1,4-diaryl-1-azabuta-1,3-diene-catalyzed complexation of cyclohexa-1,3-diene by the tricarbonyliron fragment: Development of highly efficient catalysts, optimization of reaction conditions, and proposed mechanism

Article abstract of DOI:10.1021/om980290g

The 1,4-diaryl-1-azabuta-1,3-diene-catalyzed complexation of cyclohexa-1,3-diene with either nonacarbonyldiiron or pentacarbonyliron is reported to provide high yields of the tricarbonyl(η⁴-cyclohexa-1,3-diene)iron complex. This procedure enables exploitation of both tricarbonyliron fragments of nonacarbonyldiiron for the complexation of dienes for the first time. Using 12.5 mol % of 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene and optimized reaction conditions (nonacarbonyldiiron, dimethoxyethane, reflux, 16.5 h, or pentacarbonyl-iron, dioxane, reflux, 45 h), a quantitative catalytic complexation of cyclohexa-1,3-diene is feasible with both reagents. An extensive study with a broad range of 1,4-diaryl-1-azabuta-1,3-dienes shows that the efficiency of the catalysts strongly depends on the substituents of the two aryl rings. Remarkably high activities are found for those catalysts deriving from condensation of cinnamaldehyde and ortho-methoxy-substituted arylamines. A hexacarbonyldiiron complex of 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene is obtained as a byproduct of the catalytic complexation and is structurally confirmed by X-ray crystallography. A mechanism supported by the experimental findings is proposed.

Full text of DOI:10.1021/om980290g

3916
Organometallics 1998, 17, 3916-3925
1,4-Dia r yl-1-a za bu ta -1,3-d ien e-Ca ta lyzed Com p lexa tion of
Cycloh exa -1,3-d ien e by th e Tr ica r bon ylir on
F r a gm en t: Develop m en t of High ly Efficien t Ca ta lysts,
Op tim iza tion of Rea ction Con d ition s, a n d P r op osed
Mech a n ism †
Hans-J oachim Kno¨lker,*,‡ Elke Baum,§ Peter Gonser,‡ Guy Rohde,‡ and
Herbert Ro¨ttele‡
Institut fu¨r Organische Chemie der Universita¨t Karlsruhe, Richard-Willsta¨tter-Allee,
D-76131 Karlsruhe, Germany, and Institut fu¨r Anorganische Chemie der Universita¨t
Karlsruhe, Engesserstrasse, D-76128 Karlsruhe, Germany
Received April 16, 1998
The 1,4-diaryl-1-azabuta-1,3-diene-catalyzed complexation of cyclohexa-1,3-diene with
either nonacarbonyldiiron or pentacarbonyliron is reported to provide high yields of the
tricarbonyl(η4-cyclohexa-1,3-diene)iron complex. This procedure enables exploitation of both
tricarbonyliron fragments of nonacarbonyldiiron for the complexation of dienes for the first
time. Using 12.5 mol % of 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene and optimized
reaction conditions (nonacarbonyldiiron, dimethoxyethane, reflux, 16.5 h, or pentacarbonyl-
iron, dioxane, reflux, 45 h), a quantitative catalytic complexation of cyclohexa-1,3-diene is
feasible with both reagents. An extensive study with a broad range of 1,4-diaryl-1-azabuta-
1,3-dienes shows that the efficiency of the catalysts strongly depends on the substituents of
the two aryl rings. Remarkably high activities are found for those catalysts deriving from
condensation of cinnamaldehyde and ortho-methoxy-substituted arylamines. A hexacar-
bonyldiiron complex of 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene is obtained as a
byproduct of the catalytic complexation and is structurally confirmed by X-ray crystal-
lography. A mechanism supported by the experimental findings is proposed.
In tr od u ction
Brookhart6 and tricarbonylbis(η2-cis-cyclooctene)iron
developed by Grevels7 are well-established transfer
reagents for complexations of labile diene systems by
the tricarbonyliron fragment. We recently found that
the (η4-1,4-diaryl-1-azabuta-1,3-diene)tricarbonyliron com-
plexes8 represent a novel class of tricarbonyliron trans-
fer reagents.9-11 The azadiene-tricarbonyliron com-
plexes offer several advantages over the two aforemen-
tioned reagents: they can be prepared in high yields
(70-90%) by ultrasound-promoted complexation and are
stable in the air. Most remarkable is the fact that
substoichiometric amounts of the 1,4-diaryl-1-azabuta-
The tricarbonyliron complexes of conjugated dienes
have found a broad range of applications in organic
synthesis.2 The standard procedure for the synthesis
of the tricarbonyl(η4-cyclohexa-1,3-diene)iron complex
involves either thermal or photochemical reaction of the
diene with pentacarbonyliron, nonacarbonyldiiron, or
dodecacarbonyltriiron.3 Much milder reaction condi-
tions for the complexation step can be employed by
using tricarbonyliron transfer reagents.4 (η4-Benzyli-
deneacetone)tricarbonyliron investigated by Lewis5 and
† Transition-Metal Complexes in Organic Synthesis. 45.1
‡ Institut fu¨r Organische Chemie der Universita¨t Karlsruhe.
§ Institut fu¨r Anorganische Chemie der Universita¨t Karlsruhe.
(1) Part 44: Kno¨lker, H.-J .; Reddy, K. R. Tetrahedron Lett. 1998,
39, 4007.
(4) For a brief review on tricarbonyliron transfer reagents, see:
Kno¨lker, H.-J . In Encyclopedia of Reagents for Organic Synthesis;
Paquette, L. A., Ed.; Wiley: Chichester, 1995; Vol. 1, p 333.
(5) Howell, J . A. S.; J ohnson, B. F. G.; J osty, P. L.; Lewis, J . J .
Organomet. Chem. 1972, 39, 329. Domingos, A. J . P.; Howell, J . A. S.;
J ohnson, B. F. G.; Lewis, J . Inorg. Synth. 1976, 16, 103.
(6) Graham, C. R.; Scholes, G.; Brookhart, M. J . Am. Chem. Soc.
1977, 99, 1180. Brookhart, M.; Nelson, G. O. J . Organomet. Chem.
1979, 164, 193.
(2) For reviews, see: Pearson, A. J . Acc. Chem. Res. 1980, 13, 463.
Pearson, A. J . In Comprehensive Organometallic Chemistry; Wilkinson,
G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon Press: Oxford, 1982;
Vol. 8, Chapter 58. Pearson, A. J . Metallo-organic Chemistry; Wiley:
Chichester, 1985; Chapters 7 and 8. Gre´e, R. Synthesis 1989, 341.
Kno¨lker, H.-J . In Organic Synthesis via Organometallics; Do¨tz, K. H.,
Hoffmann, R. W., Eds.; Vieweg: Braunschweig, 1991; p 119. Kno¨lker,
H.-J . Synlett 1992, 371. Pearson, A. J . Iron Compounds in Organic
Synthesis; Academic Press: London, 1994; Chapters 4 and 5. Gre´e,
R.; Lellouche, J . P. In Advances in Metal-Organic Chemistry; Liebe-
skind, L. S., Ed.; J AI Press: Greenwich, CT, 1995; Vol. 4, p 129.
Kno¨lker, H.-J . In Advances in Nitrogen Heterocycles; Moody, C. J ., Ed.;
J AI Press: Greenwich, CT, 1995; Vol. 1, p 173. Iwata, C.; Takemoto,
Y. Chem. Commun. 1996, 2497.
(7) Fleckner, H.; Grevels, F.-W.; Hess, D. J . Am. Chem. Soc. 1984,
106, 2027.
(8) Otsuka, S.; Yoshida, T.; Nakamura, A. Inorg. Chem. 1967, 6,
20. Brodie, A. M.; J ohnson, B. F. G.; J osty, P. L.; Lewis, J . J . Chem.
Soc., Dalton Trans. 1972, 2031. Yin, J .; Chen, J .; Xu, W.; Zhang, Z.;
Tang, Y. Organometallics 1988, 7, 21. Danks, T. N.; Thomas, S. E.
Tetrahedron Lett. 1988, 29, 1425. Danks, T. N.; Thomas, S. E. J . Chem.
Soc., Perkin Trans. 1 1990, 761. Semmelhack, M. F.; Cheng, C. H. J .
Organomet. Chem. 1990, 393, 237.
(3) Hallam, B. F.; Pauson, P. L. J . Chem. Soc. 1958, 642. Birch, A.
J .; Cross, P. E.; Lewis, J .; White, D. A.; Wild, S. B. J . Chem. Soc. A
1968, 332. Birch, A. J .; Chamberlain, K. B.; Haas, M. A.; Thompson,
D. J . J . Chem. Soc., Perkin Trans. 1 1973, 1882.
(9) Kno¨lker, H.-J .; Gonser, P. Synlett 1992, 517.
(10) Kno¨lker, H.-J .; Gonser, P.; J ones, P. G. Synlett 1994, 405.
(11) Kno¨lker, H.-J .; Baum, G.; Foitzik, N.; Goesmann, H.; Gonser,
P.; J ones, P. G.; Ro¨ttele, H. Eur. J . Inorg. Chem. 1998, 993.
S0276-7333(98)00290-8 CCC: $15.00 © 1998 American Chemical Society
Publication on Web 08/05/1998

Complexation of Cyclohexa-1,3-diene
Organometallics, Vol. 17, No. 18, 1998 3917
Sch em e 1. In Situ P r ep a r a tion of th e Tr a n sfer
Rea gen t
1,3-dienes can be applied in order to induce the com-
plexation of 1,3-dienes with pentacarbonyliron or nona-
carbonyldiiron.9,10 Thus, catalytic complexations of
conjugated dienes by the tricarbonyliron fragment are
feasible by using the azadienes as catalysts. Herein,
we describe the full details of the development of a
highly efficient procedure for the catalytic complexation
of cyclohexa-1,3-diene by the tricarbonyliron fragment
and preliminary results with respect to the mechanism
of this reaction.
(2b) was selected for this in situ generation of the
transfer reagent, since the corresponding tricarbonyl-
iron complex provided the best yields in its preparation
and in the transfer reaction.11
Resu lts a n d Discu ssion
Heating the azadiene 2b with nonacarbonyldiiron in
tetrahydrofuran at 50 °C for 2 h provides a deep red
solution, indicating the formation of the (η4-azadiene)-
tricarbonyliron complex. After addition of the diene 1,
the solution was refluxed for 4 h and afforded the
tricarbonyl(η4-cyclohexa-1,3-diene)iron (3) in 63% yield
based on nonacarbonyldiiron. The fact that the forma-
tion of the transfer reagent and the transfer of the metal
fragment to the diene regenerating the azadiene can be
achieved in a one-pot procedure indicated that it should
be feasible to develop an azadiene-catalyzed complex-
ation.
P r elim in a r y In vestiga tion s. The excellent results
which were achieved in the preparation of the (η4-1,4-
diaryl-1-azabuta-1,3-diene)tricarbonyliron complexes and
in their application as tricarbonyliron transfer re-
agents11 suggest performing both reactions in one pot.
Such one-pot procedures for complexations of dienes
with tricarbonyliron transfer reagents were used previ-
ously for applications of the Grevels’ reagent12 and the
benzylideneacetone complexes.13 The disadvantage of
a one-pot procedure using the Grevels’ reagent is that
the addition of the diene has to take place after the
preparation of the reagent because of the photolytical
reaction conditions. Performing a one-pot procedure via
the (η4-benzylideneacetone)tricarbonyliron complex suf-
fers from the drawback of a low yield for the preparation
of the transfer reagent itself. The thermal reaction
conditions required for the transfer of the metal frag-
ment (benzene, 60 °C) are known to afford the ben-
zylideneacetone complex in only 32% yield.5
For the complexation of ergosteryl benzoate using
nonacarbonyldiiron, a catalytic effect of p-methoxyben-
zylideneacetone was claimed by Barton.13 However,
according to the Experimental Section of that paper,13
overstoichiometric amounts of p-methoxybenzylidene-
acetone and nonacarbonyldiiron were used. In agree-
ment with an earlier report by Lewis et al. (71% yield),14
Barton and co-workers reported a yield of 59-69% for
the complexation of ergosteryl acetate and ergosteryl
benzoate using stoichiometric amounts of either the
benzylideneacetone or the p-methoxybenzylideneacetone
iron complex.13 We reinvestigated the complexation of
ergosteryl benzoate with nonacarbonyldiiron, and it was
revealed that similar yields are obtained with or without
addition of catalytic amounts of p-methoxybenzylide-
neacetone. In conclusion, there is no catalytic effect of
p-methoxybenzylideneacetone in this complexation. The
beneficial effect (80% yield), observed by adding over-
stoichiometric amounts of p-methoxybenzylideneacetone
and nonacarbonyldiiron,13 has to be ascribed to an in
situ generation of the transfer reagent.
Ca ta lytic Com p lexa tion w ith Non a ca r bon yld i-
ir on . The direct uncatalyzed complexation of dienes
with nonacarbonyldiiron can be performed at lower
reaction temperatures than with pentacarbonyliron.
Therefore, we started our investigations on catalytic
complexations by using nonacarbonyldiiron as the tri-
carbonyliron source and the azadiene 2b as the catalyst
with the objective of achieving a highly efficient exploi-
tation of the metal fragments. The best results for the
uncatalyzed complexation of cyclohexa-1,3-diene with
nonacarbonyldiiron are obtained by heating a mixture
of the diene and nonacarbonyldiiron (1.2:1) in tetrahy-
drofuran at reflux for 6 h (43% yield based on nonacar-
bonyldiiron).11,15 However, using these reaction condi-
tions, only one tricarbonyliron fragment of nonacar-
bonyldiiron is exploited for the complexation, since 1
equiv of pentacarbonyliron is formed, which does not
react further at this temperature. Thus, the yield based
on iron for this complexation is only 21.5%. Better
results (referring to higher yields based on cyclohexa-
1,3-diene) are obtained by using a large excess of nona-
carbonyldiiron. However, this procedure suffers from
the drawback that large amounts of pyrophoric iron are
formed, which is hazardous during the workup. There-
fore, we investigated the catalytic complexation with
nonacarbonyldiiron by using a slight excess of cyclohexa-
1,3-diene (1) (1:1-1.5) and 12.5 mol % of the azadiene
2b as the catalyst (Scheme 2, Table 1). All specifications
of stoichiometry and yields for the azadiene-catalyzed
complexation with nonacarbonyldiiron refer to the iron
equivalents.
The one-pot procedure of the preparation of the
azadiene iron complex in stoichiometric amounts and
subsequent transfer of the metal fragment was applied
to the complexation of cyclohexa-1,3-diene (1) (Scheme
1). 1-(4-Methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene
Catalyst 2b was chosen for our initial studies, since
it is easily prepared in large quantities and because the
corresponding tricarbonyliron complex was the best
transfer reagent for the stoichiometric complexation of
cyclohexa-1,3-diene.11 The variation of solvent and
temperature of the catalytic complexation, while care-
fully controlling the other reaction parameters, showed
(12) Franck-Neumann, M.; Briswalter, C.; Chemla, P.; Martina, D.
Synlett 1990, 637. Eilbracht, P.; Hittinger, C.; Kufferath, K.; Schmitz,
A.; Gilsing, H.-D. Chem. Ber. 1990, 123, 1089.
(13) Barton, D. H. R.; Gunatilaka, A. A. L.; Nakanishi, T.; Patin,
H.; Widdowson, D. A.; Worth, B. R. J . Chem. Soc., Perkin Trans. 1
1976, 821.
(15) Kno¨lker, H.-J .; Pannek, J .-B. Unpublished results. Pannek, J .-
B. Diplomarbeit, Universita¨t Hannover, 1989; p 46. Kno¨lker, H.-J .
Habilitationsschrift, Universita¨t Hannover, 1989; p 59.
(14) Evans, G.; J ohnson, B. F. G.; Lewis, J . J . Organomet. Chem.
1975, 102, 507.

3918 Organometallics, Vol. 17, No. 18, 1998
Kno¨lker et al.
Sch em e 2. Ca ta lytic Com p lexa tion of
Cycloh exa -1,3-d ien e (1) w ith F e2(CO)9
Increasing the amount of the catalyst up to a stoichio-
metric ratio also results in a slight decrease of the yield.
In this case, analysis of the reaction mixture shows that
the azadiene 2b is still partly present as the (η4-
azadiene)tricarbonyliron complex. Presumably, a trans-
fer of the tricarbonyliron fragment from this complex
to the free azadiene ligand becomes dominant at the end
of the reaction, when the concentration of the cyclo-
hexadiene is already very low compared to the concen-
tration of the azadiene. Thus, the intermolecular
tricarbonyliron transfer from the azadiene-tricarbonyl-
iron complex to free azadiene prevents the reaction from
going to completion if the concentration of the catalyst
is too high.
Ta ble 1. Ca ta lytic Com p lexa tion of
Cycloh exa d ien e 1 w ith F e2(CO)9 Usin g 0.125 Equ iv
of Ca ta lyst 2b: Va r ia tion of Solven t, Tem p er a tu r e,
a n d Rea ction Tim ea
solvent
T [°C]
t [h]
3, yield [%]
THF
65
80
80
85
85
16.5
16.0
16.5
6.0
63
53
66
93
98
cyclohexane
benzene
DME
The tricarbonyliron complex of p-methoxybenzylidene-
acetone was shown to transfer the metal fragment much
faster to cyclohexa-1,3-diene (1) than the (η4-azadiene)-
tricarbonyliron complexes.11 However, it could be again
confirmed that p-methoxybenzylideneacetone exhibits
no catalytic effect in the complexation of cyclohexadiene
1 with nonacarbonyldiiron. Treatment of 1 with nonacar-
bonyldiiron in the presence of 2.5 mol % of p-methoxy-
benzylideneacetone in tetrahydrofuran at reflux pro-
vided complex 3 in a yield of only 16% based on iron,
which is in the same range as for the uncatalyzed
complexation.
DME
16.5
a
Reaction conditions: 1-1.5 equiv of cyclohexa-1,3-diene (1);
yield of 3 based on iron.
that working in 1,2-dimethoxyethane (DME) at reflux
provided the best results. After a reaction time of 16.5
h, the iron complex 3 could be obtained in a yield of 98%
based on the actually available equivalents of iron (the
yield refers to a large scale, providing 11.9 g of complex
3). This result demonstrated for the first time that both
tricarbonyliron fragments present in nonacarbonyldi-
iron can be transferred to the diene in a complexation
reaction and that virtually quantitative yields are
feasible. For comparison, the uncatalyzed complexation
of 1 with nonacarbonyldiiron in dimethoxyethane (re-
flux, 16 h) afforded complex 3 in a yield of 21% based
on iron (Figure 115). Besides the economic advantage
of the catalytic method, the safety aspect has to be
pointed out. Because of the quantitative transfer of both
iron fragments, the formation of pyrophoric iron and
pentacarbonyliron does not occur.
Ca ta lytic Com p lexa tion w ith P en ta ca r bon yl-
ir on . Pentacarbonyliron is more attractive as a starting
material for the complexation because it is much cheap-
er than nonacarbonyldiiron. However, the standard
procedure for the thermal complexation of cyclohexa-
diene 1 (heating with a large excess of pentacarbonyl-
iron in di-n-butyl ether at 142 °C for 18 h) provides the
iron complex 3 in only 23% yield (single-stage proce-
dure).3 Using this method the excess pentacarbonyliron
has to be removed from the product and pyrophoric iron
is formed. Moreover, only low yields are obtained for
the complexation of labile dienes due to the high reac-
tion temperature. Therefore, we tried to develop a cata-
lytic complexation with pentacarbonyliron by using reac-
tion conditions similar to those described above for nona-
carbonyldiiron. A slight excess (1.5 equiv) of cyclohexa-
diene 1 was used in order to ensure that the iron was
quantitatively transformed to complex 3. The efficient
azadiene catalyst 2b should perhaps make it feasible
to use considerably lower reaction temperatures than
for the uncatalyzed complexation. The reaction condi-
tions leading to quantitative complexation with nona-
carbonyldiiron, using catalyst 2b in dimethoxyethane
at reflux, provided complex 3 in 54% yield when applied
to the catalytic complexation with pentacarbonyliron
(Scheme 3, Table 2). The blank experiment (reaction
of 1 with pentacarbonyliron under the same reaction
conditions but without catalyst) afforded complex 3 in
only 2% yield and indicated the high efficiency of the
catalysis, which was already achieved at this stage.
F igu r e 1. Catalytic complexation of cyclohexa-1,3-diene
(1) with Fe2(CO)9 as a function of the amount of catalyst
2b. Reaction conditions: 1.5 equiv of 1, DME, 85 °C, 16 h;
yield of 3 based on iron.
To consume the iron quantitatively for the complex-
ation, we changed the conditions and performed the
reaction in dioxane at reflux. Variations of the amount
of catalyst and of the reaction time showed again that
the use of 12.5 mol % of catalyst 2b represents the ideal
compromise between low catalyst concentration and
high turnover. Finally, it was found that the reaction
of pentacarbonyliron with the diene 1 in dioxane at
In a further series of experiments, the catalytic
complexation of cyclohexadiene 1 with nonacarbonyl-
diiron was investigated by variation of the amount of
catalyst 2b using the optimized reaction conditions
found above (dimethoxyethane, reflux, 16 h) (Figure 1).
It was shown that decreasing the amount of catalyst
leads to a drastic decrease of the yield as well, down to
a value of 21% for the uncatalyzed complexation.

Complexation of Cyclohexa-1,3-diene
Organometallics, Vol. 17, No. 18, 1998 3919
Sch em e 3. Ca ta lytic Com p lexa tion of
Cycloh exa -1,3-d ien e (1) w ith F e(CO)5
bonyliron provided 50.2 g of the tricarbonyliron complex
3 (89% yield), and the catalyst 2b can be regenerated
during the workup by crystallization.
We next investigated the efficiency of the 1,4-diaryl-
1-aza-1,3-butadiene catalysts 2 as a function of the
substituents of the two aryl rings. The azadienes 2 used
as catalysts in the present work were prepared by
condensation of the arylamines 4 with the cinnamalde-
hydes 5 as previously described for the synthesis of (η4-
1-aza-1,3-butadiene)tricarbonyliron complexes.11 To be
able to compare the results, we used a standard set of
reaction conditions (dioxane, 101 °C, 14 h, 0.125 equiv
of 2) and the scale of the reaction (4.55 mmol penta-
carbonyliron, 1.5 equiv of cyclohexa-1,3-diene 1) was
calculated to afford 1.0 g of complex 3 for 100% yield.
The reaction time of 14 h was selected because catalyst
2b provides complex 3 in a yield of 50% after this time.
Thus, the more and the less reactive catalysts can be
compared based on this result (Scheme 4, Table 3).
Ta ble 2. Ca ta lytic Com p lexa tion of
Cycloh exa d ien e 1 w ith F e(CO)5: Va r ia tion of
Solven t, Tem p er a tu r e, Rea ction Tim e, a n d Am ou n t
of Ca ta lyst 2ba
reaction conditions
1 [equiv]
2b [equiv]
3, yield [%]
DME, 85 °C, 16.5 h
DME, 85 °C, 20 h
2.5
2.5
1.5
1.5
1.5
1.5
2.5
1.5
1.5
1.5
1.5
0
0.25
0
2
54
8
dioxane, 101 °C, 45 h
dioxane, 101 °C, 5 h
dioxane, 101 °C, 5 h
dioxane, 101 °C, 5 h
dioxane, 101 °C, 20 h
dioxane, 101 °C, 63 h
dioxane, 101 °C, 14 h
dioxane, 101 °C, 37 h
dioxane, 101 °C, 45 h
0.010
0.050
0.125
0.25
0.025
0.125
0.125
0.125
5
11
21
70
67
50
91
99
The blank experiment using the reaction conditions
described above without catalyst provided complex 3 in
0.7% yield. Performing the complexation in the pres-
ence of 12.5 mol % of p-methoxybenzylideneacetone
afforded complex 3 in 4.2% yield. This outcome confirms
again that the oxadiene can be used for an in situ
formation of the tricarbonyl(η4-oxadiene)iron complex
followed by transfer of the metal fragment to 1 but
shows no catalytic effect in this reaction. In contrast
to this result, application of 12.5 mol % of the 1,4-diaryl-
1-azabuta-1,3-dienes 2a -p led to a significant increase
of the yield. The nitro derivative 2q, which could not
be transformed to the corresponding tricarbonyliron
complex for the stoichiometric complexation,11 exhibited
no catalytic effect under the applied reaction conditions.
Using the parent compound 1,4-diphenyl-1-azabuta-1,3-
diene (2a ) as the catalyst, the iron complex 3 was
obtained in 41% yield. For comparison of the activities
of the various catalysts, a standard was required and,
therefore, the relative activity of this catalyst was
defined as 1.0. It was found that acceptor substituents
in the aryl rings generally lead to a drastic decrease of
the catalytic activity, e.g., 2e (0.54), 2l (0.46), 2o (0.32),
and 2q (0). The high activity of catalyst 2p (2.02) is
explained by the chelation effect of the ortho-methoxy
group (see below) which is overriding. On the other
hand, donor substituents in positions appropriate to lead
to an increase of the electron density at the imine
a
Reaction conditions: 1.5 equiv of cyclohexa-1,3-diene (1); yield
of 3 based on iron.
reflux for 45 h in the presence of 12.5 mol % of catalyst
2b proceeds quantitatively, and the tricarbonyliron
complex 3 could be isolated in 99% yield based on
pentacarbonyliron. For comparison, the blank experi-
ment (identical conditions without catalyst) provided
complex 3 in 8% yield.
All reactions presented in Table 2 were based on a
scale of 2.0 g (10.2 mmol) of pentacarbonyliron. The
scale of the catalytic complexation of cyclohexadiene 1
using the optimal conditions (dioxane, 101 °C, 0.125
equiv of 2b) was increased up to 50 g (255 mmol) of
pentacarbonyliron (see Experimental Section). On this
scale, a large volume of carbon monoxide (more than
11 L) is formed, which mostly displaces the inert gas
(argon) using conventional laboratory equipment (flask
with reflux condenser). The resulting increase of the
partial pressure of carbon monoxide leads to an equi-
librium, which stops the catalytic cycle (see proposed
mechanism, below) and is responsible for the observed
drastic decrease of the turnover. This problem can be
simply solved by directing a light flow of argon through
the apparatus and increasing the reaction time to 4
days. By using this modified procedure, the complex-
ation of 1 on a scale of 50.0 g (255 mmol) of pentacar-
Sch em e 4. Syn th esis of th e Aza d ien e Ca ta lysts 2, a n d Ca ta lytic Com p lexa tion of Cycloh exa -1,3-d ien e (1)
w ith F e(CO)5 to th e Tr ica r bon ylir on Com p lex 3

3920 Organometallics, Vol. 17, No. 18, 1998
Kno¨lker et al.
Ta ble 3. Syn th esis of th e 1-Aza bu ta -1,3-d ien es 2, a n d Yield of th e Ca ta lytic Com p lexa tion of
Cycloh exa -1,3-d ien e (1) w ith P en ta ca r bon ylir on Usin g 0.125 Equ iv of 2a
R1
R2
R3
R4
R5
R6
R7
2, yield [%]
3, yield [%]
relative activityb
no catalyst
0.7
41
50
83
66
22
90
33
86
31
51
91
19
74
71
13
83
0
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
H
H
OMe
H
H
OMe
H
OMe
H
H
OMe
H
OMe
H
H
OMe
H
H
H
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
H
H
OMe
H
NMe2
CF3
OMe
OMe
OMe
H
OMe
H
CF3
OMe
OMe
H
H
H
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OMe
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OMe
OMe
OMe
OMe
OMe
H
CN
CN
NO2
H
H
H
H
H
H
H
H
H
H
H
H
H
OMe
H
H
H
82
100
87
40
72
81
79
73
88
95
79
59
80
86
78
99
75
≡ 1.00
1.22
2.02
1.61
0.54
2.20
0.80
2.10
0.76
1.24
2.22
0.46
1.80
1.73
0.32
2.02
0
OMe
OMe
a
b
Reaction conditions: 1 equiv of pentacarbonyliron, 1.5 equiv of cyclohexadiene 1, 0.125 equiv of 2, dioxane, 101 °C, 14 h. Relative
activity of the 1-azadiene catalysts 2 as compared to 2a (≡1.0).
Ta ble 4. Ca ta lytic Com p lexa tion of
Cycloh exa -1,3-d ien e (1) w ith P en ta ca r bon ylir on
(1 Equ iv): Va r ia tion of Ca ta lyst 2 a n d Rea ction
Tim ea
nitrogen atom result in a remarkably increased catalyst
activity. Thus, the activity of catalyst 2b (1.22) with a
para-methoxy group in the 1-aryl ring is higher com-
pared to 2a because of the inductive effect of the
methoxy group. Stronger donor groups in this position,
as the N,N-dimethylamino substituent in 2d , lead to an
even higher catalyst activity (1.61). The chemical shift
3, yield [%]
2
5 h
14 h
37 h
b
c
f
21
29
34
46
50
83
90
86
91
92
90
1
of the imine proton (H-C2) in the H NMR spectra of
the azadienes 2 represents a measure of the electron
density at the imine nitrogen atom and, thus, its
nucleophilicity.11 A much more significant increase in
catalyst activity is observed with an ortho-methoxy
group in the 1-aryl ring as for 2c (2.02). This positive
effect on catalyst activity caused by an ortho-methoxy
group is ascribed to an internal displacement of a carbon
monoxide ligand and stabilization of an intermediate
16-electron species due to chelation by coordination of
an oxygen lone pair of electrons to iron (see proposed
mechanism, below). Similar high activities are also
observed for the other catalysts with an ortho-methoxy
group in the 1-aryl ring: 2f (2.20), 2h (2.10), 2k (2.22),
2m (1.80), and 2p (2.02). The high relative catalyst acti-
vity of 2f (2.20), having an ortho- and a para-methoxy
group in the 1-aryl ring, shows that the inductive and
the chelation effect on catalyst activity are additive:
compare with the relative catalyst activities of 2a (1.00),
2b (1.22), and 2c (2.02). Thus, on catalytic complexation
of 1 with catalyst 2f, complex 3 is already obtained in
90% yield after a reaction time of 14 h.
h
a
Reaction conditions: 1.5 equiv of diene 1, dioxane, 101 °C;
0.125 equiv of 2.
Sch em e 5. Syn th esis of th e Hexa ca r bon yld iir on
Com p lex 7b
To emphasize the influence of an ortho-methoxy group
on the efficiency of the catalyst, the azadiene 2b was
compared with the azadienes 2c, 2f, and 2h using the
standard reaction conditions (dioxane, 101 °C, 1.5 equiv
of 1, 0.125 equiv of 2) at three different reaction times
(Table 4).
The higher reactivity of the three latter catalysts is
already visible after a reaction time of 5 h and becomes
even more evident after 14 h. The results of the
catalytic complexation after 37 h demonstrate again
that using a certain catalyst, the maximal accessible
yield is independent of the turnover, which has been
achieved at shorter reaction times.
catalytic complexations, especially on large scale, it was
observed that several orange and red bands of byprod-
ucts were formed with only one of them being the (η4-
azadiene)tricarbonyliron complex. We thought that
these byproducts might represent important intermedi-
ates of the catalytic cycle and that the determination
of their structures would perhaps help us to understand
the mechanism of the catalytic complexation. Therefore,
we attempted a thermal complexation of the azadiene
2b with pentacarbonyliron using conditions similar to
those used for the catalytic complexation of cyclohexa-
diene 1 (Scheme 5).
Mech a n ism of th e Ca ta lytic Com p lexa tion . Dur-
ing chromatographic workup of the crude products of
The reaction of catalyst 2b with an excess of penta-
carbonyliron provided the (η4-azadiene)tricarbonyliron

Complexation of Cyclohexa-1,3-diene
Organometallics, Vol. 17, No. 18, 1998 3921
complex 6b in 47% yield along with the hexacarbonyl-
diiron complex 7b in 5% yield. Both products could be
isolated by chromatographic separation in the glovebox.
From this result we concluded that the dinuclear iron
complex 7b is possibly formed by complexation of the
mononuclear iron complex 6b. In fact, heating of com-
plex 6b with pentacarbonyliron in dioxane at 90 °C
afforded the hexacarbonyldiiron complex 7b in 8% yield
along with 86% of recovered 6b. Alternatively, heating
of complex 6b with nonacarbonyldiiron in tetrahydro-
furan at reflux provided the dinuclear complex 7b in
6% yield along with 86% of recovered starting material
(Scheme 5).
F igu r e 2. 13C NMR spectrum of 7b at -60 °C in the metal
carbonyl region (125 MHz, toluene-d8).
The (η4-1-azabuta-1,3-diene)tricarbonyliron complex
6b was previously described and fully characterized.11
The structural assignment for the dinuclear complex 7b
is based on the following spectral data. The mass
spectrum of 7b shows the peak for the molecular ion at
m/z 517 followed by sequential loss of six carbon
monoxide ligands and two iron atoms. This typical
fragmentation pattern and the high resolution of the
peak for the molecular ion suggest that 7b represents
a hexacarbonyldiiron complex. The FT-IR spectrum
exhibits six bands in the region of 1954-2062 cm-1
,
which are characteristic of terminal carbonyl ligands.
On the other hand, no bands are found in the region of
1700-1900 cm-1, indicating that no bridging carbonyls
1
are present in the complex. In the H NMR spectrum
(in C6D6), a broad 2-proton singlet at 3.65 ppm and a
singlet for 1 proton at 4.26 ppm occur. These signals
are assigned to the three protons bound at the two
central carbon atoms (C2, C3) of the azadiene moiety.
In the 1H NMR spectrum (in C6D6) of the tricarbonyliron
complex 6b, the chemical shifts of the signals for the
azadiene protons are significantly different (2-H, δ 6.41;
3-H, δ 5.03; 4-H, δ 3.45). The chemical shifts of the
signals for the aromatic protons are almost identical in
both complexes. The major change of the signals for
the protons at the azadiene unit indicates that the
hexacarbonyldiiron fragment is coordinated to the aza-
diene. A 125 MHz 13C NMR spectrum of the dinuclear
complex 7b in toluene-d8 at 300 K shows the signals
for all 16 carbon atoms of the former azadiene ligand
but only one broad singlet at 210.96 ppm for the
carbonyl ligands. The signals for the carbon atoms of
the two aryl substituents of 7b show only a slight
change of the chemical shift compared to those of the
mononuclear complex 6b.11 However, a major change
of the chemical shift is observed for two of the signals
for the azadiene unit of complex 6b. In the 13C NMR
spectrum of complex 6b, the three CH signals for C2,
C3, and C4 appear at 105.03, 73.81, and 62.48 ppm.11
The 13C NMR and the DEPT spectrum of the dinuclear
complex 7b exhibit a signal at 76.10 ppm for a CH (C3)
but also a signal at 75.57 ppm for a CH2 (C2) and at
188.02 ppm for a C (C4). From these data it was con-
cluded that a 1,3-hydrogen shift from C4 to C2 generat-
ing a methylene group at C2 has occurred. The novel
hexacarbonyldiiron complex was assigned structure 7b
with a coordination of C4 to both iron atoms. This
arrangement rationalizes the drastic downfield shift
observed for this carbon atom in the 13C NMR spectrum.
F igu r e 3. Dynamic 13C NMR spectra of 7b in the metal
carbonyl region (125 MHz, toluene-d8).
Obviously, the fluxionality of the hexacarbonyldiiron
complex 7b leading to equivalency of all six carbonyl
ligands at room temperature is sufficiently slow on the
NMR time scale at low temperature, leading to a rigid
structure with six distinguishable carbonyl ligands. The
variable-temperature NMR spectra reveal that on warm-
ing complex 7b undergoes two successive coalescence
events (Figure 3).
We determined the activation barriers for the two
fluxional processes of complex 7b by dynamic 13C NMR
spectroscopy at 125 MHz in deuterated toluene. The
coalescence temperatures for the low-temperature pro-
cess (TC1 ) 277 K) and for the high-temperature process
(TC2 ) 287 K) could be determined from these variable-
temperature NMR spectra with a tolerance of (10 K,
and the values for δν were taken from the low-temper-
ature spectrum at 213 K (δν1 ) 535 Hz and δν2 ) 859
Hz). Using the equation16 ∆G‡ ) RTC [22.96 + ln(TC/
δν)], the free enthalpies of activation for the intramo-
lecular exchange of the carbonyl ligands at the two
‡
tricarbonyliron fragments could be calculated: ∆G1
)
12.6 ( 0.5 kcal/mol and ∆G2‡ ) 13.1 ( 0.5 kcal/mol. For
A 13C NMR spectrum of complex 7b at -60 °C shows
the six signals for the six carbonyl ligands (Figure 2).
(16) Gu¨nther, H. NMR-Spektroskopie, 3rd ed.; Thieme Verlag:
Stuttgart, 1992; p 310.

3922 Organometallics, Vol. 17, No. 18, 1998
Kno¨lker et al.
Ta ble 5. Cr ysta l Da ta a n d Str u ctu r e Refin em en t
for 7b
Crystal Data
empirical formula
C22H15Fe2NO7
fw
517.04
cryst size [mm]
cryst syst
space group
a [Å]
b [Å]
c [Å]
0.96 × 0.4 × 0.36
triclinic
P1h (No. 2)
7.8296(11)
11.475(2)
12.929(2)
77.57(2)
73.42(2)
73.657(13)
1056.7(3)
2
R [deg]
â [deg]
γ [deg]
vol [Å3]
F igu r e 4. Molecular structure of 7b (ORTEP plot at the
50% probability level, arbitrary numbering).
Z
density (calcd) [Mg/m3]
1.625
1.378
abs coeff [mm-1
]
comparison, the activation barrier for the turnstile rota-
tion of the tricarbonyliron fragment in complex 6b was
previously determined: ∆G‡ ) 13.5 ( 0.3 kcal/mol.11,17
We propose that the fluxional behavior of the hexacar-
bonyldiiron complex 7b is based on two successive coa-
lescence events, similar to the dynamic process reported
for related hexacarbonyldiruthenium complexes.18
Complex 7b is the first example of a hexacarbonyl-
diiron complex of a cinnamaldehyde imine. Related
hexacarbonyldiiron complexes of imines from aryl and
heteroaryl aldehydes19 and hexacarbonyldiruthenium
complexes of imines from R,â-unsaturated aldehydes
were previously reported.18 The structure for complex
7b was additionally confirmed by an X-ray crystal
structure determination (Figure 4, Tables 5 and 6).20
Crystallization of the dinuclear complex 7b from
pentane/ether (15:1) at -30 °C under an argon atmo-
sphere in a glovebox provided red-brown crystals suit-
able for X-ray analysis. The molecular structure of the
cluster 7b, which crystallizes in the space group P1h, is
shown in Figure 4. The Fe-Fe bond length is 2.47 Å
and almost the same as that found in nonacarbonyldi-
iron (2.46 Å). In the X-ray analysis of 7b, all hydrogen
atoms were determined by Fourier difference calculation
and refined isotropically. The structure determination
confirms the 1,3-hydrogen shift from the â-position of
the R,â-unsaturated imine to the carbon atom of the
imine group. This hydrogen shift results in a formal
reduction of the imine bond. Thus, the dinuclear cluster
7b could be formally described as follows: Fe1, which
has a formal oxidation state of 0, is η2-coordinated to
the carbon-carbon double bond and η1-coordinated to
the nitrogen atom. Fe2 is bound to the vinyl carbon
atom C13 and to the nitrogen atom and has a formal
oxidation state of +II.
F (000)
524
Data Collection
diffractometer
radiation
temp [K]
monochromator
θ range for data collection [deg] 1.66-25.00
index ranges
STOE STADI-4
Mo KR (λ ) 0.710 69 Å)
150(1)
graphite
-8 e h e 9, -13 e k e 13,
-15 e l e 15
5941
3721 [R(int) ) 0.0108]
no. of reflns collected
no. of indep reflns
Solution and Refinement
solution method
refinement method
direct methods (SHELXS-86)
full-matrix least-squares on F
(SHELXL-93)
3719/0/350
10.6:1
2
data/restraints/params
data:parameter ratio
goodness-of-fit on F
2
1.138
final R indices [I > 2σ(I)]
R indices (all data)
ext coeff
R1 ) 0.0250, wR2 ) 0.0687
R1 ) 0.0266, wR2 ) 0.0706
0.0099(10)
largest diff peak and hole [e/Å3] 0.615 and -0.360
Ta ble 6. Bon d Len gth s [Å] for 7b
Fe(1)-C(3)
Fe(1)-C(1)
Fe(1)-C(2)
Fe(1)-N
Fe(1)-C(13)
Fe(1)-C(14)
Fe(1)-Fe(2)
Fe(2)-C(5)
Fe(2)-C(6)
Fe(2)-C(4)
Fe(2)-N
Fe(2)-C(13)
C(1)-O(1)
C(2)-O(2)
C(3)-O(3)
C(4)-O(4)
C(5)-O(5)
C(6)-O(6)
C(7)-C(8)
1.791(2)
1.802(2)
1.812(2)
1.979(2)
2.132(2)
2.169(2)
2.4704(6)
1.778(2)
1.790(2)
1.834(2)
1.971(2)
1.975(2)
1.139(3)
1.135(3)
1.145(2)
1.134(2)
1.142(3)
1.141(2)
1.386(3)
C(7)-C(12)
C(8)-C(9)
1.398(3)
1.389(3)
1.385(3)
1.388(3)
1.402(3)
1.490(3)
1.391(3)
1.501(3)
1.483(2)
1.444(2)
1.373(3)
1.392(3)
1.391(3)
1.376(3)
1.373(2)
1.389(3)
1.378(3)
1.417(3)
C(9)-C(10)
C(10)-C(11)
C(11)-C(12)
C(12)-C(13)
C(13)-C(14)
C(14)-C(15)
C(15)-N
N-C(16)
C(16)-C(17)
C(16)-C(21)
C(17)-C(18)
C(18)-C(19)
C(19)-O(7)
C(19)-C(20)
C(20)-C(21)
O(7)-C(22)
(17) Kno¨lker, H.-J .; Goesmann, H.; Gonser, P. Tetrahedron Lett.
1996, 37, 6543.
(18) Mul, W. P.; Elsevier, C. J .; Polm, L. H.; Vrieze, K.; Zoutberg,
M. C.; Heijdenrijk, D.; Stam, C. H. Organometallics 1991, 10, 2247.
Imhof, W. J . Chem. Soc., Dalton Trans. 1996, 1429.
We next examined the question of whether the
tricarbonyliron complex 6b and the hexacarbonyldiiron
complex 7b are intermediates of the catalytic complex-
ation of cyclohexadiene 1 with pentacarbonyliron using
2b as the catalyst. The (η4-1-azabuta-1,3-diene)iron
complex 6b was shown to represent a highly efficient
tricarbonyliron transfer reagent in the stoichiometric
reaction with 1.9-11 It was, therefore, expected that
complex 6b is an important intermediate of the catalytic
complexation. In fact, a catalytic complexation of 1 to
afford 3 could be achieved as well by using the azadiene
complex 6b as the catalyst under conditions identical
to those described above for 2b (Scheme 6).
(19) Bagga, M. M.; Pauson, P. L.; Preston, F. J .; Reed, R. I. J . Chem.
Soc., Chem. Commun. 1965, 543. Baikie, P. E.; Mills, O. S. J . Chem.
Soc., Chem. Commun. 1966, 707. Nametkin, N. S.; Tyurin, V. D.;
Trusov, V. V.; Nekhaev, A. I.; Batsanov, A. S.; Struchkov, Y. T. J .
Organomet. Chem. 1986, 302, 243. Imhof, W. J . Organomet. Chem.
1997, 533, 31. Imhof, W. J . Organomet. Chem. 1997, 541, 109. Wang,
D.-L.; Hwang, W.-S.; Liang, L.-C.; Wang, L.-I.; Lee, L.; Chiang, M. Y.
Organometallics 1997, 16, 3109.
(20) Crystallographic data (excluding structure factors) for the
structure reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as a supplementary publica-
tion. Copies of the data can be obtained free of charge on application
to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax (+44)
1223336-033; e-mail deposit@ccdc.cam.ac.uk).

Complexation of Cyclohexa-1,3-diene
Organometallics, Vol. 17, No. 18, 1998 3923
Sch em e 6. Com p lexa tion of Cycloh exa d ien e 1
Usin g Com p lexes 6b or 7b
following mechanism for the azadiene-catalyzed com-
plexation of cyclohexadiene with pentacarbonyliron
(Scheme 7). Nucleophilic attack of the imine nitrogen
of the azadiene 2 at one of the carbonyl ligands of
pentacarbonyliron generates the (carbamoyl)tetracar-
bonyliron complex 8. Subsequent intramolecular dis-
placement of a carbonyl ligand by the carbon-carbon
double bond of the azadiene provides the (η3-allyl)-
(carbamoyl)tricarbonyliron complex 9, which isomerizes
to the [(3-4-η)-1-azabuta-1,3-diene]tetracarbonyliron
complex 10.21,22 Haptotropic migration of the tetracar-
bonyliron fragment affords the [(1-η)-1-azabuta-1,3-
diene]tetracarbonyliron complex 11.21,22 Thermally in-
duced loss of a second carbonyl ligand at the stage of
complex 11 generates the [(1-η)-1-azabuta-1,3-diene]-
tricarbonyliron complex 12, which represents the reac-
tive 16-electron intermediate of the catalytic cycle.
Complex 12 is in equilibrium with the stable 18-electron
[(1-4-η)-1-azabuta-1,3-diene]tricarbonyliron complex 6
by haptotropic migration of the tricarbonyliron frag-
ment.23,24 Further reaction of complex 6 with penta-
carbonyliron provides the hexacarbonyldiiron complex
7 and thus, as shown above, leads to loss of catalyst.
On the other hand, the vacant coordination site of the
16-electron complex 12 can be occupied by η2-coordina-
tion of cyclohexadiene 1 to afford the [(1-η)-1-azabuta-
1,3-diene]tricarbonyl[(1-2-η)cyclohexa-1,3-diene]iron com-
plex 13, which represents the crucial intermediate of
the catalytic complexation. We propose for the 18-
electron complex 13 a trigonal-bipyramidal structure
with the η2-coordinated cyclohexa-1,3-diene in an equa-
torial and the η1-coordinated 1-azabuta-1,3-diene in an
axial position. Loss of the azadiene from complex 13
regenerates the catalyst 2, and subsequent haptotropic
migration of the metal fragment provides the product
3.
No catalytic activity was detected for the hexacarbon-
yldiiron complex 7b in the complexation of cyclohexa-
diene 1. However, it was found that complex 7b acts
as a tricarbonyliron transfer reagent in a stoichiometric
reaction with 1 in benzene at reflux. In this transfer
reaction the hexacarbonyldiiron complex 7b delivers
only one of the two tricarbonyliron fragments to form
complex 3 in 64% yield. Obviously, only the tricarbon-
yliron fragment which is η2-coordinated to the carbon-
carbon double bond and η1-coordinated to the nitrogen
atom can be transferred in this process. Neither the
mononuclear tricarbonyliron complex 6b nor the free
azadiene 2b could be isolated from the reaction mixture
after the transfer was completed. The release of one
tricarbonyliron fragment is associated with decomposi-
tion of the residual iron complex. From these observa-
tions it can be concluded that the redox reaction
between the imine group and one of the iron atoms is
irreversible and that the formation of the hexacarbon-
yldiiron complex is a side reaction leading to destruction
of the catalyst. However, using the optimized reaction
conditions for the catalytic complexation described
above, this pathway is in general not significant.
On the basis of the experiments described above and
additional mechanistic investigations,21 we propose the
The proposed mechanism is supported by the follow-
ing experimental observations: (1) The azadiene-
catalyzed complexation of cyclohexa-1,3-diene with pen-
Sch em e 7. P r op osed Mech a n ism for th e Ca ta lytic Com p lexa tion

3924 Organometallics, Vol. 17, No. 18, 1998
Kno¨lker et al.
tacarbonyliron occurs quantitatively at a temperature
(dioxane, 101 °C) that is 40 °C lower than that used for
the conventional uncatalyzed complexation (di-n-butyl
ether, 143 °C). This fact strongly suggests that the
catalytic cycle is started by a nucleophilic attack of the
azadiene nitrogen at the carbon monoxide ligand (“as-
sociative mechanism”) rather than by an initial loss of
carbon monoxide and formation of a 16-electron tetra-
carbonyliron fragment (“dissociative mechanism”). (2)
The imine of benzaldehyde and p-anisidine does not
catalyze the complexation of cyclohexa-1,3-diene with
pentacarbonyliron. This result demonstrates that the
additional carbon-carbon double bond present in cata-
lyst 2b is required for the internal displacement of the
first carbon monoxide ligand. (3) The turnover of the
various azadiene catalysts 2 in the catalytic complex-
ation of 1 (see Table 3) does not correlate with the
transfer rate of the corresponding (η4-azadiene)tricar-
bonyliron complexes in the stoichiometric reaction with
1 (compare Table 1 in ref 11). In the catalytic complex-
ation, the azadienes 2c, 2f, 2h , 2k , 2m , and 2p resulting
from ortho-methoxy-substituted arylamines represent
the most efficient catalysts. The high activities observed
for these catalysts are believed to result from internal
displacement of a carbon monoxide ligand by the ortho-
methoxy group at the stage of the [(1-η)-1-azabuta-1,3-
diene]tetracarbonyliron complex 11. Stabilization of the
16-electron intermediate 12 is possible by coordination
of the oxygen to the iron atom, thus forming an
18-electron chelate complex. (4) Previous kinetic studies
on ligand exchange reactions at [(1-4-η)-1-azabuta-1,3-
diene]tricarbonyliron complexes 6 with triphenylphos-
phine also suggest a 16-electron [(1-η)-1-azabuta-1,3-
diene]tricarbonyliron complex 12 as an intermediate.25
(5) Asymmetric inductions are feasible in the catalytic
complexation of prochiral cyclohexa-1,3-dienes by using
chiral azadiene catalysts.26 This result suggests an
intermediate where both the cyclohexa-1,3-diene and
the azadiene catalyst are bound to the same metal
center as proposed for complex 13.
for the synthesis of the corresponding tricarbonyliron
complex 3. It is environmentally friendly and superior
from the safety point of view since no residual pentacar-
bonyliron is left and no pyrophoric iron is formed. The
formation of the hexacarbonyldiiron complex 7b is
shown to represent a potential pathway leading to
destruction of catalyst 2b. An extension of the present
method to the catalytic complexation of other 1,3-dienes
is feasible.9 The mechanism which is proposed for the
catalytic cycle opens up the way for the development of
novel chiral azadienes for more efficient asymmetric
catalytic complexations of prochiral cyclohexa-1,3-
dienes.
Exp er im en ta l Section
All reactions were carried out using anhydrous and degassed
solvents under an argon atmosphere. Flash chromatogra-
phy: Baker or Merck silica gel (0.03-0.06 mm). Melting
points: Bu¨chi 535. IR spectra: Bruker IFS 88 (FT-IR); ν˜ in
cm-1
.
1H NMR and 13C NMR spectra: Bruker AM-400 and
Bruker DRX-500; internal standard, tetramethylsilane or the
signal of the deuterated solvent; δ in ppm, coupling constants
J in hertz. Mass spectra: MAT-90, at an ionization potential
of 70 eV. Elemental analysis: Heraeus CHN-Rapid.
P r ep a r a tion of Tr ica r bon yl[(1-4-η)-cycloh exa -1,3-d i-
en e]ir on (3) by Com p lexa tion of Cycloh exa -1,3-d ien e (1)
via in Situ P r ep a r a tion of Stoich iom etr ic Am ou n ts of
th e Tr ica r bon ylir on Tr a n sfer Rea gen t 6b. A solution of
nonacarbonyldiiron (2.00 g, 5.50 mmol) and 1-(4-methoxyphen-
yl)-4-phenyl-1-azabuta-1,3-diene (2b) (1.40 g, 5.90 mmol) in
tetrahydrofuran (20 mL) was heated at 50 °C for 2 h.
Cyclohexa-1,3-diene (1) (473 mg, 0.56 mL, 5.9 mmol) was
added, and the reaction mixture was stirred for 4 h at reflux.
The cold mixture was filtered over a short path of Celite, which
was subsequently washed several times with diethyl ether. The
solvent was evaporated, and the residue was subjected to flash
chromatography (pentane) on silica gel to afford the tricar-
bonyliron complex 3 (759 mg, 63%) as a yellow oil. 1H NMR
(400 MHz, C6D6): δ 1.18 (br d, J ) 12 Hz, 2 H), 1.45 (br dt, J
) 12, 2 Hz, 2 H), 2.74 (dd, J ) 5, 2 Hz, 2 H), 4.63 (dd, J ) 5,
3 Hz, 2 H). For further spectroscopic data, see ref 11.
P r ep a r a tion of Tr ica r bon yl[(1-4-η)-cycloh exa -1,3-d i-
en e]ir on (3) by Ca ta lytic Com p lexa tion of Cycloh exa -
d ien e 1 w ith Non a ca r bon yld iir on Usin g Ca ta lyst 2b. A
solution of nonacarbonyldiiron (10.0 g, 27.5 mmol), 1-(4-
methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene (2b) (1.63 g,
6.87 mmol), and cyclohexa-1,3-diene (1) (6.61 g, 7.86 mL, 82.5
mmol) in DME (50 mL) was heated at reflux for 16.5 h. The
solvent was evaporated, and the residue was subjected to flash
chromatography (pentane) on silica gel to afford the tricar-
bonyliron complex 3 (11.9 g, 98%) as a yellow oil. For
spectroscopic data, see ref 11.
A detailed study of the mechanism of the azadiene-
catalyzed complexation of cyclohexa-1,3-diene by the
tricarbonyliron fragment is under active investigation
in our laboratories21 and will be reported in a forthcom-
ing publication.
Con clu sion
The azadiene-catalyzed complexation of cyclohexa-1,3-
diene 1 with pentacarbonyliron proceeds quantitatively
and represents the most efficient and convient method
P r ep a r a tion of Tr ica r bon yl[(1-4-η)-cycloh exa -1,3-d i-
en e]ir on (3) by Ca ta lytic Com p lexa tion of Cycloh exa -
d ien e 1 w ith P en ta ca r bon ylir on Usin g Ca ta lyst 2b.
A
(21) Kno¨lker, H.-J .; Gonser, P.; Hermann, H.; J ones, P. G.; Rohde,
G.; Schnepf, A. Unpublished results.
solution of pentacarbonyliron (2.00 g, 1.34 mL, 10.2 mmol),
1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-diene (2b) (302
mg, 1.27 mmol), and cyclohexa-1,3-diene (1) (1.23 g, 1.46 mL,
15.4 mmol) in dioxane (12 mL) was heated at reflux for 45 h.
The solvent was evaporated, and the residue was subjected to
flash chromatography (pentane) on silica gel to afford the
tricarbonyliron complex 3 (2.24 g, 99%) as a yellow oil. For
spectroscopic data, see ref 11.
La r ge-Sca le (50 g) P r ep a r a tion of Tr ica r bon yl[(1-4-
η)-cycloh exa -1,3-d ien e]ir on (3) by Ca ta lytic Com p lex-
a tion of Cycloh exa d ien e 1 w ith P en ta ca r bon ylir on Us-
in g Ca ta lyst 2b. A solution of pentacarbonyliron (50.0 g, 33.6
mL, 255 mmol), 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-1,3-
diene (2b) (7.70 g, 32.4 mmol), and cyclohexa-1,3-diene (1) (31.3
(22) Nesmeyanov, A. N.; Rybinskaya, M. I.; Rybin, L. V.; Gubenko,
N. T.; Bokii, N. G.; Batsanov, A. S.; Struchkov, Y. T. J . Organomet.
Chem. 1978, 149, 177. Nesmeyanov, A. N.; Rybin, L. V.; Stelzer, N.
A.; Struchkov, Y. T.; Batsanov, A. S.; Rybinskaya, M. I. J . Organomet.
Chem. 1979, 182, 399. Batsanov, A. S.; Struchkov, Y. T. J . Organomet.
Chem. 1983, 248, 101.
(23) Kno¨lker, H.-J .; Baum, G.; Gonser, P. Tetrahedron Lett. 1995,
36, 8191.
(24) Kokkes, M. W.; Beentjes, P. C. J .; Stufkens, D. J .; Oskam, A.
J . Organomet. Chem. 1986, 306, 77.
(25) Bellachioma, G.; Cardaci, G. J . Chem. Soc., Dalton Trans. 1977,
2181. Bellachioma, G.; Reichenbach, G.; Cardaci, G. J . Chem. Soc.,
Dalton Trans. 1980, 634.
(26) Kno¨lker, H.-J .; Hermann, H. Angew. Chem. 1996, 108, 363;
Angew. Chem., Int. Ed. Engl. 1996, 35, 341. Kno¨lker, H.-J .; Gonser,
P.; Koegler, T. Tetrahedron Lett. 1996, 37, 2405.

Complexation of Cyclohexa-1,3-diene
Organometallics, Vol. 17, No. 18, 1998 3925
Ta ble 7. Am ou n t of Ca ta lyst 2, a n d Yield of th e
Tr ica r bon ylir on Com p lex 3
MHz, toluene-d8, 213 K): δ 54.47 (CH3), 75.19 (CH2), 75.25
(CH), 113.55 (2 CH), 122.08 (2 CH), 129.63 (2 CH), 148.67 (C),
151.62 (C), 156.85 (C), 187.34 (C), 204.48 (CO), 209.89 (CO),
210.78 (CO), 211.30 (CO), 212.30 (CO), 214.16 (CO) (the signals
of 3 CH groups are missing due to overlapping with signals of
the solvent). 13C NMR (125 MHz, toluene-d8, 323 K): δ 54.91
(CH3), 75.68 (CH2), 76.35 (CH), 114.18 (2 CH), 122.61 (2 CH),
128.11 (CH), 128.28 (CH), 129.20 (CH), 129.47 (2 CH), 149.26
(C), 152.25 (C), 157.61 (C), 188.19 (C), 210.46 (6 CO, br). MS
(75 °C): m/z 517 (M+, 4), 489 (6), 461 (11), 433 (1), 405 (18),
377 (16), 349 (39), 293 (14), 238 (10), 237 (38), 236 (100), 115
(16). HRMS calcd for C22H15Fe2NO7, 516.9547; found, 516.9559.
Anal. Calcd for C22H15Fe2NO7: C, 51.11; H, 2.92; N, 2.71.
Found: C, 51.20; H, 3.04; N, 2.42.
(b) By Hea tin g 6b w ith P en ta ca r bon ylir on in Dioxa n e.
Pentacarbonyliron (596 mg, 0.40 mL, 3.04 mmol) was added
to a solution of complex 6b (72 mg, 0.191 mmol) in dioxane
(10 mL), and the reaction mixture was heated at 90 °C for 24
h. The solvent was evaporated, and the residue was subjected
to flash chromatography (pentane/diethyl ether 10:1) on silica
gel to afford, first, the hexacarbonyldiiron complex 7b (8 mg,
8%; brown solid, spectroscopic data, see above) followed by the
more polar fraction, the tricarbonyliron complex 6b (62 mg,
86%; red crystals).
(c) By Hea tin g 6b w ith Non a ca r bon yld iir on in THF .
Nonacarbonyldiiron (87 mg, 0.24 mmol) was added to a
solution of complex 6b (72 mg, 0.191 mmol) in tetrahydrofuran
(10 mL), and the reaction mixture was heated at 65 °C for 24
h. The solvent was evaporated, and the residue was subjected
to flash chromatography (pentane/diethyl ether 10:1) on silica
gel to afford, first, the hexacarbonyldiiron complex 7b (6 mg,
6%; brown solid, spectroscopic data, see above) followed by the
more polar fraction, the tricarbonyliron complex 6b (62 mg,
86%; red crystals).
entry
2 [mg]
3, yield [mg]
3, yield [%]
no catalyst
118
7
413
496
834
662
220
902
327
860
54
0.7
41
50
83
66
22
90
33
86
5
a
b
c
d
e
f
g
h
i
135
135
142
156
152
169
169
134
j
k
l
m
n
o
p
q
r
135
152
152
173
169
152
132
166
312
507
914
189
740
709
132
832
0
31
51
91
19
74
71
13
83
0
155
g, 37.2 mL, 391 mmol) in dioxane (200 mL) was heated at
reflux for 45 h. During the reaction a light flow of argon was
directed through the apparatus. After a reaction time of 72
h, a solution of catalyst 2b (2.00 g, 8.43 mmol) and diene 1
(4.21 g, 5.0 mL, 52.5 mmol) in dioxane (50 mL) was added.
The reaction mixture was stirred at reflux for 24 h. The
solvent was evaporated, pentane (200 mL) was added to the
residue, and the solution was filtered over Celite to provide,
after evaporation of the solvent, the tricarbonyliron complex
3 (50.2 g, 89%) as a yellow oil. For spectroscopic data, see ref
11.
P r ep a r a tion of Tr ica r bon yl[(1-4-η)-cycloh exa -1,3-d i-
en e]ir on (3) by Ca ta lytic Com p lexa tion of Cycloh exa -
1,3-d ien e (1) w ith P en ta ca r bon ylir on Usin g th e Ca ta -
lysts 2. Gen er a l P r oced u r e. Pentacarbonyliron (891 mg,
0.60 mL, 4.55 mmol), cyclohexa-1,3-diene (1) (547 mg, 0.65 mL,
6.83 mmol), and dioxane (15 mL) were added to the catalyst 2
(0.125 equiv, 0.569 mmol, Table 7). The reaction mixture was
heated at reflux for 14 h. The solvent was evaporated in vacuo,
and the residue was subjected to flash chromatography (pen-
tane) on silica gel to afford the tricarbonyliron complex 3
(Table 7) as a yellow oil. For spectroscopic data, see ref 11.
P r ep a r a tion of th e Hexa ca r bon yld iir on Com p lex 7b.
(a ) By Dir ect Com p lexa tion of 2b w ith P en ta ca r bon yl-
ir on . Pentacarbonyliron (1.19 g, 0.80 mL, 6.07 mmol) was
added to a solution of 1-(4-methoxyphenyl)-4-phenyl-1-azabuta-
1,3-diene (2b) (45 mg, 0.190 mmol) in dioxane (10 mL), and
the reaction mixture was heated at 90 °C for 24 h. The solvent
was evaporated, and the residue was subjected to flash
chromatography (pentane/diethyl ether 10:1) on silica gel to
afford, first, the hexacarbonyldiiron complex 7b (5 mg, 5%;
brown solid) followed by the more polar fraction, the tricar-
bonyl(η4-1-azabuta-1,3-diene)iron complex 6b (34 mg, 47%; red
crystals).
Ca ta lytic Com p lexa tion of Cycloh exa d ien e 1 w ith
P en ta ca r bon ylir on Usin g th e (η4-1-Aza bu ta -1,3-d ien e)-
tr ica r bon ylir on Com p lex 6b a s th e Ca ta lyst. A solution
of pentacarbonyliron (207 mg, 139 µL, 1.06 mmol), the tricar-
bonyliron complex 6b (50 mg, 0.13 mmol), and cyclohexadiene
1 (127 mg, 151 µL, 1.58 mmol) in dioxane (15 mL) was heated
at reflux for 45 h. Removal of the solvent in vacuo and flash
chromatography (pentane) of the residue on silica gel provided
the tricarbonyliron complex 3 (181 mg, 69%) as a yellow oil.
For spectroscopic data, see ref 11.
Stoich iom etr ic Com p lexa tion of Cycloh exa d ien e 1
Usin g th e Hexa ca r bon yld iir on Com p lex 7b a s th e Tr a n s-
fer Rea gen t. A solution of the hexacarbonyldiiron complex
7b (59 mg, 0.114 mmol) and cyclohexadiene 1 (17.4 mg, 20.7
µL, 0.217 mmol) in benzene (15 mL) was heated at 80 °C for
4 days. The solvent was evaporated in vacuo, and the residue
was purified by flash chromatography (pentane) on silica gel
to afford the iron complex 3 (16 mg, 64%) as a yellow oil. For
spectroscopic data, see ref 11.
Ack n ow led gm en t. This work was supported by the
Deutsche Forschungsgemeinschaft (Gerhard-Hess award)
and the Fonds der Chemischen Industrie. We are
grateful to the BASF AG, Ludwigshafen, for a generous
gift of pentacarbonyliron.
1
6b: H NMR (500 MHz, C6D6): δ 3.27 (s, 3 H), 3.45 (d, J )
9.4 Hz, 1 H), 5.03 (dd, J ) 9.4, 2.8 Hz, 1 H), 6.41 (d, J ) 2.8
Hz, 1 H), 6.64 (d, J ) 8.8 Hz, 2 H), 6.87 (d, J ) 8.8 Hz, 2 H),
6.96 (t, J ) 7.5 Hz, 1 H), 7.06 (t, J ) 7.5 Hz, 2 H), 7.14 (d, J
) 7.5 Hz, 2 H). For further spectroscopic data, see ref 11.
7b: Mp g 147 °C (dec). IR (drift): νj 2963, 2898, 2839, 2062,
2045, 2029, 1985, 1973, 1954, 1603, 1504, 1466, 1443, 1416,
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
parameters, atomic coordinates and equivalent isotropic dis-
placement parameters, bond lengths and angles, anisotropic
displacement parameters, and hydrogen coordinates and
isotropic displacement parameters for the crystal structure of
7b and figures of the 1H NMR spectrum for 6b and the variable
temperature 13C NMR spectra for 7b (13 pages). Ordering
information is given on any current masthead page.
1261, 1095, 1023, 864, 800, 764, 706 cm-1 1H NMR (500 MHz,
.
C6D6): δ 3.18 (s, 3 H), 3.65 (br s, 2 H), 4.26 (s, 1 H), 6.42 (d, J
) 8.8 Hz, 2 H), 6.79 (d, J ) 8.8 Hz, 2 H), 7.04 (t, J ) 7.6 Hz,
1 H), 7.11 (t, J ) 7.6 Hz, 2 H), 7.26 (d, J ) 7.6 Hz, 2 H). 13C
NMR (125 MHz, toluene-d8, 300 K): δ 54.78 (CH3), 75.57
(CH2), 76.10 (CH), 114.02 (2 CH), 122.48 (2 CH), 128.06 (CH),
128.24 (CH), 129.16 (CH), 129.48 (2 CH), 149.12 (C), 152.09
(C), 157.43 (C), 188.02 (C), 210.96 (6 CO, br). 13C NMR (125
OM980290G