Reaction patterns of 'tetraferrio-azaallenium', [(OC)3Cp2Fe2(μ4-C=N=C)Fe 2Cp2(CO)3]+, and its 'activated cyanide'-precursor (μ-phthalimidocarbyne)tricarbonyldicyclopentadienyldiiron(1+)

Article abstract of DOI:10.1016/s0020-1693(99)00160-7

(μ-Phthalimidocarbyne)tricarbonyldicyclopentadienyldiiron(1+) (2) which, due to the facile cleavage with cis-[Fe₂Cp₂(CN)(μ-CO)₂CO]^- (1) of its μ-CN bond to give 'tetraferrio-azaallenium' [(OC)₃Cp₂Fe₂(μ₄-C=N=C)Fe ₂Cp₂(CO)₃]⁺ (3), has been termed 'activated cyanide' reacts with pyrrolidine with ring opening to give the μ-acylisocyanide complex 5 carrying a further amido function in the o-position. Butyllithium gives rise to only traces of the corresponding mono deacylation product 6 while the main product is 3. Formation of complex 3 which is presumed to occur via partial bis-deacylation of 2 to give 1 and subsequent recombination with loss of phthalimide also dominates the reactions of 2 with CN^- and [W(CN)(CO)₅]^-. Here, however, the nucleophile also attacks the μ-carbyne carbon atom to give moderate yields of the desired μ-cyanoalkyliden complexes 7 and 10, the latter with [W]CN?[W]NC isomerization and C-C bond formation. With [Cr(CN)(CO)₅]^-, in contrast, substitution of a terminal CO ligand occurs with formation of the cyano-bridged trinuclear complex 9. The structures of complexes 7, 9 and 10 have been elucidated by X-ray analysis. Action upon 'tetraferrio-azaallenium' 3 of the primary amines H₂NCH₂CH₂R (R=Me, OMe, Ph) gives rise in low yields to the respective terminal isocyanides in compounds 11-13. Here as in the reaction with NaNH₂ to give 14, a terminal carbonyl ligand is converted into an isocyanide or, respectively, cyano ligand by an exchange of oxygen for NR (N^-). A second though less efficient approach to 14 is CO substitution in 3 by NaCN. Disubstitution of 3 to give 15 is observed with trimethylphosphite. In the reactions of 3 with diamines (H₂NCH₂CH₂NH₂, H₂NCH₂CH₂CH₂CH₂NH ₂), on the other hand, cluster fragmentation occurs with formation of the mononuclear complexes 16 and 17 containing cyclic diaminocarbene ligands. A cyclic μ-aminocarbyne diiron complex 18 is obtained in traces from 3 and pyrrolidine.

Full text of DOI:10.1016/s0020-1693(99)00160-7

www.elsevier.nl/locate/ica
Inorganica Chimica Acta 291 (1999) 266–278
Reaction patterns of ‘tetraferrio-azaallenium’,
[(OC)3Cp2Fe2(m4-CꢀNꢀC)Fe2Cp2(CO)3]+, and its ‘activated
cyanide’-precursor (m-phthalimidocarbyne)-
tricarbonyldicyclopentadienyldiiron(1+)ꢀ
Karl-Heinz Trylus, Ulrich Kernbach, Irene Bru¨dgam, Wolf Peter Fehlhammer *
Institut fu¨r Anorganische und Analytische Chemie der Freien Uni6ersita¨t Berlin, Fabeckstraße 34-36, D-14195 Berlin, Germany
Received 19 November 1998; accepted 8 April 1999
Abstract
(m-Phthalimidocarbyne)tricarbonyldicyclopentadienyldiiron(1+) (2) which, due to the facile cleavage with cis-[Fe2Cp2(CN)(m-
CO)2CO]− (1) of its m-CN bond to give ‘tetraferrio-azaallenium’ [(OC)3Cp2Fe2(m4-CꢀNꢀC)Fe2Cp2(CO)3]+ (3), has been termed
‘activated cyanide’ reacts with pyrrolidine with ring opening to give the m-acylisocyanide complex 5 carrying a further amido
function in the o-position. Butyllithium gives rise to only traces of the corresponding mono-deacylation product 6 while the main
product is 3. Formation of complex 3 which is presumed to occur via partial bis-deacylation of 2 to give 1 and subsequent
recombination with loss of phthalimide also dominates the reactions of 2 with CN− and [W(CN)(CO)5]−. Here, however, the
nucleophile also attacks the m-carbyne carbon atom to give moderate yields of the desired m-cyanoalkyliden complexes 7 and 10,
the latter with [W]CNl[W]NC isomerization and C–C bond formation. With [Cr(CN)(CO)5]−, in contrast, substitution of a
terminal CO ligand occurs with formation of the cyano-bridged trinuclear complex 9. The structures of complexes 7, 9 and 10
have been elucidated by X-ray analysis. Action upon ‘tetraferrio-azaallenium’ 3 of the primary amines H2NCH2CH2R (R=Me,
OMe, Ph) gives rise in low yields to the respective terminal isocyanides in compounds 11–13. Here as in the reaction with NaNH2
to give 14, a terminal carbonyl ligand is converted into an isocyanide or, respectively, cyano ligand by an exchange of oxygen for
NR (N−). A second though less efficient approach to 14 is CO substitution in 3 by NaCN. Disubstitution of 3 to give 15 is
observed with trimethylphosphite. In the reactions of 3 with diamines (H2NCH2CH2NH2, H2NCH2CH2CH2CH2NH2), on the
other hand, cluster fragmentation occurs with formation of the mononuclear complexes 16 and 17 containing cyclic diaminocar-
bene ligands. A cyclic m-aminocarbyne diiron complex 18 is obtained in traces from 3 and pyrrolidine. © 1999 Elsevier Science
S.A. All rights reserved.
Keywords: Crystal structures; Iron complexes; Imidocarbyne bridged complexes; Polynuclear complexes
1. Introduction
prompted us to search for simple cyano clusters which,
in addition to the terminal coordination mode, would
offer m-C,C bridging positions to the cyano ligand or its
eventual reaction products with very different implica-
tions for the stereochemistries and reactivities of the
respective groups [1]. The synthesis of cis-[Fe2Cp2-
(CN)(m-CO)2CO]− (1) [2], the elucidation of its dynam-
ics and the results of its protonation [3], alkylation,
acylation and iminoacylation studies [4] have met these
expectations at the same time considerably extending
this type of chemistry.
Our continuous interest in the development of a
cyano complex based organometallic chemistry had
ꢀ Chemistry of Hydrogen Isocyanide, Part 14. For Part 13 see Ref.
[6].
* Corresponding author. Present address: Deutsches Museum,
Museumsinsel 1, D-80538 Munich, Germany. Tel.: +49-89-217 9313;
fax: +49-89-217 9425.
E-mail address: wpf@extern.lrz-muenchen.de (W.P. Fehlhammer)
0020-1693/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0020-1693(99)00160-7

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
267
A
wholly new dimension reaching far beyond
cyanide acylation arose, however, on the attempt of
diacylation of 1 with phthaloyl dichloride. Here, in-
stead of the expected product 2 we got a unique te-
tranuclear cation, the tetrametallated azaallenium
species 3, which in spite of the very mild conditions
prevailing during the reaction must have formed by an
unprecedented aprotic cleavage of the CN triple bond
and by fixation of the formally resulting m-carbido
diiron fragment to a second cyano diiron complex
anion [5].
Scheme 1. [Fe]=Fe(h-Cp)CO; [Fe2]=Fe2(h-Cp)2(CO)3.
2. Addition of nucleophiles to the m-phthalimidocarbyne
diiron complex 2
2.1. Reactions with pyrrolidine and butyllithium
Formal transfer of [Fe2]C2+ ([Fe2]=Fe2(h-Cp)2-
(CO)3) onto secondary amines NHRR% promised a
new approach to aminocarbyne bridged diiron com-
plexes of the formula [Fe2Cp2(CO)2(m-CO)(m-
CꢀNRR%)]+. A wide range of alicyclic aminocarbyne
complexes of this type have been prepared by Man-
ning and coworkers [8] through alkylation of the re-
Concerning the mechanism, we had intuitively fa-
voured 2 as the reaction intermediate from the begin-
ning though at that time diacylation of the m-CN
ligand had not been achieved in any case. Recently,
however, we succeeded in the preparation and struc-
tural characterization of 2 and a series of related com-
pounds with N,N-diacylated cyano bridges all of
which showed dramatically elongated m-C–N bonds of
spective alkylisocyanide starting materials. A cyclic
¸¹¹¹¹º
analogue
[Fe2Cp2(CO)2(m-CO)(m-CꢀN(CH2)3CH2)]+
,
\1.30 A [6]. Chemical proof that 2 and its congeners
(cf. Section 3.3), has been synthesized in our group by
reacting Fe2Cp2(CO)4 or [Fe2Cp2(NCMe)(CO)3], re-
spectively, with 4-chlorobutyl isocyanide [9]. The
product obtained quantitatively from the reaction of 2
with pyrrolidine, however, is clearly not of this type.
The characteristic wCꢀNB band in the IR around
1590915 cm−1 is missing while the absorptions
present at 1690 m [w(CO)acyl] and 1630 versus [w(CꢀN)]
are indicative of the m-acylisocyanide grouping. Thus,
unlike 1, which in the case of the glutarimido deriva-
tive of 2 produced only traces of the bis(m-acyliso-
cyanide) 4 while no byproducts have been observed in
the reaction with 2 (see above), pyrrolidine attacks
exclusively at the C3 position causing partial deacyla-
tion of the ‘activated cyanide’ with ring opening and
formation of the ‘mixed amide’ 5. According to an IR
spectrum the same reaction pattern applies with n-
propylamine; therefore no effort was made to purify
the crude product.
were in fact the long sought after ‘activated
cyanide’—intermediates in the formation of tetrafer-
rio-azaallenium (3) has been furnished by their 1:1
reactions with the cyano diiron anion 1 rendering 3 in
practically quantitative yields. In order to obtain this
result, the nucleophile, i.e. the nitrogen end of the
terminal cyano ligand in 1, must have attacked the
carbyne carbon of 2 to give a m-alkylidene intermedi-
ate A (Nu=NꢁC–Fe2Cp2(CO)3) which subsequently
underwent some (CN)terminal“(CN)bridge rearrangement
with loss of phthalimide. From the reaction with 1 of
the glutarimido analogue of 2, however, only a minor
amount of 3 could be isolated together with traces of
a byproduct of the form 4 revealing a competing situa-
tion in ‘activated cyanides’ between the carbyne car-
bon atom (C1) and a second electrophilic centre
located on the acyl groups (C3), which is reminiscent
of another type of ‘activated (iso)cyanide’, viz.
Cr(CO)5C1ꢁN–C3Cl3 [7].
In this paper we report on the reactions of 2 and 3
with nucleophiles other than 1 which have been stud-
ied with the intention of gaining more insight in the
chemistry of these novel, obviously highly electrophilic
species in general. Particular interest was shown in the
possibility of using 2 (and perhaps also 3) as sources
of C1 fragments for the construction of new metallated
CNC (or CCN) moieties with various other cyanides
(Scheme 1).
One plausible product from the action upon 2 of
butyllithium could have been A (Nu=Bun) which, due
to the preformed phthalimide leaving group, was ex-
pected to give the secondary reactions of a m-alkyli-
dyne species [10].
A promising model for the
formation of new C–C bonds with carbanions exists
in Busetto and co-workers [Fe2Cp2(CO)3(m-CSMe)]-
CF3SO3 [11]. The actual findings, however, were
even more disappointing than those with pyrrolidine.
In addition to some unreacted starting material

268
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
only minor amounts of a monodeacylation product,
Fe2Cp2(CO)3{m-CNC(O)C6H4-o-C(O)Bun} (6), could
be isolated.
What significantly prevails in this reaction (and in
most of those to be reported in the following) is the
formation of the tetraferrio-azaallenium 3, the obvious
thermodynamic sink in the system. As outlined in Eq.
(1), for 3 to form, total, i.e. two-fold deacylation of the
‘activated cyanide’ must occur to afford free cyanodi-
iron complex anions which then recombine with re-
maining phthalimidocarbyne diiron complex cations in
the familiar way [6]. It should be noted that none of the
side products C have been observed in any case.
Scheme 2. [Fe2]=Fe2(h-Cp)2(CO)3.
Again, we have to state that with both reagents
addition to the acyl function in 2 predominates with the
familiar consequence of 3 being the main product. In
the case of [Cr(CN)(CO)5]−, the expected side product
(OC)5CrCNC(O)C6H4-o-C(O)NCCr(CO)5 has been
identified. However, there were accompanying products
(9, 10), both of which gave us a surprise. Though
analyzing as simple neutral 1:1 adducts, major inconsis-
tencies in the w(CN) and w(CO) regions of the IR and
particularly, the 1H NMR made us envisage rather
different structures for 9 and 10. In compound 9 it was
the nonequivalence of the two iron centres as indicated
by two Cp signals of equal intensity (Section 4) which
put us on the track of a possible CO substitution by
[Cr(CN)(CO)5]−, a proposal later confirmed by an
X-ray structure analysis (Section 2.3). What made us
hesitate to formulate 10 as the hoped-for m3-iso-
cyanocarbene species A (Nu=NCW(CO)5), on the
other hand, was the unusually high IR frequency of
2175 cm−1 of the presumed tungsten-coordinate iso-
cyanide. Here, Busetto’s isomers with w(CN)=2198 for
E(a) versus 2125 cm−1 for E(b) turned out to be ideal
for comparison. Verification of the obvious conclusion
that a NCW/CNW-isomerization with CC-bond forma-
tion must have occurred during or after the nucleophilic
addition to the carbyne carbon of 2 again is provided
by structure determination (see below). Notice that in
the very similar preparation of E(b) no such isomeriza-
tion was observed (Scheme 3).
(1)
2.2. Reactions with cyanide and cyanometallates
The main product of the reaction of cyanide applied
as NaCN with 2 (as PF6 salt) in 1,2-dimethoxyethane
was again tetraferrio-azaallenium 3; neither an interme-
diate B (Nu=CN) nor the expected byproduct C
(Nu=CN) have been identified in the workup proce-
dure. Instead, a neutral complex (7) could be separated
by fractional crystallization showing all the analytical,
spectroscopic (IR, MS) and structural requirements (see
Sections 4 and 2.3) for the desired primary adduct A
(Nu=CN). Interestingly, there were no indications of a
spontaneous release of the phthalimide leaving group to
end up with a cationic m-cyanocarbyne diiron complex,
nor was there any evidence of a m-dicyanocarbene
product. The latter has been obtained by Busetto and
co-workers [12] from an ylid-like storage form D which
dissociates much easier than A as no charge separation
is required.
2.3. X-ray structures of 7, 9 and 10
By employing the cyanometallates [M(CN)(CO)5]−
(M=Cr, W) we once again borrowed from Busetto’s
fine chemistry which by reaction of the carbocationic D
with (a) CN− followed by W(CO)5thf or (b)
[W(CN)(CO)5]− had brought about the two isomers E
(Scheme 2).
As part of a more general program which focuses on
the question of metal stabilization of the ‘nonexisting’
CNC in forms such as 3 (‘permetallated azaallenium’)
or 8 (‘permetallated methylisocyanide’ [13]), we have
also become interested in metallacumulenes of type F
for which the chemistry of this paragraph might open a
possible route.
The configuration of complexes 7 and 10 is cis with
the two Cp ligands sticking out on one side of the best
plane through the Fe2(m-C)2 rhomboid, and the two
terminal CO groups on the other (Figs. 1 and 2).
Identical values (163(1)°) are found for the interplanar
angles Fe1Fe2(m-C)/Fe1Fe2(m-C%), which describe the
fold of the central four-membered ring along the
Fe1···Fe2 axis.
The m-alkylidene carbon is quasi-tetrahedral with
C–N and C–C single bonds (Tables 1 and 3) and
N–(m-C)–C angles between the freely adjusting phthal-
imido and cyano or, respectively, CNW(CO)5 sub-
stituents of 105(1) and 104(2)°, i.e. close to the ideal

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
269
Fig. 2. Molecular structure of complex 10.
two calculations with exchanged C and N scattering
factors which are clearly in favour of [Fe2]C–CꢁN–W;
for the corresponding U11 values (see caption).
Scheme 3. [Fe]=Fe(h-Cp)CO.
value. The planar phthalimido group is practically per-
pendicular to the N,(m-C),C plane which in turn is
perpendicular to Fe1Fe2(m-Ccarbene). The remaining
bonding parameters are all in the usual range.
Some effort has been made to furnish unequivocal
proof of the nitrile structure of complex 10. The usual
strategy, viz. comparison of the effects on the tempera-
ture factors of several cycles of refinement for the two
atomic arrangements in question [Fe2]C–NꢁC–W or
[Fe2]C–CꢁN–W, has also been successful in the present
case. Fig. 3 depicts the thermal ellipsoids resulting from
Fig. 3. ORTEP drawing of the central linkage after two cycles of
refinement with (a) [Fe2]C–C–N–[W] (U11(C) 4.48(4), U11(N)
4.35(5)) and (b) [Fe2]C–N–C–[W] (U11(N) 1.66(8), U11(C) 7.86(5))
atomic sequences.
Fig. 1. Molecular structure of complex 7.

270
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
Table 1
a
,
Selected bond distances (A) and bond angles (°), best planes and interplanar angles (°) of 7
,
Bond distances (A)
Fe1–Fe2
Fe1–C4
Fe2–C2
Fe2–C4
Fe1–C(Cp) b
O1–C1
O3–C3
N1–C4
N1–C11
N2–C5
O11–C11
C11–C13
C13–C14
C14–C15
C16–C17
2.513(3)
1.97(2)
1.75(2)
1.98(2)
2.12(2)
1.14(3)
1.21(2)
1.45(2)
1.44(2)
1.12(3)
1.16(2)
1.49(3)
1.42(3)
1.44(3)
1.41(3)
Fe1–C3
Fe1–C1
Fe2–C3
1.87(2)
1.75(2)
1.92(2)
Fe2–C(Cp) b
O2–C2
2.12(2)
1.13(3)
N1–C12
C4–C5
1.41(2)
1.51(3)
1.24(2)
1.42(2)
1.39(2)
1.40(3)
1.39(3)
O12–C12
C12–C18
C13–C18
C15–C16
C17–C18
Bond angles (°)
Fe1–Fe2–C2
Fe1–Fe2–C4
Fe2–Fe1–C1
Fe2–Fe1–C4
Fe1–C3–Fe2
Fe1–C1–O1
Fe1–C4–N1
Fe2–C2–O2
Fe2–C4–N1
C1–Fe1–C3
C2–Fe2–C3
C3–Fe1–C4
C4–N1–C11
C4–C5–N2
99.1(6)
50.3(5)
98.7(7)
50.6(6)
83.0(8)
178(2)
124(1)
177(2)
123(1)
87(1)
86.6(9)
98.4(8)
124(1)
178(2)
105(1)
125(2)
102(1)
110(1)
132(2)
124(2)
107(1)
113(2)
125(2)
121(2)
Fe1–Fe2–C3
Fe2–Fe1–C3
47.6(6)
49.4(6)
Fe1–C4–Fe2
Fe1–C3–O3
Fe1–C4–C5
Fe2–C3–O3
Fe2–C4–C5
C1–Fe1–C4
C2–Fe2–C4
C3–Fe2–C4
C4–N1–C12
79.1(7)
141(2)
112(1)
136(2)
113(1)
90.5(8)
92.5(8)
96.4(8)
125(1)
N1–C4–C5
N1–C11–O11
N1–C11–C13
C11–N1–C12
O11–C11–C13
C11–C13–C14
C12–C18–C13
C13–C14–C15
C14–C13–C18
C15–C16–C17
N1–C12–O12
N1–C12–C18
121(2)
109(1)
O12–C12–C18
C11–C13–C18
C12–C18–C17
C13–C18–C17
C14–C15–C16
C16–C17–C18
129(2)
111(1)
133(2)
121(2)
123(2)
118(2)
Best planes
I: Fe1, C3, Fe2
II: Fe1, C4, Fe2
III: C5, C4, N1
IV: N1, C12, C18, C13, C11
V: C13, C14, C15, C16, C17, C18
Interplanar angles (°)
I/II: 17(1)
II/IV: 137(1)
III/IV: 84(1)
IV/V: 0.9(5)
a E.s.d.s are given in parentheses, see Fig. 1 for atomic numbering.
b Average value.
The core of complex 9, in contrast, is configurated trans
exhibiting an asymmetry which reflects its asymmetric
substitution (Fig. 4, Table 2). Why [Cr(CN)(CO)5]−
chooses to attack at the iron atom in the second place,
i.e. after having largely deacylated the ‘activated cyanide’
remains in the dark. Steric reasons can definitely be ruled
out as is obvious from the tungsten case 10 and the
quantitative reaction between 2 and [Fe2Cp2(CN)-
(CO)3]−. Moreover, the minimal steric requirements of
[M(CN)(CO)5]− have impressively been demonstrated by
its extensive use as a ligand in hexacoordinated supercom-
plexesofthetype[M%{NCM(CO)5}6]x−6 (M=e.g. Si(IV),
Cr(III)) (‘octahedro octahedra’) [14].
Actually, most of the stereochemical peculiarities
in 9 (as compared to, e.g. 7 and 10 or better 2 and
the other ‘activated cyanides’ [6]) can be traced back
to the presence of the electron-rich pentacarbonyl
(cyano)chromate in the place of a p-acidic CO ligand.

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
271
Fig. 4. ORTEP drawing of compound 9.
The afore-mentioned asymmetry of the Fe2(m-C)2 core
with the unusually short Fe1–C2 bond of pronounced
double bond character—the 180(1) pm correspond well
with the 180.3(5) pm for the ¸F¹eꢀ¹C¹¹b¹o¹nd¹¹inº Berke’s
lished a highly electrophilic character of the carbon
atoms, we expected nucleophilic attack on 3 to occur at
these positions with formation of imine-type structures
and the possibility of obtaining the bis(m-methyl-
iden)amine structures H (Eq. (2)), provided the CꢀNꢀC
framework was sufficiently accessible in spite of its
steric shielding by the substituents on iron as apparent
from a space-filling model of the cation of 3 (see Fig. 1
in Ref. [5]).
ferra-cumulene (OC)4FeꢀCꢀCꢀCC(But)2OC(O)O [15]—
finds its logical consequence in an extremely elongated
C2–N2 bond of 137(1) (!) pm, which sets up a new
record within the family of ‘activated cyanides’ [6]. The
bonding which resembles that in formamides (C–N:
136 pm) is perhaps best represented by resonance for-
mula G.
(2)
3.1. Addition of primary amines. Isocyanide deri6ati6es
Reactions between 3 and primary amines require
distinctly higher temperatures (Section 4). The only
products which could be isolated in low yields besides
unreacted starting material were the mono-isocyanide
derivatives of 3, 11–13, i.e. nucleophilic attack occurred
solely at the carbon atom of a terminal CO ligand. It is
very likely that steric reasons have to be put forward
for this chemistry forced entirely to the periphery (see
above).
Methods for the synthesis of isocyanides at metal
complexes have been extensively studied in our group
including the conversion of CO into CNR ligands
[17–19]. The latter generally proceeds in two steps via
intermediate carbamoyl (carboxamido) species which is
3. The nucleophilic reaction pattern of tetraferrio-aza-
allenium (3)
The agreement of the measured CꢀN bond lengths in
3 with those calculated by geometry optimization for
the model and parent compounds Li2CꢀNꢀCLi2+ and
H2CꢀNꢀCH2+ as well as the continual preference of the
cumulene over the tautomeric bent 2-azaallylium form
shows, that contrary to lone pair donor substituents
(–OR, NR2) [16], no unusual electronic effects are
exerted by the organometallic substituents in tetrafer-
rio-azaallenium ions. As calculations on azaallenium
(‘bisiminium’) compounds in general have also estab-

272
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
easily dehydrated with phosgen in analogy to the
preparation of isocyanides from formamides (Eq. (3))
[19].
coordination centres. Due to the low solubility of the
complexes no 13C NMR spectra could be obtained.
[M]CꢁO+RHN2“[M]C(ꢀO)NHRÇ− “[M]CꢁNR
(3)
At the elevated temperatures applied in the present
case, dehydration must have occurred spontaneously
as no carbamoyl complexes have been detected.
Number, positions and intensities of the characteris-
1
tic bands and signals in the IR and H NMR spectra,
respectively, are in accord with the assigned structures
(Section 4). Compared with the related iso-
cyanide
complexes
[Fe2Cp2(CO)(CNMe)(m-CO)2]
(2138 cm−1
)
and [Fe2Cp2(CO)(CNMe)(m-CO)(m-
CN(H)Me)]+ (2189 cm−1), the w(CN) IR frequencies
of complexes 11–13 (215395 cm−1) reflect (a) the
cationic overall charge which (b) is spread across four
3.2. Cyano and phosphite deri6ati6es of 3
Tetraferrio-azaallenium is strikingly inert. Even the
small, negatively charged nucleophiles NH2− and
CN− do not react at ambient temperature. Heating
mixtures of 3 and sodium amide suspended in dioxane
for several hours to above 100°C resulted in very
modest yields of the neutral cyano cluster 14. The
starting complex is recovered almost quantitatively.
Again the terminal carbonyl carbon atom proves to be
the most reactive since it’s the most easily accessible
site in the molecule. A vast number of similar trans-
formations of carbonyl to cyano ligands has been car-
ried out by reaction in liquid ammonia [20c]. They are
known to proceed via unsubstituted carbamoyl species
([M]–C(ꢀO)HN2) which subsequently are dehydrated.
A major improvement was reached with the introduc-
tion of ‘silazid’ (NaN(SiMe3)2) as reagent which al-
lowed these reactions to be conducted under aprotic
(anhydrous) conditions [20a,b].
Table 2
,
Selected bond distances (A) and bond angles (°), best planes and
interplanar angles (°) of 9 a
,
Bond distances (A)
Fe1–Fe2
Fe1–N1
Fe1–C2
Fe1–C4
Fe1–C(Cp) b
Fe2–C2
Fe2–C3
Fe2–C4
Fe2–C(Cp)
Cr–C1
Cr–C13
Cr–C11
Cr–C12
Cr–C14
Cr–C15
N1–C1
2.508(3)
1.92(1)
1.80(1)
1.89(1)
2.12(2)
1.86(1)
1.77(1)
1.93(2)
2.10(2)
2.03(2)
1.86(2)
1.86(2)
1.92(2)
1.89(2)
1.89(2)
1.18(2)
1.37(1)
1.45(2)
1.45(2)
C3–O3
C4–O4
C6–O6
C9–O9
C13–O13
C11–O11
C12–O12
C14–O14
C15–O15
C6–C7
C7–C8
C7–C16
C8–C9
1.15(2)
1.18(2)
1.19(2)
1.17(2)
1.16(2)
1.16(2)
1.12(2)
1.15(2)
1.16(2)
1.50(2)
1.37(2)
1.37(2)
1.49(2)
1.44(2)
1.42(2)
1.37(3)
1.42(2)
Identical conditions have been applied for the reac-
tion of 3 with NaCN and practically identical results
have been achieved (Section 4). Neither here nor in
the case of NaNH2 have we made any move to im-
prove the yields. Yet knowing that substitutions are
possible, we started to look at some more. Of the
potential ligands employed, however, only P(OMe)3
gave minimum amounts of a doubly phosphite-substi-
C8–C19
C16–C17
C17–C18
C18–C19
N2–C2
N2–C6
N2–C9
Bond angles (°)
Fe1–C2–Fe2
Fe1–C4–Fe2
C1–Cr–C13
C12–Cr–C15
C11–Cr–C14
C2–N2–C6
86.6(5)
82.0(5)
177.5(8)
176.6(7)
178.3(8)
125(1)
C6–C7–C8
C7–C8–C9
C6–N2–C9
C8–C7–C16
C7–C16–C17
C16–C17–C18
107(1)
112(1)
110.2(9)
126(1)
113(2)
124(2)
1
tuted derivative 15. Two singlets in the H NMR for
the Cp ligands clearly indicate that the substitutions
occurred at iron atoms related by a two-fold rotation
axis, the only symmetry element present in 3 (see
above and [5]). No reaction was observed with the
phosphines PPh3 and PBun3 .
N2–C6–O6
125(1)
Best planes
I: Fe2, C2, Fe1
II: Fe2, C4, Fe1
III: N2, C6, C9, C8, C7
IV: C7, C8, C19, C18, C17, C16
Interplanar angles (°)
I/II: 176.1(6)
III/IV: 179.6(4)
I/III: 150.3(4)
a E.s.d.s are given in parentheses, see Fig. 4 for atomic numbering.
b Average value.

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
273
ton and 13C NMR spectra were recorded on a Bruker AC
250andaBrukerSY270spectrometer. Silicagel(100–200
mm) from ICN and aluminium oxide (N, Akt. I) from
Woelm–Pharma were used for chromatography.
Complexes 2 (as PF6 salt) [6], 3 (as BF4 salt) [5],
NEt4[Cr(CN)(CO)5] and NEt4[W(CN)(CO)5] [21] were
prepared according to published procedures.
4.1. Reaction of 2 with pyrrolidine 5
A suspension of 0.35 g (0.55 mmol) of 2-PF6 in 10 ml
of THF was cooled by means of an ice-bath. To this
pyrrolidine (0.11 ml, 1.11 mmol) was dropped quickly
causing the suspension to clear up to a deep red solution.
For the reaction to be completed the mixture was stirred
for another 3 h at room temperature (r.t.). After removal
of the solvent the crude product was purified by chro-
matography on a silica charged column (2×15 cm) with
THF/ether (1/1) as eluent. The first red phase containing
the ring-opened product was collected, reduced to about
half of its volume and layered with light petroleum
(40–60°C). Within 24 h the product crystallized as deep
red thin rods. Yield 0.11 g (36%). M.p. 165–167°C. Anal.
Calc. for C26H22Fe2N2O5 (554.16 g mol−1): C, 56.35; H,
4.00; N, 5.06. Found: C, 56.32; H, 4.09; N, 5.01%. IR
(CH2Cl2, cm−1): 2001vs, 1995vs, 1957m [w(CO)term];
1794m-s [w(m-CO)]; 1691m,br [w(CꢀO)acyl]; 1630vs [w(m-
3.3. Nucleophilic additions with cluster fragmentation
For the reactions with diamines (1,2-diaminoethane,
1,4-diaminobutane)wechoseconditionsidenticaltothose
for the amine reactions (Section 2.1). In the first step the
reactions appear to proceed in the same fashion forming
terminal isocyanide ligands I which, however, are ‘func-
tional’, i.e. carry further NH2 groups in their side chains.
Functional isocyanides of this type are unstable both in
the free and coordinated state, immediately undergoing
cyclization to give (carbenoid) N,N%-heterocycles. In fact,
we observe species (16, 17) with this kind of ligand along
with a typical degradation product, [Fe2Cp2(CO)4] (Fp2),
yet they are mononuclear iron(II) compounds. In the end
it is probably the presence of three strong electron donors
in this array of four different ligands about the tetrahedral
center which causes the cluster to break up into fragments.
Offivesecondaryaminestestedforreactionswith3only
pyrrolidine gave rise, besides Fp2, to an identifiable
product 18. Actually, this is the first and only case in which
the carbido species [Fe2Cp2(CO)3m-C]2+ has formally
been transferred to a substrate (see Section 2.1).
1
CN)]. H NMR (l, CD3CN): 8.25–7.40 (4H, Ph); 5.04
(s, 10H, Cp); 3.61 (t, 2H, CH2, JHH 4.48 Hz); 3.21 (t, 2H,
CH2, JHH 4.48 Hz); 1.94 (m, 4H, CH2). 13C{H} NMR
(l, CDCl3): 268.8 (m-CO); 263.4 (m-CN); 209.8 (CO)term
;
155.4 (\CꢀN–CꢀO); 140.8 (\NCꢀO); 132.7, 131.7,
131.0, 128.5, 127.3, 127.1 (Ph); 88.1 (Cp); 48.5, 45.6
CH2N). MS (EI), m/z (rel. int.(%)): 554 (0.1) (M+); 526
(0.1) [M–CO]+; 498 (0.6) [M–2CO]+; 470 (0.1) [M–
3CO]+; 186 (59) [FeCp2]+.
4.2. Reaction of 2 with n-butyllithium
A suspension of 0.30 g (0.47 mmol) of 2-PF6 in 5 ml
of THF was cooled to −60°C. To this 0.2 ml of
n-butyllithium (15% in n-hexane) were dropped within 6
min, and the mixture was allowed to warm to r.t. over
a period of 24 h. The precipitate was separated and
extracted with 4 ml of dichloromethane leaving 3-PF6 as
a red powder. From the dichloromethane solution major
amounts of the educt 2-PF6 were recovered.
Some more 3-PF6 was precipitated from the red THF
phase on addition of 4 ml of diethyl ether, and the clear
supernatant solution was taken to dryness. The residue
was then chromatographed on silica (2×15 cm-column)
with light petroleum/ether (1/2) as eluent. A first pale
yellow fraction contained phthalimine (5 mg), the second
orange–red one contained 6 which after removal of the
solvent remained as an orange–red powder.
4. Experimental
All operations were carried out under an atmosphere
of pure argon using Schlenk-tube and vacuum techniques.
The organic solvents were dried, distilled and stored under
argon. IR spectra were run on Perkin–Elmer model 983
and Beckman model 4240 spectrometers. Microanalyses
(C, H, N) were obtained with a Heraeus CHN-Rapid
analyzer. Melting points were taken on a Gallenkamp
MFB-595 apparatus and are uncorrected. Mass spectra
(EI, pos-FAB) were obtained with Finnigan (Varian)
MAT711(excitationenergy80eV), MAT112SandMAT
CH5-DF (neutral xenon source at 3 keV, co-solvent
DMSO, matrix 3-nitrobenzyl alcohol) instruments. Pro-

274
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
3-PF6: Yield 48 mg (15%). M.p. 215–216°C. IR
Table 3
(Nujol mull, cm−1): 2005s, 1963m [w(CO)term]; 1834m
,
Selected bond distances (A) and bond angles (°), best planes and
interplanar angles (°) of 10 a
1
[w(m-CO)]; 1789s [w(CꢀNꢀC)]; 845s [w(PF6)]. H NMR
(l, CD3CN): 5.37, 5.12 (2s, Cp). MS (pos-FAB), m/z
(rel. int. (%)): 690 (5) (M+); 634 (2) [M–2CO]+; 324(5)
[Fe2(CN)Cp2(CO)2]+.
6: Yield 1 mg (0.3%). M.p. 65–66°C (dec.). IR (Nujol
mull, cm−1): 1994s, 1956m [w(CO)term]; 1774m [w(m-
CO)]; 1731s [w(CꢀO)ketone]; 1680m [w(CꢀO)acyl]; 1655m
[w(m-CN)]. MS (pos-FAB), m/z (rel. int. (%)): 542 (0.9)
[M+H]+; 324 (4) [Fe2(CN)Cp2(CO)2]+.
,
Bond distances (A)
W1–N2
W1–C6
W1–C7
W1–C9
W1–C10
W1–C8
Fe1–Fe2
Fe1–C1
Fe1–C2
Fe1–C4
Fe1–C(Cp) b
Fe2–C1
Fe2–C2
Fe2–C3
Fe2–C(Cp) b
N1–C1
2.15(2)
1.96(3)
1.93(2)
2.01(4)
1.92(2)
1.82(3)
2.512(6)
1.95(3)
1.88(2)
1.76(2)
2.11(3)
2.01(2)
1.88(3)
1.75(2)
2.12(3)
1.47(2)
1.40(4)
1.41(3)
1.18(3)
C2–O2
C3–O3
C4–O4
C6–O6
C7–O7
C9–O9
C10–O10
C8–O8
C11–O11
C12–O12
C1–C5
C11–C18
C12–C13
C13–C14
C13–C18
C14–C15
C15–C16
C16–C17
C17–C18
1.19(3)
1.15(3)
1.11(3)
1.18(4)
1.21(3)
1.15(5)
1.20(3)
1.24(4)
1.22(3)
1.18(3)
1.45(3)
1.44(3)
1.53(3)
1.40(3)
1.35(4)
1.50(4)
1.33(4)
1.44(3)
1.40(4)
4.3. Reaction of 2 with NaCN
A mixture of 0.43 g (0.68 mmol) of 2-PF6 and
0.033 g (0.68 mmol) of NaCN was stirred in 1,2-
dimethoxyethane (15 ml) for 14 h. The deep-red suspen-
sion was filtered through a G4 frit, the solid was
recrystallized from THF/light petroleum and identified
as the tetraferrio-azaallenium species 3-PF6 by its IR
spectrum. Yield 40 mg.
N1–C11
N1–C12
N2–C5
Bond angles (°)
N2–W1–C8
C6–W1–C9
C7–W1–C10
N1–C1–C5
Fe1–C1–Fe2
Fe1–C2–Fe2
Fe1–C4–O4
Fe2–C3–O3
N1–C11–C18
N1–C11–O11
N1–C12–C13
N1–C12–O12
The red filtrate was cooled to −78°C causing precip-
itation of 20 mg of a colourless crystalline material
which according to the IR spectrum was sodium hexa-
fluorophosphate. On addition of light petroleum (3 ml)
product 7 crystallized in deep-red lumps within 13 h at
0°C. Yield 5 mg (14%). M.p. 120–123°C. Anal. Calc.
for C23H14Fe2N2O5 (510.07 g mol−1): C, 54.16; H, 2.77;
N, 5.49. Found: C, 51.86; H, 3.09; N, 5.17%. IR
(CH2Cl2, cm−1): 2166m [w(CN)term]; 2007vs, 1971m
[w(CO)term]; 1805s,br, 1777w [w(m-CO)]; 1712s, 1702s
177(1)
176(1)
176(1)
104(2)
78.9(8)
84(1)
176(2)
175(2)
111(2)
123(2)
105(2)
125(2)
C1–N1–C11
127(2)
108(2)
107(2)
110(2)
110(2)
121(2)
123(2)
123(2)
115(3)
C11–N1–C12
C11–C18–C13
C12–C13–C18
C13–C14–C15
C13–C18–C17
C14–C15–C16
C15–C16–C17
C16–C17–C18
C41–O44–C411
C42–C41–O44
116(3)
106(3)
1
[w(CꢀO)acyl]. H NMR (l, CD3CN): 7.80 (mc, 4H, Ph);
5.10 (s, 10H, Cp). 13C{H} NMR (l, CD2Cl2): 264.1
(m-CO); 210.2 (CO)term; 134.5 (CN); 132.0, 123.7, 123.3
(Ph); 92.3 (Cp). MS (EI), m/z (rel. int. (%)): 482 (4)
[M–CO]+; 454 (9) [M–2CO]+; 426 (66) [M–3CO]+;
186 (100) [FeCp2]+; 147 (16) [phthalimine]+.
Best planes
I: Fe1, C1, Fe2
II: Fe2, C2, Fe1
III: N1, C11, C12, C13, C18
IV: C13, C14, C15, C16, C17
Interplanar angles (°)
I/II: 163(1)
III/IV: 179.3(7)
a E.s.d.s are given in parentheses, see Fig. 2 for atomic numbering.
b Average value.
4.4. Reaction of 2 with tetraethylammonium pentacar-
bonylcyanochromate, NEt4[Cr(CN)(CO)5]. Complex 9
In total 7 ml of dichloromethane were added to a dry
mixture of the reactants NEt4[Cr(CN)(CO)5] (0.28 g,
0.79 mmol) and 2-PF6 (0.50 g, 0.79 mmol), and the
suspension was stirred at 32°C for 12 h. The solvent
was removed in an oil vacuum, and the copper-
coloured solid was extracted with 3 ml of ether. After
filtration through a G4 frit small, deep violet–red par-
allelepipeds together with needle-like fragments crystal-
lized from the deep-red solution at −30°C. Yield 15
mg (2%). M.p. 143°C (dec.). IR (KBr, cm−1): 2137m-w
[w(CN)term]; 2060m, 1991sh, 1977vs, 1938vs, 1908vs,
1887vs [w(CO)term]; 1820s [w(m-CO)]; 1713m-s
[M–5CO]+; 399 (29) [M–8CO–CrCNH]+; 268 (13)
[Fe2Cp2CN]+.
After purification by recrystallization from dichloro-
methane the residue proved to be pure 3-PF6.
4.5. Reaction of 2 with tetraethylammonium pentacar-
bonylcyanotungstate, NEt4[W(CN)(CO)5]. Complex 10
Mixing of the dry components 2-PF6 (0.48 g, 0.76
mmol) and NEt4[W(CN)(CO)5] (0.37g, 0.77 mmol) was
followed by the addition of 7 ml of dichloromethane
and warming of the reaction mixture to 35°C. After 13
h the solvent was removed in an oil vacuum, and the
1
[w(CꢀO)acyl]. H NMR (l, CD3CN): 8.32–8.14 (mc, 4H,
Ph); 5.00 (s, 5H, Cp); 4.82 (s, 4.9H, Cp). MS (pos-
FAB), m/z (rel. int. (%)): 674 (1.7) (M+); 534 (6)

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
275
4.8. Reaction of 3 with 2-phenylethylamine. Complex
13
resulting copper-coloured solid was extracted with 4 ml
of ether. Within 18 h at −30°C, the product crystal-
lized in deep-red rhomboids and blocks. Yield 70 mg
To a stirred suspension of 3-BF4 (0.27 g, 0.35 mmol)
(11%).
M.p.
149–151°C.
Anal.
Calc.
for
C28H14Fe2N2O10W (833.96 g mol−1): C, 40.33; H, 1.69;
N, 3.36. Found: C, 41.35; H, 2.81; N, 3.24%. IR (KBr,
cm−1): 2175m [w(CN)term]; 2068m, 2008s, 1910vs,br,
1907vs,br [w(CO)term]; 1809s [w(m-CO)]; 1713m-s
in 5 ml of 1,4-dioxane was added phenylethylamine
(0.05 ml, 0.40 mmol) within 3 min. The suspension was
then heated to 114°C for 11 h. When it cooled to r.t.
the suspension was filtered through a G4 frit to afford
190 mg of pure 3-BF4. Treatment with 5 ml of ether
followed by 7 ml of light petroleum as described above
(Sections 4.6 and 4.7) resulted in 5 mg (16%) of a
wine-red powder. M.p. 110–111°C. IR (KBr, cm−1):
1
[w(CꢀO)acyl]. H NMR (l, CD3CN): 7.82–7.16 (m, 4H,
Ph); 5.20, 4.85 (2s, 10H, Cp). 13C{H} NMR (l,
CD3CN): 262.3 (m-CO); 211.1 (CO)term
;
202.0
(WCO)ax; 197.3 (WCO)eq; 169.3 (CꢀO)acyl; 137.7
(CNW); 135.5, 132.2, 121.4 (Ph); 107.7, 92.6 (Cp). MS
(pos-FAB), m/z (rel. int. (%)): 834 (1) (M+); 806 (8)
[M–CO]+; 722 (4) [M–4CO]+; 638 (4) [M–7CO]+;
610 (14) [M–8CO]+.
2149m-w
[w(CN)term];
2055w,
2001s,
1965s-m
[w(CO)term]; 1763vs [w(CꢀN1ꢀC)]; 1083m, 1048s
[w(BF4)]. MS (pos-FAB), m/z (rel. int. (%)): 793 (1)
(M+); 737 (0.5) [M–2CO]+.
4.9. Reaction of 3 with sodium cyanide. Complex 14 by
substitution
4.6. Reaction of 3 with n-propylamine. Complex 11
n-Propylamine (0.016 ml, 0.2 mmol) was added to a
suspension of 3-BF4 (0.16 g, 0.2 mmol) in 3 ml of
1,4-dioxane, and the mixture was stirred at 108°C for
12 h. After cooling to r.t. the red solution was filtered
and mixed with 5 ml of ether to remove unreacted
3-BF4. Upon layering with another 5 ml of ether the
product separated overnight as a red powder. Yield 30
mg (15%). M.p. 115–116°C. IR (KBr, cm−1): 2158m
[w(CN)term]; 2055w, 2001s, 1964m-w [w(CO)term]; 1810sh
[w(m-CO)]; 1768vs [w(CꢀNꢀC)]; 1048vs [w(BF4)]. 1H
NMR (l, CD3CN): 5.33, 5.21, 5.12 (3s, 20H, Cp). MS
(pos-FAB), m/z (rel. int. (%)): 731 (20) (M+); 703
[M–CO]+; 675 (11) [M–2CO]+; 470 (3) [M–5CO–
FeCp]+.
The dry reactants 3-PF6 (0.63 g, 0.75 mmol) and
NaCN (0.04 g, 0.81 mmol) were thoroughly mixed,
then 1,4-dioxane (10 ml) was added. The suspension
was stirred for 15 h at 128°C after which time the
solvent was removed in vacuo. From the red solid the
product was extracted with a total of 30 ml of
dichloromethane. The burgundy-red opalescent solu-
tion was filtered through a G4 frit which was charged
with a thin layer of silica. Addition to the solution of
an equal amount of ether causes the microcrystalline
product to separate overnight. Yield 20 mg (4%). M.p.
245°C (dec.). IR (KBr, cm−1): 2082m [w(CN)term];
2001s, 1966s [w(CO)term]; 1812w [w(m-CO)]; 1757vs, br
[w(CꢀNꢀC)]; (CH2Cl2, cm−1): 2090w [w(CN)term];
2005m, 1975m [w(CO)term]; 1813w [w(m-CO)]; 1764s
1
[w(CꢀNꢀC)]. H NMR (l, CD3CN): 5.17 (s, 10H, Cp),
4.7. Reaction of 3 with 2-methoxyethylamine. Complex
12
4.87 (s, 10H, Cp). MS (pos-FAB), m/z (rel. int. (%)):
689 (0.9) ({M+H}+); 661 [{M+H}–CO]+; 633
[{M+H}–2CO]+. MS (neg-FAB), m/z (rel. int. (%)):
660 (14) [M–CO]−; 632 (14) [M–2CO]−; 604 (19)
[M–3CO]−; 483 (12) [M–3CO–FeCp]−; 362 (9) [M–
3CO–Fe2Cp2]−.
A mixture of 0.40 g (0.52 mmol) of 3-BF4 and 0.05
ml (0.52 mmol) of 2-methoxyethylamine in 8 ml of
1,4-dioxane was stirred at 108°C for 16 h after which
time it was allowed to cool to r.t. The solid that had
separated from the solution was identified as unreacted
starting material (140 mg). A second crop of 3-BF4
formed on addition of 12 ml of ether to the deep red
solution and was filtered off. Layering with 10 ml of
light petroleum of the solution caused crystallization of
the product as a fine red–violet powder. Yield 10 mg
(2%). M.p. 110–115°C. IR (KBr, cm−1): 2155m-w,
2091w [w(CN)term]; 2055w, 2001s, 1961w [w(CO)term];
1768vs [w(CꢀNꢀC)]; 1051s [w(BF4)]. 1H NMR (l,
CDCl3): 5.28 (s, 5H, Cp), 5.08 (s, 10H, Cp), 4.84 (s, 5H,
Cp); 3.72 (s, 3H, OMe); 3.48 (mc, 4H, 2CH2). MS
(pos-FAB), m/z (rel. int. (%)): 747 (39) (M+); 719 (3)
[M–CO]+; 691 (26) [M–2CO]+; 663 (3) [M–3CO]+;
634 (14) [M–CO–C4H7NO].
4.10. Reaction of 3 with sodium amide. Complex 14 by
CO“CN− intercon6ersion
Sodium amide (0.08 g, 1.93 mmol) was added in
small doses to a suspension of 3-BF4 (1.00 g, 1.29
mmol) in 17 ml of 1,4-dioxane and the mixture was
stirred for 21 h at 104°C. The still hot suspension was
then filtered through a G4 frit. After cooling to r.t. the
filtrate was mixed with ether (15 ml) in order to remove
some dissolved 3-BF4. Layering of the solution with
light petroleum (10 ml) overnight resulted in the precip-
itation of a violet powder which was washed with more
light petroleum and dried in vacuo. Yield 80 mg (9%).
For further data, cf. Section 4.9.

276
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
Table 4
Crystallographic data collection parameters of 7, 9 and 10
7
9
10
Crystal data
Formula
C
510.07
0.60×0.25×0.32
monoclinic
23H14Fe2N2O5
C27H14CrFe2N2O9
674.12
0.275×0.075×0.05
triclinic
C28H14Fe2N2O10W+C4H10
908.08
0.35×0.28×0.15
monoclinic
O
Molecular weight (g mol−1
Crystal size
)
Crystal system
(
Space group
P21/c (no. 14)
P1 (no. 2)
C2/c (no. 15)
Cell parameters
a (pm)
b (pm)
c (pm)
h (°)
1408.3(2)
1224.1(1)
1182.2(1)
90.0
94.38(3)
90.0
2332.0(5)
4
1.453
1059.8
1208.3
1244.1
105.15
94.81
115.35
1354.26
2
3374.1(4)
1121.3(2)
1921.5(3)
90.0
122.33(1)
90.0
6143.02
8
1.963
i (°)
k (°)
U (106 pm3)
Z
Dc (g cm−3
)
1.653
Data collection and refinement parameters
Radiation, u (pm)
Temperature (K)
graphite-monochromated Mo Ka, 71.069
293
293
293
Scan mode
2q limits (°)
ꢀ–2q scan
452q540
2136
1700
284
0.080
0.101
1.0/−1.1
ꢀ scan
452q550
3767
2230
370
0.057
0.062
0.9/−1.5
ꢀ–2q scan
452q544
3016
1884
198
0.070
0.042
1.9/−3.4
Total no. reflections
No. data [I52|(I)]
No. parameters
a
R
b
Rw
Highest residual electron-density peaks (10−6 e pm−3
)
a R=SꢀFoꢁ−ꢁFcꢀ/SꢁFoꢁ.
2
2
b Rw=[S(wꢀFoꢁ−ꢁFcꢀ )/S(wꢁFoꢁ )]1/2, w=1.0 (unit weights).
4.12. Reactions of 3 with diamines. Complexes 16 and
4.11. Reaction of 3 with trimethylphosphite. Complex
17
15
A slight excess of 1,2-diaminoethane (0.05 ml, 0.61
mmol) was added dropwise to a suspension of 3-BF4
(0.40 g, 0.5 mmol) in 1,4-dioxane (8 ml), and the
mixture was heated to about 110°C. After 15 h the
red suspension was allowed to cool to r.t. and was
filtered (G4). The filtrate was slightly concentrated
and layered with ether whereupon the yellow-ochre
product (16) crystallized.
The reaction with 1,4-diaminobutane was carried
out in the same way to give 17. The main product
(97 mg) was Fp2.
16: Yield 20 mg (16%). M.p. 132°C. IR (KBr,
cm−1): 3230s [w(NH)]; 2083vs [w(CN)term]; 1931vs,
1896vw [w(CO)term]; 1514s [w(N–C–N)=l(NH)]. MS
(pos-FAB), m/z (rel. int. (%)): 246 (4) [M+H]+; 217
(6) [M–CO]+; 191 (100) [M–CO–CN]+; 71 (37)
[C3H6N2]+.
To 3-BF4 (0.20 g, 0.26 mmol) suspended in 1,4-
dioxane (4 ml) trimethylphosphite (0.1 ml, 0.8 mmol)
was added dropwise within a few minutes. The red
suspension was kept stirring at 116°C for 14 h after
which time it was allowed to cool and filtered
through a G4 frit. The filtrate was layered with 5 ml
of light petroleum. Within 14 h a violet powder had
separated which was washed with a small amount of
ether and dried in vacuo. Yield 1 mg (0.4%). M.p.
120–121°C. Anal. Calc. for C32H38BF4Fe4NO10P2
(968.78 g mol−1): C, 39.67; H, 3.95; N, 1.44. Found:
C, 40.98; H, 5.60; N, 1.39%. IR (KBr, cm−1): 1994s,
1962vs [w(CO)term]; 1748vs, br [w(CꢀNꢀC)]; 1177s,
1121s [w(C–O)]; 1049s [w(BF4)]; 1018s [w(P–O)]. 1H
NMR (l, CDCl3): 5.02, 4.88 (2s, 20H, Cp); 3.58 (s,
18H, Me). MS (pos-FAB), m/z (rel. int. (%)): 882
(52) (M+); 730 (10) [M–CO–P(OMe)3]+; 245 (33)
[FeCp{P(OMe)3}]+.
17: Yield 10 mg (8%). M.p. 100–101°C. IR (KBr,
cm−1): 3404m, 3274m [w(NH)]; 2076s [w(CN)term];

K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
277
Acknowledgements
1947s, 1881vw [w(CO)term]; 1535m, 1516m-s [w(N–C–
N)+l(NH)]. MS (pos-FAB), m/z (rel. int. (%)): 274
(100) [M+H]+; 245 (81) [M–CO]+; 219 (91) [M–
CO–CN]+; 99 (48) [C5H11N2]+.
We thank the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie for financial
support. We are also grateful to Professor Dr J.
Fuchs and Professor Dr H. Hartl for their help with
the X-ray structures, and to Manuel A. Fehlhammer
for the computer drawings.
4.13. Reaction of 3 with pyrrolidine. Complex 18
A suspension of 3-BF4 (0.30 g, 0.40 mmol) in 7 ml
of 1,4-dioxane was reacted with pyrrolidine (55 mg,
0.8 mmol) at 105°C. After 13 h the solvent was re-
moved and the red residue was washed several times
with light petroleum. By treating the residue with
light petroleum/ether (1/4) 40 mg of Fp2 were ex-
tracted. Recrystallization from THF/ether gave
the known product. Yield 2.1 mg (2%). M.p. 190–
194°C. IR (KBr, cm–1): 2018vs, 1999s [w(CO)term];
1818s [w(m-CO)]; 1591s [w(m-CNB)]; 1077s, 1051s
[w(BF4)]. MS (pos-FAB), m/z (rel. int. (%)): 408 (13)
[M+]; 380 (3) [M–CO]+; 352 (5) [M–2CO]+; 324 (3)
[M–3CO]+.
References
[1] W.P. Fehlhammer, M. Fritz, Chem. Rev. 93 (1993) 1243.
[2] W.P. Fehlhammer, A. Schro¨der, F. Schoder, J. Fuchs, A. Vo¨lkl,
B. Boyadjiev, S. Schro¨lkamp, J. Organomet. Chem. 411 (1991)
405.
[3] W.P. Fehlhammer, F. Schoder, G. Beck, S. Schro¨lkamp, Z.
Anorg. Allg. Chem. 619 (1993) 1171.
[4] W.P. Fehlhammer, A. Schro¨der, W. Sperber, J. Fuchs, Chem.
Ber. 125 (1992) 1087.
[5] W.P. Fehlhammer, A. Schro¨der, J. Fuchs, E.-U. Wu¨rthwein,
Angew. Chem. 104 (1992) 604; Angew. Chem., Int. Ed. Engl. 31
(1992) 590.
[6] K.-H. Trylus, A. Schro¨der, R. Thiel, I. Bru¨dgam, W.P. Fehlham-
mer, Inorg. Chim. Acta 269 (1998) 23.
4.14. X-ray structure determinations
[7] G. Beck, W.P. Fehlhammer, Angew. Chem. 100 (1988) 1391;
Angew. Chem., Int. Ed. Engl. 27 (1988) 1344.
Single crystals of 7 were obtained by layering a
solution in 1,2-dimethoxyethane with light petroleum,
those of 9 and 10 by cooling their etheral solution to
−30°C. The latter two were mounted in glass capil-
laries to prevent oxidative decomposition. The crystal
symmetries and approximate cell parameters were de-
termined by means of Weissenberg photographs.
The intensity data of compound 9 were collected
on a STOE four-circle diffractometer, those of 7 and
10 on a Nonius CAD 4. All crystal structures were
solved by direct methods and refined by full-matrix
least-squares calculations based on Sw(ꢁFoꢁ−ꢁFcꢁ)2 ini-
tially with isotropic, then with anisotropic thermal
parameters for all non-hydrogen atoms.
Calculations were carried out on a VAX computer
using the programs SHELX-86 [22], XTAL 3.0 [23], DI-
FABS [24] and SCHAKAL-88 [25]. Table 4 contains ad-
ditional crystal and data collection parameters;
selected bond lengths and angles are listed in Tables
1–3. For further information, see Section 5.
[8] (a) S. Willis, A.R. Manning, F.S. Stephens, J. Chem. Soc.,
Dalton Trans. (1979) 23. (b) S. Willis, A.R. Manning, F.S.
Stephens, J. Chem. Soc., Dalton Trans. (1980) 186. (c) S. Willis,
A.R. Manning, J. Chem. Soc., Dalton Trans. (1981) 322. (d)
A.R. Manning, P. Murray, J. Chem. Soc., Dalton Trans. (1986)
2399. (e) G. Cox, C. Dowling, A.R. Manning, P. McArdle, D.
Cunningham, J. Organomet. Chem. 438 (1992) 143.
[9] F. Schoder, Ph.D. Thesis, University of Erlangen-Nu¨rnberg,
1985.
[10] W.P. Fehlhammer, H. Stolzenberg, in: G. Wilkinson, F.G.A.
Stone, E.W. Abel (Eds.), Comprehensive Organometallic Chem-
istry, vol. 4, Pergamon, Oxford, 1982, pp. 531, 532.
[11] S. Bordoni, L. Busetto, C. Camiletti, V. Zanotti, V.G. Albano,
M. Monari, F. Prestopino, Organometallics 16 (1997) 1224.
[12] V.G. Albano, S. Bordoni, D. Braga, L. Busetto, A. Palazzi, V.
Zanotti, Angew. Chem. 103 (1991) 895; Angew. Chem., Int. Ed.
Engl. 30 (1991) 847.
[13] W.P. Fehlhammer, F. Degel, H. Stolzenberg, Angew. Chem. 93
(1981) 184; Angew. Chem., Int. Ed. Engl. 20 (1981) 214.
[14] (a) E. Ba¨r, W.P. Fehlhammer, D.K. Breitinger, J. Mink, Inorg.
Chim. Acta 82 (1984) L17. (b) M. Fritz, D. Rieger, E. Ba¨r, G.
Beck, J. Fuchs, G. Holzmann, W.P. Fehlhammer, Inorg. Chim.
Acta 198–200 (1992) 513.
[15] H. Berke, U. Gro¨ßmann, G. Huttner, L. Zsolnai, Chem. Ber.
117 (1984) 3432.
[16] R. Kupfer, E.-U. Wu¨rthwein, M. Nagel, R. Allmann, Chem.
Ber. 118 (1985) 643.
5. Supplementary material
[17] (a) W.P. Fehlhammer, A. Mayr, B. Olgemo¨ller, Angew. Chem.
87 (1975) 290; Angew. Chem., Int. Ed. Engl. 14 (1975) 369. (b)
W.P. Fehlhammer, A. Mayr, M. Ritter, Angew. Chem. 89 (1977)
660; Angew. Chem., Int. Ed. Engl. 16 (1977) 641. (c) W.P.
Fehlhammer, A. Mayr, J. Organomet. Chem. 191 (1980) 153.
[18] (a) W.P. Fehlhammer, G. Christian, A. Mayr, J. Organomet.
Chem. 199 (1980) 87. (b) W.P. Fehlhammer, A. Mayr, G.
Christian, J. Organomet. Chem. 209 (1981) 57. (c) W.P.
Fehlhammer, P. Hirschmann, A. Vo¨lkl, J. Organomet. Chem.
302 (1986) 379.
Further details of the crystal structures may be ob-
tained upon request from the Fachinformations-
zentrum Karlsruhe, Gesellschaft fu¨r wissenschaftlich-
technische Information mbH, D-76344 Eggenstein-
Leopoldshafen (Germany), giving reference to the de-
pository number CSD 59463 and citing the authors
and this paper.

278
K.-H. Trylus et al. / Inorganica Chimica Acta 291 (1999) 266–278
[19] W.P. Fehlhammer, A. Mayr, Angew. Chem. 87 (1975) 776;
Angew. Chem., Int. Ed. Engl. 14 (1975) 757.
[20] (a) U. Wannagat, H. Seyffert, Angew. Chem. 77 (1965) 457;
Angew. Chem., Int. Ed. Engl. 4 (1965) 438. (b) R.B. King,
Inorg. Chem. 6 (1967) 25. (c) H. Behrens, Adv. Organomet.
Chem. 18 (1980) 1.
[22] G. Sheldrick, SHELX-86 program, Institut fu¨r Anorganische
Chemie der Universita¨t Go¨ttingen, Germany, 1986.
[23] S.R. Hall, J.M. Stewart, XTAL 3.0, University of Western Aus-
tralia and University of Maryland.
[24] N. Walker, D. Stuart, Acta Crystallogr., Sect. A 39 (1983)
158.
8
[21] W.P. Fehlhammer, W.A. Herrmann, K. Ofele, in: G. Brauer
[25] E. Keller, SCHAKAL, program for the graphic representation of
molecular and crystallographic models, Chemisches Institut der
Albert-Ludwig-Universita¨t, Freiburg, Germany, 1988.
(Ed.), Handbuch der pra¨parativen anorganischen Chemie, vol.
III, 3rd ed., Ferdinand Enke, Stuttgart, Germany, 1981.
.
.