Reaction of unconjugated dienes with [Fe(R2P(CH2)nPR2)] species

Article abstract of DOI:10.1021/om9610004

The effect of varying the bidentate ligand upon the reaction of unconjugated dienes with (R₂P(CH₂)_nPR₂)Fe⁰ species has been investigated. Depending upon the nature of the substituents at phosphorus and upon the length of the hydrocarbon chain joining the phosphorus atoms, either (η²:η²-diene)Fe(R₂P(CH ₂)_nPR₂) or (η⁵-dienyl)Fe(R₂P(CH₂)_nPR ₂)H species are formed. Extended Hu?ckel MO-calculations have been used to assess the importance of electronic factors in influencing the course of reaction, and it can be shown that distortion of the tetrahedral geometry of the (diene)Fe species causes a destabilization.

Full text of DOI:10.1021/om9610004

1612
Organometallics 1997, 16, 1612-1620
Rea ction of Un con ju ga ted Dien es w ith
[F e(R2P (CH2)n P R2)] Sp ecies†
S. Geier, R. Goddard, S. Holle, P. W. J olly,* C. Kru¨ger, and F. Lutz
Max-Planck-Institut fu¨r Kohlenforschung, D-45466 Mu¨lheim an der Ruhr, Germany
Received November 27, 1996X
The effect of varying the bidentate ligand upon the reaction of unconjugated dienes with
(R2P(CH2)nPR2)Fe0 species has been investigated. Depending upon the nature of the
substituents at phosphorus and upon the length of the hydrocarbon chain joining the
phosphorus atoms, either (η2:η2-diene)Fe(R2P(CH2)nPR2) or (η5-dienyl)Fe(R2P(CH2)nPR2)H
species are formed. Extended Hu¨ckel MO-calculations have been used to assess the
importance of electronic factors in influencing the course of reaction, and it can be shown
that distortion of the tetrahedral geometry of the (diene)Fe species causes a destabilization.
Sch em e 1
In tr od u ction
Recently we reported that the structure of the com-
pounds formed upon reacting Fe(iPr2P(CH2)nPiPr2)Cl2
(n ) 1-3) with active Mg and acyclic dienes depends
upon the nature of the diene and the length of the
methylene chain joining the two P atoms of the biden-
tate ligand.1 For example, 1,5-hexadiene reacts with
the Fe(iPr2P(CH2)nPiPr2)fragment to give the (η5-2,4-
hexadien-1-yl)iron hydride 1 in the presence of the
ethylene-bridged ligand and to give the (η2:η2-1,5-
hexadiene)Fe species 2 in the presence of the trimeth-
ylene-bridged ligand (Scheme 1). Furthermore, neither
compound shows any tendency to interconvert below the
decomposition temperature.
We have now extended our investigations to include
other substituted diphosphines and unconjugated dienes.
Here we report the synthetic results and discuss the
effect of electronic factors upon the course of reaction.
In a subsequent publication we will present a theoretical
interpretation on the basis of molecular modeling.
(diisopropylphosphino)ethane-stabilized zerovalent-iron
fragment to give (η5-dienyl)Fe-hydrido complexes (1 and
3-5; eq 1) as yellow to orange solids. The yield
Resu lts a n d Discu ssion
The reactions investigated fall into three groups,
which will be discussed separately: (1) the reaction of
(iPr2PC2H4PiPr2)Fe0 species with a variety of noncon-
jugated dienes, (2) the reaction of the same dienes with
the (iPr2PC3H6PiPr2)Fe0 species, and (3) the reaction of
1,5-hexadiene with a series of R2P(CH2)nPR2-stabilized
iron species as well as with (Et3P)2Fe0. In all three
cases, the (bidentate phosphine)Fe0 fragment was pre-
pared in situ by reducing FeCl2‚nTHF with active Mg
(prepared by vacuum pyrolysis of MgH2)2 in the pres-
ence of the ligand at low temperature.
decreases with the separation of the double bonds, being
ca. 70% for 1,5-hexadiene and ca. 20% for 1,8-nonadiene.
The propyl-substituted compound 4 could not be isolated
analytically pure and has been identified by comparison
Rea ction of (iP r 2P C2H4P iP r 2)F e0 w ith Non con -
ju ga ted Dien es. 1,5-Hexadiene, 1,6-heptadiene, 1,7-
octadiene, and 1,8-nonadiene all react with the bis-
1
of the spectroscopic data (31P NMR δ 119.4, 113.7; H
NMR δ(FeH) - 18.22, J (P,H) ) 73.0 Hz) with those for
1, 3 and 5. In contrast, no stable product could be be
isolated in the presence of 1,4-pentadiene (n ) 1),
presumably because strain prevents the complexation
of both double bonds to one metal center.
Compounds 3 and 5 have been identified by compar-
ing the spectroscopic data with those for 1, whose crystal
structure has been confirmed by X-ray diffraction.1 Of
* To whom correspondence should be addressed.
† Part of the doctoral thesis of S.G., submitted to the Ruhr-
Universita¨t Bochum in J anuary 1996.
X Abstract published in Advance ACS Abstracts, March 15, 1997.
(1) Gabor, B.; Goddard, R.; Holle, S.; J olly, P. W.; Kru¨ger, C.; Mynott,
R.; Wisniewski, S. Z. Naturforsch. 1995, 50B, 503.
(2) Bartmann, E.; Bogdanovic´, B.; J anke, N.; Liao, S.; Schlichte, K.;
Spliethoff, B.; Treber, J .; Westeppe, U.; Wilczok, U. Chem. Ber. 1990,
123, 1517.
S0276-7333(96)01000-X CCC: $14.00 © 1997 American Chemical Society

Reaction of Dienes with [Fe(R2P(CH2)nPR2)]
Organometallics, Vol. 16, No. 8, 1997 1613
Sch em e 2
is a compound (7) containing 5-methyl-1,5-heptadiene
as the result of a CO-induced isomerization of the diene.
The structure of 7 follows from the NMR spectroscopic
data.
Rea ction of (iP r 2P C3H6P iP r 2)F e0 w ith Non con -
ju ga ted Dien es. The dienes mentioned above all react
with the bis(diisopropylphosphino)propane-stabilized
zerovalent-iron species to give the (η2:η2-diene)Fe(iPr2-
PC3H6PiPr2) compounds 2 and 8-10 (eq 3) as dark green
paramagnetic compounds. The compounds decompose
diagnostic use are the absorptions at ca. 1900 cm-1 in
the IR spectra and at ca. -18 ppm in the 1H NMR
spectra, which we assign to the Fe-H moiety. Mecha-
nistically relevant is the observation that, irrespective
of the separation of the double bonds, the product
invariably contains a dienyl group substituted in the
5-position (and not, for example, in the 1,5-positions).
The driving force for the reaction is presumably the
formation of a stable 18-electron compound, and it is
assumed to proceed by initial coordination of the diene
to the metal atom followed by H-transfer from a meth-
ylene group to the metal and sequential Fe-H addi-
tion-elimination steps. This is illustrated in Scheme
2 for 1,5-hexadiene. The decrease in yield with increas-
ing alkene separation is probably the result of the
difference in the stability of the initial (η2:η2-diene)Fe
species and in the increasing importance of side reac-
tions during the isomerization steps.
The suggested initial C-H activation shown above is
supported by results with 3-methyl-1,6-heptadiene: the
dark green, paramagnetic (η2:η2-diene)Fe0 species 6 is
formed in good yield (eq 2) and not a complex related to
3. Furthermore, 6 shows no tendency to rearrange
above -10 to 0 °C without rearranging to the corre-
sponding (η5-dienyl)Fe-H derivative.
The crystal structures of 2 and 8 have been estab-
lished by X-ray diffraction. The hexadiene molecule in
2 was shown previously1 to adopt two orientations at
the metal which differ in the cisoid (Cs, local symmetry)
arrangement (70% occupancy) or transoid (C2, local
symmetry) arrangement (30% occupancy) of the double
bonds. The heptadiene molecule in 8 also adopts two
arrangements (occupancy 75%:25%) but in contrast to
2 it is not possible to determine whether the diene
occurs solely in the cisoid configuration with a different
orientation at the metal atom (occupancy 100%) or as a
mixture of cisoid and transoid arrangements (75%:25%).
The molecular structure of the main conformer is shown
in Figure 1 (minor component dashed), and selected
structural data are listed in Table 1. The iron atom in
8 has a distorted-tetrahedral geometry: a decrease in
the P-Fe-P angle from the tetrahedral value is ac-
companied by an increase in the D1-Fe-D2 angle
(where D1 and D2 are the central points of the double
bonds). The preferred chair conformation of the diene
presumably reflects a more strain-free complexation, an
effect which has also been noted for a number of
zerovalent nickel-diene compounds.3,4
below the decomposition point (0 °C). Apparently the
presence of a methyl group in the 3-position prevents
the hydrocarbon chain joining the two double bonds
from approaching the metal atom and suppresses
further reaction.
The structure of 6 has been established by a crystal
structure determination. Although disorder in both the
bidentate ligand and in the diene prevented refinement
of the data (R > 10%), the tetrahedral environment
around the metal atom and the conformation of the
organic ligand could be confirmed. In addition, the mass
spectrum of 6 contains fragments typical for 3-methyl-
1,6-heptadiene.
The iron atom in 2 and 8-10 is both coordinatively
and electronically unsaturated, and a facile reaction
with CO is observed. As discussed above for 7, these
reactions are probably accompanied by isomerization of
the diene, and in all cases the 31P NMR spectra indicate
(3) Proft, B.; Po¨rschke, K. R.; Lutz, F.; Kru¨ger, C. Chem. Ber. 1991,
124, 2667.
(4) Dreher, E.; Gabor, B.; J olly, P. W.; Kopiske, C.; Kru¨ger, C.;
Limberg, A.; Mynott, R. Organometallics 1995, 14, 1893.
Careful identification of 6 was necessary because the
product of the further reaction with CO is a mixture of
three diamagnetic adducts. The main component (46%)

1614 Organometallics, Vol. 16, No. 8, 1997
Geier et al.
hexadiene (which lead to formation of 1 and 2), we have
also investigated the reactions where n ) 1, 4, or 5. In
the presence of bis(diisopropylphosphino)methane (n )
1), a dark black-brown reaction mixture is formed at
low temperatures, while the reaction with bis(diisopro-
pylphosphino)pentane (n ) 5) leads to the formation of
a green compound which decomposes in solution even
at -78 °C and could not be obtained analytically pure;
therefore, it was not investigated further. The reaction
in the presence of bis(diisopropylphosphino)butane was
more successful, and the (η2:η2-1,5-hexadiene)Fe com-
plex 11 was obtained in good yield. This compound
shows no tendency to react further below the decompo-
sition point while, here again, the reaction with CO
leads to isomerization of the diene and to the formation
of a mixture of adducts, among which the (η4-cis,trans-
2,4-hexadiene)Fe0 complex 12 was the main component
(67%; 31P NMR) (eq 4). The geometry of the diene in
F igu r e 1. Molecular structure of (η2:η2-1,6-heptadiene)Fe
(iPr2PC3H6PiPr2) (8), showing the disorder in the heptadi-
ene ligand (minor component (25%) represented by dashed
lines). D1 and D2 are the midpoints of the C16A-C17A
and C21A-C22A bonds.
Ta ble 1. Str u ctu r a l Da ta for
(η2:η2-1,6-C7H12)F e(iP r 2P C3H6P iP r 2) (8)a
12 follows from a comparison of the 13C NMR data with
those of CpCo(η4-cis,trans-2,4-hexadiene).7
Bond Lengths (Å)
Fe-P1
2.283(1)
2.109(3)
2.066(4)
1.424(4)
1.485(4)
1.548(5)
1.847(3)
1.527(4)
1.885(3)
1.867(3)
Fe-P2
2.393(1)
2.076(3)
2.065(3)
1.408(4)
1.499(5)
1.515(5)
1.839(3)
1.541(4)
1.878(3)
1.884(3)
Fe-C16A
Fe-C21A
C16A-C17A
C17A-C18
C19A-C20
P1-C1
Fe-C17A
Fe-C22A
C21A-C22A
C18-C19A
C20-C21A
P2-C3
Having varied the length of the hydrocarbon chain
between the two phosphorus atoms, we next varied the
substituent R in the (R2PC2H4PR2)Fe0 species. The
reaction where R is isopropyl leads to the formation of
1, and in addition, we have investigated reactions where
the substituent is cyclopentyl (Cyp), cyclohexyl (Cy), tert-
butyl, or ethyl. The course of reaction with 1,5-hexa-
diene was found to depend upon the nature of the
substituent. In the case of the smallest substituent,
ethyl, no reaction of the Fe(Et2PC2H4PEt2) moiety with
the diene was observed and the only product which
could be isolated was Fe(Et2PC2H4PEt2)2.8 In the other
cases, an increase in the size of the substituent led to
the suppression of C-H activation. Thus, whereas the
product of the reaction in the presence of bis(dicyclo-
pentylphosphino)ethane (R ) Cyp) is (η5-2,4-hexadien-
1-yl)Fe(Cyp2PC2H4PCyp2)H, in the presence of bis-
(dicyclohexylphosphino)ethane a mixture of the (η5-2,4-
hexadien-1-yl)- and (η2:η2-1,5-hexadiene)iron complexes
is formed, while the bis(di-tert-butylphosphino)ethane-
stabilized system leads to the formation of (η2:η2-1,5-
hexadiene)Fe(tBu2PC2H4PtBu2) (13) in 46% yield. In-
terestingly, although both types of compounds are
formed in the reaction involving Cy2PC2H4PCy2, they
do not interconvert and the (diene)Fe species decompose
above 0 °C.
C1-C2
C2-C3
P1-C4
P1-C5
P2-C6
P2-C7
Bond Angles (deg)
P1-Fe-P2
92.6(4)
107.8
101.9
D1-Fe-D2
124.1
116.8
107.8
P1-Fe-D1
P1-Fe-D2
P2-Fe-D1
P2-Fe-D2
C16A-Fe-C17A
C16A-C17A-C18
C17A-C18-C19A
C19A-C20-C21A
Fe-P2-C3
39.7(1)
122.2(3)
115.3(3)
114.0(3)
112.3(1)
113.5(2)
C21A-Fe-C22A
C22A-C21A-C20
C18-C19A-C20
Fe-P1-C1
38.9(1)
117.9(3)
111.0(3)
113.9(1)
114.5(2)
113.4(2)
P1-C1-C2
P2-C3-C2
C1-C2-C3
a
The data are from the isomer having 75% occupancy. Esd’s
are given in parentheses.
that a mixture of products is obtained. These have not
been investigated further, since attempted displacement
of the organic group by further reaction with CO under
mild conditions was unsuccessful. The CO-induced
isomerization of iron-complexed dienes is a well-known
phenomenon, and examples relevant here are the reac-
tion between iron pentacarbonyl and 1,5-hexadiene to
The reaction of 13 with CO takes a different course
from that described above for the related compounds 6
and 11: substitution occurs, presumably due to the
lability of the diene-iron bond (eq 5).
5
give (η4-1-ethyl-, 1,3-butadiene)Fe(CO)3 and between
(η2:η2-1,5-cyclooctadiene)Fe(iPr2PC3H6PiPr2) and CO to
give (η4-1,3-cyclooctadiene)Fe(iPr2PC3H6PiPr2)CO.6
Rea ction of (R2P (CH2)n P R2)F e0 w ith 1,5-Hexa -
d ien e. In addition to the reactions of the iPr2P(CH2)nPi-
Pr2-stabilized species (where n ) 2, 3) with 1,5-
We have also investigated the reactions of the (Et3P)2-
Fe0 fragment with 1,5-hexadiene: (η2:η2-1,5-hexadiene)-
(7) Cibura, K. Doctoral dissertation, Ruhr-Universita¨t Bochum,
1985. 13C NMR (d8-toluene): δ 80.1 (C-3), 77.4 (C-4), 45.6 (C-2), 40.6
(C-5), 22.1 (C-6), 15.6(C-1) (cis,trans-2,4-hexadiene).
(5) Mahler, J . E.; Pettit, R. J . Am. Chem. Soc. 1963, 85, 3955.
(6) Frings, A.; J onas, K. Unpublished results. Frings, A. Doctoral
Thesis, Ruhr-Universita¨t Bochum, 1988.
(8) Field, L.; Baker, M. V. Organometallics 1986, 5, 821.

Reaction of Dienes with [Fe(R2P(CH2)nPR2)]
Organometallics, Vol. 16, No. 8, 1997 1615
Fe(PEt3)2 (14) is formed in good yield. The crystal
structure of 14 has been determined by X-ray diffrac-
tion, and the molecular structure is shown in Figure 2
and selected structural data in Table 2. In contrast to
(η2:η2-1,5-hexadiene)Fe(iPr2PC3H6PiPr2) (2),1 the diene
chain in 14 is not disordered and adopts exclusively a
cisoid (Cs, local symmetry) conformation. This confor-
mation, which is the less preferred one in 2, is presum-
ably a result of the ca. 12° larger P-Fe-P angle. The
ligands in 14 are arranged in an almost ideal tetrahe-
dron around the metal atom, and the geometry re-
sembles that observed for (η2-CH2dCH2)2Fe(PEt3)2.9
The experimental and structural data presented
above suggest that the course of reaction between an
unconjugated diene and a (R2P(CH2)nPR2)Fe0 species is
controlled by both the steric requirements of the sub-
stituents at phosphorus and the bite angle (P-Fe-P)
formed by the chelating bisphosphine with the metal:
bulky substituents tend to encapsulate the metal atom,
preventing the CH2 groups bridging the double bonds
from approaching, while the accessibility of the metal
atom is directly dependent upon the bite angle. These
observations are summarized in Figure 3. In an at-
tempt to give these effects a quantitative basis, we have
made use of both molecular modeling and extended
Hu¨ckel calculations. The latter are discussed below,
while the application of molecular modeling will be
presented in a later publication.
F igu r e 2. Molecular structure of (η2:η2-1,5-hexadiene)Fe-
(PEt3)2 (14). D1 and D2 are the midpoints of the C13-C14
and C17-C18 bonds.
Exten d ed Hu1 ck el MO Ca lcu la tion s. The results
shown in Figure 4 confirm that the EHT calculations
are able to qualitatively reproduce the d-orbital splitting
expected for the pseudooctahedral geometry formed in
(η5-2,4-hexadien-1-yl)Fe(iPr2PC2H4PiPr2)H (1) as well as
for the distorted-tetrahedral geometry found in (η2:η2-
1,5-hexadiene)Fe(iPr2PC3H6PiPr2) (2) and (η2-CH2dCH2)2-
Fe(PEt3)2. The calculations were carried out by replac-
F igu r e 3. Schematic illustration of the products obtained
when acyclic dienes react with Fe(R2P(CH2)nPR2) frag-
ments in the examples studied (Td ) tetrahedral, 16e (η2:
η2-diene)Fe(R2P(CH2)nPR2); Oh ) octahedral, 18e (η5-
dienyl)Fe(R2P(CH2)nPR2)H).
i
ing the Pr groups by H atoms and the Et groups by Me
Ta ble 2. Str u ctu r a l Da ta for
groups, and the atomic positions were taken from the
crystal structure data as described in the Experimental
Section. No attempt was made to optimize the geom-
etry.
(η2:η2-1,5-C6H10)F e(P Et3)2 (14)a
Bond Lengths (Å)
Fe-P1
2.280(1)
2.069(2)
2.086(2)
1.406(3)
1.523(3)
1.517(3)
Fe-P2
2.263(1)
2.120(2)
2.091(2)
1.409(3)
1.516(4)
Fe-C13
Fe-C18
C13-C14
C14-C15
C16-C17
Fe-C14
Fe-C17
C18-C17
C15-C16
The octahedral separation (∆E) for 1 is 3.76 eV (338
nm). The tetrahedral separation is much smaller, and
as a measure we have chosen the energy difference
between the lowest and highest d orbitals and the
following values are found: 2.06 eV (600 nm) for 2a ,
2.32 eV (535 nm) for 2b, and 2.16 eV (574 nm) for the
bis(η2-ethylene)Fe complex. The MO schemes correctly
reflect the magnetic behavior of these complexes: the
(η5-2,4-hexadien-1-yl)FeH-species (1) is formally an Fe-
(II) complex with six d electrons in the valence shell,
and the low-spin configuration is confirmed by NMR
spectroscopy, while the tetrahedral complexes are for-
mally Fe0 species with eight d electrons and the result-
ing paramagnetism with two unpaired electrons is
Bond Angles (deg)
P1-Fe-P2
105.1(1)
109.2
115.8
D1-Fe-D2
102.2
108.8
115.6
P1-Fe-D1
P1-Fe-D2
P2-Fe-D2
P2-Fe-D2
C13-Fe-C14
C13-C14-C15
C14-C15-C16
39.2(1)
121.9(2)
109.1(2)
C17-Fe-C18
C18-C17-C16
C15-C16-C17
39.4(1)
122.3(2)
109.5(2)
a
Esd’s are given in parentheses.
supported by the value of 3.6 µB determined experimen-
tally for the magnetic susceptibility of (η2:η2-1,6-
heptadiene)Fe(iPr2PC3H6PiPr2) (8).
(9) Hoberg, H.; J enni, K.; Angermund, K.; Kru¨ger, C. Angew. Chem.
1987, 99, 141; Angew. Chem., Int. Ed. Engl. 1987, 99, 899.

1616 Organometallics, Vol. 16, No. 8, 1997
Geier et al.
F igu r e 4. The d orbital separation in selected octahedral and tetrahedral iron complexes according to EHMO calculations.12
Geometries are based upon the crystal structure data and the assignment of the MO’s upon the largest contribution from
the corresponding AO’s (see Experimental Section for definition of coordinate system).
Inspection of the structural data available for the 16e
tetrahedral iron complexes indicates that the introduc-
tion of methylene groups either between the two alkene
groups or as bridging groups between the two P atoms
leads to a decrease in the P-Fe-P angle from the near-
tetrahedral value observed for (η2-CH2dCH2)2Fe(PEt3)2
(P-Fe-P ) 106.2°):9
(η2:η2-1,5-hexadiene)Fe(PEt3)2 (14)
(η2:η2-1,5-hexadiene)Fe(iPr2PC3H6PiPr2) (2)
(η2-CH2dCH2)2Fe(iPr2PC3H6PiPr2)
105.1°
95.0° 1
92.6° 6
This distortion apparently becomes so large in going
from a trimethylene-bridged to an ethylene-bridged
bisphosphine that in the latter case the methylene
groups in 1,5-hexadiene can interact with the metal and
(η5-2,4-hexadien-1-yl)Fe(iPr2PC2H4PiPr2)H (1) is formed.
In order to determine what effect a change in the
P-Fe-P angle has on the d-orbital splitting, we carried
out a series of extended Hu¨ckel calculations on the
hypothetical compound (η2:η2-1,5-hexadiene)Fe(PH3)2
with the diene arranged in a transoid (C2), conformation.
The geometries were generated by molecular modeling
using the crystal structure of (η2:η2-1,5-hexadiene)Fe-
(iPr2PC3H6PiPr2) (2, transoid-hexadiene)1 as the basis,
and the P-Fe-P angles were varied from 75 to 130° in
steps of 5°. The resulting correlation of the P-Fe-P
angle and the separation of the metal d orbitals (∆E,
the separation between the highest and lowest d orbit-
als) is shown in Figure 5.
F igu r e 5. Correlation between the d orbital separation
(∆E) and the P-Fe-P angle (θ) in the hypothetical
compound (η2:η2-1,5-hexadiene)Fe(PH3)2 (d8 system), in
which the diene is arranged in a transoid (C2) conformation.
ultimately to a distorted-square-planar geometry in
which the diene double bonds are rotated out of the
plane. Decreasing the P-Fe-P angle leads to increased
exposure of the metal orbitals.
To check the relevance of the curve shown in Figure
5, we have calculated the P-Fe-P angle for a series of
model (η2:η2-1,5-hexadiene)Fe(iPr2P(CH2)nPiPr2) (n )
1-5) compounds (generated from valence force field
calculations or minimized crystal structure data). The
calculated d-orbital separations (∆E) are listed in Table
3 and support the trend shown in Figure 5: ∆E is a
A broad minimum is observed between 95° and 105°,
confirming that any distortion of a tetrahedral arrange-
ment causes an increase in the separation of the d
orbitals and an increase in the total energy of the
system. An increase in the P-Fe-P angle will lead

Reaction of Dienes with [Fe(R2P(CH2)nPR2)]
Organometallics, Vol. 16, No. 8, 1997 1617
Ta ble 3. Cor r r ela tion betw een th e Ca lcu la ted
P -F e-P An gle for (η2:η2-1,5-Hexa d ien e)-
F e(iP r 2P (CH2)n P iP r 2) Com p lexes (n ) 1-5) a n d d
Or bita l Sep a r a tion (∆E)
mation of the diene shown is actually 0.87 kcal/mol less
stable then a crossed (transoid, C2) arrangement of the
double bonds). The antibonding metal dxy orbital in-
teracts with the p orbitals of the P atom while the other
metal d orbitals are essentially unchanged. The MO’s
of the hexadiene interact strongly with the dxy-acceptor
θ(P-Fe-P),
deg (R ) iPr)
∆E, eV
(R ) H)
n
1
2
3
4
74.1
85.1
2.63
2.58
2.36
2.31
2.15
2.41
2
orbital and the dyz- and dz orbitals. The d-orbital
separation which results is that expected for a distorted-
tetrahedral field, while the similarity in the energies
of the dxy and dyz orbitals (0.6 eV ) 13.8 kcal/mol) results
in the species having two unpaired electrons (the small
difference in energy between these orbitals is more than
compensated by the spin-pairing energy, and a high-
spin configuration results). Saillard and Hoffmann have
applied EHT calculations to C-H activation in square-
planar or distorted-square-planar complexes10 and shown
that a necessary prerequisite for oxidative addition of
a C-H bond to a metal atom is a destabilization of an
occupied metal d orbital (thus moving it closer in energy
to σ* of the C-H bond with which it interacts) and
hybridization away from the fragment (thus providing
better overlap with σ* of C-H). The MO scheme shown
in Figure 6 indicates that, in our tetrahedral complexes,
the highest occupied orbitals also have a suitable energy
and orientation to interact with the unfilled σ* C-H
MO’s of the methylene groups. Furthermore, distortion
from ideal tedrahedral geometry will lead to an increase
in the energy of the dyz orbital (see Figure 5) and result
in better overlap with the σ* C-H orbital.
94.8 (95.0a)
98.7
(η2:η2-1,5-hexadiene)Fe(PEt3)2
101.8 (102.3a)
113.7
5
a
Experimentally determine value.
Con clu sion s
The results of the EHT calculations discussed here,
and the molecular modeling which will be presented in
a later publication, suggest that the two approaches
complement each other. The course of the reaction
between 1,5-hexadiene and the Fe(R2P(CH)nPR2) species
is controlled in the first instance by steric factors, i.e.
encapsulation of the metal by the bidentate ligand, and
only when their significance is reduced can electronic
factors become important and the reaction proceeds to
the pseudooctahedral (η5-2,4-hexadien-1-yl)FeH species.
The experimental and theoretical results indicate
F igu r e 6. MO correlation scheme (EHT) for the hypo-
thetical compound (η2:η2-1,5-hexadiene)Fe(H2PC2H4PH2)
with the hexadiene molecule in a cisoid conformation.
i
minimum close to the tetrahedral angle and increases
on going to larger or smaller values.
that, in the case of the Pr-substituted bisphosphine
ligands, the limiting case occurs upon going from the
trimethylene-bridged ligand (iPr2PC3H6PiPr2) to the
ethylene-bridged ligand (iPr2PC2H4PiPr2). In the case
of the ethylene-bridged R2PC2H4PR2 ligands, the change
from mainly steric to mainly electronic control occurs
with Cy2PC2H4PCy2, since in the presence of this ligand
both types of compounds, i.e. (η2:η2-1,5-hexadiene)Fe(Cy2-
PC2H4PCy2) and (η5-2,4-hexadien-1-yl)Fe(Cy2PC2H4-
PCy2)H, are formed. The sterically less demanding
iPr2PC2H4PiPr2 or sterically more demanding tBu2-
PC2H4PtBu2 ligands lead exclusively to the formation
of (η5-2,4-hexadien-1-yl)Fe(iPr2PC2H4PiPr2)H or (η2:η2-
1,5-hexadiene)Fe(tBu2PC2H4PtBu2).
The calculations are supported by the experimental
results. Compounds such as (η2:η2-1,5-hexadiene)Fe-
(PEt3)2 (14) which have a near-tetrahedral P-Fe-P
angle (and a low ∆E) are thermally stable and show no
tendency to react further in spite of being 16e com-
plexes. The introduction of bridging CH2 groups be-
tween the P atoms leads to a decrease in the P-Fe-P
angle and destabilizes the compound (∆E increases). As
i
i
one goes from Pr2PC3H6PiPr2 to Pr2PC2H4PiPr2, the
destabilization is apparently so large that the resulting
energy increase is sufficient to compensate for a rehy-
bridization of the metal d orbitals toward an octahedral
field, leading to an increased exposure of the appropriate
partially filled d orbitals, which are then able to interact
with the σ* C-H orbitals of the methylene bridge of the
diene, and C-H cleavage can occur. The nature of this
interaction is discussed below.
Exp er im en ta l Section
The general experimental conditions and instrumentation
have been described in an earlier publication.11 The prepara-
tion of (η5-2,4-hexadien-1-yl)Fe(iPr2PC2H4PiPr2)H (1) and (η2:
η2-1,5-hexadiene)Fe(iPr2PC3H6PiPr2) (2) have been reported
earlier.1
An MO correlation diagram for the hypothetical
compound (η2:η2-1,5-hexadiene)Fe(H2PC2H4PH2) is shown
in Figure 6, which has been constructed with the aid of
force field calculations and conformational analysis to
generate an energetic minimum (the cisoid (Cs) confor-
(10) Saillard, J .-Y.; Hoffmann, R. J . Am. Chem. Soc. 1984, 106, 2006.
(11) Gabor, B.; Holle, S.; J olly, P. W.; Mynott, R. J . Organomet.
Chem. 1994, 466, 201.

1618 Organometallics, Vol. 16, No. 8, 1997
Geier et al.
FeCl2‚nTHF was prepared by extracting anhydrous FeCl2
(Ventron) with THF in a Soxhlet apparatus, evaporating the
extract, and drying the residue under high vacuum for 48 h.
The composition was determined by elemental analysis.
Activated Mg was prepared by the vacuum pyrolysis of MgH2.2
1,5-Bis(d iisop r op ylp h osp h in o)p en t a n e. A solution of
iPr2PLi (10.39 g, 84 mmol) in THF (150 mL) was stirred, and
a solution of 1,5-dibromopentane (10.35 g, 6 mL, 45 mmol) in
THF (10 mL) was added slowly at 0 °C. The reaction mixture
was warmed to room temperature, and the solvent was
removed under vacuum. The resulting oily residue was taken
up in hexane (150 mL) and filtered to remove LiBr. Evapora-
tion of the solvent under high vacuum gave the compound as
a viscous liquid. Yield: 8.33 g (65.2%). Anal. Calcd for
(η5-2,4-Hexa d ien -1-yl)F e(Cyp 2P C2H4P Cyp 2)H was pre-
pared as described above by reacting FeCl2‚nTHF (n ) 1.5)
with bis(dicyclopentylphosphino)ethane, 1,5-hexadiene, and
active Mg in THF at -35 °C. The compound is a yellow solid
which decomposes above 0 °C. Yield: 39%. Anal. Calcd for
C28H50FeP2: C, 66.7; H, 10.0; Fe, 11.1; P, 12.3. Found: C, 64.9;
H, 9.8; Fe, 10.7; P, 11.4. IR (KBr): ν(FeH) 1875 m. MS (70
°C): m/e 504 (M+), 500 (M+ - 4H), 418 (M+ - 4H/C6H10), 350.
1H NMR (d8-toluene, -30 °C): δ 5.42 (br, H-3), 5.10 (br, H-4),
4.25 (br, H-2), 2.56 (br, H-5), 2.43 (br, H-1syn), -0.20 (br,
H-1anti), -18.02 (dd, FeH, J (P,H) ) 72.5). 13C NMR (d8-toluene,
-30 °C): δ 93.3 (C-4), 90.5 (C-2), 80.4 (C-3), 56.5 (C-5, J (P,C)
) 6.3/9.9), 46.1 (C-1, J (P,C) ) 10.1), 22.8 (C-6). 31P NMR (d8-
toluene, -30 °C): δ 111.1, 109.5 (numbering scheme shown
below).
C
17H38P2: C, 67.1; H, 12.6; P, 20.4. Found: C, 67.0; H, 12.7;
P, 20.1. 1H NMR (d8-toluene): δ 1.70-0.94. 13C NMR (d8-
toluene): δ 33.9/28.6/22.4 (t, CH2, J (P,C) ) 11.6/19.2/19.2),
23.9/20.4/19.2 (d/q/q, i-C3H7, J (P,C) ) 14.8/15.7/10.5). 31P NMR
(d8-toluene): δ 2.87 (s).
(η5-2,4-Hexa d ien -1-yl)F e(iP r 2P C2H4P iP r 2)H (3). FeCl2‚
nTHF (n ) 1.43; 0.72 g, 3.12 mmol) was suspended in THF
(60 mL) and treated with bis(diisopropylphosphino)ethane
(0.82 g, 0.98 mL, 3.12 mmol). The reaction mixture was stirred
at room temperature for 1 h and cooled to -78 °C, and 1,6-
heptadiene (0.30 g, 0.42 mL, 3.12 mmol) and active Mg (0.08
g, 3.29 mmol) were added. The reaction mixture was stirred
for 24 h at -35 °C, the resulting yellow-brown suspension was
evaporated to dryness at -30 °C, and the residue extracted
with precooled pentane (200 mL) at -30 °C. The extract was
evaporated to dryness at -35 °C, and the residue dissolved in
precooled pentane (100 mL). The solution was concentrated
to half its volume and cooled to -78 °C to give a red-brown
solid which was recrystallized from pentane to give the
compound as an orange powder, which was washed with a
small amount of precooled pentane and dried under high
vacuum. Further product was obtained by concentrating the
(η5-2,4-Hexa d ien -1-yl)F e(Cy2P C2H4P Cy2)H was prepared
as described above by reacting FeCl2‚nTHF (n ) 1.5) with bis-
(dicyclohexylphosphino)ethane, 1,5-hexadiene, and active Mg
in THF at -35 °C. The reaction leads to the formation of a
small amount of a green compound, which is presumably (η2:
η2-1,5-hexadiene)Fe(Cy2PC2H4PCy2), and the title compound,
which is formed as a yellow-brown solid that decomposes
rapidly above ca. 0 °C. Yield: 37.0%. Anal. Calcd for C32H58
-
FeP2: C, 68.6; H, 10.4; Fe, 10.0; P, 11.1. Found: C, 67.0; H,
10.3; Fe, 9.7; P, 10.8. IR (KBr): ν(FeH) 1920 w. MS (90 °C):
m/e 560 (M+), 556 (M+ - 4H), 474 (M+ - 4H/C6H10). 1H NMR
(d8-toluene, -30 °C): δ 5.44 (br, H-3), 5.09 (br, H-4), 4.32 (br,
H-2), 2.63 (br, H-5), 2.42 (br, H-1syn), -0.09 (br, H-1anti), -18.20
(dd, FeH, J (P,H) ) 68.7). 13C NMR (d8-toluene, -30 °C): δ
94.0 (C-4), 90.8 (C-2), 80.4 (C-3), 57.1 (C-5), 48.5 (C-1), 22.2
(C-6). 31P NMR (d8-toluene, -30 °C): δ 110.4, 103.7 (number-
ing scheme shown above).
mother liquor. Yield: 0.28 g (22%). Anal. Calcd for C21H44
-
FeP2: C, 60.9; H, 10.7; Fe, 13.5; P, 15.0. Found: C, 61.0; H,
10.6; Fe, 13.6; P, 14.9. IR (KBr): ν(FeH) 1850 m. MS (60
°C): m/e 414 (M+), 318 (M+ - C7H12). 1H NMR (d8-toluene,
-30 °C): δ 5.35 (br, H-3), 4.96 (br, H-4), 4.16 (br, H-2), 2.93
(br, H-5), 2.43 (br, H-1syn), -0.09 (br, H-1anti), -18.18 (dd, Fe-
H, J (P,H) ) 72.0). 13C NMR (d8-toluene, -30 °C): δ 92.4 (C-
4), 90.6 (C-2), 80.1 (C-3), 65.4 (C-5), 48.4 (C-1), 31.7 (C-6). 31P
NMR (d8-toluene, -30 °C): δ 119.6, 113.5 (numbering scheme
shown below).
(η2:η2-1,6-H ep t a d ien e)F e(iP r 2P C3H 6P iP r 2) (8). FeCl2‚
nTHF (n ) 1.38; 0.87 g, 3.85 mmol) was suspended in THF
(50 mL) at room temperature and stirred for 1 h with
bis(diisopropylphosphino)propane (1.06 g, 1.23 mL, 3.85 mmol).
The solution was cooled to -78 °C and treated with 1,6-
heptadiene (0.37 g, 0.52 mL, 3.85 mmol) and active Mg (0.1 g,
4.11 mmol). The reaction mixture was stirred for 30 h at -40
°C to give a green suspension, which was evaporated to
dryness at -35 °C under high vacuum. The residue was
extracted with precooled pentane (200 mL) at -30 °C; the
extract was filtered and concentrated to 60 mL. The solution
was filtered, the residue was extracted with diethyl ether (40
mL), and the solutions combined and cooled to -78 °C. The
resulting dark green precipitate was recrystallized from a
pentane-diethylether mixture (1:1) at -30 to -78 °C to give
the compound as dark green crystals. A second fraction was
obtained from the mother liquor. The compound decomposes
above 0 °C. Yield: 1.00 g (60.6%). Anal. Calcd for C22H46
-
FeP2: C, 61.7; H, 10.8; Fe, 13.0; P, 14.5. Found: C, 61.5; H,
10.9; Fe, 13.0; P, 14.5. IR (KBr): ν 1195 s, 1010 s. MS (50
°C): m/e 428 (M+), 332 (M+ - C7H12). Magnetic susceptibil-
ity: 3.6 µB. Crystal structure determination: see Figure 1.
(η2:η2-1,7-Octa d ien e)F e(iP r 2P C3H6P iP r 2) (9) was pre-
pared as described above for 8 as a green, somewhat sticky
solid by reacting FeCl2‚nTHF (n ) 1.38) with the bidentate
ligand, 1,7-octadiene, and active Mg in THF at -78 to -35
°C. The compound decomposes rapidly at temperatures above
0 °C. Yield: 20.1%. Anal. Calcd for C23H48FeP2: C, 62.4; H,
10.9; Fe, 12.6; P, 14.0. Found: C, 62.6; H, 10.9; Fe, 12.5; P,
14.1. IR (KBr): ν 2955 s, 2925 s. MS (60 °C): m/e 442 (M+),
332 (M+ - C8H14).
(η5-2,4-Non a d ien -1-yl)F e(iP r 2P C2H 4P iP r 2)H (5). Pre-
pared as described above by reacting FeCl2‚nTHF (n ) 1.43)
with bis(diisopropylphosphino)ethane, 1,8-nonadiene, and ac-
tive Mg in THF at -35 °C. The compound was isolated as a
yellow-brown powder. Yield: 18.7%. Anal. Calcd for
C23H48FeP2: C, 62.4; H, 10.9; Fe, 12.6; P, 14.0. Found: C, 62.5;
H, 10.9; Fe, 12.5; P, 13.9. IR (KBr): ν(FeH) 1890 m. MS (60
°C): m/e 442 (M+), 318 (M+ - C9H16). 1H NMR (d8-toluene,
-30 °C): δ 5.37 (br, H-3), 5.02 (br, H-4), 4.17 (br, H-2), 2.47
(br, H-5), 2.02 (br, H-1syn), -0.06 (br, H-1anti), -18.17 (dd, FeH,
J (P,H) ) 72.2). 13C NMR (d8-toluene, -30 °C): δ 92.9 (C-4),
90.6 (C-2), 79.9 (C-3), 63.3 (C-5), 48.5 (C-1), 38.3 (C-6), 37.5
(C-7), 22.9 (C-8), 14.4 (C-9). 31P NMR (d8-toluene, -30 °C): δ
119.8, 113.9, J (P,P) ) 36.0 (numbering scheme shown above).
(η2:η2-1,8-Non a d ien e)F e(iP r 2P C3H6P iP r 2) (10) was pre-
pared as described above for 8 as a green, sticky solid by

Reaction of Dienes with [Fe(R2P(CH2)nPR2)]
Organometallics, Vol. 16, No. 8, 1997 1619
reacting FeCl2‚nTHF (n ) 1.38) with the bidentate ligand, 1,8-
nonadiene, and active Mg in THF at -78 to -35 °C. The
compound decomposes rapidly above -10 °C. Yield: 20%. The
thermal instability prevented the obtaining of satisfactory
analytical data; however, an Fe:P ratio of 1:2 was confirmed.
IR (KBr): ν 2950 s, 2920 s, 2870 s. MS (60 °C): m/e 456 (M+),
332 (M+ - C9H16).
(η4-(2Z,4E)-hexadiene)Fe(iPr2PC4H8PiPr2)CO (12) as a yellow
powder which is stable at room temperature. Yield: 23.5%.
Anal. Calcd for C23H46FeOP2: C, 60.5; H, 10.2; Fe, 12.2; P,
13.6. Found: C, 60.0; H, 9.9; Fe, 11.9; P, 13.3. IR (KBr): ν-
(CO) 1870 vs. MS (80 °C): m/e 456 (M+), 428 (M+ - CO), 374
(M+ - C6H10), 346 (M+ - C6H10/CO). 1H NMR (d8-toluene, -30
°C): δ 5.13 (m, H-3), 4.49 (m, H-4), 1.54 (d, H-6, J (5,6) ) 5.6),
1.36 (d, H-1, J (1,2) ) 6.4), 0.11 (m, H-2), -0.05 (m, H-5). 13C
NMR (d8-toluene, -30 °C): δ 223.4 (FeCO), 83.7 (C-3, J (C,H)
) 159), 77.5 (C-4, J (C,H) ) 161), 49.7 (C-2, J (C,H) ) 152,
J (P,C) ) 6.4), 41.9 (C-5, J (C,H) ) 153, J (P,C) ) 7.6/11.5), 21.9
(C-6, J (P,C) ) 4.8), 17.3 (C-1, J (P,C) ) 3.7), 33.6-26.4/28.3-
23.5/20.8-19.0 (iPr2PC4H8PiPr2). 31P NMR (d8-toluene, -30
°C): δ 61.3, 58.0, J (P,P) ) 10.9 (numbering scheme is given
below; only the main isomer (80%, 31P NMR) has been
identified).
(η2:η2-1,5-Hexa d ien e)F e(iP r 2P C4H8P iP r 2) (11) was pre-
pared as described above by reacting FeCl2‚nTHF (n ) 1.43)
with bis(diisopropylphosphino)butane, 1,5-hexadiene, and ac-
tive Mg in THF at -35 °C. The compound is a green solid
which decomposes above -10 °C. Yield: 81%. Anal. Calcd
for C22H46FeP2: C, 61.7; H, 10.8; Fe, 13.0; P, 14.5. Found: C,
60.9; H, 10.5; Fe, 13.4; P, 13.8. IR (KBr): ν 2955 s, 2925 s,
2890 s, 2870 s. MS (90 °C): m/e 428 (M+), 346 (M+ - C6H10),
304, 262.
(η2:η2-1,5-Hexa d ien e)F e(tBu 2P C2H4P tBu 2) (13) was pre-
pared as described above by reacting FeCl2‚nTHF (n ) 1.5)
with bis(di-tert-butylphosphino)ethane, 1,5-hexadiene, and
active Mg in THF at -35 °C. The compound is dark green
and decomposes above 0 °C. Yield: 46%. Anal. Calcd for
C24H50FeP2: C, 63.2; H, 11.0; Fe, 12.2; P, 13.6. Found: C, 63.1;
H, 10.0; Fe, 12.4; P, 13.6. IR (KBr): ν 1475 m, 1385 m, 1365
m, 1175 m, 1020 m, 845 m, 810 m, 770 m, 650 m, 640 m. MS
(70°C): m/e 456 (M+), 374 (M+ - C6H10).
(η2:η2-1,5-Hexa d ien e)F e(P Et3)2 (14) was prepared as de-
scribed above by reacting FeCl2‚nTHF (n ) 1.38) with trieth-
ylphosphine, 1,5-hexadiene, and active Mg in THF at -35 °C.
The compound is dark violet-black. Yield: 82%. Anal. Calcd
for C18H40FeP2: C, 57.8; H, 10.8; Fe, 14.9; P, 16.6. Found: C,
57.6; H, 10.6; Fe, 15.0; P, 16.7. IR (KBr): ν 1460 m, 1420 m,
1375 m, 1030 m, 760 s, 695 m. MS (30 °C): m/e 374 (M+), 292
(M+ - C6H10), 264 (M+ - C6H10/C2H4), 256 (M+ - PEt3).
Crystal structure: see Figure 2.
(η2:η2-3-Meth yl-1,6-h ep ta d ien e)F e(iP r 2P C2H4P iP r 2) (6).
The compound was prepared as described above by reacting
FeCl2‚nTHF (n ) 1.50) with bis(diisopropylphosphino)ethane,
3-methyl-1,6-heptadiene, and active Mg in THF at -30 °C. The
compound is formed as a dark green solid which decomposes
Rea ction of 13 w ith CO. Compound 13 was reacted as
described above with CO in pentane at -30 °C to give (OC)3Fe-
(tBu2PC2H4PtBu2) as an almost colorless solid. Yield: 55%.
Anal. Calcd for C21H40FeO3P2: C, 55.0; H, 8.8; Fe, 12.2; P,
13.5. Found: C, 54.1; H, 8.7; Fe, 12.5; P, 13.2. IR (KBr): ν-
(CO) 1955 s, 1860 vs. MS (80 °C): m/e 458 (M+), 430 (M+
-
CO), 402 (M+ -2CO), 374 (M+ - 3CO). 1H NMR (d8-toluene):
δ 1.39 (d, CH2, J (P,H) ) 14.0), 1.17 (d, Me, J (P,H) ) 12.0). 13
C
NMR (d8-toluene): δ 37.7/30.5 (tBu), 24.1 (CH2). 31P NMR
(d8-toluene): δ 132.6.
Rea ction of 6 w ith CO. Compound 6 was reacted with
CO in pentane at -30 °C as described above to give an oily
yellow product in quantitative yield which was shown by 31P
NMR to consist mainly (44%) of (η2:η2-5-methyl-1,5-hepta-
rapidly above 0 °C. Yield: 68.3%. Anal. Calcd for C22H46
-
FeP2: C, 61.7; H, 10.8; Fe, 13.0; P, 14.5. Found: C, 61.8; H,
10.8; Fe, 12.8; P, 14.6. IR (KBr): ν(dCH) 3040 w. MS (40
°C): m/e 428 (M+), 318 (M+ - C8H14).
diene)Fe(iPr2PC2H4PiPr2)CO (7). Anal. Calcd for C23H46
-
FeOP2: C, 60.5; H, 10.2; Fe, 12.2; P, 13.6. Found: C, 60.4; H,
10.2; Fe, 12.4; P, 13.5. IR (KBr): ν(CO) 1975 w, 1870 br s.
MS (60 °C): m/e 456 (M+), 428 (M+ - CO), 346 (M+ - C8H14),
318 (M+ - CO/C8H14). 1H NMR (d8-toluene): δ 4.94 (m, H-2),
3.45/3.08 (m, H-3/4), -0.34 (m, H-6), -1.18 (m, H-1Z). 13C
NMR (d8-toluene): δ 219.1 (FeCO), 98.0 (C-5), 77.7 (C-2), 51.1
(C-6, J (P,C) ) 5.0/10.1), 39.2 (C-4), 36.3 (C-3), 34.2 (C-1, J (P,C)
) 7.8/12.2, assignment provisional). 31P NMR (d8-toluene): δ
95.7, 54.4, J (P,P) ) 40.7) (44%) (numbering scheme is given
above). The 31P NMR spectrum indicates the presence of two
other species: δ 104.2, 98.4, J (P,P) ) 11.5 (23%); 98.5, 97.7,
J (P,P) ) 18.2 (22%).
Rea ction of 8 w ith CO. Compound 8 (0.35 g, 0.8 mmol)
was dissolved in pentane (40 mL) at -30 °C and treated with
excess CO. The solution changed from dark green to yellow.
The solvent was removed at room temperature and the
resulting yellow-brown, oily residue dissolved in pentane (2
mL). The solution was filtered and evaporated to give the
product as a yellow, viscous oil. The spectroscopic data
indicate that a mixture of isomers is formed. Yield: 0.31 g
(85%). Anal. Calcd for C23H46FeOP2: C, 60.5; H, 10.2; Fe,
12.2; P, 13.6. Found: C, 60.3; H, 10.2; Fe, 12.3; P, 13.7. IR
(KBr): ν(CO) 1965 s, 1905 s, 1870 s, 1815 s. MS (40 °C): m/e
456 (M+), 428 (M+ - CO), 360 (M+ - C7H12), 332 (M+ - CO/
C7H12). 1H NMR (d8-toluene): δ 5.0-4.5/2.46 to -0.2. 13C
NMR (d8-toluene): δ 200.4 (s, CO), 89-79/58-55/39-14. 31P
NMR (d8-toluene): δ 70.0-51.9.
Rea ction of 10 w ith CO. Compound 10 was reacted with
CO in pentane at -30 °C as described above to give a yellow
viscous oil. The spectroscopic data indicate that a mixture of
two compounds is formed. The yield is quantitative. Anal.
Calcd for C25H50FeOP2: C, 62.0; H, 10.4; Fe, 11.5; P, 12.8.
Found: C, 61.9; H, 10.4; Fe, 11.7; P, 12.7. IR (KBr): ν(CO)
1905 w, 1870 s. MS (65 °C): m/e 484 (M+), 456 (M+ - CO),
360 (M+ - C9H16). 332 (M+ - CO/C9H16). 1H NMR (d8-
toluene): δ ca. 4.6/2.4 to -0.6. 13C NMR (d8-toluene): δ 85-
76/56-50/38-14. 31P NMR (d8-toluene): δ 59.4, 56.7, J (P,P)
) 6.0 (compound I, 20%); 56.9, 53.0, J (P,P) ) 9.4 (compound
II, 66%).
Exten d ed Hu1 ck el Ca lcu la tion s. The EHT calculations12
were based on crystal structure and valence force field modeled
structures using the program ICON (version 8, 1974) and a
Micro VAX II computer system. The s and p AO’s (Slater type)
of the C, H, Fe, and P atoms were calculated with the Coulomb
potentials (IE), ú values, and Hu¨ckel coefficients (ci) shown in
Table 4. The Fe d orbitals are of the double-ú type.13 The
Cartesian coordinates for the atomic positions were determined
from crystal structure data and modeled structures using the
Ta ble 4
atom IEs, eV
ús
1.3
1.625 -11.4
1.600 -14.0
IEp, eV
úp
IEd, eV
úd
cia
H
C
P
-13.6
-21.4
1.625
1.600
-18.6
Fe
-9.10 1.575
-5.32 0.975 -12.60 5.35 0.5366
1.80 0.6678
Rea ction of 11 w ith CO. Compound 11 was reacted in
toluene with CO at -30 °C as described above to give
a
ψd ) c1φ(ú1) + c2φ(ú2).

1620 Organometallics, Vol. 16, No. 8, 1997
Geier et al.
octahedral complex (η5-2,4-hexadien-1-yl)Fe(iPr2PC2H4PiPr2)
was oriented with the Fe atom in the 0, 0, 0 position and the
P atoms on the x axis and in the x,y plane. The calculations
were simplified by replacing the alkyl groups in the bisphos-
phine by H atoms and the Et groups in PEt3 complexes by Me
groups. No attempt was made to optimize the geometry.
Cr ysta l Str u ctu r e Deter m in a tion s of 8 a n d 14. Low-
temperature data sets were collected with crystals mounted
in Lindemann glass capillaries under argon and cooled by a
cold stream of N2 gas on an Enraf-Nonius CAD-4 diffracto-
meter using graphite-monochromated Mo KR X-radiation.
Crystal data and details of data collection and structure
refinement are given in Table 5. Structures were solved using
the program SHELXS-8615 and refined on F2 using all the data
(SHELXL-93).16 For 8 the atoms C16, C17, C19, C21, and C22
were disordered over two positions (A:B ) 0.75:0.25 oc-
cupancy), and all C atoms in the heptadiene were refined with
isotropic atomic displacement parameters. All other non-H
atoms were refined anisotropically while the H atoms were
calculated and allowed to ride.
Ta ble 5. Cr ysta llogr a p h ic Da ta for 8 a n d 14
8
14
empirical formula
mol wt
C22H46FeP2
428.4
C18H40FeP2
374.3
cryst color
cryst size, mm
a, Å
black
black
0.32 × 0.49 × 0.49 0.46 × 0.42 × 0.56
22.489(8)
15.388(3)
18.536(7)
132.53(2)
4727.3(3)
1.20
9.164(7)
14.396(9)
16.035(7)
92.75(4)
2112.9(2)
1.18
b, Å
c, Å
â, deg
V, Å3
D
calcd, g cm-3
T, °C
λ(Mo), Å
space group (No.)
-173
-173
0.710 69
C2/c (15)
8
0.710 69
P21/n (14)
4
Z
µ
abs, cm-1
7.74
8.60
no. of rflns, measd
no. of indep rflns
no. of obsd rflns
(I > 2σ(I))
no. of variables
wR1 (obsd)
5551 ((h,+k,+1)
6333 ((h,+k,+l)
5383
6129
4908
4908
196
0.050
350
0.045
Su p p or tin g In for m a tion Ava ila ble: Two X-ray crystal-
lographic files (for 8 and 14), in CIF format, are available.
Access information is given on any current masthead page.
wR2 (obsd)
0.125
0.127
weighting factorsa
a ) 0.046,
b ) 23.860
0.001
a ) 0.0905,
b ) 1.2173
0.001
max ∆/σ
OM9610004
rest electron dens, e Å-3 0.92
0.60
a
Structures were refined using all data, where the function
minimized was ∑[w(Fo2 - Fc2)2], with w-1 ) [σ2(Fo2) + (aP)2 + bP]
and P [max(Fo2,0) + 2Fc2]/3.
(12) Hoffmann, R. J . Chem. Phys. 1963, 39, 1397. Hoffmann, R.;
Lipscomb, W. N. J . Chem. Phys. 1962, 36, 2176.
(13) Richardson, J . W.; Nieuwport, W. C.; Powell, R. R.; Edgell, L.
F. J . Chem. Phys. 1962, 36, 1057.
routine QCPE from SYBYL.14 The distorted-tetrahedral (di-
ene)Fe(bisphosphine) complexes were so oriented that one P
atom occupies the basis position (0, 0, 0), the second P atom
lies on the x axis, and the Fe atom lies in the x,y plane. The
(14) SYBYL 6.0 and 6.2, TRIPOS Associates, Inc., St. Louis, MO.
(15) Sheldrick, G. M. SHELXS 86, Program for the Solution of
Crystal Structures; University of Go¨ttingen, Go¨ttingen, Germany,
1986.
(16) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.