TLC test
(silica 60, Merck; dichloromethane/acetonitrile, 25:1), which
1
(KBr): ν˜ ϭ 1754 (m), 1719 (s), 1695 (s) cmϪ1 (CϭO). Ϫ H NMR
(CDCl3, 400 MHz): δ ϭ 8.14/8.09 (2 s, 1 H, NH), 3.39Ϫ3.31 (m,
2 H, OCϪCHϪCH), 3.303/3.298 (2s, 3 H, NϪCH3), 2.102/2.100
(2 dq, 2 H, 5J ϭ 2.2 Hz, 3J ϭ 7.5 Hz, CϵCϪCH2ϪCH3), 1.444/ showed how much allyl acetate, AAMBS, Methohexital and MBS
1.440 (2d, 3 H, 3J ϭ 7.0 Hz, CHϪCH3), 1.040/1.038 (2 t, 3 H, 3J ϭ
were present in a specific fraction (yellow spots after dipping into
7.5 Hz, CϵCϪCH2ϪCH3). Ϫ 13C{1H} NMR (CDCl3, 62.9 MHz): a dilute KMnO4 solution). All the fractions, which contained Me-
δ ϭ 169.3/168.7 (HNϪCOϪCC3), 167.5/166.8 (H3CϪNϪCOϪ thohexital, were combined to prevent an enrichment of the dia-
58
Eur. J. Inorg. Chem. 1999, 51Ϫ59

Asymmetric Catalysis, 125
FULL PAPER
stereomers during the chromatography. After complete elution of
Methohexital, the unreacted starting material MBS was eluted with
dichloromethane/acetonitrile (3:1). The yield of Methohexital and
unreacted MBS was determined by weighing. After about 1Ϫ2 d
the colourless oil of Methohexital crystallized to a colourless solid.
Rα1/β2 ϭ 1.46, Rβ1/α1 ϭ 3.81, Rα2/β1 ϭ 6.88, Rα2/α1 ϭ 9.93,
Rβ1/β2 ϭ 5.36.
3,5-Diallyl-1-methyl-5-(1Ј-methylpent-2Ј-ynyl)barbituric
Acid
(AAMBS): By-product of the catalysis under standard reaction
conditions. AAMBS was separated by chromatography as de-
scribed above. Yield 0Ϫ3%, light yellow liquid. Ϫ MS (EI, 70 eV);
m/z (%): 302.2 (41) [Mϩ], 287.2 (31) [Mϩ Ϫ CH3], 273.1 (19) [Mϩ
Ϫ C2H5], 261.1 (100) [Mϩ Ϫ C3H5], 221.0 (41) [Mϩ Ϫ C6H9], 81.0
(66) [C6H9ϩ], 79.0 (41), 40.9 (84) [C3H5ϩ].
Methohexital: Maximum yield 1.19 g (94%, Table 1, entry 7), m. p.
1
61Ϫ65°C. Ϫ IR (KBr): ν˜ ϭ 1717 (s), 1690 (s) cmϪ1 (CϭO). Ϫ H
NMR (CDCl3, 400 MHz): δ ϭ 8.04/8.02 (2 s, 1 H, NH), 5.553/
3
3
3
5.547 (2 tdd, J ϭ 7.3 Hz, Jcis ϭ 10.1 Hz, Jtrans ϭ 17.0 Hz, 1 H,
3
H2CϭCHϪCH2), 5.149 (md, J ϭ 17.0 Hz, 1 H, CHϭCHHtrans),
[1]
H. Brunner, C. Krumey, J. Mol. Catal., in press.
5.084 (md, 3J ϭ 10.1 Hz, 1 H, CHϭCHcisH), 3.293/3.286 (2 s, 3
H, NϪCH3), 3.087/3.081 (2 tq, 5J ϭ 2.3 Hz, 3J ϭ 7.0 Hz, 1 H,
H3CϪCHϪCϵCϪCH2), 2.864/2.852 (2 dd, 3J ϭ 7.3 Hz, 2J ϭ 13.0
[2]
H. Brunner, I. Deml, W. Dirnberger, B. Nuber, W. Reißer, Eur.
J. Inorg. Chem. 1998, 43.
[3]
I. Deml, Dissertation, Universität Regensburg, 1998.
[4]
W. Reißer, Dissertation, Universität Regensburg, 1997.
3
2
Hz, 1 H, H2CϭCHϪCHH), 2.657/2.648 (2 dd, J ϭ 7.3 Hz, J ϭ
[5]
H. Auterhoff, J. Knabe, H.-D. Höltje, Lehrbuch der pharmazeu-
13.0 Hz, 1 H, H2CϭCHϪCHH), 2.102/2.099 (2 dq, 5J ϭ 2.3 Hz,
tischen
Chemie,
Wissenschaftliche
Verlagsgesellschaft,
3J ϭ 7.5 Hz, 2 H, HCϪCϵCϪCH2ϪCH3), 1.301/1.291 (2 d, J ϭ
Stuttgart, 1994.
3
[6]
7.0 Hz, 3 H, HCϪCH3), 1.038/1.034 (2 t, 3J ϭ 7.5 Hz, 3 H,
CϵCϪCH2ϪCH3). Ϫ 13C{1H} NMR (CDCl3, 62.9 MHz): δ ϭ
171.6/170.8 (HNϪCOϪCC3), 169.9/169.1 (H3CϪNϪCOϪCC3),
150.4 (HNϪCOϪNϪCH3), 131.2 (H2CϭCH), 120.6 (H2CϭCH),
86.4 (HCϪCϵC), 77.6 (CϵCϪCH2), 59.8 (OCϪCϪCO), 38.0
(H2CϭCHϪCH2), 36.0/35.6 (HCϪCϵC), 27.6/27.5 (NϪCH3),
16.0 (H3CϪCH), 13.8 (CH2ϪCH3), 12.1 (CϵCϪCH2). All the sig-
nals appear doubled because of the two Methohexital dia-
stereomers. Ϫ MS (EI, 70 eV); m/z (%): 262.1 (22) [Mϩ], 247.1 (30)
[Mϩ Ϫ CH3], 233.1 (32) [Mϩ Ϫ C2H5], 221.0 (81) [Mϩ Ϫ C3H5],
181.0 (31) [Mϩ Ϫ C6H9], 81.0 (100) [C6H9ϩ], 79.0 (60), 41.2 (80)
[C3H5ϩ]. Ϫ C14H18N2O3 (262.31): calcd. C 64.11, H 6.92, N 10.68;
found C 63.91, H 6.78, N 10.63.
W. Forth, D. Henschler, W. Rummel, Pharmakologie und Toxi-
kologie, Bibliographisches Institut
Mannheim, 1987, p. 484.
& F. A. Brockhaus,
[7]
[8]
S. Ebel, Synthetische Arzneimittel, VCH, Weinheim, 1979, p.
101.
C. Lehmann, Das Ultrakurznarkotikum Methohexital, Springer-
Verlag, Berlin, 1972.
[9]
Bundesverband der Pharmazeutischen Industrie, Rote Liste
1996, Editio Cantor Verlag, Aulendorf, 1996, Nr. 65010.
A. Kleemann, J. Engel, Pharmazeutische Wirkstoffe, Thieme
Verlag, Stuttgart, 1982, p. 578.
S. Ebel, H. J. Roth, Lexikon der Pharmazie, Thieme Verlag,
Stuttgart, 1987, p. 430.
W. R. Gibson, W. J. Doran, W. C. Wood, E. E. Swanson, J.
Pharmacol. Exp. Ther. 1959, 125, 23.
W. J. Doran, J. Org. Chem. 1960, 25, 1737.
Eli Lilly, United States Patent, 2872448, 03.02.1959, prior.
04.04.1956; Chem. Abstr. 1959, 53, 13185D.
L. Brandsma, Preparative Acetylenic Chemistry, Elsevier, Am-
sterdam, 1988, p. 165.
J. V. Allen, G. J. Dawson, C. G. Frost, J. M. J. Williams, S. J.
Coote, Tetrahedron 1994, 50, 799.
M. Peer, J. C. de Jong, M. Kiefer, T. Langer, H. Rieck, H.
Schell, P. Sennhenn, J. Sprinz, H. Steinhagen, B. Wiese, G.
Helmchen, Tetrahedron 1996, 52, 7547.
M. Janura, Dissertation, Universität Regensburg, 1997.
H. Brunner, W. Pieronczyk, Angew. Chem. 1979, 91, 655; An-
gew. Chem. Int. Ed. Engl. 1979, 18, 620.
[10]
[11]
[12]
[13]
[14]
GC Analysis of Methohexital: 20 mg of Methohexital was dissolved
in 1 mL of dichloromethane. Ϫ Method a: 30-m Durabond-DB-225
column (25% cyanopropyl-, 25% phenyl-, 50% methylpolysiloxane;
0.25 mm inner diameter, 0.25 µm film thickness; from J & W)
linked with an SiO2 glass connector of Restek to a 30-m Chiraldex
B-DA column (dipentylated β-cyclodextrin; 0.25 mm inner diam-
eter, 0.125 µm film thickness; from astec), column temp. 145°C, H2
pressure 2.07 bar, injector temp. 250°C, detector temp. 250°C
(flame ionisation). Retention times: α1 ϭ 145.42 min, α2 ϭ 148.78
min, β1 ϭ 151.75 min, β2 ϭ 156.64 min; dead time t0 (CH2Cl2) ϭ
[15]
[16]
[17]
[18]
[19]
[20]
H. Brunner, W. Pieronczyk, B. Schönhammer, K. Streng, I.
Bernal, J. Korp, Chem. Ber. 1981, 114, 1137.
T. P. Dang, H. B. Kagan, Chem. Commun. 1971, 481.
H. B. Kagan, T. P. Dang, J. Am. Chem. Soc. 1972, 94, 6429.
E. L. Eliel, S. H. Wilen, L. N. Mander, Stereochemistry of Or-
ganic Compounds, John Wiley & Sons, Inc., New York, 1994,
p. 395.
[21]
[22]
[23]
1.76 min. Separation factors: αα2/α1 ϭ 1.02, αβ2/α2 ϭ 1.02, αβ1/β2
ϭ
1.03. Resolutions: Rα2/α1 ϭ 1.66, Rβ2/α2 ϭ 1.53, Rβ1/β2 ϭ 2.52. Ϫ
Method b: A 30-m Chiraldex B-PM column (coated with permeth-
ylated β-cyclodextrin; 0.25 mm inner diameter; from astec), column
temp. 125°C, H2 pressure 1.75 bar, injector temp. 260°C, detector
temp. 260°C (flame ionisation). Retention times: β2 ϭ 149.62 min,
α1 ϭ 153.62 min, β1 ϭ 164.30 min, α2 ϭ 188.38 min; dead time t0
[24]
[25]
H. B. Kagan, J.-C. Fiaud, Top. Stereochem. 1988, 18, 249.
T. Tervo, F. Joo, K. T. Huikuri, I. Toth, A. Palkama, Pain 1979,
6, 57.
[26]
Publication in preparation.
(CH2Cl2) ϭ 0.64 min. Separation factors: αα1/β2 ϭ 1.03, αβ1/α1
ϭ
Received May 28, 1998
1.07, αα2/β1 ϭ 1.15, αα2/α1 ϭ 1.23, αβ1/β2 ϭ 1.10. Resolutions:
[I98166]
Eur. J. Inorg. Chem. 1999, 51Ϫ59
59