Transition metal complexes in organic synthesis, part 49. Development of novel chiral catalysts for the asymmetric catalytic complexation of prochiral cyclohexa-1,3-dienes by the tricarbonyliron fragment - Mechanism of the asymmetric catalysis and involvement of a dinuclear iron cluster

Article abstract of DOI:10.1055/s-1999-2626

Planar-chiral tricarbonyliron-diene complexes are obtained quantitatively in up to 73% ee by asymmetric catalytic complexation of prochiral cyclohexa-1,3-dienes. A series of novel chiral 1-azabuta-1,3-diene catalysts is investigated. The mechanism is shown to involve at least in some cases dinuclear iron cluster compounds which lead to a lower asymmetric induction.

Full text of DOI:10.1055/s-1999-2626

LETTER
421
Transition Metal Complexes in Organic Synthesis, Part 49.1
Development of Novel Chiral Catalysts for the Asymmetric Catalytic
Complexation of Prochiral Cyclohexa-1,3-dienes by the Tricarbonyliron
Fragment - Mechanism of the Asymmetric Catalysis and Involvement of a
Dinuclear Iron Cluster
Hans-Joachim Knölker,a* Helmut Goesmann,b Holger Hermann,a Daniela Herzberg,a Guy Rohde a
a Institut für Organische Chemie, Universität Karlsruhe, Richard-Willstätter-Allee, D-76131 Karlsruhe, Germany
b Institut für Anorganische Chemie der Universität, Engesserstraße, D-76128 Karlsruhe, Germany
Fax: +49(721)698-529; E-mail: knoe@ochhades.chemie.uni-karlsruhe.de
Received 15 January 1999
Abstract: Planar-chiral tricarbonyliron-diene complexes are ob-
tained quantitatively in up to 73% ee by asymmetric catalytic com-
plexation of prochiral cyclohexa-1,3-dienes. A series of novel chiral
1-azabuta-1,3-diene catalysts is investigated. The mechanism is
shown to involve at least in some cases dinuclear iron cluster com-
pounds which lead to a lower asymmetric induction.
Key words: asymmetric catalysis, chirality, cluster, diene complex-
es, iron compounds
Tricarbonyl(h4-cyclohexa-1,3-diene)iron complexes rep-
resent versatile starting materials for stereoselective or-
ganic synthesis.2 For applications to enantioselective
synthesis an easy access to enantiopure tricarbonyliron
complexes is required. The synthesis of optically active
tricarbonyliron complexes was usually achieved by dias-
tereoselective complexation of enantiopure dienes,3 the
separation of racemic complexes by enzymatic reactions4
or via diastereomeric intermediates,5 and the enantiose-
lective complexation of prochiral dienes by chiral tricar-
bonyliron transfer reagents.6 Using tricarbonyliron
transfer reagents mild reaction conditions can be em-
ployed for the complexation of dienes by the metal frag-
ment.7 Recently, we reported that (h4-1-azabuta-1,3-
diene)tricarbonyliron complexes represent a novel class
of highly efficient tricarbonyliron transfer reagents.8,9
They offer the special advantage that the free 1-azabuta-
1,3-dienes can be used as catalysts for the complexation of
dienes with either pentacarbonyliron or nonacarbonyldi-
iron.8,10 Using chiral azabutadienes the first asymmetric
catalytic complexation of prochiral dienes became feasi-
ble.11 In the present paper we describe the development of
novel catalysts and an optimization of the asymmetric ca-
talysis based on mechanistic considerations.
1a (90% yield) and (S)-1b (46% yield), respectively. The
1-pyranosylazadiene D-2 was prepared by acetic acid cat-
alyzed
reaction
of
2,3,4,6-tetra-O-pivaloyl-b-D-
galactopyranosylamine13 with cinnamaldehyde in dry iso-
propanol at room temperature (53% yield). Combination
of the chiral a,b-unsaturated aldehyde (1R)-(-)-myrtenal
with o-anisidine provided the second type of chiral azabu-
tadiene catalyst, (R)-3 (75% yield). The third type of
chiral azabutadiene was synthesized starting from a chiral
bicyclic ketone by a three-component-coupling consisting
of aldol condensation with p-methoxybenzaldehyde fol-
lowed by condensation with p-anisidine to the unsaturated
imine. Thus, (1R)-(+)-nopinone provided (R)-4 (76%
overall yield) and (1R)-(+)- and (1S)-(-)camphor gave
(R)-5 and (S)-5, respectively (69% overall yield).
Three different types of chiral azadienes 1-5 were synthe-
sized and used as catalysts for the asymmetric complex-
ation (Scheme 1). The first type was obtained by reaction
of chiral amines with cinnamaldehyde. Condensation of
(S)-2-amino-2’-methoxy-1,1’-binaphthyl12 and of the cor-
responding 2’-isopropoxy derivative with cinnamalde-
hyde at room temperature in the presence of 4 Å
molecular sieves afforded the axially chiral azadienes (S)-
The asymmetric catalytic complexation of 1-methoxycy-
clohexa-1,3-diene (6a) as the prochiral substrate using the
chiral azadiene catalysts (S)-1a to (S)-5 was performed
under a standard set of reactions conditions (4 eq penta-
carbonyliron, 0.25 eq of catalyst, benzene, 80°C) (Table
1). The pure enantiomers of complex 7a are stable under
Synlett 1999, No. 4, 421–425 ISSN 0936-5214 © Thieme Stuttgart · New York

422
H.-J. Knölker et al.
LETTER
clear iron cluster (R)-8 in up to 31% yield dependent on
the reaction time and the scale. The chiral hexacarbonyl-
diiron complex (R)-8 was obtained more easily by direct
reaction of (R)-3 with pentacarbonyliron using conditions
identical to those of the catalytic complexation (Scheme
2). The structural assignment for (R)-8 is based on spectral
data,15 an elemental analysis,15 and has been confirmed by
an X-ray crystal structure determination.16
In the course of our studies of the catalytic complexation
of cyclohexa-1,3-diene we recently described the first
synthesis of the hexacarbonyldiiron complex of a cinnam-
aldehyde imine.10 Hexacarbonyldiiron complexes of aryl
and heteroaryl aldehyde imines were previously report-
ed.17 We could demonstrate that the dinuclear iron cluster
resulting from the reaction of pentacarbonyliron and 1-p-
anisyl-4-phenyl-1-azabuta-1,3-diene transfers one tricar-
bonyliron fragment to cyclohexa-1,3-diene in a stoichio-
metric process, but does not show a catalytic activity in
the complexation of cyclohexa-1,3-diene with pentacar-
bonyliron.10 In the present case however, the
hexacarbonyldiiron complex (R)-8 is a catalyst for the
complexation of 1-methoxycyclohexa-1,3-diene (6a) with
pentacarbonyliron (Table 1). This difference in behavior
of the cluster (R)-8 is believed to arise from the ortho-
methoxy group of the aryl ring which may be capable of
stabilizing reactive intermediates of the catalytic cycle by
the reactions conditions.11 Thus, using a certain amount of
catalyst, quantitative yields are obtained in most of the
cases by simple extension of the reaction time while the
asymmetric induction remains constant. In order to come
to reasonable reaction times we applied 25 mol-% of the
catalyst, which can be recycled during workup. The enan-
tiomeric excess of the planar chiral tricarbonyliron com-
plex 7a was determined by separation of the two
enantiomers (S)-7a and (R)-7a at a permethylated b-cy-
clodextrin column.14
The axially chiral 1-binaphthyl azadienes (S)-1a and (S)- chelation. It was clearly shown that the catalytic activity
1b provided complex 7a in excellent yields and led pref- and the asymmetric induction of the cluster (R)-8 is lower
erentially to the formation of the R enantiomer in 25% ee than obtained with the azadiene (R)-3. Using the cluster
and 32% ee, respectively. From these results we conclud- (R)-8 as catalyst under the standard reaction conditions for
ed that an increase of the sterical demand of the 2-alkoxy 2 d, complex 7a was obtained in 47% yield with 11% ee
substituent in the binaphthyl moiety leads to a slightly im- of the R enantiomer and 70% of (R)-8 was reisolated.
proved asymmetric induction, however at the expense of
catalyst activity. A similar degree of asymmetric induc-
tion (28% ee of the R enantiomer) was observed using the
1-pyranosylazadiene D-2. In this case the catalyst activity
The characteristic structural feature of the third type of
chiral azadiene catalysts is a fixed s-cis conformation of
the azadiene because the bicyclic terpenoid framework is
annulated at the 2,3-position. Complexation of 6a with
is significantly lower. After the same reaction time of 2
pentacarbonyliron in the presence of catalytic amounts of
days, the yield of complex 7a was only 38% and the cor-
(R)-4 led quantitatively to complex 7a with 38% ee of the
responding (h4-1-azabuta-1,3-diene)tricarbonyliron com-
S enantiomer. The camphor-derived chiral azadienes pro-
vided the highest asymmetric inductions for the complex-
plex of D-2 was isolated in 31% yield as by-product.
In compound (R)-3, representing the second class of chiral ation of 6a. Reaction of 6a with pentacarbonyliron for 12
azadiene catalysts investigated for asymmetric induction, days using the (R)-camphor derived chiral azadiene (R)-5
the chiral auxiliary is annulated at the 3,4-position of the afforded quantitatively complex (S)-7a with 73% ee. Cat-
azadiene. Reaction of 6a with pentacarbonyliron and cat- alyst (S)-5 provided complex (R)-7a in the same ee as de-
alyst (R)-3 provided complex 7a after 3 d in 95% yield termined by chiral HPLC.
with 33% ee of the R enantiomer. As a by-product of this
asymmetric catalytic complexation we isolated the dinu-
Synlett 1999, No. 4, 421–425 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Transition Metal Complexes in Organic Synthesis, Part 49
423
The azadiene (S)-5 was used for the asymmetric catalytic
complexation of a series of prochiral cyclohexa-1,3-
dienes 6 with pentacarbonyliron to the tricarbonyliron
complexes 7 (Table 2). The asymmetric inductions as de-
termined by chiral HPLC were significant, although no
optimization for each single case was executed.
We propose the following mechanism to rationalize the
asymmetric catalytic complexation of prochiral cyclo-
hexadienes, e.g. 6a, by using chiral azadienes 9 (Scheme
3). Nucleophilic attack of the imine 9 at a carbonyl ligand
of pentacarbonyliron generates the (carbamoyl)tetra-
carbonyliron complex 10 which is converted to the (h3-
allyl)(carbamoyl)tricarbonyliron complex 11 by intramo-
lecular ligand displacement. Complex 11 isomerizes by a
sequence of two haptotropic migrations via the intermedi-
ate (h2-olefin)tetracarbonyliron complex 12 to the (h1-
imine)tetracarbonyliron complex 13. A thermally induced
loss of a second carbon monoxide provides the (h1-im-
ine)tricarbonyliron complex 14. The 16-electron complex
14 represents the reactive intermediate of the catalytic cy-
cle. The vacant coordination site can be filled in an in-
tramolecular process by haptotropic migration (h1 to h4)
of the metal fragment generating the (h4-azadiene)tricar-
bonyliron complex 15. Alternatively, one of the double
bonds of cyclohexadiene 6a could bind to the iron atom.
This coordination presumably takes place at the more
electron-rich double bond and leads to the 18-electron in-
termediate 16. We assume for complex 16 a trigonal-bipy-
ramidal structure with the (1-2-h)-1-methoxycyclohexa-
Synlett 1999, No. 4, 421–425 ISSN 0936-5214 © Thieme Stuttgart · New York

424
H.-J. Knölker et al.
LETTER
ganic Synthesis, Vol. 1, (Eds.: M. Beller, C. Bolm), Wiley-
VCH, Weinheim, 1998, chap. 3.13, p. 534.
1,3-diene ligand in equatorial and the (1-h)-1-azabuta-
1,3-diene ligand in axial position. At this stage of the cat-
alytic cycle the enantioselection is achieved. The coordi-
nation of the metal center to one of the two enantiotopic
faces of the methoxy-substituted double bond leads to di-
astereoisomeric complexes which are differentiated. Loss
of the azadiene from this intermediate regenerates the cat-
alyst 9. A haptotropic migration (h2 to h4) of the tricarbo-
nyliron fragment provides complex 7a. This final
isomerization occurs with retention of configuration,
since the metal remains bound to the same enantiotopic
face of the prochiral ligand. The course of this first cata-
lytic cycle is supported by previous work on (h4-1-aza-
buta-1,3-diene)tricarbonyliron complexes18 and our own
mechanistic studies.8c,10,19
Reaction of the chiral (h4-azadiene)tricarbonyliron com-
plex 15 with additional pentacarbonyliron affords the
chiral diiron cluster 17, which also represents a catalyst
for the asymmetric complexation of 6a with pentacarbon-
yliron. However, our initial results indicate that the asym-
metric induction via this second catalytic cycle is lower.
This observation is of high importance for the future de-
sign of novel more efficient chiral catalysts and for opti-
mizing the reaction conditions of the asymmetric catalytic
complexation.
(3) P. W. Howard, G. R. Stephenson, S. C. Taylor, J. Chem. Soc.
Chem. Commun. 1988, 1603; P. W. Howard, G. R. Stephen-
son, S. C. Taylor, J. Chem. Soc. Chem. Commun. 1990, 1182;
A. J. Pearson, W. J. Youngs, Organometallics 1994, 13, 4; H.-
G. Schmalz, E. Hessler, J. W. Bats, G. Dürner, Tetrahedron
Lett. 1994, 35, 4543.
(4) N. W. Alcock, D. H. G. Crout, C. M. Henderson, S. E. Tho-
mas, J. Chem. Soc. Chem. Commun. 1988, 746; J. A. S.
Howell, M. G. Palin, G. Jaouen, S. Top, H. E. Hafa, J. M.
Cense, Tetrahedron: Asymmetry 1993, 4, 1241; M. Uemura,
H. Nishimura, S. Yamada, Y. Hayashi, K. Nakamura, K.
Ishihara, A. Ohno, Tetrahedron: Asymmetry 1994, 5, 1673.
(5) A. J. Birch, B. M. R. Bandara, Tetrahedron Lett. 1980, 21,
2981; A. Monpert, J. Martelli, R. Grée, R. Carrié, Tetrahedron
Lett. 1981, 22, 1961; M. Franck-Neumann, C. Briswalter, P.
Chemla, D. Martina, Synlett 1990, 637; S. Nakanishi, H.
Yamamoto, Y. Otsuji, H. Nakazumi, Tetrahedron: Asymmetry
1993, 4, 1969.
(6) A. J. Birch, W. D. Raverty, G. R. Stephenson, Tetrahedron
Lett. 1980, 21, 197; A. J. Birch, W. D. Raverty, G. R. Stephen-
son, Organometallics 1984, 3, 1075.
(7) For a brief review on tricarbonyliron transfer reagents, see:
H.-J. Knölker, in Encyclopedia of Reagents for Organic Syn-
thesis, Vol.1, (Ed.: L. A. Paquette), Wiley, Chichester, 1995,
p. 333.
(8) a) H.-J. Knölker, P. Gonser, Synlett 1992, 517; b) H.-J. Knöl-
ker, P. Gonser, P. G. Jones, Synlett 1994, 405; c) H.-J. Knöl-
ker, G. Baum, P. Gonser, Tetrahedron Lett. 1995, 36, 8191;
d) H.-J. Knölker, H. Goesmann, P. Gonser, Tetrahedron Lett.
1996, 37, 6543.
Experimental Procedure
(9) H.-J. Knölker, G. Baum, N. Foitzik, H. Goesmann, P. Gonser,
P. G. Jones, H. Röttele, Eur. J. Inorg. Chem. 1998, 993.
(10) H.-J. Knölker, E. Baum, P. Gonser, G. Rohde, H. Röttele, Or-
ganometallics 1998, 17, 3916.
(11) H.-J. Knölker, H. Hermann, Angew. Chem. 1996, 108, 363;
Angew. Chem. Int. Ed. Engl. 1996, 35, 341.
(12) T. Hattori, M. Shijo, S. Kumagai, S. Miyano, Chem. Express
1991, 6, 335; T. Hattori, H. Hotta, T. Suzuki, S. Miyano, Bull.
Chem. Soc. Jpn. 1993, 66, 613; P. Metz, B. Hungerhoff, J.
Org. Chem. 1997, 62, 4442.
A
solution of 6a (220 mg, 2.00 mmol), pentacarbonyliron
(1.05 mL, 1.57 g, 8.01 mmol), and (R)-5 or (S)-5 (188 mg,
0.5 mmol) in anhydrous and degassed benzene (30 mL) was heated
at reflux for 12 d under an argon atmosphere. The cold reaction
mixture was filtered through a short path of Celite, which was sub-
sequently washed several times with diethyl ether. Evaporation of
the solvent and flash chromatography (pentane) of the residue on
silica gel afforded the complexes (S)-7a ([a]20 = +108.0, c = 1.00,
D
CHCl3) or (R)-7a (([a]20 = -106.2, c = 1.07, CHCl3), respectively
D
1
(498 mg, 99%) as yellow oils. H NMR (500 MHz, CDCl3): d =
(13) H. Kunz, W. Sager, Angew. Chem. 1987, 99, 595; Angew.
Chem. Int. Ed. Engl. 1987, 26, 557; W. Pfrengle, H. Kunz, J.
Org. Chem. 1989, 54, 4261; H. Kunz, W. Sager, D. Schanzen-
bach, M. Decker, Liebigs Ann. Chem. 1991, 649.
(14) H.-J. Knölker, P. Gonser, T. Koegler, Tetrahedron Lett. 1996,
37, 2405.
1.66-1.72 (m, 2 H), 1.75-1.84 (m, 1 H), 2.24 (m, 1 H), 2.95 (m, 1 H),
3.46 (s, 3 H), 5.04 (dd, J = 6.3, 4.5 Hz, 1 H), 5.32 (d, J = 4.5 Hz, 1
H).
Acknowledgement
(15) (R)-8: mp 123-124°C; [a]20 = +1418.4 (c = 1.0, CHCl3); IR
D
This work was supported by the Deutsche Forschungsgemeinschaft
(Gerhard-Hess-Förderpreis) and the Fonds der Chemischen Indu-
strie. We are grateful to Professor H. Kunz, Universität Mainz, for
a generous gift of 2,3,4,6-tetra-O-pivaloyl-b-D-galactopyranosyl-
amine. We thank the BASF AG, Ludwigshafen, for constant supply
with pentacarbonyliron.
(drift): = 3000, 2927, 2060, 2021, 2011, 1969, 1590, 1487,
1279, 1253, 1119, 1025 cm-1; 1H NMR (500 MHz, CDCl3): d
= 0.74 (d, J = 8.6 Hz, 1 H), 1.17 (s, 3 H), 1.35 (s, 3 H), 2.08
(m, 1 H), 2.43 (m, 2 H), 3.21 (d, J = 18.0 Hz, 1 H), 3.32 (d, J
= 18.0 Hz, 1 H), 3.47 (br s, 1 H), 3.83 (s, 3 H), 4.25 (br s, 1 H),
6.73 (d, J = 8.0 Hz, 1 H), 6.78 (t, J = 8.0 Hz, 1 H), 6.97 (m, 1
H), 7.08 (t, J = 8.0 Hz, 1 H); 13C NMR and DEPT (125 MHz,
CDCl3): d = 22.07 (CH3), 26.35 (CH3), 32.66 (CH2), 41.45
(C), 43.33 (CH), 43.61 (CH2), 44.72 (CH), 54.48 (CH3), 78.15
(CH2), 110.67 (CH), 117.12 (C), 120.26 (CH), 126.18 (CH),
126.78 (CH), 145.75 (C), 153.44 (C), 168.07 (C), 206.48
(CO), 209.79 (CO), 211.92 (CO), 212.01 (CO); analysis calcd.
for C23H21Fe2NO7: C 51.62, H 3.96, N 2.62; found: C 51.50,
H 4.09, N 2.97.
References and Notes
(1) Part 48: H.-J. Knölker, H. Goesmann, R. Klauss, Angew.
Chem. 1999, 111; Angew. Chem. Int. Ed. Engl. 1999, 38, in
print.
(2) For reviews, see: A. J. Pearson, Iron Compounds in Organic
Synthesis, Academic Press, London, 1994, chap. 4 and 5; R.
Grée, J.-P. Lellouche in Advances in Metal-Organic Chemi-
stry, Vol. 4, (Ed.: L. S. Liebeskind), JAI Press, Greenwich
(CT), 1995, p. 129; H.-J. Knölker in Transition Metals for Or-
(16) X-ray crystal structure analysis of (R)-8: C46H42Fe4N2O14, mo-
noclinic (twin), space group C2, a = 34.426(7), b = 8.971(2),
c = 15.005(3) Å, V = 4634.1(16) Å3, Z = 4, T = 200(2) K, rcalcd
= 1.534 g cm-3, m = 1.296 mm-1, l = 0.71073 Å, q range: 2.71-
Synlett 1999, No. 4, 421–425 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
25.00°; 7662 independent reflections; refinement method:
Transition Metal Complexes in Organic Synthesis, Part 49
425
Struchkov, J. Organomet. Chem. 1986, 302, 243; W. Imhof, J.
Organomet. Chem. 1997, 533, 31; D.-L. Wang, W.-S. Hwang,
L.-C. Liang, L.-I. Wang, L. Lee, M. Y. Chiang, Organometal-
lics 1997, 16, 3109.
full-matrix least squares on F2; final R indices [I>2s(I)]: R1 =
0.0650, wR2 = 0.1430, abs. configuration by refinement of the
Flack parameter: c = 0.00(3), maximal residual electron den-
sity: 0.905 e Å-3. Programs: G. M. Sheldrick, SHELXS-86
(Göttingen 1986), SHELXL-93 (Göttingen 1993); E. Keller,
SCHAKAL-97 (Freiburg i.Br., 1997). Crystallographic data
(excluding structure factors) for the structure reported in this
paper have been deposited with the Cambridge Crystallogra-
phic Data Centre as supplementary publication no. CCDC-
112629. Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge CB2 1 EZ,
UK (fax: (+44) 1223-336-033; e-mail: deposit@ccdc.cam.
ac.uk).
(18) G. Bellachioma, G. Cardaci, J. Chem. Soc. Dalton Trans.
1977, 2181; A. N. Nesmeyanov, M. I. Rybinskaya, L. V. Ry-
bin, N. T. Gubenko, N. G. Bokii, A. S. Batsanov, Y. T. Struch-
kov, J. Organomet. Chem. 1978, 149, 177; A. N.
Nesmeyanov, L. V. Rybin, N. A. Stelzer, Y. T. Struchkov, A.
S. Batsanov, M. I. Rybinskaya, J. Organomet. Chem. 1979,
182, 399; G. Bellachioma, G. Reichenbach, G. Cardaci, J.
Chem. Soc. Dalton Trans. 1980, 634; A. S. Batsanov, Y. T.
Struchkov, J. Organomet. Chem. 1983, 248, 101; M. W.
Kokkes, P. C. Beentjes, D. J. Stufkens, A. Oskam, J. Organo-
met. Chem. 1986, 306, 77.
(17) M. M. Bagga, P. L. Pauson, F. J. Preston, R. I. Reed, J. Chem.
Soc. Chem Commun. 1965, 543; P. E. Baikie, O. S. Mills, J.
Chem. Soc. Chem. Commun. 1966, 707; N. S. Nametkin, V. D.
Tyurin, V. V. Trusov, A. I. Nekhaev, A. S. Batsanov, Y. T.
(19) H.-J. Knölker, P. Gonser, H. Hermann, P. G. Jones, G. Rohde,
A. Schnepf, unpublished results.
Synlett 1999, No. 4, 421–425 ISSN 0936-5214 © Thieme Stuttgart · New York