The fluxionality and crystal structures of [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2 P}2O) and [Fe3(CO)9 (dppm){P(OMe)3}]

Article abstract of DOI:10.1139/v01-081

The crystal structures of [Fe₃(CO)₁₀L₂] (L₂ = dppm, (Ph₂P)₂NH, {(EtO)₂P}₂O) and [Fe₃(CO)₉(dppm){P(OMe)₃}] (dppm = (Ph₂P)₂CH₂) have been determined. [Fe₃(CO)₁₀L₂] (L₂ = dppm, (Ph₂P)₂NH, {(EtO)₂P}₂O) adopt a stereochemistry in the crystal with the bidentate ligand across an unbridged Fe-Fe edge, resulting in one phosphorus ligand being unusually in the axial position. In contrast, [Fe₃(CO)₉(dppm){P(OMe)₃}] adopts a stereochemistry in the crystal with the bidentate ligand across the bridged Fe-Fe edge. The structures of the [Fe₃(CO)₁₀L₂] (L₂ = dppm, (Ph₂P)₂NH, {(EtO)₂P}₂O) support the Mann concerted bridge - opening bridge - closing mechanism of carbonyl fluxionality rather than the Johnson libration mechanism. The fluxionality of the compounds in the ¹³C NMR spectra is explained by a combination of the concerted bridge - opening bridge - closing, merry-go-round, and trigonal twist mechanisms, and ΔG? values were determined.

Full text of DOI:10.1139/v01-081

760
The fluxionality and crystal structures of
[Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O)
and [Fe3(CO)9(dppm){P(OMe)3}]
Harry Adams, Sylvana C.M. Agustinho, Krystyna Chomka, Brian E. Mann,
Stephen Smith, Clare Squires, and Sharon E. Spey
Abstract: The crystal structures of [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O) and
[Fe3(CO)9(dppm){P(OMe)3}] (dppm = (Ph2P)2CH2) have been determined. [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH,
{(EtO)2P}2O) adopt a stereochemistry in the crystal with the bidentate ligand across an unbridged Fe–Fe edge, result-
ing in one phosphorus ligand being unusually in the axial position. In contrast, [Fe3(CO)9(dppm){P(OMe)3}] adopts a
stereochemistry in the crystal with the bidentate ligand across the bridged Fe–Fe edge. The structures of the
[Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O) support the Mann concerted bridge – opening bridge – closing
mechanism of carbonyl fluxionality rather than the Johnson libration mechanism. The fluxionality of the compounds in
the 13C NMR spectra is explained by a combination of the concerted bridge – opening bridge – closing, merry-go-
round, and trigonal twist mechanisms, and ∆G‡ values were determined.
Key words: iron carbonyls, crystal structures, 13C NMR spectroscopy, fluxionality, mechanisms.
Résumé : On a déterminé les structures cristallines des complexes [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O)
et [Fe3(CO)9(dppm)[P(OMe)3]. Dans le cristal, le [Fe(CO)10L2 (L2 = dppm, (Ph2P)2NH, [(EtO)2P]2O) adopte une stéréo-
chimie comportant un ligand bidentate à travers une arête non pontée Fe–Fe; il en résulte qu’un ligand phosphoré se
trouve de façon inhabituelle en position axiale. Par ailleurs, à l’état cristallin, le [Fe3(CO)9(dppm){P(OMe)3}] adopte une
stéréochimie avec un ligand bidentate à travers une arête Fe–Fe pontée. Les structures des complexes [Fe3(CO)10L2] (L2 =
dppm, (Ph2P)2NH, {(EtO)2P}2O) sont en accord le mécanisme d’ouverture et fermeture concertées de pont de Mann pour
la fluxion du carbonyle plutôt que le mécanisme de libration de Johnson. La fluxion des composés dans les spectres RMN
du 13C est expliquée par les mécanismes d’ouverture et de fermeture concertées de pont et de carrousel et par des méca-
nismes trigonaux déformés et on a déterminé les valeurs de ∆G‡.
Mots clés : fer-carbonyle, structures cristallines, spectroscopie RMN du 13C, fluxion, mécanismes.
[Traduit par la Rédaction] Adams et al. 774
Introduction
fluxionality of [Fe3(CO)12] on the NMR timescale, it is nec-
essary to introduce ligands that do not readily move between
iron atoms, and hence, block dynamic pathways.
There have been a number of papers reporting dynamic
NMR investigation of [Fe3(CO)12] derivatives. In 1981,
Johnson and co-workers (3) examined the 13C NMR spec-
trum of [Fe3(CO)11{P(OR)3}] and concluded that at –90°C it
shows CO signals in an intensity ratio of 6:4:1. They ex-
[Fe3(CO)12] is highly fluxional, and even at –150°C a sin-
glet is observed in the 13C NMR spectrum (1). Cotton and
Hunter (1) proposed that the fluxionality is a result of the
merry-go-round mechanism, see Scheme 1. In 1976, John-
son (2) proposed a mechanism going via a [Fe3(CO)10(µ3-
CO)2] intermediate and (or) transition state 1 as a result of
rotation of the Fe(CO)4 moieties. To slow down the
plained the fluxionality on the basis of an icosahedron º
cube octahedron
icosahedron rearrangement. Subse-
º
Received August 31, 2000. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on June 30, 2001.
quently Farrar and Lunniss (4) synthesized a number of di-
and trisubstituted [Fe3(CO)12] derivatives. Unfortunately,
their interpretation of the fluxionality was flawed as it re-
quired the activation energy of the merry-go-round mecha-
nism to vary wildly between similar derivatives.
In 1989, the fluxionality of [Fe3(CO)12–n{P(OMe)3}n] was
reexamined (5). It was found that at –101°C, the intensity of
the carbonyl signals of [Fe3(CO)11{P(OMe)3}] are 5:5:1, not
6:4:1 as previously reported (3). As a result, a new mecha-
nism for the fluxionality of [Fe3(CO)12] and its derivatives
was proposed (5). The mechanism involves the rotation of
the Fe3 triangle around the S10 axis of the carbonyl
icosahedron defined by carbonyls C(6)O and C(10)O, see
Dedicated to Professor B.R. James on the occasion of his
65th birthday.
H. Adams, S.C.M. Agustinho,1 K. Chomka, B.E. Mann,2
S. Smith, C. Squires, and S.E. Spey. The Department of
Chemistry, The University of Sheffield, Sheffield, S3 7HF,
U.K.
1Permanent address: Institito de Quimica de São Carlos,
Universidade de São Paulo, Cx Postal 780, CEP 13.560-970,
São Carlos, São Paulo, Brazil.
2Corresponding author (telephone: +44 (0) 114 222-9332; fax
+44 (0) 114 273-8673; e-mail: B.Mann@sheffield.ac.uk).
Can. J. Chem. 79: 760–774 (2001)
© 2001 NRC Canada
DOI: 10.1139/cjc-79-5/6-760

Adams et al.
761
Scheme 1.
Scheme 2.
Scheme 2.3 As the carbonyl bridge opens on one Fe–Fe
edge, another closes on a second Fe–Fe edge, and hence,
this mechanism is described as the concerted bridge – open-
ing bridge – closing mechanism. The halfway stage of this
carbonyl exchange has been observed in the crystal structure
of [Fe3(CO)10{1,2-(Me2As)2C6H4}] (6). This mechanism,
when applied to [Fe3(CO)11{P(OMe)3}], causes the ex-
change of carbonyls C(1)O, C(2)O, C(7)O, C(11)O, and
C(12)O as one set of five carbonyls, and C(3)O, C(4)O,
C(5)O, C(8)O, and C(9)O as the second set of five carbon-
yls, leaving C(6)O as the unique carbonyl.
Subsequently Johnson and Bott (7) revised the Johnson
mechanism, and showed how rotations of the Fe(CO)4
groups about the axis going through the iron and the oppo-
site edge of the Fe3 triangle, called librations, could lead to
the same exchange.
way stage for the exchange. The structure lies almost exactly
on the ideal pathway predicted by the Mann concerted
bridge – opening bridge – closing mechanism with the line
joining C(6)O and C(10)O lying only 2.6° from the plane of
the iron triangle. Although Johnson (11) published a rebuttal
of the Mann concerted bridge – opening bridge – closing
mechanism, he failed to address this key point. Johnson also
claimed priority, but his mechanism is different going via 1,
which does not lie on the pathway mapped out by the Bürgi–
Dunitz analysis (9).
The majority of crystal structures of [Fe3(CO)12] and its
derivatives have structures that are close to that found in
[Fe3(CO)12] and shed little light on the differences between
the two mechanisms. In view of the distortion produced by
1,2-(Me2As)2C6H4, a number of compounds containing
bidentate ligands have been synthesized. [Fe3(CO)10(dppe)]
showed only minor distortions from the [Fe3(CO)12] struc-
ture with the ligand bridging the Fe(µ-CO)2Fe edge of the
iron triangle (12, 13). To increase the distortions, ligands
with a smaller bite angle were examined, and it is this that is
reported in this paper.
An
investigation
involving
[Fe3(CO)10(CN-t-
Bu){P(OMe)3}] confirmed that the rotation of the two
groups of five carbonyls is concerted (8).
To differentiate between the two mechanisms, an analysis
of X-ray crystal structures was performed (9). A Bürgi–
Dunitz analysis (10) was applied to then known crystal
structures and a web movie was produced.4
Experimental
The Johnson libration mechanism and the Mann concerted
bridge – opening bridge – closing mechanism differ in the
pathway followed by the iron triangle during fluxional path-
way. However, the difference is best seen in terms of the lo-
cation of the carbonyls with respect to the iron triangle. In
the Johnson libration mechanism all the equatorial carbonyls
(C(5)O, C(6)O, C(7)O, C(10)O) move out of the plane de-
fined by the iron triangle. In the Mann concerted bridge –
opening bridge – closing mechanism, C(5)O and C(7)O
move out of the plane defined by the iron triangle, while
C(6)O and C(10)O remain in the plane. The crystal structure
of [Fe3(CO)10{1,2-(Me2As)2C6H4}] (6) represents the half
[Fe3(CO)12] was synthesized according to the literature
(14). Me3NO·2H2O, Ph2PCH2PPh2 (Ph2PCH2PPh2 = dppm),
(EtO)2POP(OEt)2, and P(OMe)3 were purchased from
Aldrich. Ph2PNHPPh2 was purchased from Strem Chemi-
cals, Inc. Me3NO·2H2O was dehydrated by Dean–Stark dis-
tillation using toluene.
Synthesis of [Fe3(CO)10L2] (L2 = dppm, Ph2PNHPPh2,
(EtO)2POP(OEt)2)
[Fe3(CO)12] (0.5 g, 0.992 mmol) was placed in a nitrogen
purged three-neck round-bottom flask and dissolved in
3 There is an identical axis through C(5)O and C(7)O.
4 To observe movie go to http://www.shef.ac.uk/misc/personal/ch1bem/Fe3-carbonyl12A.html.
© 2001 NRC Canada

762
Can. J. Chem. Vol. 79, 2001
CH2Cl2 (150 cm3). The flask was cooled to –78°C in a solid
CO2–acetone bath. The phosphorus ligand (0.992 mmol)
was slowly added to the flask with stirring, followed by
Me3NO (0.15 g, 2.0 mmol). The flask was left in the solid
CO2–acetone bath and allowed to warm to room tempera-
ture. The reaction was then monitored by TLC; all the
[Fe3(CO)12] had been consumed in 4 h at room temperature.
The solvent was removed and the product purified by chro-
matography on a 30 cm silica gel column, eluting with hex-
ane–CH2Cl2 (3:1). The dark green compounds were
recrystallized from acetonitrile.
Lorentz and polarization effects and for absorption by semi
empirical methods based on symmetry-equivalent and re-
peated reflections (minimum and maximum transmission co-
efficients 0.8812 and 0.9745), 6546 independent reflections
exceeded the significance level |F|/σ(|F|) > 4.0. The structure
was solved by direct methods and refined by full-matrix
least-squares methods on F2. Hydrogen atoms were placed
geometrically and refined with a riding model (including
torsional freedom for methyl groups) and with Uiso con-
strained to be 1.2Ueq (1.5 for methyl groups) of the carrier
atom. Refinement converged at a final R = 0.0598 (wR2 =
0.2215, for all 10688 data, 984 parameters, mean and maxi-
mum δ/σ: 0.000, 0.001) with allowance for the thermal ani-
sotropy of all non-hydrogen atoms. Minimum and maximum
final electron density –0.834 and 1.540 e Å–3.
[Fe3(CO)10(dppm)]
Yield: 47%. MS m/z: 832. IR (CH2Cl2) (cm–1): 2114,
2048, 2008. 13C NMR (–105°C) δ: 229.9, 225.5, 218.6,
205.9, 131.5 (t, NP,C = 48 Hz), 130.5 (ortho), 129.8 (para),
127.7 (meta), 47.8, (JP,C = 19 Hz, CH2). 31P NMR (CD2Cl2)
δ: 47.
A weighting scheme w = 1/[σ2(F2) + (0.1390P)2 + 0.00P]
where P = (F2+ 2F2)/3 was used oin the latter stages of re-
o
finement. Complex cscattering factors were taken from the
program package SHELXTL (15) as implemented on the
Viglen Pentium computer.
[Fe3(CO)10(Ph2PNHPPh2)]
Yield: 25%. IR (CH2Cl2) (cm–1): 2062, 1998, 1815. 13C
NMR (CD2Cl2) δ: 220.4, 136.7 (t, NP,C = 55 Hz, ipso), 131.1
(para), 130.8 (NP,C = 12 Hz, ortho), 128.9 (NP,C = 10 Hz,
meta). 31P NMR (CD2Cl2) δ: 99. Anal. calcd. for
C34H21NO10P2Fe3: C 46.46, H 2.70, N 1.68; found: C 45.78,
H 3.25, N 1.69.
Crystallographic data for
[Fe3(CO)10(Ph2PNHPPh2)]·0.5CH2Cl2
Crystal data for C34.5H22ClNO10P2Fe3: MW = 875.47,
crystallizes from deuterated dichloromethane as dark green
blocks, crystal dimensions 0.38 × 0.25 × 0.12 mm3.
Monoclinic, space group C2/c, a = 22.223(15), b =
20.984(17), and c = 16.229(15) Å, β = 106.46(6)°, U =
7258(10) Å3, Z = 8, Dc = 1.602 Mg m–3, Mo Kα radiation
(λ = 0.71073 Å), µ (Mo Kα) = 1.402 mm–1, F(000) = 3528.
Three-dimensional, room temperature X-ray crystal data
were collected in the range 2.72 < 2θ < 50.20° on a Siemens
P4 diffractometer by the omega scan method. Of the 7266
reflections measured, all of which were corrected for Lo-
rentz and polarization effects, and for absorption by semi-
empirical methods based on symmetry-equivalent and re-
peated reflections (minimum and maximum transmission co-
efficients 0.6178 and 0.8498), 3620 independent reflections
exceeded the significance level |F|/σ(|F|) > 4.0. The structure
was solved by direct methods and refined by full-matrix
least-squares methods on F2. Hydrogen atoms were placed
geometrically and refined with a riding model and with Uiso
constrained to be 1.2Ueq of the carrier atom. Refinement
converged at a final R = 0.0908 (wR2 = 0.2657 for all 6338
data, 465 parameters, mean and maximum δ/σ: 0.000, 0.000)
with allowance for the thermal anisotropy of all non-
hydrogen atoms. Minimum and maximum final electron den-
sity –0.709 and 1.402e Å–3 (the latter 1.28 Å from Fe(1)). A
weighting scheme w = 1/[σ2(F2) + (0.1456P)2 + 0.00P]
where P = (F2 + 2F2)/3 was usoed in the latter stages of re-
[Fe3(CO)10{(EtO)2POP(OEt)2}]
Yield: 21%. IR (CH2Cl2) (cm–1): 2069, 2009, 1955, 1909,
1750, 1711. 13C NMR (CD2Cl2) (RT) δ: 218 (br), 63.2 (t,
NP,C = 7.2 Hz, OCH2), 15.8 (t, NP,C = 7.0 Hz, CH3). 31P
NMR (CD2Cl2) δ: 155. Anal. calcd. for C18H20O15P2Fe3:
C 30.63, H 2.86; found: C 29.43, H 3.31.
Synthesis of [Fe3(CO)9(dppm){P(OMe)3}]
The same method as for [Fe3(CO)10(dppm)] was followed
with the exception that [Fe3(CO)10(dppm)] was used in place
of [Fe3(CO)12] and 1 mol equiv of Me3NO was used. Yield:
18%. MS m/z: 928. IR (CH2Cl2) (cm–1): 2037, 1998, 1940,
1890, 1772, 1717. 13C NMR (CD2Cl2) (RT) δ: 227.9, 223.4,
135.0 (t, NP,C = 45 Hz, ipso), 132.1 (t, NP,C = 10 Hz, ortho),
130.5 (para), 128.8 (t, NP,C = 9 Hz, meta), 52.9 (JP,C = 7 Hz,
OMe), 43.9, (JP,C = 22 Hz, CH2). 31P NMR (CD2Cl2) δ:
166.4, 48.6. Anal. calcd. for C37H31O12P3Fe3: C 47.93,
H 3.26; found: C 47.70, H 3.37.
Crystallographic data for [Fe3(CO)10(dppm)]·1.5NCCH3
Crystal data for C76H53Fe6N3O20P4: MW = 1787.19, crys-
tallizes from acetonitrile as black blocks, crystal dimensions
0.10 × 0.02 × 0.02 mm3. Triclinic, space group P1, a =
11.3686(13), b = 17.2410(19), and c = 20.148(2) Å, α =
82.753(2), β = 83.366(2), and γ = 72.853(2)°, U = 3730.6(7)
Å3, Z = 2, Dc = 1.591 Mg m–3, Mo Kα radiation (λ =
0.71073 Å), µ (Mo Kα) = 1.297 mm–1, F(000) = 1812.
Data collected were measured on a Bruker Smart CCD
area detector with Oxford Cryosystems low temperature sys-
tem. Cell parameters were refined from the setting angles of
64 reflections (θ range 1.02 < 23.35°).
o
o
finement. Complex scattering factors were taken from the
program package SHELXTL (15) as implemented on the
Viglen Pentium computer.
Crystallographic data for
[Fe3(CO)10{(EtO)2POP(OEt)2}]
Crystal data for C18H20Fe3O15P2: MW = 705.83, crystal-
lizes from pet ether as black blocks, crystal dimensions 0.20
× 0.10 × 0.06 mm. Monoclinic, space group P21/n (a non-
standard setting of P21/c Ch, No. 14), a = 9.175(5), b =
18.294(11), and c = 16.584(10) Å, β = 104.489(13)°, U =
Reflections were measured from a hemisphere of data col-
lected of frames each covering 0.3 degrees in omega. Of the
16769 reflections measured, all of which were corrected for
© 2001 NRC Canada

Adams et al.
763
2695(3) Å3, Z = 4, Dc = 1.740 Mg m–3, Mo Kα radiation
(λ = 0.71073 Å), µ (Mo Kα) = 1.781 mm–1, F(000) = 1424.
Data collected were measured on a Bruker Smart CCD
area detector with Oxford Cryosystems low temperature sys-
tem. Cell parameters were refined from the setting angles of
23 reflections (θ range 1.69 < 28.47°).
all non-hydrogen atoms with the exception of the solvent
molecules. Minimum and maximum final electron density –
0.613 and 1.004 e Å–3. Two solvents of crystallization were
found to be disordered and refined as 61.8% acetonitrile and
38.2% ethanol. A weighting scheme w = 1/[σ2(F2) +
o
(0.0225P)2 + 0.00P] where P = (F2 + 2F2)/3 was used in
c
the latter stages of refinement. Coomplex scattering factors
were taken from the program package SHELXTL (15) as
implemented on the Viglen Pentium computer.
Reflections were measured from a hemisphere of data col-
lected of frames each covering 0.3 degrees in omega. Of the
17546 reflections measured, all of which were corrected for
Lorentz and polarization effects and for absorption by semi
empirical methods based on symmetry-equivalent and re-
peated reflections (minimum and maximum transmission co-
efficients 0.7172 and 0.9007) 2572 independent reflections
exceeded the significance level |F|/σ(|F|) > 4.0. The structure
was solved by direct methods and refined by full-matrix
least-squares methods on F2. Hydrogen atoms were placed
geometrically and refined with a riding model (including tor-
sional freedom for methyl groups) and with Uiso constrained
to be 1.2Ueq (1.5 for methyl groups) of the carrier atom. Re-
finement converged at a final R = 0.0766 (wR2 = 0.2053 for
all 6439 data, 343 parameters, mean and maximum δ/σ:
0.000, 0.000) with allowance for the thermal anisotropy of
all non-hydrogen atoms. Minimum and maximum final elec-
tron density –0.762 and 0.890e Å–3.
NMR Investigations
NMR spectra were recorded on a Bruker AMX400 NMR
spectrometer in CD2Cl2. Temperatures were calibrated using
a Comark electronic thermometer, placing a thermocouple in
an NMR tube containing ethanol and placing this tube in the
probe. EXSY spectra were determined qualitatively using
the Bruker pulse sequence “NOESYTP” which had been
1
modified to include H composite pulse decoupling.
∆G‡ values were obtained from line widths. In most cases
the simple equation, k = π∆ν was used below coalescence,
where ∆ν is the line broadening due to exchange. This was
estimated from the spectrum. In the case of rotation about
the C(7)O–P(1) axis in [Fe3(CO)10{(EtO)2POP(OEt)2], only
a broad signal was obtained for C(1)O, C(3)O, C(4)O, and
C(12)O at –106°C. It was assumed that in the absence of ro-
tation about the C(7)O–P(1) axis they would give two sig-
nals of equal intensity separated by 15 ppm. The equation
given below was used to determine the rate.
A weighting scheme w = 1/[σ2(F2) + (0.0824P)2 + 0.00P]
where P = (F2 + F2)/3 was used oin the latter stages of re-
o
finement. Complex cscattering factors were taken from the
program package SHELXTL (15) as implemented on the
Viglen Pentium computer.
π(νA −νB)2
[1]
k =
Crystallographic data for [Fe3(CO)9(dppm){P(OMe)3}]
·0.618NCCH3·0.382CH3CH2OH
2∆ν
Crystal
data
for
C39H35.14N0.62O12.38P3Fe3
(C37H31Fe3O12P3 and two disordered solvents – acetonitrile
and ethanol): MW = 971.04, crystallizes from acetonitrile –
pet. ether as black plates, crystal dimensions 0.26 × 0.15 ×
0.02 mm3. Monoclinic, space group P21/n (a non-standard
setting of P21/c C25h, No. 14), a = 11.1541(16), b =
10.3350(14), and c = 35.384(5) Å, β = 91.678(3)°, U =
4077.3(10) Å3, Z = 4, Dc = 1.582 Mg m–3, Mo Kα radiation
(λ = 0.71073 Å), µ (Mo Kα) = 1.235 mm–1, F(000) = 1982.
Data collected were measured on a Bruker Smart CCD
area detector with Oxford Cryosystems low temperature sys-
tem. Cell parameters were refined from the setting angles of
98 reflections (θ range 1.15 to 28.33°).
Reflections were measured from a hemisphere of data col-
lected of frames each covering 0.3 degrees in omega. Of the
23908 reflections measured, all of which were corrected for
Lorentz and polarization effects and for absorption by semi
empirical methods based on symmetry-equivalent and re-
peated reflections (minimum and maximum transmission co-
efficients 0.7396 and 0.9757) 4145 independent reflections
exceeded the significance level |F|/σ(|F|) > 4.0. The structure
was solved by direct methods and refined by full-matrix
least-squares methods on F2. Hydrogen atoms were placed
geometrically and refined with a riding model (including tor-
sional freedom for methyl groups) and with Uiso constrained
to be 1.2Ueq (1.5 for methyl groups) of the carrier atom. Re-
finement converged at a final R = 0.0648 (wR2 = 0.1180 for
all 9780 data, 522 parameters, mean and maximum δ/σ:
0.000, 0.000) with allowance for the thermal anisotropy of
Results and discussion
Crystallographic sudies
The crystal structures of [Fe3(CO)10L2] (L2 = dppm,
(Ph2P)2NH, {(EtO)2P}2O)
The crystal structures of [Fe3(CO)10L2] (L2 = dppm,
(Ph2P)2NH, {(EtO)2P}2O) are all similar and are shown in
Figs. 1–4. Crystal intensity data, and collection and struc-
tural refinement information are summarized in Table 1. Po-
sitional parameters and isotropic thermal parameters are
given for the four compounds in Tables 2–5. Key distances
and angles are given in Table 6.
The phosphorus ligand substitutes an equatorial carbonyl
on Fe(1) and an axial carbonyl on Fe(3). This contrasts with
the structure of [Fe3(CO)10(dppe)] where the ligand bridges
the Fe(µ-CO)2Fe edge of the iron triangle. For monodentate
phosphorus ligands there is a strong preference for them to
go into an equatorial position (12), and this is the first exam-
ple where a phosphorus group goes into an axial position.
As a result the structure distorts towards a geometry where
this phosphorus is equatorial. This can be seen in a view
down the C(6)O–C(10)O axis, see Fig. 5, where P(2) is
markedly out of the plane of the Fe3 triangle. The group of
five ligands at the front of the projection, C(1)O, C(2)O,
C(7)O, C(11)O, and P(1), have rotated clockwise from the
position found in [Fe3(CO)12]. As a result of this rotation,
carbonyls C(1)O and C(2)O have become semi-bridging.
© 2001 NRC Canada

764
Can. J. Chem. Vol. 79, 2001
Fig. 3. The molecular structure of [Fe3(CO)10
{(EtO)2POPO(OEt)2}] as an ORTEP diagram showing the num-
bering system.
Fig. 1. The molecular structure of [Fe3(CO)10(dppm)] as an
ORTEP diagram showing the numbering system. There are two
independent molecules in the unit cell.
Fig. 4. The molecular structure of [Fe3(CO)9(dppm){P(OMe)3] as
an ORTEP diagram showing the numbering system.
Fig. 2. The molecular structure of [Fe3(CO)10(Ph2PNHPPh2)] as
an ORTEP diagram showing the numbering system.
There has been a smaller clockwise rotation for C(3)O, P(1),
C(4)O, C(9)O, and C(8)O.
In view of the strong tendency of phosphorus ligands to
go into equatorial positions, it might be expected that the
chelating ligands would prefer to coordinate in the positions
occupied by C(5)O and C(6)O, or C(7)O and C(10)O. The
only example of a similar bidentate ligand coordinating to a
single iron in [Fe3(CO)12] derivatives is [Fe3(CO)10{1,2-
(Me2As)2C6H4}] (6). It is possible that such a derivative is
formed for [Fe3(CO)10L2] (L2
= dppm, (Ph2P)2NH,
{(EtO)2P}2O), but the compounds decompose. This leaves
coordination in the positions of C(7)O and C(10)O. This ge-
ometry is adopted by [Fe3(CO)10{(Ph2PC5H4-η5)2Fe}] (16),
[Fe3(CO)10(dppe)] (13), and [Fe3(CO)10{1,2-(Me2As)2-
perfluorocyclobut-1-ene}] (17). For these compounds, the
P···P–As···As distances are 4.797, 3.867, and 4.079–4.083 Å,
respectively. This is a long distance for the ligands dppm,
(Ph2P)2NH, and {(EtO)2P}2O to bridge. The distance can be
© 2001 NRC Canada

Adams et al.
765
Table 1. Crystal data and structure refinement for [Fe3(CO)10L2] (L2 =dppm, Ph2PNHPPh2, (EtO)2POP(OEt)2), and
[Fe3(CO)9(dppm){P(OMe)3}].
[Fe3(CO)10{(EtO)2
POP(OEt)2}]
[Fe3(CO)10(dppm)] [Fe3(CO)10(Ph2PNHPPh2)]
C76H53Fe6N3O20P4 C34.50H22ClFe3N O10P2
[Fe3(CO)9(dppm){P(OMe)3}]
Empirical formula
FW
Temperature (K)
Wavelength (Å)
Crystal system
Space group
C18H20FeO15P2
705.83
150(2)
0.71073
Monoclinic
P21/n
C39H35.14Fe3N0.62O12.38P3
1787.19
150(2)
875.47
293(2)
0.71073
Monoclinic
C2/c
971.04
150(2)
0.71073
Monoclinic
P21/n
0.71073
Triclinic
P1
a (Å)
b (Å)
c (Å)
α (°)
β (°)
γ (°)
Volume (Å3)
11.3686(13)
17.2410(19)
20.148(2)
82.753(2)
83.366(2)
72.853(2)
3730.6(7)
2
22.223(15)
20.984(17)
16.229(15)
90
106.46(6)
90
7258(10)
8
1.602
9.175(5)
18.294(11)
16.584(10)
90
104.489(13)
90
2695(3)
4
1.740
11.1541(16)
10.3350(14)
35.384(5)
90
91.678(3)
90
4077.3(10)
4
1.582
Z
Density (Mg m–3) (calcd.)
Absorption coefficient (mm–1)
F(000)
1.591
1.297
1812
1.402
3528
1.781
1424
1.235
1982
Crystal size (mm3)
θ range for data collection (°)
Index ranges
0.10 × 0.02 × 0.02 0.38 × 0.25 × 0.12
0.20 × 0.10 × 0.06 0.26 × 0.15 × 0.02
1.02–23.35
–12 ≤ h ≤ 11,
–19 ≤ k ≤ 11,
–22 ≤ l ≤ 21
16769
1.36–25.10
–1 ≤ h ≤ 26,
–1 ≤ k ≤ 24,
–19 ≤ l ≤ 18
7266
1.69–28.47
–9 ≤ h ≤ 12,
–15 ≤ k ≤24,
–22 ≤ l ≤ 21
17546
1.15–28.33
–14 ≤ h ≤ 14,
–12 ≤ k ≤ 13, –27 ≤ l
≤ 46
Reflections collected
23908
Independent reflections
10688 (R(int)
= 0.0865)
98.6
6338 (R(int)
= 0.1085)
97.9
6439 (R(int)
= 0.1297)
94.4
9780 (R(int) = 0.1526)
Completeness to θ = 23.35° (%)
Absorption correction
96.3
Semi-empirical
Semi-empirical
Semi-empirical
Semi-empirical
Max. and min. transmission
Refinement method
0.9745 and 0.8812 0.8498 and 0.6178
0.9007 and 0.7172 0.9757 and 0.7396
Full-matrix least-
squares on F2
10688/0/984
1.036
Full-matrix least-squares
on F2
6338/0/465
Full-matrix least-
squares on F2
6439/0/343
Full-matrix least-squares
on F2
9780/5/522
Data/restraints/parameters
GoF on F2
1.033
0.901
0.812
Final R indices [I > 2σ(I)]
R1 = 0.0598, wR2
= 0.1697
R1 = 0.0908, wR2
= 0.2199
R1 = 0.0766, wR2
= 0.1727
R1 = 0.0648, wR2 = 0.0961
R indices (all data)
R1 = 0.0924, wR2
= 0.2215
R1 = 0.1587, wR2
= 0.2657
1.402 and –0.709
R1 = 0.1855, wR2
= 0.2053
0.892 and –0.762
R1 = 0.1740, wR2 = 0.1180
Largest diff. peak and hole (e Å–3) 1.540 and –0.834
1.004 and –0.613
reduced by bridging between the positions occupied by
C(5)O and C(12)O. In the case of [Fe3(CO)12] at 100 K (18),
the C(7)O···C(10)O carbon—carbon distance is 3.948 Å, while
the C(5)O···C(12)O carbon—carbon distance is 3.480 Å. The
distance is reduced further by decreasing the torsion angle
from 46.9° for C(5)-Fe(1)-Fe(3)-C(12) found in [Fe3(CO)12]
at 100 K (18), to between 29.5 and 32.5° for P(1)-Fe(1)-
Fe(3)-P(2). The P···P distances are L2 = dppm (3.067(2),
3.053(2) Å), Ph2PNHPPh2 (2.978(4) Å), and (EtO)2POP(OEt)2
(2.836(3) Å).
bridged by the dppm ligand. It would be wrong to attribute
this change of geometry to the P(OMe)3 ligand as it will be
shown later by 13C NMR spectroscopy that both geometries
are present in solution. Hence, the geometry observed in the
crystal structure is more a function of crystal packing than
of intramolecular factors.
Comparison of the concerted bridge – opening bridge –
closing and the libration mechanisms
The molecules [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH,
{(EtO)2P}2O) provide excellent examples to compare the
concerted bridge – opening bridge – closing and the
libration mechanisms. The L2 ligand bridges between the
C(5)O and C(12)O coordination sites. The P···P distance is
rather short and this will cause a torque to try to rotate the
iron triangle within the icosahedron of ligands. If the
The crystal structure of [Fe3(CO)9(dppm){P(OMe)3}]
The crystal structure of [Fe3(CO)9(dppm){P(OMe)3}] is
shown in Fig. 4, and key distances and angles are given in
Table 1. In contrast with [Fe3(CO)10L2] (L2 = dppm,
(Ph2P)2NH, {(EtO)2P}2O) it is the Fe(µ-CO)2Fe edge that is
© 2001 NRC Canada

766
Can. J. Chem. Vol. 79, 2001
Table 2. Atomic coordinates (1 x 104) and equivalent isotropic
Table 2 (continued).
displacement parameters (1 x 103, Å2) for [Fe3(CO)10(dppm)].
O(2A)
O(3A)
O(4A)
O(6A)
O(7A)
O(8A)
O(9A)
O(10A)
O(11A)
C(1A)
C(2A)
C(3A)
C(4A)
C(6A)
C(7A)
C(8A)
C(9A)
C(10A)
C(11A)
C(13A)
C(14A)
C(15A)
C(16A)
C(17A)
C(18A)
C(19A)
C(20A)
C(21A)
C(22A)
C(23A)
C(24A)
C(25A)
C(26A)
C(27A)
C(28A)
C(29A)
C(30A)
C(31A)
C(32A)
C(33A)
C(34A)
C(35A)
C(36A)
C(37A)
N(1S)
5776(4)
7260(3)
9130(3)
6054(2)
6745(3)
8869(3)
8859(3)
6613(3)
9468(3)
9667(3)
9281(4)
7722(4)
8553(4)
6668(4)
7078(4)
8644(4)
8606(4)
7235(4)
9075(4)
9189(4)
7967(3)
8439(4)
8836(4)
8767(4)
8299(5)
7904(4)
6372(4)
5697(3)
4947(4)
4843(4)
5498(4)
6257(4)
6725(3)
6629(4)
5905(5)
5250(5)
5324(4)
6077(4)
8230(3)
7740(4)
8097(4)
8936(4)
9404(4)
9048(4)
7691(3)
2080(4)
3225(5)
2579(5)
2204(7)
1657(7)
1013(7)
2858(4)
2370(5)
1755(5)
5750(2)
3234(2)
4916(2)
2926(2)
4840(2)
2945(2)
4280(3)
6230(2)
4554(2)
4464(3)
5399(3)
3580(3)
4631(3)
3374(3)
4671(3)
3483(3)
4318(3)
5807(3)
4760(3)
3562(3)
3708(3)
3202(3)
2546(3)
2380(3)
2889(3)
4168(3)
4012(3)
4001(3)
4146(3)
4291(3)
4298(3)
6022(3)
6532(3)
6916(3)
6781(4)
6276(4)
5893(3)
6294(3)
6844(3)
7354(3)
7307(3)
6757(3)
6256(3)
5014(3)
898(3)
37(1)
38(1)
36(1)
51(1)
44(1)
57(1)
58(1)
43(1)
43(1)
34(2)
29(1)
29(1)
29(1)
34(2)
33(2)
40(2)
41(2)
29(1)
33(2)
27(1)
34(2)
40(2)
39(2)
50(2)
39(2)
27(1)
31(1)
37(2)
38(2)
33(1)
32(1)
28(1)
39(2)
54(2)
63(3)
55(2)
43(2)
26(1)
37(2)
44(2)
36(2)
45(2)
39(2)
26(1)
57(2)
66(2)
55(2)
105(3)
77(3)
102(4)
54(2)
51(2)
59(2)
Atom
x
y
z
Ueq
24(1)
3140(4)
5104(4)
5314(5)
7614(4)
6210(4)
7475(4)
3550(4)
1516(4)
4533(5)
5115(5)
3475(5)
4676(5)
4760(6)
6872(6)
6025(6)
6751(6)
3619(5)
2367(6)
1116(5)
–44(5)
–820(6)
–393(6)
796(6)
1553(6)
1982(5)
2979(6)
2790(6)
1628(6)
616(6)
Fe(1)
Fe(2)
Fe(3)
P(1)
P(2)
O(1)
6012(1)
4261(1)
6280(1)
7907(1)
7643(1)
4188(4)
5706(4)
4981(4)
6880(4)
4850(5)
2315(4)
2612(4)
3774(5)
6284(4)
8447(4)
4830(5)
5424(5)
5397(5)
6533(5)
5332(6)
3098(6)
3325(6)
3954(6)
6273(5)
7608(6)
7177(5)
7747(6)
7386(8)
6517(8)
5969(7)
6299(6)
8663(6)
9832(6)
7380(1)
6742(1)
6520(1)
7523(1)
7199(1)
7792(3)
5021(3)
8927(3)
5812(3)
8203(3)
6233(3)
8407(3)
6164(3)
5619(3)
5240(3)
7308(4)
5721(4)
8304(4)
6391(4)
7874(4)
6428(4)
7775(4)
6406(4)
5973(4)
5748(4)
8262(3)
8831(4)
9620(4)
9837(4)
9282(5)
8488(4)
6679(4)
6226(4)
5812(4)
5844(4)
6250(5)
6660(4)
6930(3)
7102(4)
6700(5)
6149(4)
5973(4)
6362(4)
8531(3)
8615(4)
9378(4)
870(1)
608(1)
–190(1)
769(1)
–654(1)
–774(2)
617(2)
47(2)
32(1)
25(1)
25(1)
23(1)
37(1)
39(1)
36(1)
39(1)
50(1)
50(1)
56(1)
59(1)
50(1)
45(1)
31(1)
33(1)
29(1)
29(1)
32(1)
35(2)
39(2)
43(2)
30(1)
33(2)
28(1)
40(2)
57(2)
62(3)
55(2)
41(2)
31(1)
48(2)
57(2)
62(2)
55(2)
49(2)
27(1)
44(2)
52(2)
41(2)
44(2)
39(2)
25(1)
33(1)
32(1)
43(2)
36(2)
32(1)
27(1)
23(1)
30(1)
25(1)
24(1)
23(1)
41(1)
O(2)
O(3)
O(4)
O(6)
O(7)
O(8)
1694(2)
2053(2)
119(2)
724(3)
2022(2)
O(9)
O(10)
O(12)
C(1)
–1327(2)
298(2)
–431(3)
491(3)
C(2)
C(3)
341(3)
C(4)
C(6)
C(7)
1354(3)
1591(3)
293(3)
C(8)
671(3)
C(9)
1480(3)
–891(3)
128(3)
C(10)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
C(28)
C(29)
C(30)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(37)
Fe(1A)
Fe(2A)
Fe(3A)
P(1A)
P(2A)
O(1A)
–1033(3)
–884(3)
–1186(3)
–1643(4)
–1804(3)
–1492(3)
–1340(3)
–1296(3)
–1835(4)
–2415(4)
–2470(3)
–1941(3)
1415(3)
2073(3)
2594(3)
2441(3)
1803(3)
1294(3)
787(3)
676(3)
669(3)
762(3)
883(3)
896(3)
–40(3)
4086(1)
4355(1)
5134(1)
4195(1)
5617(1)
4337(2)
801(5)
2975(5)
3757(5)
4144(7)
3774(8)
3009(8)
2586(7)
1255(5)
922(6)
10537(7)
10061(8)
8804(7)
8118(7)
8934(5)
8545(6)
9295(7)
10410(6)
10773(6)
10028(5)
8108(5)
9312(6)
9514(6)
8505(7)
7328(6)
7149(6)
8717(5)
4029(1)
5732(1)
3686(1)
2162(1)
2391(1)
4249(4)
46(6)
–481(6)
–175(6)
694(6)
1334(5)
8887(6)
7242(7)
8176(7)
5164(8)
6125(10)
7358(9)
1106(6)
1781(7)
2659(7)
C(1S)
C(2S)
N(2S)
C(3S)
C(4S)
N(3S)
C(5S)
C(6S)
1538(4)
1176(4)
2540(4)
2446(4)
2305(5)
4202(3)
3903(4)
3509(4)
10052(4)
9976(4)
9221(4)
7183(3)
7584(1)
8268(1)
8480(1)
7388(1)
7757(1)
9982(3)
Note: Ueq is defined as one third of the trace of the orthogonalized Uij
tensor.
© 2001 NRC Canada

Adams et al.
767
Table 3. Atomic coordinates (1 × 104) and equivalent isotropic
displacement parameters (1 × 103, Å2) for
[Fe3(CO)10(Ph2PNHPPh2)].
Table 4. Atomic coordinates (1 × 104) and equivalent isotropic
displacement parameters (1 × 103, Å2) for
[Fe3(CO)10{(EtO)2POP(OEt)2}].
Atom
x
y
z
Ueq
36(1)
Atom
x
y
z
Ueq
35(1)
Fe(1)
Fe(2)
Fe(3)
P(1)
P(2)
N(1)
O(1)
O(2)
O(3)
O(4)
O(6)
O(7)
O(8)
O(9)
O(10)
O(11)
C(1)
C(2)
C(3)
2838(1)
2087(1)
1769(1)
3105(1)
1731(1)
2456(4)
2076(4)
1058(4)
3445(4)
2237(4)
3889(4)
962(5)
1398(1)
1757(1)
781(1)
715(1)
459(1)
322(3)
566(4)
1932(3)
662(3)
2214(3)
2286(4)
2247(4)
1973(5)
3012(4)
198(4)
–377(4)
944(5)
1573(4)
945(5)
1887(5)
1927(5)
2040(6)
1905(5)
2508(6)
432(5)
73(5)
2995(1)
1458(1)
2232(1)
4081(2)
3528(2)
4175(5)
504(5)
2398(5)
1911(5)
4031(5)
3274(6)
203(6)
Fe(1)
Fe(2)
Fe(3)
P(1)
P(2)
O(1)
O(2)
O(3)
O(4)
O(6)
2009(1)
–545(1)
–488(1)
3261(2)
381(2)
–1141(7)
–2097(7)
2573(7)
1305(6)
4383(7)
–3545(8)
1057(8)
–177(9)
–3381(7)
992(7)
1220(1)
684(1)
2570(1)
1606(1)
3110(1)
3780(1)
3912(1)
2673(4)
1875(4)
3035(4)
1980(4)
1694(4)
565(4)
46(1)
38(1)
33(1)
36(1)
39(2)
64(2)
57(2)
59(2)
65(2)
74(2)
78(3)
101(4)
83(3)
80(3)
65(2)
54(3)
45(2)
45(2)
43(2)
47(3)
53(3)
67(3)
59(3)
46(2)
46(2)
45(2)
56(3)
79(4)
83(5)
91(5)
71(4)
36(2)
43(2)
53(3)
58(3)
63(3)
54(3)
38(2)
57(3)
73(4)
68(3)
59(3)
44(2)
44(2)
68(4)
79(4)
87(5)
99(5)
79(4)
138(11)
172(3)
43(1)
36(1)
34(1)
35(1)
49(2)
51(2)
46(2)
46(2)
62(2)
70(2)
60(2)
69(2)
67(2)
47(2)
37(1)
60(2)
40(1)
41(1)
39(1)
47(2)
43(2)
37(2)
37(2)
40(2)
52(3)
43(2)
53(3)
47(2)
33(2)
45(2)
51(2)
106(5)
66(3)
56(3)
88(4)
49(2)
63(3)
1134(1)
1592(1)
2057(1)
–489(3)
2205(3)
–322(3)
2730(3)
1160(4)
377(4)
O(7)
O(8)
O(9)
–422(4)
1820(4)
900(4)
885(4)
412(4)
3027(5)
2278(5)
596(4)
528(7)
2328(6)
1225(6)
2083(5)
1032(7)
2268(6)
2311(6)
3604(6)
3190(7)
678(7)
O(10)
O(11)
O(12)
O(13)
O(14)
O(15)
O(16)
C(1)
C(2)
C(3)
C(4)
C(6)
3493(5)
4409(4)
3502(3)
4758(4)
4424(3)
3912(3)
4243(3)
2434(6)
2239(5)
2849(5)
2216(5)
2026(5)
985(7)
85(3)
2477(4)
2035(6)
1402(5)
3182(5)
2444(5)
3478(6)
1399(6)
2673(6)
2208(6)
1053(5)
2228(5)
3663(5)
4248(5)
4702(7)
4591(8)
4019(8)
3546(7)
3469(4)
3663(5)
3950(6)
4041(5)
3845(6)
3569(6)
1347(5)
1643(6)
1326(8)
717(7)
–12(6)
–61(8)
3909(6)
4680(6)
2208(6)
–833(9)
2832(3)
2168(4)
967(3)
2107(3)
2074(3)
69(6)
1728(5)
265(5)
2137(5)
1188(5)
512(5)
C(4)
C(6)
C(7)
C(8)
–1421(10)
2294(9)
930(9)
C(9)
2040(7)
1623(7)
2162(6)
4035(6)
3949(7)
3956(9)
4049(8)
4143(10)
4126(8)
5167(5)
5301(6)
6134(7)
6822(7)
6687(6)
5876(6)
4157(6)
5014(7)
5501(8)
5125(9)
4263(8)
3788(7)
3526(6)
3323(8)
3276(9)
3415(10)
3656(12)
3669(10)
2500
1510(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
C(28)
C(29)
C(30)
C(31)
C(32)
C(33)
C(34)
C(35)
C(1S)
Cl(1S)
3461(10)
–2375(12)
452(10)
–253(12)
–2254(12)
484(9)
C(7)
C(8)
C(9)
77(4)
13(5)
1173(5)
934(6)
262(6)
–182(8)
–799(8)
1421(5)
999(4)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
3355(6)
3906(5)
3976(6)
3363(6)
5315(7)
5981(6)
5278(5)
5908(7)
3457(6)
3706(7)
496(5)
–1008(6)
–546(5)
1027(4)
1650(4)
1868(5)
1463(6)
838(5)
431(10)
188(11)
3509(4)
4110(5)
1669(7)
1942(6)
1109(6)
849(8)
–252(19)
–848(13)
4793(11)
4054(13)
4704(10)
6270(11)
2773(5)
3066(6)
632(5)
948(4)
Note: Ueq is defined as one third of the trace of the orthogonalized Uij
tensor.
1093(5)
1440(5)
1638(6)
1515(5)
1176(4)
–320(4)
–862(4)
–1440(6)
–1493(6)
–982(6)
–379(5)
1186(14)
1607(3)
libration mechanism is the lowest energy pathway, then the
ligands P(1), C(6)O, C(7)O, and C(10)O will all be out of
the plane of the iron triangle. If the concerted bridge – open-
ing bridge – closing mechanism is the lowest energy path-
way, then P(1) and C(7)O are permitted to move out of the
plane of the iron triangle, but C(6)O and C(10)O should re-
main in it. Examination of the data in Table 7 shows that
C(6)O and C(10)O do remain in the plane of the iron trian-
gle, while P(1) and C(7)O move substantially out of the
plane. This is strong evidence for the concerted bridge –
opening bridge – closing mechanism over the libration
mechanism.
425(6)
744(5)
1350(5)
1635(7)
1325(8)
772(8)
497(7)
769(6)
5000
5336(3)
3398(5)
Note: Ueq is defined as one third of the trace of the orthogonalized Uij
tensor.
Although it has been shown that the two groups, each
composed of five ligands, rotate about the S10 axis of the
© 2001 NRC Canada

768
Can. J. Chem. Vol. 79, 2001
Table 5. Atomic coordinates (1 × 104) and equivalent isotropic
displacement parameters (1 × 103, Å2) for
[Fe3(CO)9(dppm){P(OMe)3}].
Table 5 (continued).
C(2SA)
N(1SA)
C(1SB)
C(2SB)
O(1SB)
9025(18)
9711(8)
9672(14)
9080(30)
9940(20)
8284(14)
10221(9)
8022(14)
8600(30)
7740(20)
239(5)
621(3)
70(5)
378(8)
–268(5)
109(4)
62(3)
30(4)
104(5)
199(11)
Atom
x
y
z
Ueq
22(1)
Fe(1)
Fe(2)
Fe(3)
P(1)
4709(1)
5083(1)
3058(1)
4724(1)
2181(1)
6159(1)
3164(3)
4646(3)
3937(4)
5902(4)
3052(4)
5702(3)
6850(3)
6301(4)
7301(3)
7289(3)
1257(3)
2332(4)
3570(5)
4369(4)
4224(5)
5436(5)
3701(5)
6502(5)
6229(6)
5817(5)
6460(5)
8171(5)
1955(5)
2615(5)
3374(4)
1368(5)
1962(5)
1404(6)
234(5)
2027(1)
301(1)
284(1)
839(1)
1413(1)
1079(1)
1841(1)
1436(1)
720(1)
18(1)
19(1)
18(1)
19(1)
22(1)
26(1)
25(1)
37(1)
41(1)
45(1)
25(1)
28(1)
35(1)
35(1)
28(1)
35(1)
42(1)
22(1)
17(1)
26(2)
27(2)
29(2)
30(2)
39(2)
22(1)
22(1)
38(2)
21(1)
29(2)
17(1)
18(1)
28(2)
36(2)
32(2)
31(2)
21(1)
19(1)
27(1)
33(2)
32(2)
28(2)
25(1)
16(1)
24(1)
27(1)
27(1)
22(1)
19(1)
18(1)
23(1)
28(1)
27(2)
23(1)
23(1)
45(3)
Note: Ueq is defined as one third of the trace of the orthogonalized Uij
tensor.
–1213(1)
–1216(1)
3340(1)
1685(3)
–1855(3)
4006(4)
61(4)
P(2)
P(3)
O(1)
O(2)
1807(1)
866(1)
icosahedron concertedly (7), the crystal structures of
[Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O) show
that it is possible to distort the icosahedron by slightly rotat-
ing one set of five ligands with respect to the other. This can
be seen qualitatively in Fig. 5, where the L2 ligand closes
the angle between the two sets of ligands, C(1)O, C(11)O,
P2, C(2)O, and C(7)O with C(3)O, P(1), C(4)O, C(9)O, and
C(8)O. This can be seen most easily by comparing the pro-
jected angle between P(1)-Fe(1)-Fe(3)-P(2) with the angles
between P(1)-Fe(1)-Fe(2)-C(11) and C(4)-Fe(1)-Fe(2)-P(2).
The distortion is localized to that side of the icosahedron
and is barely apparent around C(7). This can be verified by
comparing the projected angles between C(8)-Fe(2)-C(7)
and C(7)-Fe(2)-C(9) in Fig. 5.
O(3)
O(4)
1377(1)
377(1)
O(5)
O(6)
O(7)
O(8)
2953(4)
4803(3)
3176(3)
2311(4)
–932(4)
3478(3)
2373(4)
–573(4)
1078(5)
–896(5)
3185(5)
771(5)
244(1)
714(1)
335(1)
1866(1)
1169(1)
1002(1)
1091(1)
313(1)
1558(2)
1026(2)
1175(2)
567(2)
O(9)
O(10)
O(11)
O(12)
C(1)
C(2)
C(3)
NMR spectra and fluxionality
C(4)
For [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O)
the 13C carbonyl NMR spectrum in CD2Cl2 consists of a sin-
glet at approximately δ 223. On cooling to ca. –40°C two
new signals appear at approximately δ 206 and 230 giving
intensity ratios of 2:6:2, respectively. These two signals are
not observed at room temperature as they are exchanging at
an intermediate rate of exchange, and hence, give a signal
too broad to observe. Because at room temperature the sig-
nal at δ 223 is broadened, presumably by exchange with
these two signals, it is possible that all three signals are ex-
changing by a single dynamic process. Pairwise exchange
between the signals at approximately δ 206 and 230 was
demonstrated using exchange spectroscopy (EXSY), see
Fig. 6. It will be shown later that these signals are due to
C(2)O–C(8)O and C(7)O–C(9)O, and the free energies of
activation were derived for this pairwise exchange for all
three compounds, see Table 8.
C(5)
C(6)
C(7)
C(8)
2568(5)
5861(5)
3185(6)
1534(5)
–420(5)
2482(5)
1561(5)
–235(5)
–2143(5)
–2552(5)
–3272(5)
–4316(5)
–4635(5)
–3920(5)
–2877(5)
–672(5)
–21(5)
479(2)
643(2)
–26(2)
1690(2)
1263(2)
1033(2)
1081(2)
609(2)
1695(2)
1204(2)
935(2)
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
C(28)
C(29)
C(30)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(37)
C(1SA)
755(2)
833(2)
–363(5)
184(5)
1150(4)
106(5)
–741(5)
–563(5)
468(5)
1307(5)
4471(4)
4270(5)
3996(5)
3893(5)
4093(4)
4382(4)
5887(5)
6886(4)
7857(5)
7808(5)
6827(5)
5872(5)
9427(10)
1094(2)
1272(2)
1789(2)
1665(2)
1924(2)
2307(2)
2432(2)
2173(2)
2333(2)
2585(2)
2956(2)
3087(2)
2847(2)
2473(2)
1888(2)
2114(2)
2146(2)
1951(2)
1724(2)
1689(2)
450(3)
A similar exchange is observed between C(2)O–C(8)O
and C(9)O in {Fe3(CO)9(dppm){P(OMe)3}], and the activa-
tion energy for this is also given in Table 8.
338(5)
79(5)
–535(5)
–904(5)
–837(5)
–1857(5)
–1640(5)
–369(5)
635(5)
On further cooling of [Fe3(CO)10(dppm)], the 13C NMR
signal at ca. δ 223 broadened and split into two signals at δ
225.4 and 218.6, in a 2:4 ratio (Fig. 7). In the case of
[Fe3(CO)10{(EtO)2POP(OEt)2}], the 13C NMR signal at δ
223 split into two very broad signals at δ 223 and 206, in a
4:2 intensity ratio (Fig. 8). In the case of
[Fe3(CO)10(Ph2PNHPPh2)], the low temperature limit was
not achieved and the central signal of intensity 6 produced a
broad hump, as seen in Fig. 9.
420(5)
–2451(5)
–2163(5)
–3010(5)
–4176(5)
–4471(5)
–3618(5)
9372(9)
The variable temperature behaviour of the 13C NMR spec-
tra of [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O)
proved difficult to explain until the low temperature 13C
NMR spectrum of [Fe3(CO)9(dppm){P(OMe)3}] was re-
corded. The presence of the P(OMe)3 ligand acted as a
block, and the low temperature 13C NMR spectrum consisted
© 2001 NRC Canada

Adams et al.
769
Table 6. Selected bond distances and angles in [Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O) and
[Fe3(CO)9(dppm){P(OMe)3}].
[Fe3(CO)10
[Fe3(CO)10
{(Ph2P)2NH}]
[Fe3(CO)10
{(EtO)2P}2O]
[Fe3(CO)9(dppm)
{P(OMe)3}]
[Fe3(CO)12]
at 100 K
(dppm)]a
Bond distances (Å)
Fe(1)—Fe(2)
2.6804(12)
2.6707(12)
2.6966(11)
2.6889(11)
2.6099(11)
2.6046(11)
2.2279(17)
2.2224(17)
2.2608(17)
2.2599(17)
1.893(6)
2.301(6)
2.314(6)
1.887(6)
2.162(7)
1.870(6)
1.846(6)
2.175(6)
1.859(5)
2.687(3)
2.679(3)
2.602(2)
2.218(3)
2.234(3)
1.833(12)
2.309(11)
2.215(11)
1.860(10)
1.705(8)
1.681(8)
2.978(4)
2.671(2)
2.665(2)
2.614(2)
2.158(3)
2.175(3)
1.843(9)
2.229(10)
2.412(9)
1.839(10)
1.634(6)
1.629(6)
2.836(3)
2.7262(11)
2.7313(11)
2.5176(11)
2.677
2.683
2.551
Fe(1)—Fe(3)
Fe(2)—Fe(3)
Fe(1)—P(1)
Fe(3)—P(2)
2.2411(16)
1.952(6)
1.994(5)
1.953(6)
1.918(5)
1.847(5)
1.859(5)
Fe(2)—C(1)
2.103
1.986
1.969
2.109
Fe(2)—C(2)
Fe(3)—C(1)
Fe(3)—C(2)
P(1)—C or NH or O
P(2)—C or NH or O
1.849(5)
1.838(5)
1.829(6)
3.067(2)
P(1)···P(2)
Fe(2)—P(1)
3.1392(19)
2.2220(16)
3.053(2)
Bond angles (°)
P(1)-C-P(2)
P(1)-Fe(1)-Fe(2)-P(2)
112.1(3)
30.0
123.2(4)
30.6
120.8(3)
29.5
115.8(3)
32.5
112.2(3)
1.0
aTwo independent molecules in each unit cell. Note that the two molecules are mirror images of each other to the first approximation.
This results in reversal of bond lengths involving C(1)O and C(2)O.
by a ligand, and ∆G‡ is less than 16 kJ mol–1. In general, the
merry-go-round mechanism is just stopped at –100°C with
∆G‡ around 40 kJ mol–1. However, with the dppm,
Ph2PNHPPh2, and (EtO)2POP(OEt)2 ligands used in this pa-
per, the ligands are trying to open the Fe(µ-CO)2Fe bridge as
shown by the asymmetric nature of the bridge, and the bar-
rier for the merry-go-round mechanism could be lowered.
Fig. 5. A projection of the crystal structure of
[Fe3(CO)10(dppm)] looking along the S10 axis from C(6)O to
C(10)O.
There are two types of the concerted bridge – opening
bridge – closing mechanism possible in these complexes.
The rotation may occur about the C(6)O–C(10)O axis, see
Scheme 3, or about the C(7)O–P(1) axis, see Scheme 4.
Rotation about the C(6)O–C(10)O axis in Scheme 3 has
the effect of exchanging the role of the two phosphorus at-
oms and making the carbonyl ligands pairwise exchange.
There is no apparent reason for this dynamic process to be
of much higher energy than in [Fe3(CO)12] and is expected
to be fast on the NMR timescale at –100°C. Substitution of
C(7)O by P(OMe)3 in [Fe3(CO)9(dppm){P(OMe)3}] will re-
sult in the blockage of this pathway, as there is a substantial
barrier to the phosphorus ligands going into the axial posi-
tion.
of five signals with an intensity ratio of 2:2:2:2:1, see
Fig. 10.
Previous work has shown that only two dynamic pro-
cesses need to be considered at ca. –100°C (5). In
[Fe3(CO)12] and its derivatives, the concerted bridge – open-
ing bridge – closing mechanism is very fast unless blocked
The ease of rotation about the C(7)O–P(1) axis will de-
pend on the accessibility of the isomer with the chelating
phosphorus ligand along the Fe(µ-CO)2Fe edge. This should
© 2001 NRC Canada

770
Can. J. Chem. Vol. 79, 2001
Table 7. Deviations (in Å) of chosen atoms from the Fe3 plane for [Fe3(CO)10L2] (L2 =
dppm, (Ph2P)2NH, {(EtO)2P}2O), and [Fe3(CO)9(dppm){P(OMe)3}].
[Fe3(CO)10
(dppm)]a
[Fe3(CO)10
{(Ph2P)2NH}]
[Fe3(CO)10
{(EtO)2P}2O]
[Fe3(CO)9(dppm)
{P(OMe)3}]
[Fe3(CO)12]
at 100 K
C(5)
C(6)
C(7)
C(10)
P(1)
P(2)
P(3)
0.072
–0.195
0.103
0.139
–0.025
–0.240
–0.024
0.200
0.051
–0.584
0.025
0.165
–0.953
0.023
–0.323
0.054
0.182
–1.009
–0.117
0.092
0.137
0.067
–0.968
aTwo independent molecules in the unit cell.
Table 8. Free energies of activation (∆G‡) for fluxional processes in [Fe3(CO)10L2] (L2 = Ph2PCH2PPh2,
Ph2PNHPPh2, (EtO)2POP(OEt)2), and [Fe3(CO)10(Ph2PCH2PPh2){P(OMe)3}].
C(7)O—P(1) rotation
(kJ mol–1)
Merry-go-round
(kJ mol–1)
C(2)O, C(8)O and C(7)O,
C(9)O exchange (kJ mol–1)
[Fe3(CO)10{dppm}]
<24
33.7 ± 1.0
46.4 ± 1.0
52.2 ± 1.0
49.5 ± 1.0
42.8 ± 1.0
[Fe3(CO)10{Ph2PNHPPh2}]
[Fe3(CO)10{(EtO)2POP(OEt)2}]
[Fe3(CO)10{dppm}{P(OMe)3}]
33 ± 1.0
ca. 28
37.3 ± 1.0
Fig. 7. A partial 100.62 MHz 13C NMR spectrum of
Fig. 6. A partial 100.62 MHz 13C NMR 2D EXSY NMR spec-
[Fe3(CO)10(dppm)] in CD2Cl2 at –105°C.
trum of [Fe3(CO)10(dppm)] in CD2Cl2 at –36.2°C
The effect of the merry-go-round mechanism is given in
Scheme 5. At ca. –40°C, the 13C carbonyl spectrum of
[Fe3(CO)10L2] (L2 = dppm, (Ph2P)2NH, {(EtO)2P}2O) con-
sists of a 2:6:2 set of signals. No single mechanism produces
six exchanging carbonyls, nor does a combination of two
mechanisms. To achieve this pattern all three mechanisms
shown in Schemes 3–5 have to be combined. This produces
the exchanging sets of carbonyls: C(2)O and C(8)O; C(1)O,
C(3)O, C(4)O, C(6)O, C(10)O, and C(12)O; C(7)O and
C(9)O.
On cooling to around –105°C, the carbonyl 13C NMR sig-
nals split to give a 2:2:4:2 pattern for [Fe3(CO)10(dppm)] and
a 2:4:2:2 pattern for [Fe3(CO)10{(EtO)2POP(OEt)2}]. The be-
haviour is markedly different for the two compounds with
the new signal of intensity 2 being at δ 225.6 for
[Fe3(CO)10(dppm)] and δ 206 for [Fe3(CO)10{(EtO)2POP(OEt)2].
The behaviour is substantially different and suggests for one
the mechanism in Scheme 4 is freezing out, while for the
other it is the mechanism in Scheme 5 that is freezing out.
Combining the mechanisms in Schemes 3 and 4 produces
be accessible for [Fe3(CO)10(dppm)] as this bonding ar-
rangement is found in the crystal structure of
[Fe3(CO)9(dppm){P(OMe)3}]. However, as the P···P dis-
tance shortens on going to [Fe3(CO)10(Ph2PNHPPh2)] and
[Fe3(CO)10{(EtO)2POP(OEt)2}], the barrier to this pathway
should increase. As C(7)O lies on the axis of rotation, sub-
stitution by P(OMe)3 should not block the pathway based on
this rotation, but it will block the pathway based on rotation
about the C(9)O–P(2) given in Scheme 4.
© 2001 NRC Canada

Adams et al.
771
Fig. 8. A partial 100.62 MHz 13C NMR spectrum of
Fig. 10. A partial 100.62 MHz 13C NMR spectrum of
[Fe3(CO)10{(EtO)2POP(OEt)2}] in CD2Cl2 at –106°C.
[Fe3(CO)9(dppm){P(OMe)3}] in CD2Cl2 at –97°C.
Fig. 9. A partial 100.62 MHz 13C NMR spectrum of
[Fe3(CO)10(Ph2PNHPPh2)] in CD2Cl2 at –105°C.
Scheme 3.
Scheme 4.
four sets of exchanging carbonyls: C(2)O and C(8)O; C(3)O
and C(12)O; C(1)O, C(4)O, C(6)O, and C(10)O; C(7)O and
C(9)O. Combining the mechanisms in Schemes 3 and 5 pro-
duces the exchanging sets of carbonyls: C(2)O and C(8)O;
C(1)O, C(3)O, C(4)O, and C(12)O; C(6)O and C(10)O;
C(7)O and C(9)O. This analysis works very well for
[Fe3(CO)10{(EtO)2POP(OEt)2] with the mechanism in
Scheme 4 freezing out. Figure 11 shows the estimated chem-
ical shifts for the carbonyls at the various positions of
[Fe3(CO)12]. This can be used to estimate the 13C chemical
shifts for the carbonyls of [Fe3(CO)10{(EtO)2POP(OEt)2] as-
suming that the fluxional mechanisms are according to
Schemes 3 and 5. The calculated shifts (δ) based on Fig. 11
are as follows: 233.5 for C(2)O, C(8)O; 227.8 for C(1)O,
C(3)O, C(4)O, C(12)O; 207 for C(6)O, C(10)O; and 209 for
C(7)O, C(9)O, which are in tolerable agreement with the ob-
served shifts (Table 9). A perfect agreement would not be
expected as the effects of substitution and distortion from an
ideal geometry would produce deviations from the predicted
values.
When this approach is applied to [Fe3(CO)10(dppm)] with
the mechanism in Scheme 5 frozen out, there is a problem.
The predicted chemical shifts (δ), based on the crystal struc-
ture, for the four sets of exchanging carbonyls are: 233.5 for
C(2)O, C(8)O; 218.5 for C(3)O, C(12)O; 222 for C(1)O,
C(4)O, C(6)O, C(10)O; and 209 for C(7)O, C(9)O. The
agreement is very poor for C(3)O and C(12)O. However, as
the crystal structure of [Fe3(CO)9(dppm){P(OMe)3}] shows
that the geometry with the chelate ligand across the Fe(µ-
CO)2Fe edge is accessible the chemical shifts would be very
different. It follows from the position of the ligands in
Scheme 3, that if this geometry was exclusively present, the
chemical shifts would be δ: 222 C(2)O, C(8)O; 252 C(3)O,
C(12)O; 215 C(1)O, C(4)O, C(6)O, C(10)O; and 203 C(7)O,
C(9)O. Hence, to account for the observed shifts, it is neces-
sary to conclude that isomers 2 and 3 are present in solution
in a ratio of approximately 3:1.
The low temperature 13C NMR spectrum of
[Fe3(CO)9(dppm){P(OMe)3}] is shown in Fig. 10. By anal-
ogy with [Fe3(CO)10(dppm)], only the mechanism in
Scheme 4 should be operating as the presence of the
P(OMe)3 blocks the mechanism shown in Scheme 3. This
leads to the following exchange: C(1)O º C(10)O, C(4)O º
C(6)O, C(3)O º C(12)O, C(2)O º C(8)O, C(6)O º C(10)O,
© 2001 NRC Canada

772
Can. J. Chem. Vol. 79, 2001
Scheme 5.
Fig. 11. Estimated 13C NMR chemical shifts for carbonyls on
[Fe3(CO)12] based on the reported chemical shifts for
[Fe3(CO)12–n{P(OMe)3}n], n = 2, 3.
only previous example of it being stopped is
[Fe3(CO)11(CNCF3)] and related complexes, where the
CNCF3 is tightly held into a µ2-bridging position (19). In the
cases of the [Fe3(CO)10L2] (L2 = dppm, Ph2PNHPPh2,
(EtO)2POP(OEt)2) complexes, rotation about the S10 axis
through C(6)O and C(10)O is fast on the NMR time scale
even at –105°C. However, the barrier to rotation about the
S10 axis through C(7)O and P(1) is increased because of the
strain produced by the chelating ligand in 3. The formation
of a carbonyl bridge on the same edge as the chelate is
bonding increases the ideal separation of the two phosphorus
atoms. The barrier increases from <24 kJ mol–1 in
[Fe3(CO)10(dppm)] to 33
±
2.0 kJ mol–1 in
[Fe3(CO)10{(EtO)2POP(OEt)2}] correlating with the P···P
distance decreasing from 3.053(2)–3.067(2) to 2.836(3) Å.
The merry-go-round mechanism
The strain produced by the chelate ring has a tendency to
open the Fe(µ-CO)2Fe bridge. As the strain increases from
[Fe3(CO)10(dppm)] to [Fe3(CO)10{(EtO)2POP(OEt)2}], the
Fe(µ-CO)2Fe bridge opens more easily and ∆G‡ drops from
33.7 ± 1.0 to ca. 26 kJ mol–1. The clearest demonstration of
and C(9)O is left unaffected. If the structure in solution is 2,
then the predicted chemical shifts are δ: 227.5 C(1)O, C(10)O;
216.5 C(4)O, C(6)O; 218.5 C(3)O, C(12)O; 233.5 C(2)O,
C(8)O; and 215 C(9)O. If the structure in solution is 3, then
the predicted chemical shifts are δ: 215 C(1)O, C(10)O; 215
C(4)O, C(6)O; 252 C(3)O, C(12)O; 222 C(2)O, C(8)O; 211
C(7)O and C(9)O. Comparison of these shifts with the spec-
trum and the observation of JP,C on the signal at δ 232 makes
the probable assignments δ: 225 C(1)O, C(10)O; 218 C(4)O,
C(6)O; 233 C(3)O, C(12)O; 232 C(2)O, C(8)O; 213 C(9)O.
The 13C chemical shifts suggest that 2 is the major isomer in
solution despite the crystal structure being 3.
the
merry-go-round
mechanism
occurs
with
[Fe3(CO)10(dppm){P(OMe)3}], where an EXSY NMR spec-
trum clearly shows C(1)O, C(4)O exchanging with C(3)O,
C(12)O and C(6)O, C(12)O as would be expected when the
merry-go-round mechanism is applied to isomer 3, see
Fig. 12. No direct exchange is detected between the signals
due to C(3)O, C(12)O and C(6)O, C(12)O. The exchange
mechanism in Scheme 5 does not produce this exchange,
and the merry-go-round mechanism has to operate around
the edge Fe(1)–Fe(3), see Scheme 6.
The concerted bridge – opening bridge – closing
mechanism
Usually the concerted bridge – opening bridge – closing
mechanism is extremely facile and cannot be stopped. The
For both the concerted bridge – opening bridge – closing
and
merry-go-round
mechanism,
∆G‡
for
[Fe3(CO)10(Ph2PNHPPh2)] is expected to be between the
© 2001 NRC Canada

Adams et al.
773
Table 9. 13C NMR chemical shifts of the carbonyls of [Fe3(CO)10L2] (L2 = dppm, Ph2PNHPPh2,
(EtO)2POP(OEt)2) and [Fe3(CO)10(Ph2PCH2PPh2){P(OMe)3}] at ca. –100°C in CD2Cl2.
C(1)O,
C(4)O
C(2)O,
C(8)O
C(3)O,
C(12)O
C(6)O,
C(10)O
C(7)O,
C(9)O
[Fe3(CO)10(Ph2PCH2PPh2)]
[Fe3(CO)10(Ph2PNHPPh2)]
[Fe3(CO)10{(EtO)2POP(OEt)2}]
[Fe3(CO)10(Ph2PCH2PPh2){P(OMe)3}]
218.6a
219.1
224a
229.9
230.5
229.8
231.9 (24)
225.5
219.1
224a
218.6a
219.1
206
205.9
206.2
206.6
213.2
233.7
225.1
218.1
aAveraged.
Fig. 12. A partial 100.62 MHz 13C EXSY NMR spectrum of
[Fe3(CO)10(dppm){P(OMe)3}] in CD2Cl2 at –96°C with a mixing
delay of 0.002 s.
Scheme 6.
Conclusions
The distortions observed in the crystal structures induced
in the structure of [Fe3(CO)12] by substitution with the small
bite-angle
ligands
(Ph2PCH2PPh2,
Ph2PNHPPh2,
(EtO)2POP(OEt)2) follow from the predictions of the Mann
concerted bridge – opening bridge – closing mechanism
rather than the Johnson libration mechanism.
The dynamic processes observed in the 13C NMR spectra
can be fully explained using either the Mann concerted
bridge – opening bridge – closing mechanism or the Johnson
libration mechanism, the merry-go-round mechanism, and
the trigonal twist mechanism.
values
determined
for
[Fe3(CO)10(dppm)]
and
[Fe3(CO)10{(EtO)2POP(OEt)2}]. Hence, both values are ex-
pected to be less than 30 kJ mol–1. As a result, neither mech-
anism is frozen out at –105°C and the resulting signal is
broad. No value could be determined for ∆G‡ as it is un-
known if one or both mechanisms are freezing out.
The changes in ∆G‡ from those previously observed with
monodentate ligands are predictable as a consequence of us-
ing small bite-angle ligands.
Trigonal twist mechanism
References
All four complexes reported here show exchange between
C(2)O, C(8)O and C(7)O, C(9)O, or just C(9)O in the case
of [Fe3(CO)10(dppm){P(OMe)3}]. At the onset of the ex-
change, the exchange is localized to C(2)O, C(8)O and
C(7)O, C(9)O for [Fe3(CO)10L2] (L2 = dppm, Ph2PNHPPh2,
(EtO)2POP(OEt)2), although this is not known for
[Fe3(CO)10(dppm){P(OMe)3}]. The most likely mechanism
to fit this exchange is the trigonal twist mechanism. It ap-
pears to have been facilitated in these compounds as in
[Fe3(CO)11{P(OMe)3}] it is believed to occur with ∆G‡ =
1. F.A. Cotton and D.L. Hunter. Inorg. Chim. Acta, 11, L9
(1974).
2. B.F.G. Johnson. J. Chem. Soc. Chem. Commun. 703 (1976).
3. R.E. Benfield, P.D. Gavens, B.F.G. Johnson, M.J. Mays, S.
Aime, L. Milone, and D. Osella. J. Chem. Soc. Dalton Trans.
1535 (1981).
4. D.H. Farrar and J.A. Lunniss. J. Chem. Soc. Dalton Trans.
1249 (1987).
5. H.A. Adams, N.A. Bailey, G.W. Bentley, and B.E. Mann. J.
Chem. Soc. Dalton Trans. 1831 (1989).
62 kJ mol–1 and in [Fe3(CO)10{P(OMe)3}2] with ∆G‡
=
60 kJ mol–1. It would, therefore, appear that substitution of
the carbonyls facilitates the dynamic process.
6. A. Bino, F.A. Cotton, P. Lahuerta, P. Puebla, and R. Usón.
Inorg. Chem. 19, 2357 (1980).
© 2001 NRC Canada

774
7. B.F.G. Johnson and A. Bott. J. Chem. Soc. Dalton Trans. 2437
Can. J. Chem. Vol. 79, 2001
14. W. McFarlane and G. Wilkinson. Inorg. Synth. 8, 181 (1966).
15. G.M. Sheldrick. SHELXL. An integrated system for solving
and refining crystal structures from diffraction data (Revision
5.1), Bruker AXS Ltd.
(1990).
8. H. Adams, A.G. Carr, B.E. Mann, and R. Melling. Polyhedron,
14, 2771 (1995).
9. B.E. Mann. J. Chem. Soc. Dalton Trans. 1457 (1997).
10. (a) H.B. Bürgi and J.D. Dunitz. Acc. Chem. Res. 16, 153
(1983); (b) T.A.D. Heyde. Structure correlation. Vol 1 and 2.
Edited by H.B. Bürgi and J.D. Dunitz. VCH, Weinheim. 1994.
Chap. 8.
11. B.F.G. Johnson. J. Chem. Soc. Dalton Trans. 1472 (1997).
12. E. Stein and F.Y. Fujiwara. J. Organomet. Chem. 525, 31 (1996).
13. H. Adam, S.C.M. Agusthino, B.E. Mann, and S. Smith. J.
Organomet. Chem. 607, 175 (2000).
16. T.-J. Kim, S.-C. Kwon, Y.-H. Kim, N.H. Heo, M.M. Teeter,
and A. Yamano. J. Organomet. Chem. 426, 71 (1991).
17. P.J. Roberts, B.R. Penfold, and J.R. Trotter. Inorg. Chem. 9,
2137 (1970).
18. D. Braga, F. Grepioni, L.J. Farrugia, and B.F.G. Johnson. J.
Chem. Soc. Dalton Trans. 2911 (1994).
19. D. Lentz and R. Marschall. Organometallics, 10, 1487 (1991).
© 2001 NRC Canada