вход по аккаунту


Diastereoselective Strecker Synthesis of -Aminonitriles on Carbohydrate Templates.

код для вставкиСкачать
trisaccharide-3"-uloside 19['l is obtained, which, upon normal methyl branching reactions (MeLi, MeMgCI), gives
the wrong, namely the D-arabino-configurated, terminal
component (cf. [8]).
Thus, a controlled branching reaction was necessary,
which was first worked out for the monosaccharide. Reductive debromination (h, 91%) of 10 and subsequent oxidation (b, 75%) affords the deoxygenated 3-uloside 11 .Ish1
The methylene group can be introduced either by Wittig
reaction (j,64%) to give 12171
or by Peterson olefination (k,
46%) to give the derivative 13.l7lAll attempts to carry out a
Sharpless epoxidationly"l of the exocyclic allylic alcohol
system of 13 were u n s u c c e ~ s f u l Apparently,
the chiral
elements in the saccharide derivatives 12 and 13 already
direct the epoxidation with rn-chloroperbenzoic acid (1,
83%) completely stereospecifically to give the (3R)-configurated epoxides 14l7] and 15, respectively. Reductive
opening of the epoxide results (c, 75%) in exclusive formation of the 3-C-methyl-branched and D-ribo-configurated
glycoside 16,"' the E building block.
Whereas the Wittig olefination was unsuccessful in the
case of the trisaccharide-uloside 19, the Peterson reaction
(k, 25%) gave a moderate yield of the exocyclic 3" olefin
20.[71Reaction of 20 with peracid (I, 65%) leads to exclusive, stereospecific formation of the epoxide 22,17'which,
after reductive opening (c, 68%), gives the E - D - C trisaccharide component 2l1'] of 1 . Comparison of the ' H NMR spectra of 21 and the trisaccharide having a terminal 3-C-methyl-branched D-arabino building block,LX1as
well as NOE measurements, confirms the structure of 21.
Received: January 30, 1987 [Z 2078 IE]
German version: Angew. Chem. 99 (1987) 591
[ I ] W. A. Remers: The Chemistry o/Antitumor Anfrbiolics. Wiley, New York
[2] a) J. Thiem, B. Meyer, Tetrahedron 37 (1981) 551; b) J. Thiem, G.
Schneider, Angew. Chem. 95 (1983) 54: Angew. Chem. Int. Ed. Engl. 22
(1983) 58; c) J. Thiem, G. Schneider, V. Sinnwell, Liebigs Ann. Chem.
1986. 814.
[3] J. Thiem, H. Karl, J. Schwentner, Synfhesis 1978. 696.
141 a) J. Thiem, Nuchr. Chem. Tech. Lab. 32 (1984) 6; b) ACS Symp. Ser.. in
press; c) J. Thiem, M. Gerken, J. Org. Chem. SO (1985) 954.
[ 5 ] a) J. Thiem, P. Ossowski, U. Ellermann, Liebigs Ann. Chem. 1981, 2228:
b) J. Thiem, P. Ossowski, J. Curbohydr. Chem. 3 (1984) 287; c) J. Thiem,
A. Prahst, I . Lundt, Liebigs Ann. Chem. 1986. 1044.
161 a) K. Bock, C. Pedersen, J. Thiem, Curbohydr. Res. 73 (1979) 85; b) J.
Thiem, M. Gerken, J . Curbohydr. Chem. / (1982/1983) 229; c) J. Thiem,
M . Gerken, K. Bock, Liebigs Ann. Chem. 1983. 462: d) 1. Lundt, J. Thiem,
A. Prahst, J . Org. Chem. 49 (1984) 3063.
171 Selected analytical data ([a]?;'
for c=0.2-1.3 in CH2CI2: ' H - N M R at 300
MHz in CDCl2): 4 : [a]= +63.8"; 6=4.99 (d. J(1,2)=1.2 Hz; I-H);
m.p.= 132"C.-5: [a]=+297.6O; 6=6.59 (d, J(1,2)=3.5 Hz; l-H).-7:
[f~]=+36.4";6=4.60(d,J(1,2)=8.4Hz; I-H),3.97(ddd,J(3,3-OH)=3.8
Hz: 3-H).--8: [a]=+ 3 8 S 0 ; 6(C,D,,)=4.49 (d, J(1.2)=8.6 Hz; I-H), 4.69
(d,J(1',2')=86 Hz; l'-H).-9:[a]= -7.4";6=4.5S(dd,J(1,2a)=9.2 Hz;
I-H), 4.57 (dd. J(1',2a')=9.1 Hz: l'-H).- 12: [a]=-70.6"; 6=4.62 (dd,
J(1.2a)=9.2, J(1,2e)=2.6 Hz; I-H), 4.92, 5.28 (each m, each 1 H ;
H?('=).- 13: [ a ] =- 11.7"; 6=4.90, 5.05 (each br. s, each 1 H ; H,C=):
m . p . = 9 6 T - 14: [a]=-33.7';
6=2.61, 2.72 (each d, each 1 H;
[a]=-53.7'; 6=4.79, 4.94 (each br. d, each
I H. H?C=), 3.45 ( d d z t , 5(4".4"-0H)=7.0, J(4",5")=9.0 Hz; 4"-H).2 1 : [ a ] =-58.3"; 6=4.83 (dd. J(1",2a")=9.4, J(1",2e")=2.2 Hz; I"-H),
3.69 (dq, J(4",5")=9.4, J(5",6")=6.4 Hr; 5"-H), 1.21 ( s , 3 H: 3"-CH3).22. [m]=-66.9": 6=2.47, 2.91 (each d, each I H, J(A,B)=4.5 Hz;
C H,-O(epoxide)).
[8] J. Thiem, M. Gerken, B. Schottmer, J. Weigand, Curbohydr. Res.. in
191 a ) K. B. Sharpless, S. S. Woodard, M. G. Finn, Pure Appl. Chem. 58(1983)
1923; h) in accordance with the stereochemistry of the allylic alcohol
fragment in 13. the formation of the (3R) epoxide 15 was expected to
occur via attack from the Re side with terf-butyl hydroperoxide. titanium
isopropoxide, and ( R . R ) diethyl (+)-tartrate.
Angrw Chrm. Int Ed. Engl. 26 (1987) No. 6
Diastereoselective Strecker Synthesis
of a-Aminonitriles on Carbohydrate Templates**
By Horst Kunz* and Wilfn'ed Sager
Dedicated to Professor Hans Paulsen
on the occasion of his 65th birthday
Chiral amino acids are of great biological and economic
importance. This is also true of the nonproteogenic ( R ) enantiomers, which are found as components of many microbial products, in particular antibiotics. The diastereoselective syntheses of amino acids"] are usually carried out via
organometallic intermediates: for example, the approach
via metalated bislactim ethers of cyclic dipeptides.'" Besides the isolation from protein hydrolysates and the enzymatic reactionsJ3] the Strecker synthesis is interesting from
an economic viewpoint and has already been performed
diastereoselectively with I-phenylethylamine14-"' and with
Optical inductions of 50-75% were achieved.[51The very high
inductions (98% de) obtained in one caseIx1proved to be a
consequence of fractionation^.^^^ Highly enriched or even
pure diastereomers are obtained only when one of the
diastereomers precipitates out." "'I This applies to the N 4-phenyl- 1,3-dioxan-5-yl-substituteda-methyl-a-aminonitriIes.l71
Using the concept that the chirality of the carbohydrates
can be exploited for diastereoselective reactions," 'I we
have now developed a Strecker synthesis with glycosyl amines as chiral auxiliaries."*' The 2,3,4,6-tetra-O-pivaIoyl-PD-galactopyranosylamine 3 proved to be especially effective. It can be obtained from penta-O-pivaloyl-fi-r>-galactopyranose 1 by reaction with trimethylsilyl azide/tin
tetrachloride["] to give the galactosyl azide 2 followed by
its hydrogenation."' 14]
Piv = Me&-CO
N-Glycosylaminonitriles are formed in high yields from
aldehydes, sodium cyanide, and glacial acetic acid in 2propanol, the ( a R ) diastereomers"" being favored in a ratio of 5 : 1. However, reaction times of two weeks are necessary. This results in partial anomerization, presumably at
the stage of the Schiff base intermediate, giving u p to 20%
of the a-configurated glycosylaminonitriles. The long reaction times can be avoided and the anomerization largely
suppressed by first synthesizing the Schiff bases 4 from 3
and the respective aldehyde in 2-propanol or heptane in
the presence of acetic acid. The aldimines 4, formed in
[*] Prof. Dr. H. Kunz, DipLChem. W. Sager
lnstitut fur Organische Chemie der Universitat
J.-J.-Becher-Weg 18-20, D-6500 Mainz (FRG)
This work was supported by the Fonds der Chemischen lndustrie
0 VCH Verlugsgesdlschuff mbH. 0-6940 Weinheim. I987
0570-0833/87/0606-0557 $! 02.81)/0
high yields, react with trimethylsilyl cyanidet5.'. "I i' n the
presence of zinc chloride in 2-propanol or tin tetrachloride in tetrahydrofuran (THF) to give nearly quantitative
yields of the corresponding N-glycosyl-a-aminonitriles 5
(Scheme I).
bohydrate. Charge transfer from the C = N TC orbital into
the o* orbital of the ring C - 0 bond favors the conformation A for the Schiff base 4. Conformation A is confirmed
by a strong NOE in the 'H-NMR spectrum between H-I
and the proton introduced by the aldehyde. Because of the
large axial substituent at C-4, the Lewis acids presumably
complex and attack the aldimines 4 in the conformation A
from below. This is revealed in the 'H-NMR spectrum of
the complex by a pronounced broadening of the signals of
the axial protons H-I, H-3, and H-5, which project below
the ring. The cyanide generated from the silyl cyanide in
the polar solution can then approach from the opposite,
free side, i.e., from the side of the ring oxygen, resulting in
preferential formation of the ( a R ) diastereomer. We conclude from this hypothesis that it should be possible, by
suitably changing the reaction conditions, to reverse the
sense of induction. Preliminary experiments substantiate
ZnCI, in 2-propanol
or SnCI4 in THF
4 (> 90%)
5 (quantitative)(aR) : (as) w 7 - 13: 1
Scheme I .
The reactions are complete after only a few minutes at
room temperature and in the presence of equimolar
amounts of Lewis acids as catalysts. At lower concentrations of catalyst ( 5 mol%), they require about 12 h. The
( a R ) diastereomers 5 are preferentially formed in a ratio
of about 7 : I . At lower temperatures and corresponding
longer reaction times, the diastereoselectivity can be increased to more than 10 : 1. The ratios of diastereomers
were determined by analytical HPLC (diode array detector
190-370 nm) and 400-MHz 'H-NMR spectroscopy for the
quantitatively produced mixture of products obtained by
hydrolysis. The results of the two methods are in agreement. The HPLC chromatograms reveal that the a-anomeric a-aminonitriles are only formed in trace amounts or
at most in amounts of u p to 5%. The results of the diastereoselective Strecker synthesis are shown in Table 1. Recrystallizing once from heptane affords the ( a R ) diastereomers of the N-glycosyl-a-aminonitriles 5 , which are
pure by elemental analysis and NMR spectroscopy, in
yields of between 75 and 90%.
We ascribe the high selectivity of this Strecker reaction
to steric and stereoelectronic effects arising from the car-
Interesting chiral products can be prepared from the
5, which are thereby selectively formed in high yields. For example, acid hydrolysis yields D-amino acids such as 6. These d o not contain
the L enantiomers according to thin layer chromatography
on "Chiral plate."!'"' The measured optical rotations are
also in agreement with reported
After hydrolysis,
the pivaloylated galactose moieties can be recovered to a
large extent (70-90%) by extraction with dichloromethane.
According to the procedure described here, enantiomerically pure o-amino acids and their derivatives are thus
available by using inexpensive carbohydrate templates to
Table I . Diastereoselective Strecker synthesis by reaction of the aldimines 4 with trimethylsilyl cyanide according to Scheme I .
a Anomer
Pure (aR)-5 [IS]
(yield [Yo])
5a (78)
5a (87)
5b (80)
5e (91)
5d (75)
5d (84)
5e (84)
- 10
5f (74)
- 10
5g (86)
(mol Oh)
time [h]
(uR)-5 :(aS)-5 1151
ZnCI-, (100)
SnCI4 (130)
ZnCI, (5)
S K I , ( I 30)
- 30
only (R)
ZnCl-, (5)
6.5 : I
SnCI, (130)
- 78-
- 30
SnCI, (130)
- 10
SnCI, (130)
SnCI, (130)
6.5 : I
0 VCH Verlagrqereilsch~ftmhH. 0-6940 Wemherm. 1987
S 02.50/0
Angew. Chem. Ini. Ed. Engl. 26 11987) No. 6
We wish to report here the synthesis of tricyclo['~']nona-4,6,8-triene 4, involving a novel thermal
rearrangement of 1 -chlorobicyclo[6. I .O]nona-2,4,6-triene
6 (94%)
direct the stereochemistry. The intermediacy of organometallic compounds, the exclusion of oxygen, and tedious
separations and workup are thereby avoided.
Received: February 4, 1987:
supplemented: March 18, 1987 [Z 2090 I € ]
German version: Angew. Chem. 99 (1987) 595
a ) Review: J. Martens, Top. Curr. Chem. 125 (1984) 165: b) D. A. Evans,
A. E. Weber, J . Am. Chem. Soc. 108 (1986) 6757.
U Schollkopf, Top1 Curr. Chem 109 (1983) 65.
Review: A. Kleemann, W. Leuchtenberger. B. Hoppe, H.Tanner in: UIImann 'c Encyclopedia of Industrial Chemistry. Vol. A2, VCH Verlagsgeselischaft, Weinheim 1985, p. 57.
K. Harada, T. Okawara, J . Org. Chem. 38 (1973) 707.
J. Ojirna, S. Inaba, Chem. Lett. 1975, 737.
D. M. Stout, L. A. Black, W. L. Matier. J . Org. Chem. 48 (1983) 3098.
K. Weinges, H. Blackholm, Chem. Ber. 113 (1980) 3098.
M S. Patel, M. Worsley, Can. J. Chem. 48 (1970) 1881.
K. Weinges, H. Brachmann, P. Stahnecker, H. Rodewald, M. Nixdorf,
H. Irngartinger, Liebigs Ann. Chem. 1985. 566, see the two-step procedure discussed there.
See also P. K. Subramanian, R. W. Woodard, Synth. Commun. 16 (1986)
337: the product crystallizes from methanol; the optical induction is reported for the substance recrystallized a second time from methanol.
H. Kunz, B. Miiller, D. Schanzenbach, Angew. Chem. 99 (1987) 297: Angeu'. Chem. I n t . Ed. Engl. 26 (1987) 294.
H. Kunz, W. Sager, W. Pfrengle, M. Decker, unpublished results; German patent application P3624376.0 (18 July 1986).
H. Paulsen, Z. Gyorgydeak, M. Friedmann, Chem Ber. 107 (1974)
Photochemical chlorination of cis-bicyclo[6. I.O]nona2,4,6-triene 5 with tert-butyl hypochlorite (1.5 equivalents)
in rather concentrated CCI, solution (ca. 0.8 M) at - 30°C
gave 8 in 50-70Yo yield after chromatographic purification
(SOz, pentane, OOC). Although 8 (Table 1) is fairly stable
below O'C, it underwent thermal rearrangement around
and above room temperature in nonpolar solvents to give
7 and indene 11
in an almost constant ratio of 7:3.L4J In a polar solvent
(CH,CN) or on adsorption to silica gel, 7 further rearranged to l-[(E)-2-chlorovinyl]cycloheptatriene 9."."I
Scheme 1 explains these rearrangements based on the results of trapping experiments and reported mechanistic
pathways.['.'] Notably, 7, although its formation is symrnetry allowed, is a new type of product that has not been
previously observed in the extensively studied thermal
rearrangement of bicyclo[6.1 .O]nona-2,4,6-triene~.~'IThe
chlorine atom seems to play a role in changing the thermodynamics of these rearrangements.
M. Decker, Diplomarheit. Universitat Mainz 1983.
The R / S descriptor refers to the u-C atom of the aminonitrile moiety:
the configuration of the carbohydrate moiety remains unchanged.
W Lidy, W. Sundermeyer, Chem. Ber. 106 (1973) 587.
D. A. Evans, L. K. Truesdale. G. L. Carroll, J. Chem. Sac. Chem. Commuti. 1973. 55.
H. Kunz. W. Sager, W. Pfrengle, unpublished results.
K . Giinther, J. Martens, M. Schickedanz, Angew. Chem. 96 (1984) 514;
Angeur. Chem In!. Ed. Engl. 23 (1984) 506; see brochure "Chiral plate,"
Macherey & Nagel, Diiren 1985/86.
6 : [a'];;=- 135.1" (c=0.5, I N HCI): according to [21] [a],,= - 1 8 "
( c = I . I N HCI).
K. Yokozeki, K. Mitsugi, C. Eguchi, H. Iwagami, DOS 285245 (December 14, 1978), Ajinomoto Inc., Tokyo.
Synthesis of Tricyclo[ 1731n~na-4,6,8-triene,
a Norcaradiene Structure Constrained by the
Antiaromaticity of Cyclobutadiene
By Takeshi Kawase. Masahiko Iyoda, and Masaji Oda*
Cycloheptatriene-norcaradiene tautomerism has long
attracted the attention of organic chemists."' Although cycloheptatriene itself is thermodynamically more stable
than norcaradiene by about 4 kcal mol-',[21the tautomeric
equilibrium is sensitive to the steric and electronic effects
of substituents. Klarner and co-workers have shown that
the derivatives of SH-cyclobutacycloheptene 1 exist exclusively in the norcaradiene form 2 owing to the strong antiaromaticity of cyclobutadiene.[31However, no synthesis
of the 3H-isomer 3 or its tautomer 4 has been described.
Prof. Dr M Oda, Dr. M. lyoda. Dr. T. Kawase
Department of Chemistry, Faculty of Science, Osaka University
Toyonaka. Osaka 560 (Japan)
Anyeu. Cliem In1 Ed. Engl. 26 (1987) No. 6
Scheme I.
Having 7 in hand, we expected it to be a good precursor
for 3 or 4, even though the lability of 7 made its isolation
in pure form difficult. Treatment of a mixture of 7 and 11
with tBuOK (1.5 equivalents) in tetrahydrofuran under air
at room temperature gave 4 as sole hydrocarbon product
in about 50% yield (from 8). The presence of oxygen is
advantageous for destroying indene through oxidation of
the indenide ion. Compound 4 can be formed in principle
either by direct 1,7-elimination of hydrogen chloride or by
1,2-elimination followed by valence isomerization; however, the former process seems more likely since the reluctant
1,2-elimination would result in the formation of a cyclobutadiene.I9l
The norcaradiene 4 tended to decompose giving oligomeric substances on removal of solvents. A fairly clean solution for N M R spectra could be obtained, however, by
0 VCH VerlagsgesellschaJl mbH. 0- 6930 Weinheim. 1987
0570-0833/87/0606-0559 $ 02.50/0
Без категории
Размер файла
328 Кб
diastereoselective, synthesis, strecker, aminonitriles, carbohydrate, template
Пожаловаться на содержимое документа