A cw EPR and ENDOR investigation on a series of Cr(i) carbonyl complexes with relevance to alkene oligomerization catalysis: [Cr(CO)4L] + (L = Ph2PN(R)PPh2, Ph2P(R)PPh 2)

Article abstract of DOI:10.1039/c0dt00127a

The preparation and characterisation of the Cr(i) complexes [Cr(CO) ₄L]⁺ (L = Ph₂PN(R)PPh₂, Ph ₂P(R)PPh₂), which are used as pre-catalysts for the selective oligomerization of ethylene, are reported. The electronic properties and structural features of these complexes in frozen solution have been established via continuous wave X-band Electron Paramagnetic Resonance (cw-EPR) and continuous wave ¹H, ¹⁴N and ³¹P Electron Nuclear Double Resonance (cw-ENDOR) spectroscopy. The EPR spectra are dominated by the g anisotropy, with notably large ^PA couplings from the two equivalent ³¹P nuclei. The spin Hamiltonian parameters (g (g _xx = g_yy) > g_e > g_∥ (g_zz)) are consistent with a low-spin d⁵ system possessing C_2v symmetry, with a SOMO where the metal contribution is primarily d_xy for all complexes. The isotropic Fermi contact term ( ^Pa_iso, determined by EPR and ENDOR) was found to be largest for complexes containing ligands e, d, f and g, indicating that the ³¹P 3 s character in the SOMO is higher for the PNP type ligands than the PCP type. Subtle structural differences in the complexes were also identified through variations in the Δg shifts (identified by EPR), and through differences in the phenyl ring conformations (identified by ¹H ENDOR). Attempts to correlate trends in EPR-derived parameters with data measured for catalysis using these pre-catalysts are also made, but no clear connections were found. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c0dt00127a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
A cw EPR and ENDOR investigation on a series of Cr(I) carbonyl complexes
with relevance to alkene oligomerization catalysis: [Cr(CO)4L]+ (L =
Ph2PN(R)PPh2, Ph2P(R)PPh2)†
Lucia E. McDyre,a Tracy Hamilton,a Damien M. Murphy,*a Kingsley J. Cavell,a William F. Gabrielli,b
Martin J. Hantonc and David M. Smithc
Received 12th March 2010, Accepted 9th June 2010
First published as an Advance Article on the web 21st July 2010
DOI: 10.1039/c0dt00127a
The preparation and characterisation of the Cr(I) complexes [Cr(CO)4L]+ (L = Ph2PN(R)PPh2,
Ph2P(R)PPh2), which are used as pre-catalysts for the selective oligomerization of ethylene, are
reported. The electronic properties and structural features of these complexes in frozen solution have
been established via continuous wave X-band Electron Paramagnetic Resonance (cw-EPR) and
continuous wave 1H, 14N and 31P Electron Nuclear Double Resonance (cw-ENDOR) spectroscopy. The
EPR spectra are dominated by the g anisotropy, with notably large PA couplings from the two
equivalent 31P nuclei. The spin Hamiltonian parameters (g^ (gxx = gyy) > ge > gꢀ (gzz)) are consistent
with a low-spin d5 system possessing C2v symmetry, with a SOMO where the metal contribution is
primarily dxy for all complexes. The isotropic Fermi contact term (Paiso, determined by EPR and
ENDOR) was found to be largest for complexes containing ligands e, d, f and g, indicating that the 31
P
3 s character in the SOMO is higher for the PNP type ligands than the PCP type. Subtle structural
differences in the complexes were also identified through variations in the Dg shifts (identified by EPR),
and through differences in the phenyl ring conformations (identified by 1H ENDOR). Attempts to
correlate trends in EPR-derived parameters with data measured for catalysis using these pre-catalysts
are also made, but no clear connections were found.
membered metallocycles; the 7-membered metallocycle decom-
Introduction
poses to give 1-hexene.8,9 Chromium catalysed trimerization with
1 : 1 C2D4 and C2H4 gave only the even numbered isotomers
C6D12, C6D8H4, C6D4H8 and C6H12 in the ratio 1 : 3 : 3 : 1, in
total agreement with the metallocycle mechanism.10 However, an
important unknown in this mechanism is the oxidation state of
Cr during the catalytic cycle. The active catalyst is generated
in situ from a Cr(III) compound, added ligand, and cocatalyst
(commonly, MAO - methylaluminoxane). Cr(III)–Cr(V),9 Cr(I)–
Cr(III)11,12 and Cr(II)–Cr(IV)13,14 couples have been suggested as the
principle oxidation states involved. Cr(V) and Cr(I) complexes are
rare and generally unstable; Cr(V) complexes typically have oxygen
or halide ligands, whereas Cr(I) complexes with isocyanides and
bipy ligands are known.15
At present, a Cr(I)–Cr(III) couple is favoured. However, this
is by no means certain, and it may be that different couples
operate with different catalyst systems or operating conditions.
X-ray photoelectron spectroscopy studies on the Phillips catalyst
support a Cr(I)–Cr(III) couple16 and elegant studies by Ko¨hn and
coworkers lend further weight to a Cr(I)–Cr(III) mechanism.17
Several molecular modelling studies have been undertaken;14,18–21
in one study the Cr(II)-Cr(IV) system was assumed,14 while in others
Ta18 and Ti19-21 systems were considered. The main aim of the
latter study was to understand why insertion of a third ethylene
occurs in preference to liberation of 1-butene, and then why
additional insertions did not occur. Catalytic 1-octene formation
is extremely new, however, and a creditable mechanism has only
recently been published.22 The proposed mechanism is able to
explain the observed by-products, and there is a self-consistency
in the explanations.
A current focus in olefin oligomerization is the design of highly
selective catalysts for the formation of single chain length a-
olefins such as the Phillips Cr based catalyst system, for the highly
selective trimerization of ethylene into 1-hexene.1 Recently, new Cr
complexes containing P–N–P, P–S–P and S–N–S bidentate and
tridentate ligands have been developed as catalysts for 1-hexene
formation,2 and even more excitingly, 1-octene formation (i.e.
selective tetramerization of ethylene).3 Chromium based catalysts
dominate the field, although active Ti and Ta trimerization cata-
lysts have also been reported.4,5 In notable recent studies, Wass and
coworkers have furthermore demonstrated the trimerization and
co-trimerization of substrates other than ethylene.6,7 For example,
ethylene/styrene cotrimerization and isoprene trimerization have
been reported.6,7
It is clear that an entirely different mechanism, to the traditional
Cossee-Arlman mechanism, must be operating and a mechanism
involving Cr-metallocycles is favoured.1,8,10 Evidence to support
this mechanism is growing, with the isolation of Cr 5- and 7-
aSchool of Chemistry, Cardiff University, Main Building, Park Place, Cardiff,
UK CF10 3AT. E-mail: murphydm@cardiff.ac.uk
bSasol Technology (Pty) Ltd, R&D Division, 1 Klasie Havenga Road,
Sasolburg, 1947, South Africa
cSasol Technology (UK) Ltd, Purdie Building, North Haugh, St Andrews,
UK KY16 9ST
† Electronic supplementary information (ESI) available: Figure S1: FSED
EPR spectra; Figure S2a–S2g: Experimental and simulated cw EPR spec-
tra; Figure S3a–S3g: Additional 1H ENDOR data; Figure S4: Additional
14N ENDOR data; Figure S5: Additional 31P ENDOR data; Table S1:
Catalysis data. See DOI: 10.1039/c0dt00127a
7792 | Dalton Trans., 2010, 39, 7792–7799
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Undoubtedly, one of the most versatile and important analytical
tools for characterising the above catalytic systems remains mag-
netic resonance techniques. In the case where paramagnetic oxi-
dation states are involved, the commonly used NMR methods are
rendered less effective or difficult to apply. Determination of ligand
structure near a paramagnetic centre has historically been difficult
because the standard techniques of spectral assignment and NOE-
based distance constraints used for diamagnetic compounds are
not readily applied to paramagnetically perturbed resonances.
However, this type of information can be readily obtained by EPR
and the related hyperfine techniques of ENDOR, HYSCORE,
ESEEM and ELDOR detected NMR. Similar to NMR, these
EPR techniques can be used to study the Cr based catalyst systems
under a variety of conditions (variable temperatures, variable
pressures, in solution, etc), providing information not only on the
principle oxidation states involved but also a structural description
on the complexes in solution. To date most of the available EPR
literature pertaining to Cr complexes has focussed on Cr(III)23,24
and Cr(V)25 compounds, and to a much lesser extent on low spin
Cr(I).26 Bruckner et al.,24 have recently monitored the structure
and valence state of a Cr(III) oligomerization system via in situ
EPR. They found that upon addition of MMAO acting as a
co-catalyst to activate Cr(acac)3/PNP, the Cr(III) EPR intensity
decreased, with a cocomittant increase of new signals between
3200–3600 G. This axial signal with gꢀ = 2.0127 and g^ = 1.9868
was similar to that observed from low spin Cr(I) complexes.26,27
Thus it was shown that Cr(III) is reduced to Cr(I) in the presence
of MMAO. As the reduction of Cr(III) is much faster than Cr(I)
formation, it was postulated that the major species formed may be
an antiferromagnetic Cr(I) dimer or a Cr(II) species.24
Scheme 1
in the EPR tube under an argon atmosphere. The resulting
solutions were deep blue in colour for all [Cr(CO)4a–g]+ complexes.
The cw-EPR spectrum for [Cr(CO)4b]+ is shown in Fig. 1. The
low temperature FSED pulsed EPR and room temperature cw
EPR spectra were also recorded, but no improved resolution was
observed compared to the low temperature cw-measurement (see
Electronic Supplementary Information; ESI, Figure S1 & S2h†).
The spectrum shown in Fig. 1 can be approximately described
as possessing an axial g matrix with well resolved superhyperfine
structure in both the perpendicular and parallel components. The
corresponding EPR simulation is also displayed in Fig. 1 and the
resulting spin Hamiltonian parameters are listed in Table 1.
Whilst EPR offers valuable insights into the redox and elec-
tronic properties of the Cr complexes in the catalytic reaction,
ENDOR (Electron Nuclear DOuble Resonance) provides further
complementary information on the structure of the paramagnetic
complex. This information can be accessed via analysis of the
hyperfine coupling tensor from remote ligand nuclei.28,29 The
paucity of literature pertaining to ENDOR studies of Cr(I)
compounds is remarkable (and there are none using the pulsed
hyperfine methods). In this study we will therefore use cw-EPR and
ENDOR to study the electronic properties and ligand structure on
the series of paramagnetic Cr(I) carbonyl complexes [Cr(CO)4L]+
(L = Ph2PN(R)PPh2, Ph2P(R)PPh2, abbreviated hereafter as P–
N–P and P–C–P ligands respectively) which can be used as model
catalytic systems for the selective oligomerization of ethylene.
Fig.
1 Experimental and simulated cw-EPR spectra (140 K) of
[Cr(CO)4b]+ recorded in dichloromethane–toluene at a microwave fre-
quency of 9.371 GHz. The angular dependency curves calculated for the
Cr(I) g matrix and the 31P A matrix are shown in the lower trace.
Results and discussion
EPR Spectroscopy: the g matrix
The ligands L used for the synthesis of the [Cr(CO)4L]+ complexes
are shown in Scheme 1, labelled a–g. Syntheses of chromium (0)
and chromium(I) complexes of these ligands (1a–g and 2a–g for
Cr(0) and Cr(I) respectively) followed published procedures30–32
and are described in the experimental section. Cr(0) complexes are
quite stable and can be freely handled under an inert atmosphere;
however, care must be taken in the manipulation of the Cr(I)
complexes, which are air and thermally sensitive. For EPR analysis,
each complex 2a–g was dissolved in dry dichloromethane–toluene
Each component of the g matrix is split into an unmistakable
1 : 2 : 1 triplet pattern arising from the superhyperfine interaction
with two equivalent 31P nuclei (I = 12 ) in the P–C–P based ligand
b. Since the natural abundance of 53Cr (I = 3/2) is only ca. 9.5%,
coupled with the large linewidths associated with the 31P hyperfine
pattern, no anisotropic hyperfine interaction associated with 53Cr
was detected in the frozen solution spectrum.
The EPR spin Hamiltonian parameters (g and A) for any
paramagnetic complex will depend on the coordination state and
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 7792–7799 | 7793
©

View Article Online
Table 1 Spin Hamiltonian parameters obtained by simulation for the compounds [Cr(CO)4a–g]+
Complex
Ligand
g^
gꢀ
PA^/Ga
PA^/MHz
PAꢀ/Ga
PAꢀ/MHz
a
iso/G
a
iso/MHz
% s character
2e
2c
2a
2d
2g
2f
PNP
2.089
2.084
2.083
2.077
2.072
2.068
2.063
1.983
1.989
1.989
1.985
1.988
1.988
1.987
29.0
25.5
24.8
27.7
27.0
27.0
24.9
84.8
74.4
72.3
80.5
78.3
78.1
72.0
24.0
25.0
24.5
25.5
25.5
25.5
24.5
66.6
69.6
68.2
70.8
71.0
71.0
68.1
27.3
25.3
24.8
27.0
26.5
26.5
24.7
78.7
72.8
70.9
77.3
75.9
75.8
70.7
0.573
0.532
0.518
0.534
0.556
0.556
0.520
PCCP
PCCP
PNP
PNP
PNP
2b
PCCCP
a A values 0.2 G.
symmetry of the metal centre. For the six coordinate [Cr(CO)4b]+
complex, the strong ligand fields are expected to cause a large
splitting between the t2g orbitals and the eg orbitals resulting
in a low spin (LS) d5 state (S = 12 ). High spin (HS) or
intermediate spin d5 states (S = 5/2 or 3/2) could only occur
in the presence of a weak ligand field, which is not expected in
the [Cr(CO)4b]+ case. Furthermore, the LS Cr(I) complex could
exist in two possible electronic ground states: a (dxz,dyz)4(dxy)1 or
(dxy)2(dxz,dyz)3 configuration. Simple crystal field arguments would
suggest that the expected ground state for the [Cr(CO)4b]+ complex
is (dxz,dyz)4(dxy)1.26 It is proposed that p-back donation to CO helps
to stabilize dxz and dyz relative to the dxy based HOMO. This is
borne out by the observed g matrix. For a SOMO where the metal
contribution is primarily dxy, the components of the g matrix are
given by the following equations:26,27
example, is greatest for [Cr(CO)4e]+ and smallest for [Cr(CO)4b]+
(see Fig. 2 and Table 1).
(cxy )2(cxmz )2
(1a)
(1b)
(1c)
gxx − ge = 2l
gyy − ge = 2l
∑
E0 − Em
m≠0
(cxy )2(cymz )2
E0 − Em
Fig.
2
cw-EPR spectra (140 K) of [Cr(CO)4a–g]+ recorded in
dichloromethane–toluene. The spin Hamiltonian parameters obtained by
simulation are shown in Table 1.
∑
m≠0
(cxy )2(cm
)
2
2
2
x
−y
Despite these clear differences in the Dg shift, caused by the
extent of tetragonal distortion in the complexes, no obvious
correlation emerges between the observed spectral shifts and
the ligand type (i.e., P–N–P or P–C–P based ligands). Some
correlations were identified with respect to the PA hyperfine values,
and these will be discussed later.
gzz − ge = 8l
∑
E0 − Em
m≠0
In these equations ge is the free electron g value, l is the spin–orbit
coupling constant for the free Cr(I) ion (-185 cm-1), cxy is the
m
LCAO coefficient, E0 is the energy of the SOMO and m sums over
the other MOs with energies Em. Spin–orbit coupling mixes in dxz,
2
2
dyz and dx
character, but for a low spin d5 system, dxz and dyz
–y
ENDOR Spectroscopy
will lie just below the SOMO, while dx 2 will be empty and much
2
–y
1
higher in energy. As a result the above equations predict that the
gxx and gyy values should be significantly higher than ge while gzz
should be slightly less than ge. These trends are indeed observed
experimentally, with g^ (gxx = gyy) = 2.063 and ge > gꢀ (gzz) = 1.987
(Table 1), agreeing with a dxy ground state of [Cr(CO)4b]+.
The cw-EPR spectra for all the remaining complexes, 2a–g,
are shown in Fig. 2. In all cases, axial g matrices (g^ > ge > gꢀ)
are observed and the corresponding spin Hamiltonian parameters
for each complex are listed in Table 1 (the individual simulated
spectra are given in the ESI†). Similar to the above discussion on
[Cr(CO)4b]+, it appears that the ground state in all the complexes
can therefore be described as dxy. It should be noted however, that
the resolution of the spectra, and indeed the spin Hamiltonian
values, are found to be highly dependent on the ligand type
(Table 1). The difference in g values (defined as Dg = g^ - gꢀ), for
The H, 14N and 31P cw-ENDOR spectra were also recorded for
all complexes in frozen (deuterated) dichloromethane–toluene at
10K. Analysis ofthese ENDOR spectrawas basedontheobserved
orientational selection in the EPR spectra. It should be recalled
that the powder EPR spectra (shown in Fig. 1 and 2 above)
reflect an average of all molecular orientations of the complex
with respect to the external field (B). If the g (and A) matrix is
known, one can then easily associate distinct sets of molecular
orientations with a given resonant field value. For example when
the applied field (B) = B^ the specific orientations corresponding
to the xy molecular direction (g = g^) are chosen. Similarly when
B = Bꢀ, specific orientations corresponding to the z molecular
direction (g = gꢀ) are selected. These two resonant field positions
(sometimes referred to as ‘single-crystal like’ positions in the
powder pattern29,33,34) are indicated in Fig. 1. This set of molecular
7794 | Dalton Trans., 2010, 39, 7792–7799
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
orientations can then be selected at fixed magnetic field settings
for nuclear resonance (in the ENDOR experiment). The result is
a hyperfine spectrum that contains only one part of the powder
pattern, or at least very few molecular orientations.
˚
closer to Cr (ca. 4.5 A), their interaction would be expected to
be predominantly dipolar in character and once again unlikely to
account for the observed couplings.
Whilst the current [Cr(CO)4a–g]+ complexes are dominated by
the g anisotropy at X-band, the PA coupling is appreciable at
this frequency (Table 1) and this complicates the choice of field
position for the ENDOR measurements. In this case, when the
applied field (B) = B^ at least two sets of specific orientations
are selected (as seen in the angular dependency plot for the Cr
P
g matrix combined with the A superhyperfine lines in the lower
trace of Fig. 1). However, when the ENDOR spectra are recorded
at the field position labeled a in Fig. 1, only a single orientation
(corresponding to qH = 90◦) is selected: effectively a field position
analogous to g = g^. Similarly when the ENDOR spectra are
recorded at the field position labeled b in Fig. 1, only a single
orientation (corresponding to qH = 0◦) is selected: effectively
a field position analogous to g = gꢀ. Although the ENDOR
spectra were measured at several (mixed) field positions, these
latter two unique field positions greatly simplify the analysis of
the subsequent ENDOR data.
Scheme 2
According to the published crystal structure of [Cr(CO)4g]+,31
the two sets of phenyl groups in the complex are twisted with
respect to each other. As a result, the two protons in the ortho-
position of each phenyl ring are structurally inequivalent. This
results in substantially different Cr ◊ ◊ ◊ Ho-phenyl distances; i.e.,
for each phenyl ring one of the ortho-protons has a shorter
1
1H ENDOR. The 1H ENDOR spectra of each complex
recorded at the field position a (effective g = g^ position) are
shown in Fig. 3 for comparison. The spectra contain a matrix
19F peak (labeled * in Fig. 3; nN = 14.0272 MHz for 19F at
3500 G) which arises from the [Al(OC(CF3)3)4]- counter ion
used in the preparation of the Cr(I) complex (see experimental
section). All of the spectra appear to be qualitatively similar,
containing couplings from weakly interacting protons. The outer
ENDOR peaks are quite broad, and this is usually indicative of
a distribution of proton environments, producing a minor strain
on the HA values. The weakly coupled protons, responsible for the
outer lines in the ENDOR spectra, most likely originate from the
phenyl substituents on ligand a–g, since they are common to all
complexes and closest to the Cr centre (see Scheme 2). The methyl
and methine protons in the P–N–P ligand backbone (d–g) are,
1
Cr ◊ ◊ ◊ Ho-phenyl distance compared to the other. The four shortest
1
Cr ◊ ◊ ◊ Ho-phenyl distances from the crystal structure are reported
31
˚
to be 3.30, 3.54, 3.71 and 4.18 A, and these distances are all
easily within range of weakly coupled nuclei detectable by cw-
ENDOR. Since the experimental 1H hyperfine tensor will contain
both isotropic (arising from spin polarisation) and anisotropic
(arising from dipole–dipole interactions) terms, the observed
differences in Fig. 3, must therefore arise either from changes to
the relative conformation of the phenyl groups or due to changes
in the electronic spin delocalisation in the complexes. In order to
examine these changes in more depth, detailed simulations of the
ENDOR data were performed.
The 1H ENDOR spectra recorded at several field positions for
[Cr(CO)4g]+, as a representative example, are shown in Fig. 4 along
with the associated simulations. Two distinct proton environments
account for the spectra.
˚
for example, more than 5 A away from the Cr centre (too remote
to account for the observed hyperfine couplings). Although the
methine protons in the P–C–P ligand backbone (a and b) are
Fig. 4 Experimental and simulated cw 1H ENDOR spectra (10 K)
of [Cr(CO)4g]+ recorded in deuterated dichloromethane–toluene. The
angular selective spectra were obtained at the magnetic field positions
(B in Gauss) and corresponding g values shown in the Figure.
Fig. 3 cw 1H ENDOR spectra (10 K) of [Cr(CO)4a–g]+ recorded in
deuterated dichloromethane–toluene. The spectra were obtained at the
effective field position corresponding to g = g^ for each system. * = 19F
matrix peak from the [Al(OC(CF3)3)4]- counter ion.
However, owing to the broad linewidths observed, only the
nuclei with the largest couplings can be simulated with any
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 7792–7799 | 7795
©

View Article Online
accuracy (i.e., the error associated with the simulated hyperfine
studies devoted to the activation of the complex. In that case,
the structural and electronic changes in the complex may be better
monitored via the changes to the 14N A and Q values via pulsed
hyperfine methods.
couplings for the second set of protons, responsible for the inner
1
˚
peaks, is too large and suggests a Cr ◊ ◊ ◊ H distance of >4 A).
The nucleus responsible for the largest couplings contains a small
aiso contribution (-0.05 MHz), and the resulting Aꢀ term (3.4
MHz) gives an estimated Cr ◊ ◊ ◊ H distance of 3.58 A (calculated
31P couplings. The 31P couplings are sufficiently large, that they
are clearly visible in the EPR spectra (Fig. 1 and 2). However,
they are also present in the ENDOR spectra, as shown in
Fig. 6 for [Cr(CO)4f]+. Owing to the higher resolving powers of
ENDOR, in principle their couplings, and possibly orientations,
can be determined with more accuracy. The isotropic 31P hyperfine
couplings should arise from 31P 3 s character in the SOMO, from
polarisation of inner shell P s orbitals by spin density on the
metal or in P 3p orbitals. These summed contributions can then
be analysed by EPR to account for the observed spin densities.
However, as discussed by Rieger et al.,26 reliable interpretation
of the 31P hyperfine matrix should be treated carefully as the
anisotropies are often small and the g matrix anisotropy is much
greater than the AP matrix (hence observed spectral features
correspond to orientations of the magnetic field along one of
the g matrix principal axes). This will depend on the symmetry
of the complex. Assuming an approximate C2v symmetry for the
[Cr(CO)4L]+ complexes (since the C2 axis runs from Cr through
the N atom bridging the two P atoms, with one vertical mirror
plane containing the two P atoms and another containing the two
CO molecules) the g matrix axes are necessarily along the x, y
and z molecular axes, with the Cr–P vectors lying approximately
midway between the x and y axes (Scheme 2). However, at the low
X-band frequency used in this work, the symmetry of the complex
can be treated as axial for analysis purposes.
1
˚
using a simple point-dipole approximation based on the principal
hyperfine components of A1 = -1.5, A2 = -2.0, A3 = 3.35 MHz).
Analysis of the ENDOR spectra for the remaining complexes
reveals small aiso values in all cases and this suggests that the
variations observed in Fig. 3 for the different ligands must origi-
nate from slight differences in the phenyl group conformations. As
1
mentioned earlier, distinctive Cr ◊ ◊ ◊ H distances of 3.30, 3.54, 3.71
+
˚
and 4.18 A, were identified in the single crystal of [Cr(CO)4g] .
It is highly likely that the main proton observed in the frozen
1
˚
solution ENDOR spectra, with the Cr ◊ ◊ ◊ H distance of 3.58 A,
represents an averaged distribution of the single crystal distances.
This would certainly account for the unusually broad linewidths
of the ENDOR spectra.
14N couplings. The 14N ENDOR spectra were also recorded for
the P–N–P containing ligands ([Cr(CO)4d–g]+). As the 14N nuclei
are not directly coordinated to Cr, their couplings are expected to
be weak. In such weak coupling cases, pulsed hyperfine techniques
such as ESEEM and HYSCORE are ideally suited to extract the
full hyperfine (A) and quadrupolar (Q) matrices. Nevertheless, a
good estimate can also be extracted from the more poorly resolved
cw ENDOR spectra. The 14N ENDOR spectra for [Cr(CO)4f]+,
recorded at four different field positions are shown in Fig. 5.
Fig. 6 Experimental and simulated cw 31P ENDOR spectra (10 K) of
[Cr(CO)4c]+ recorded in deuterated dichloromethane–toluene. The angular
selective spectra were obtained at the magnetic field positions (B in Gauss)
and corresponding g values shown in the Figure.
Fig. 5 Experimental cw 14N ENDOR spectra (10 K) of [Cr(CO)4f]+
recorded in deuterated dichloromethane–toluene. The angular selective
spectra were obtained at the magnetic field positions and corresponding g
values of (a) 3220 G, g = 2.10 (b) 3260 G, g = 2.08 (c) 3370 G, g = 2.01
2
and (d) 3394 G, g = 1.99. * = H matrix peak.
The cw 31P ENDOR spectra were simulated at a number of
field positions for [Cr(CO)4f]+ (Fig. 6). Although the observed
linewidths were broad, the matrix is largely axially symmetric with
As deuterated solvents were used, a deuterium matrix peak is
clearly visible in the spectra (labeled * in Fig. 5; nN = 2.2876 MHz
for 2H at 3500 G). The remaining features in the spectra are
attributed to the superimposed A and Q terms. To a first and very
crude approximation, the values obtained were A1 = 4.38 MHz,
A2 = 4.41 MHz, A3 = 6.65 MHz and Q1 = 0.13 MHz, Q2 =
0.19 MHz, Q3 = 0.32 MHz.
~
A1 = 70 1, A2 = 72 1 and A3 = 83 1 MHz (aiso = 75 MHz
=
26.75 G). This isotropic value is similar to that observed by EPR
(see Table 1, aiso = 26.5 G, allowing for the larger error in the EPR
spectra). Interestingly the PAꢀ value is calculated as 8 MHz, giving
˚
an estimated Cr ◊ ◊ ◊ P bond length of 2.02 A. Clearly this is an
However, since the quadrupole value contains important in-
formation on the electronic structure of the complex, these 14N
spectra will become more important and meaningful in later
under estimation of the distance compared to the known crystal
31
structures of [Cr(CO)4g]+
(Cr ◊ ◊ ◊ P bond lengths of 2.26 and
0 36
˚
˚
2.66 A) and [Cr(CO)4(diphos)] (Cr ◊ ◊ ◊ P distance of 2.36 A). This
7796 | Dalton Trans., 2010, 39, 7792–7799
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
discrepancy arises primarily from the dominant aiso contribution to
the hyperfine tensor, resulting in a higher degree of error associated
with the dipolar term, and due to the limitations in the point-
dipole approximation at such short electron-nuclear distances. For
example, with a hyperfine tensor of A1 = 70, A2 = 72 and A3 =
characterised via cw-EPR and ENDOR spectroscopies. Subtle
differences were identified between the complexes, most notably
in the shifts to the g components and the changes to the 31P 3 s
character. The spin Hamiltonian parameters were found to be
consistent with a low-spin d5 system of C2v symmetry, possessing
a SOMO where the metal contribution is primarily dxy. The
isotropic Fermi contact term (Paiso) was found to be largest for
˚
81 MHz, then Aꢀ = 6.0 MHz giving a Cr ◊ ◊ ◊ P distance of 2.2 A.
For these reasons, it is qualitatively more meaningful to compare
in detail the aiso values rather than the anisotropic PA terms.
It should be noted that the small degree of anisotropy observed
in the PA matrix is analogous to the experimental and calculated
values for a series of Cr(I) carbonyl phosphine and phosphonite
complexes as reportedbyCummings et al.26b The [Cr(CO)4(dppe)]+
complex had reported g values of 2.09, 2.08 and 1.988 and
PA values of 66, 66 and 68 MHz. In all cases, the calculated
anisotropies were small while the predicted aiso values were
analogous to those experimentally observed in this work.
As seen in Table 1, the aiso values appear to be larger for the
P–N–P type ligands compared to the P–C–P ligands. These results
indicate that the electronic 31P 3 s character in the SOMO is higher
for the P–N–P type ligands (2e > 2g ~ 2f > 2d) compared to the
P–C–P type ligands (2b ~ 2a); although 2c has a slightly higher aiso
values in this expected trend. A similar correlation does not exist
with respect to the g values (Table 1). Since the g matrix depends
on the energy of the SOMO, which in turn is affected by the extent
of tetragonal distortion in the complex, then any perturbation or
changes in the coordination and symmetry of the system will be
manifested in the g^ components. Therefore ligands a–g, appear
to alter the electronic properties of the complexes [Cr(CO)4L]+ in
a very subtle and delicate manner.
complexes containing ligands d, f, e and g, indicating that the 31
P
3 s character in the SOMO is higher for the P–N–P type ligands
than the P–C–P types. Observed changes in the g matrix did not
however follow the same trends of ligand type, indicating that g
is dependent not just on the energy of the SOMO but also on
the structural differences in ligand which influence the extent
of tetragonal distortion in the complexes. Structural differences
1
in the [Cr(CO)4a–g]+ complexes were also revealed though H
ENDOR, where the observed spectral changes were attributed to
variations in the phenyl ring conformations as a function of ligand
type. These EPR and ENDOR results reveal that the ligands a–g
impart very subtle electronic and structural alterations to this
class of complex, but that these parameters do not correlate with
any trend in catalytic data at least for the parent pre-catalyst prior
to activation.
Experimental section
General procedures
All manipulations were performed using standard Schlenk tech-
niques under an argon atmosphere, or under a nitrogen atmo-
sphere in a MBraun UNILAB glovebox with less than 0.1 ppm
water and O2. Solvents were dried using a Braun Solvent Pu-
rification System, and degassed prior to use. Ligands a, b and c
were purchased from Aldrich and used as received; d and f were
prepared according to a literature procedure;35 e was prepared
according to a literature procedure;37 g was prepared according
to a literature procedure.39 The corresponding chromium(0)
1a–1g and chromium(I) compounds 2a–2g were prepared accord-
ing to literature procedures.30,31 Ag[Al(OC(CF3)3)4] was prepared
according to a literature procedure.32
Structure function relations
The ligands chosen for study were done so on the basis
that catalytic ethylene tetramerization data has been previously
reported,31,37,38 and furthermore that they represent significantly
different variations of the diphosphine scaffold capable of enabling
this reaction. Pertinent catalytic data was tabulated38 and a
correlation searched for between the measured EPR-derived
parameters (specifically the g values, phosphorus spin densities
and % s orbital character (Fermi contact term), see Table 1) and
various parameters of significance to the tetramerization reactions
(activity, % C6, % C8, % 1-C6, % 1-C8, % C6-cyclics, 1-C8:1-C6
ratio and PE formation). However, no meaningful correlation in
trends could be identified between any combination of parameters.
Given the significant pertubation to the chromium environment
when [Cr(CO)4L]-[Al(OC(CF3)3)4] is activated for catalysis using
excess trialkylaluminium and then placed under elevated pressure
of ethylene, this lack of correlation is perhaps not surprising, and
suggests that more meaningful analysis can only be achieved by
studying species that have been activated in the first instance, and
secondarily under pressure of ethylene. For this reason ongoing
work will extend this initial study of the precatalyst to monitor the
changes to electronic and structure properties after activation of
the catalysts.
Instruments
NMR spectra were recorded at 298 K on Bruker Avance AMX
400 or Jeol Eclipse 300 spectrometers. Chemical shift values are
given relative to residual solvent peak. ESI-MS were performed
on a Waters LCT Premier XE instrument. Infra-red spectra
were recorded using a JASCO FT/IR-660 Plus spectrometer and
analysed in solution (dichloromethane).
EPR/ENDOR measurements
Each complex 2a–g was dissolved in 200 ml DCM–toluene in the
EPR tube and a frozen solution produced by placing the tube in
liquid nitrogen. Each spectrum was recorded at 140 K (EPR) or
10 K (ENDOR).
Instruments. All continuous-wave (cw) EPR spectra were
recorded on an X-band Bruker EMX spectrometer operating at
100 kHz field modulation, 10 mW microwave power and equipped
with a high sensitivity cavity (ER 4119HS). EPR computer
Conclusions
Reactive Cr(I) bis(diphenylphosphine) species, labelled
[Cr(CO)4a–g]+, were prepared and the paramagnetic complexes
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 7792–7799 | 7797
©

View Article Online
simulations were performed using the SimEPR32 program.40
g
NMR (CD2Cl2, 121.7 MHz, 298 K): d (ppm) 83.33 (s); 13C NMR
(CD2Cl2, 125.8 MHz, 298 K): d (ppm) 127.39 (meta-C6H5), 127.60
(para-C6H5), 128.93, 129.65, 131.33 (C6H4), 131.37 (ortho-C6H5),
135.60 (ipso-C6H5); High Resolution ESIpos-MS (MeCN): found
610.0564 (calc 610.0555 dev: 1.4 ppm); IR (CH2Cl2): n = 1893 (s)
(CO), 1916 (s) (CO), 2012 (s) (CO) cm-1.
values were determined using a DPPH standard. All cw ENDOR
measurements were performed on a Bruker ESP300E series
spectrometer operating at 12.5 kHz field modulation and equipped
with an ESP360 DICE ENDOR unit in an EN-801 ENDOR
cavity. The spectra were recorded at 10 K, using 8 dB RF power
from an ENI A-300 RF amplifier, 251 kHz modulation depth
and 4mW microwave power. The ENDOR spectra were simulated
using an in-house programme based on the resonance expressions
given in.34 The EPR linewidths used in the ENDOR simulations
were 5 G.
[Cr(CO)4(Ph2PN(Et)PPh2)] (1d). Yellow solid (350 mg, 50%);
1H NMR (CDCl3, 400.8 MHz, 298 K): d (ppm) 0.75 (t, 3H,
CH3, JHH = 7.3 Hz), 3.00 (m, 2H, CH2), 7.41 (m, 20H, C6H5);
1
31P { H} NMR (CDCl3, 121.7 MHz, 298 K): d (ppm) 114.36 (s);
13C NMR (CDCl3, 125.8 MHz, 298 K): d (ppm) 15.12 (CH3),
43.99 (CH2), 127.52 (meta-C6H5), 129.65 (para-C6H5), 130.86
(ortho-C6H5), 135.55 (ipso-C6H5), 221.24 (cis-CO), 227.22 (trans-
CO); High Resolution ESIpos-MS (MeCN): found 577.0656 (calc
577.0664 dev: -1.4 ppm); IR (CH2Cl2): n = 1891 (s) (CO), 1915
(s) (CO), 2007 (s) (CO) cm-1.
Representative synthesis of chromium(0) tetracarbonyl species
[Cr(CO)4(Ph2PN(iPr)PPh2)] (1g). Toluene (20 ml) was added
to chromium hexacarbonyl [Cr(CO)6], (350 mg, 1.59 mmol) and g
(500 mg, 1.17 mmol) and the stirred mixture was heated under
reflux for 48 h. The solution was cooled to 0 ◦C and filtered
to remove excess [Cr(CO)6]. Solvent was removed under reduced
pressure and the product extracted into dichloromethane (5 ml).
Methanol (10 ml) was added to precipitate the product, which was
isolated by filtration and dried in vacuo to yield the yellow solid
[Cr(CO)4(Ph2PN(iPr)PPh2)] (1 g):
[Cr(CO)4(Ar2PN(Me)PAr2)] Ar = 2-C6H4(Et) (1e). Yellow
1
solid (350 mg, 53%); H NMR (CD2Cl2, 400.8 MHz, 298 K):
d (ppm) 0.85 (br s, 12H, CH3), 2.46 (s, 3H, CH3), 2.61 (br s, 8H,
1
CH2), 7.32 (m, 16H, Ar-H); 31P { H} NMR (CD2Cl2, 121.7 MHz,
298 K): d (ppm) 103.4 (br s); 13C NMR (CD2Cl2, 125.8 MHz,
298 K): d (ppm) 13.27 (CH3), 26.03 (CH2), 33.55 (CN), 124.89
(meta-C6H5), 129.05 (para-C6H5), 134.47 (ortho-C6H5), 144.59
(ipso-C6H5), 219.76 (cis-CO), 227.72 (trans-CO); IR (CH2Cl2):
n = 1864 (s) (CO), 1895 (s) (CO), 2006 (s) (CO) cm-1; High
Resolution ESIpos-MS (MeCN): found 675.1746 (calc 675.1759
dev: -1.9 ppm).
1
Yellow solid (260 mg, 38%); H NMR (CDCl3, 400.8 MHz,
298 K): d (ppm) 0.62 (d, 6H, CH3, JHH = 6.8 Hz), 3.52 (sept, 1H,
CH, JHH = 7.0 Hz), 7.41 (m, 12H, meta-, para-C6H5), 7.69 (m,
8H, ortho-C6H5); 31P {1H} NMR (CDCl3, 121.7 MHz, 298 K): d
(ppm) 112.70 (s); 13C NMR (CDCl3, 125.8 MHz, 298 K): d (ppm)
22.54 (CH3), 54.79 (CH), 127.39 (meta-C6H5), 129.52 (para-C6H5),
130.86 (ortho-C6H5), 136.09 (ipso-C6H5), 221.89 (cis-CO), 227.40
(trans-CO); High Resolution ESIpos-MS (MeCN): found 591.0796
(calc 591.0820 dev: -4.1 ppm); IR (CH2Cl2): n = 1887 (s) (CO),
1923 (s) (CO), 2006 (s) (CO) cm-1.
1
[Cr(CO)4(Ph2PN(iBu)PPh2)] (1f). 31P { H} NMR (CD2Cl2,
121.7 MHz, 298 K): d (ppm) 115.86 (s). The supply of ligand was
limited, therefore the entire complex 1f was converted to the Cr(I)
complex as priority was given to EPR/ENDOR measurements.
All other Cr(0) complexes were synthesised using an analogous
method to give 1a–f. Analytical data are provided below.
32
Synthesis of Ag[Al(OC(CF3)3)4
. LiAlH4 (1.0 g, 0.026 mol)
was suspended in hexane (60 ml), cooled to 253 K and HOC(CF3)3
(15 ml, 0.11 mol) added slowly. The mixture was stirred for
45 min then heated under reflux overnight using a condenser set
at 253 K. The solution was filtered, the product washed with
hexane and solvent removed in vacuo to yield the white solid
Li[Al(OC(CF3)3)4] (20.0 g, 80%); 19F NMR ((CD3)2SO, 250 MHz,
298 K); d (ppm) -75.06.
Li[Al(OC(CF3)3)4] (10.0 g, 0.01 mol) and AgF (1.7 g, 0.013mol)
were suspended in CH2Cl2 (50 ml) in the dark and mixed in an
ultrasonic bath overnight. The solution was filtered and the solvent
removed in vacuo to yield the white solid Ag[Al(OC(CF3)3)4] (8.3 g,
77%).
[Cr(CO)4(Ph2PCH2CH2PPh2)] (1a). Yellow solid (300 mg,
42%); 1H NMR (CD2Cl2, 400.8 MHz, 298 K): d (ppm) 2.00 (t, 4H,
CH2CH2 J = 4.1 Hz), 7.20–7.35 (m, 16H, ortho- and meta-C6H5),
7.50 (m, 4H, para-C6H5); 31P {1H} NMR (CD2Cl2, 121.7 MHz,
298 K): d (ppm) 80.35 (s); 13C NMR (CD2Cl2, 125.8 MHz, 298 K):
d (ppm) 27.30 (CH2CH2), 127.76 (meta-C6H5), 130.36 (para-
C6H5), 131.88 (ortho-C6H5), 137.63 (ipso-C6H5), 219.61 (cis-CO),
228.30 (trans-CO); High Resolution ESIpos-MS (MeCN): found
562.0542 (calc 562.0555 dev: -2.3 ppm); IR (CH2Cl2): n = 1870
(s) (CO), 1902 (s) (CO), 2005 (s) (CO) cm-1.
[Cr(CO)4(Ph2PCH2CH2CH2PPh2)]
(400 mg, 57%); H NMR (CDCl3, 400.8 MHz, 298 K): d (ppm)
(1b). Yellow
solid
1
Representative synthesis of chromium(I) tetracarbonyl species
1.88 (m, 2H, CH2), 2.34 (m, 4H, CH2), 7.32 (m, 20H, C6H5);
1
31P { H} NMR (CDCl3, 121.7 MHz, 298 K): d (ppm) 42.38
Cr(CO)4(Ph2PN(iPr)PPh2][Al(OC(CF3)3)4] (2g). Complex 1g
(100 mg, 0.17 mmol) and Ag[Al(OC(CF3)3)4] (220 mg, 0.23 mmol)
were dissolved in dichloromethane (5 ml) to give a dark blue solu-
tion. The Schlenk tube was covered with foil to reduce exposure of
the reaction mixture to light. The solution was stirred at room tem-
perature overnight, then filtered and the solvent removed in vacuo
to yield the blue solid [Cr(CO)4(Ph2PN(iPr)PPh2)][Al(OC(CF3)3)4]
(s); 13C NMR (CDCl3, 125.8 MHz, 298 K): d (ppm) 18.57
(CH2), 29.64 (CH2), 127.31 (meta-C6H5), 128.42 (para-C6H5),
130.78 (ortho-C6H5), 136.72 (ipso-C6H5), 220.70 (cis-CO), 225.07
(trans-CO); High Resolution ESIpos-MS (MeCN): found 576.0717
(calc 576.0711 dev: 1.0 ppm); IR (CH2Cl2): n = 1885 (s) (CO),
1913 (s) (CO), 2005 (s) (CO) cm-1.
[Cr(CO)4(Ph2PBzPPh2)] (1c). Yellow solid (320 mg, 47%); 1H
NMR (CD2Cl2, 400.8 MHz, 298 K): d (ppm) 7.30 (m, 20H,
(2g): Dark blue powder (105 mg, 40%); High Resolution ESIpos-
MS (MeCN): found 591.0824 (calc 591.0820 dev: 0.6 ppm); IR
(CH2Cl2): n = 1964 (s) (CO), 2032 (s) (CO), 2086 (s) (CO)cm-1.
1
ortho-, meta-C6H5, C6H4), 7.45 (m, 4H, para-C6H5); 31P { H}
7798 | Dalton Trans., 2010, 39, 7792–7799
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
All other Cr(I) complexes were synthesised via an analogous
method (using either 100 mg or 50 mg Cr(0) complex) to give
2a–f: Analytical data is provided below.
6 L. E. Bowen and D. F. Wass, Organometallics, 2006, 25, 555.
7 L. E. Bowen, M. Charernsuk and D. F. Wass, Chem. Commun., 2007,
2835.
8 R. Emrich, O. Heinemann, P. W. Jolly, C. Kru¨ger and G. P. J. Verhovnik,
Organometallics, 1997, 16, 1511.
9 N. Meijboom, C. J. Schaverien and A. G. Orpen, Organometallics, 1990,
9, 774.
10 T. Agapie, S. J. Schofer, J. A. Labinger and J. E. Bercaw, J. Am. Chem.
Soc., 2004, 126, 1304.
11 R. M. Manyik, W. E. Walker and T. P. Wilson, J. Catal., 1977, 47,
197.
[Cr(CO)4(Ph2PCH2CH2PPh2)][Al(OC(CF3)3)4]
(2a). Dark
purple powder (90 mg, 66%); High Resolution ESIpos-MS
(MeCN): found 562.0562 (calc 562.0555 dev: 1.2 ppm); High
Resolution ESIneg-MS (MeCN): found 966.9030 (calc 966.9037
dev: -0.7 ppm); IR (CH2Cl2): n = 1971 (s) (CO), 2034 (s) (CO),
2085 (s) (CO) cm-1.
12 R. D. Ko¨hn, M. Haufe, S. Mihan and D. Lilge, Chem. Commun., 2000,
1927.
[Cr(CO)4(Ph2PCH2CH2CH2PPh2)][Al(OC(CF3)3)4]
(2b).
13 D. H. Morgan, S. L. Schwikkard, J. T. Dixon, J. J. Nair and R. Hunter,
Adv. Synth. Catal., 2003, 345, 939.
14 W. J. van Rensburg, C. Grove, J. P. Steynberg, K. B. Stark, J. J. Huyser
and P. J. Steynberg, Organometallics, 2004, 23, 1207.
15 F. A. Cotton, G. Wilkinson, C. A. Murillo, M. Bochmann, Advanced
Inorganic Chemistry, Wiley, 1999, 1355.
Dark blue powder (145 mg, 54%); High Resolution ESIpos-MS
(MeCN): found 576.0706 (calc 576.0711 dev: -0.8 ppm); High
Resolution ESIneg-MS (MeCN): found 966.9084 (calc 966.9037
dev: 4.8 ppm); IR (CH2Cl2): n = 1954 (s) (CO), 2046 (s) (CO),
2086 (s) (CO) cm-1.
16 Y. Fang, Y. Liu, Y. Ke, C. Guo, N. Zhu and Y. Hu, Appl. Catal., A,
2002, 235, 33.
17 R. D. Ko¨hn, M. Haufe, G. Kociok- Ko¨hn, P. Wasserscheid and W.
Keim, Angew. Chem. Int. Ed., 2000, 39, 4337.
18 Z. Yu and K. N. Houk, Angew. Chem., Int. Ed., 2003, 42, 808.
19 A. N. J. Blok, P. H. M. Budzelaar and A. W. Gal, Organometallics,
2003, 22, 2564.
20 T. J. M. de Bruin, L. Magna, P. Raybaud and H. Toulhoat,
Organometallics, 2003, 22, 3404.
21 S. Tobisch and T. Ziegler, Organometallics, 2003, 22, 5392.
22 M. J. Overett, K. Blann, A. Bollmann, J. T. Dixon, D. Haasbroek, E.
Killian, H. Maumela, D. S. McGuinness and D. H. Morgan, J. Am.
Chem. Soc., 2005, 127, 10723.
[Cr(CO)4(Ph2PBzPPh2)][Al(OC(CF3)3)4] (2c). Dark blue
powder (65 mg, 50%); High Resolution ESIpos-MS (MeCN):
found 610.0540 (calc 610.0555 dev: -2.4 ppm); IR (CH2Cl2): n =
1969 (s) (CO), 2032 (s) (CO), 2086 (s) (CO) cm-1.
[Cr(CO)4(Ph2PN(Et)PPh2)][Al(OC(CF3)3)4] (2d). Dark blue
powder (120 mg, 45%); High Resolution ESIpos-MS (MeCN):
found 577.0648 (calc 577.0664 dev: -2.7 ppm); IR (CH2Cl2): n =
1968 (s) (CO), 2036 (s) (CO), 2089 (s) (CO) cm-1.
23 T. Weyhermuller, T. K. Paine, E. Bothe, E. Bill and P. Chaudhuri, Inorg.
Chim. Acta, 2002, 337, 344.
24 A. Bruckner, J. K. Jabor, A. E. C. McConnell and P. B. Webb,
Organometallics, 2008, 27, 3849.
25 M. Branca, M. Fruianu, S. Sau and M. A. Zoroddu, J. Inorg. Biochem.,
1996, 62, 223; M. Branca, G. Micera and D. Sanna, Inorg. Chem., 1993,
32, 578.
[Cr(CO)4(Ar2PN(Me)PAr2)][Al(OC(CF3)3)4] Ar = 2-C6H4(Et)
(2e). Dark blue powder (150 mg, 62%); High Resolution ESIpos
-
MS (MeCN): found 675.1773 (calc 675.1759 dev: 2.0 ppm); IR
(CH2Cl2): n = 1975 (s) (CO), 2022 (s) (CO), 2052 (s) (CO), 2082
(s) (CO) cm-1.
26 P. H. Rieger, Coord. Chem. Rev., 1994, 135–136, 203; A. L. Rieger
and P. H. Rieger, Organometallics, 2002, 21, 5868; D. A. Cummings, J.
McMaster, A. L. Rieger and P. H. Rieger, Organometallics, 1997, 16,
4362.
27 B. R. McGarvey, Coord. Chem. Rev., 1998, 170, 75.
28 A. Schweiger, G. Jeschke, Principles of pulse electron paramagnetic
resonance, Oxford University Press, 2001.
29 D. M. Murphy and R. D. Farley, Chem. Soc. Rev., 2006, 35, 249.
30 L. E. Bowen, M. F. Haddow, A. G. Orpen and D. F. Wass, Dalton
Trans., 2007, 1160.
31 A. J. Rucklidge, D. S. McGuinness, R. P. Tooze, A. M. Z. Slawin, J. D. A.
Pelletier, M. J. Hanton and P. B. Webb, Organometallics, 2007, 26,
2782.
[Cr(CO)4(Ph2PN(iBu)PPh2)] Al(OC(CF3)3)4 (2f). The supply
of complex was limited, therefore priority of use of 2f was given
to EPR/ENDOR measurements.
Acknowledgements
L. McDyre and T. Hamilton would like to thank Sasol Technology
(Pty) Ltd. for funding. We are grateful to the referees for their
extensive and helpful comments, and particularly for providing
further confirmation on the magnitude and sign of the g matrix.
32 I. Krossing and A. Reisinger, Coord. Chem. Rev., 2006, 250,
2721.
33 B. M. Hoffman, J. Martinsen and R. A. Venters, J. Mag. Reson., 1984,
59, 110.
Notes and references
34 G. C. Hurst, T. A. Henderson and R. W. Kreilick, J. Am. Chem. Soc.,
1985, 107, 7294; T. A. Henderson, G. C. Hurst and R. W. Kreilick,
J. Am. Chem. Soc., 1985, 107, 7299.
35 M. J. Overett, K. Blann, A. Bollmann, J. T. Dixon, F. M. Hess, E.
Killian, H. Maumela, D. H. Morgan, A. Neveling and S. Otto, Chem.
Commun., 2005, 620.
36 M. J. Bennett, F. A. Cotton and M. D. LaPrade, Acta Crystallogr., Sect.
B: Struct. Crystallogr. Cryst. Chem., 1971, 27, 1899.
37 K. Blann, A. Bollmann, H. de Bod, J. T. Dixon, E. Killian, P.
Nongodlwana, M. C. Maumela, H. Maumela, A. E. C. McConnell,
D. H. Morgan, M. J. Overett, M. Pre´torius, S. Kuhlmann and P. J.
Wasserscheid, J. Catal., 2007, 249, 244.
1 M. E. Lashier, EP 0780353A1, 1997, Phillips Petroleum; J. T. Dixon,
M. J. Green, F. M. Hess and D. H. Morgan, J. Organomet. Chem., 2004,
689, 3641.
2 D. S. McGuinness, P. Wasserscheid, W. Keim, C. Hu, U. Englert, J. T.
Dixon and J. J. C. Grove, Chem. Commun., 2003, 334; J. T. Dixon,
J. J. C. Grove, P. Wasserscheid, D. S. McGuinness, F. M. Hess, H.
Maumela, D. H. Morgan, A. Bollmann, WO 03053891A1 2003; D. S.
McGuinness, P. Wasserscheid, W. Keim, D. Morgan, J. T. Dixon, A.
Bollmann, H. Maumela, F. Hess and U. Englert, J. Am. Chem. Soc.,
2003, 125, 5272; J. T. Dixon, P. Wasserscheid, D. S. McGuinness, F. M.
Hess, H. Maumela, D. H. Morgan, A. Bollmann, WO 03053890A1,
2003 (Sasol Technology).
38 M. J. Overett, K. Blann, A. Bollmann, R. de Villiers, J. T. Dixon,
E. Killian, M. C. Maumela, H. Maumela, D. S. McGuinness, D. H.
Morgan, A. Rucklidge and A. M. Z. Slawin, J. Mol. Catal. A: Chem.,
2008, 283, 114.
39 M. S. Balakrishna, T. K. Prakasha and S. S. Krishnamurthy,
J. Organomet. Chem., 1990, 390, 203.
40 T. P. P. Spalek and Z. Sojka, J. Chem. Inf. Model., 2005, 45, 18.
3 M. J. Overett, K. Blann, A. Bollmann, J. T. Dixon, F. M. Hess, E.
Killian, H. Maumela, D. H. Morgan, A. Neveling and S. Otto, Chem.
Commun., 2005, 622.
4 P. J. W. Deckers, B. Hessen and J. H. Teuben, Angew. Chem., Int. Ed.,
2001, 40, 2516.
5 C. Andes, S. B. Harkins, S. Murtuza, K. Oyler and A. Sen, J. Am. Chem.
Soc., 2001, 123, 7423.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 7792–7799 | 7799
©