Intramolecular formation of a CrI(bis-arene) species via TEA activation of [Cr(CO)4(Ph2P(C3H 6)PPh2)]+: An EPR and DFT investigation

Article abstract of DOI:10.1021/om2006062

Activation of the catalytically relevant complex [Cr(CO) ₄(1)]⁺ (1 = Ph₂P(C₃H ₆)PPh₂) by Et₃Al (TEA) leads to formation of the Cr(I) bis-arene complex [Cr(1-bis-η⁶-arene)]⁺, as revealed by EPR and DFT calculations. This bis-arene complex is formed by intramolecular rearrangement and coordination of Cr(I) to the ligand phenyl groups in aliphatic solvents following loss of CO, preventing release of Cr(I) into solution. By comparison in aromatic solvents (toluene), the [Cr(bis-tolyl)]⁺ complex is preferentially formed.

Full text of DOI:10.1021/om2006062

COMMUNICATION
pubs.acs.org/Organometallics
Intramolecular Formation of a CrI(bis-arene) Species via TEA
Activation of [Cr(CO)4(Ph2P(C3H6)PPh2)]+: An EPR and DFT
Investigation
Lucia McDyre,† Emma Carter,† Kingsley J. Cavell,*,† Damien M. Murphy,*,† James A. Platts,†
Katharine Sampford,† Benjamin D. Ward,† William F. Gabrielli,‡ Martin J. Hanton,‡ and David M. Smith‡
†School of Chemistry, Cardiff University, Cardiff CF10 3AT, U.K.
‡Sasol Technology (U.K.) Ltd, Purdie Building, North Haugh, St Andrews, U.K.
S Supporting Information
b
ABSTRACT: Activation of the catalytically relevant complex
[Cr(CO)4(1)]+ (1 = Ph2P(C3H6)PPh2) by Et3Al (TEA) leads to
formation of the Cr(I) bis-arene complex [Cr(1-bis-η6-arene)]+, as
revealed by EPR and DFT calculations. This bis-arene complex is
formed by intramolecular rearrangement and coordination of Cr(I) to
the ligand phenyl groups in aliphatic solvents following loss of CO,
preventing release of Cr(I) into solution. By comparison in aromatic
solvents (toluene), the [Cr(bis-tolyl)]+ complex is preferentially
formed.
r-based complexes are key catalytic species in the industrially
important selective trimerization/tetramerization of ethylene.1
Scheme 1. Structure of the CrI(CO)4 Bis(diphenylphosphino)-
propane and Bis(dicyclohexylphosphino)propane Complexes
C
Although Cr(III) complexes are commonly used as catalyst
precursors, lower oxidation state Cr species, e.g. Cr(I), are
thought to be the active catalysts leading to production of
predominantly 1-hexene and 1-octene.2 The bis(phosphine)
family of ancillary ligands, such as [Cr(CO)4(1)]+ (Scheme 1),
are frequently employed for this important process;2 the activity
and selectivity of the catalytic reaction can be tuned by varying
the properties of the ligand.2 It is also known that the choice of
solvent is an important variable in the catalytic reactions; e.g.,
deactivation of the catalysts is known to occur in aromatic
solvents compared to aliphatic solvents.3
Although the structures of these Cr precatalysts have been
investigated by different techniques, the nature of the catalytically
active complexes remains unclear. Furthermore, the oxidation
state of the chromium center which exists during the catalytic
reaction remains ambiguous, with Cr(I/III) and Cr(II/IV) redox
cycles having been implicated, while possible changes to the spin
state during the catalytic cycle adds further complexity to the
problem.4 The Cr precatalysts are not reactive for the trimeriza-
tion/tetramerization reaction and must therefore be activated
before use. Activation of the Cr(III) or Cr(I) precursor com-
plexes prior to catalysis is usually achieved by addition of a
cocatalyst, such as triethylaluminium (Et3Al) or methylalumi-
noxane (MAO).5 One role of the cocatalyst is thought to be
alkylation of the metal center following loss of the coordinating
carbonyl groups.2a,4 Hence, investigations into the structure of
the Cr complex after activation are crucial for a further under-
standing of this industrially important reaction.
Received: July 8, 2011
Published: August 15, 2011
r
2011 American Chemical Society
4505
dx.doi.org/10.1021/om2006062 Organometallics 2011, 30, 4505–4508
|

Organometallics
COMMUNICATION
Figure 1. X-band CW-EPR spectra of [Cr(CO)4(1)]+ dissolved in
dichloromethane solvent after activation with Et3Al/toluene (1/10 Cr/
Et3Al (equiv)) at (a) 140, (b) 160, (c) 180, (d) 200, (e) 220, (f) 240, and
(g) 260 K. The EPR spectrum of the starting [Cr(CO)4(1)]+ (140 K) is
shown in the inset.
Figure 2. X-band CW-EPR spectra (185 K) of [Cr(CO)4(1)]+ dis-
solved in dichloromethane, following activation with (a) Et3Al/toluene,
(c) Me3Al/toluene, and (d) Et3Al/hexane. An EPR simulation is shown
as a dashed line in (b).
Table 1. Computed vs. Experimental hfc Values (MHz, top)
and g Values (bottom) for the [Cr(1-bis-η6-arene)]+ Complex
It is well-known that low-valent chromium has a strong
tendency to form η6-arene complexes.6 Indeed, it was recently
reported that a solvent-derived CrI(bis-arene) complex, [Cr(η6-
CH3C6H5)2]+ or [Cr(bis-tolyl)]+, was formed by activation of a
Cr(acac)3/Ph2PN(i-Pr)PPh2 system via MAO addition in an
aromatic solvent.7 Its formation was attributed to demetalation of
the complex and interaction of Cr(I) with the aromatic (toluene)
solvent.7 A range of substituted Cr(I) bis-arene complexes has
been previously studied by electron paramagnetic resonance
(EPR) spectroscopy, and all produce very characteristic and
distinct EPR spectra.8 Furthermore, bridged bis-arene Cr(I)
complexes have also been reported, including those formed with
the bis(triphenylphosphino) ligands.9 Hence, one must consider
the likelihood that analogous Cr(I) bis-arenes could conceivably
be formed during the activation of the chromium bis(phosphine)
complexes for selective trimerization/tetramerization of ethy-
lene. To examine this, we applied EPR techniques coupled with
DFT to explore the TEA activation of [Cr(CO)4(1)]+.
Haiso
HA(^, ||)
a
A
(^, ||)
Paiso
PA(^, ||)
Cr
Cr
iso
ref 8ca
DFT
9.72
10.22
8.72
50.85
75.6
1.4
10.91
9.65
10.15
12.57
43.722
61.245
57.514
12.406
73.0
À4.594
1.889
À6.536
À9.135
nr
exptlb
nrd
50.5
<3.78
nr
nr
giso
g^
g||
ref 8ca
DFT
exptlc
1.9865
1.9856
1.988
1.9785
1.9771
1.979
2.0023
2.0027
2.003
a Literature values for [Cr(bis-benzene)]+ complex. b Errors (0.15 MHz.
c Errors (0.0005. d nr = not resolved experimentally.
The continuous-wave (CW) EPR/ENDOR spectra for a series
of [Cr(CO)4PP]+ complexes (including [Cr(CO)4(1)]+) were
recently reported by us;10 the EPR spectra were dominated by a
pronounced ganisotropy and large PAcouplings, typical for low-spin
d5 Cr(I) with a SOMO of largely dxy character.10 A representative
CW-EPR spectrum10 for the precursor [Cr(CO)4(1)]+ complex is
shown in Figure 1 (inset). These bidentate diphosphine ligands
containing chromium are highly effective for ethylene trimerization
and tetramerization following activation.1 Upon addition of an
Et3Al/toluene solution to the [Cr(CO)4(1)]+ complex at 298 K,
the blue coloration of the precursor complex was immediately lost.
The solutions were rapidly frozen to 77 K after activation. Full
details on sample preparation, activation, and EPR measurements
are given in the Supporting Information.
The resulting EPR spectra of the activated system are shown
in Figure.1. The low-temperature spectrum (140 K) was broad
and poorly resolved, possessing a quasi-axial profile due to small
amounts of unreacted [Cr(CO)4(1)]+ (Figure 1a). As the
temperature was raised, a well-resolved isotropic signal pro-
gressively emerged (Figure 1bÀg). This new signal was con-
sistently observed in these experiments following Et3Al
addition, regardless of Et3Al concentration (see the Supporting
Information, Figure S1).
For clarity, an expanded view of the isotropic signal
(Figure 1g) is shown in Figure 2a. This spectrum (hereafter
labeled signal I) was simulated on the basis of a giso value of 1.988,
with 10 equivalent protons (1H; I = 1/2, Haiso = 9.65 MHz) and a
weak 53Cr hyperfine pattern (9.5% abundance, I = 3/2, Craiso
=
50.5 MHz) (Figure 2b). The low-temperature spectrum (shown
in Figure 1a) was simulated using an axial g tensor (Table 1;
simulation shown in the Supporting Information, Figure S2) but
without any resolved 1H hfc’s. The isotropic spectrum does not
result from an anisotropic summation of hyperfine couplings
originating from a coordinated ethyl fragment; in that case much
larger Haiso values are predicted. Furthermore, an identical EPR
spectrum was obtained following activation of [Cr(CO)4(1)]+
using Me3Al/toluene (Figure 2c); coordinated ethyl or methyl
groups are therefore not responsible for the hfc’s. Consequently,
this new signal can only be assigned to a [Cr(bis-arene)]+ type
complex, to account for the 10 equivalent protons, by analogy
4506
dx.doi.org/10.1021/om2006062 |Organometallics 2011, 30, 4505–4508

Organometallics
COMMUNICATION
Scheme 2. Structure of the Intramolecular Formed
CrI(bis-arene) Complex, Formed by Et3Al Activation of
[Cr(1)]+ in Aliphatic Solvents (Hexane), and the Bis-Tolyl
Complex Formed in Toluene
with the reported EPR spectra of similar CrI bis-arene com-
plexes.8 Note that any magnetic inequivalency in these protons
was not resolved in the EPR spectrum. The reversed g tensor of
the new signal (g|| > g^), in comparison to that for the
Figure 3. DFT structure of the [Cr(1-bis-η6-arene)]+ complex.
precatalyst complex ([Cr(CO)4(1)]+; g^ = 2.063 > g||
=
attempts to isolate this complex. As a result, DFT calculations
were carried out and the optimized structure of [Cr(1-bis-η6-
arene)]+ is shown in Figure 3; the calculated spin Hamiltonian
parameters are given in Table 1. The g, CrA, and HA values were in
good agreement with the EPR data. However, the calculated
DFT value of Paiso = 4.594 MHz was not observed experimen-
tally, where the intrinsic line width of the CW-EPR spectrum was
1.987)10 and the g|| value of 2.0023, is also consistent with a
bis-arene assignment possessing a SOMO of predominantly dz2
character.8a
As stated earlier, activation of a CrIII system (Cr(acac)3/PNP)
by MAO produces the [Cr(bis-tolyl)]+ complexes in toluene, as
the liberated CrI ions form the bis-arene sandwich complex with
the solvent.7 This interpretation can also account for the spectra
shown in Figure 2a,c, since toluene was present. However, when
[Cr(CO)4(1)]+ was activated using Et3Al/hexane, a virtually
identical EPR spectrum was obtained (Figure 2d, labeled signal
II). Clearly the EPR spectrum shown in Figure 2d is characteristic
of a [Cr(bis-arene)]+ complex but cannot arise from coordina-
tion of Cr(I) by the hexane solvent and must therefore be ligand
derived. In this case the complex is formed by an intramolecular
rearrangement of Cr(I) in the activated and decarbonylated
[Cr(CO)4(1)]+ system (labeled [Cr(1-bis-η6-arene)]+ in
Scheme 2). Coordination of Cr(I) by the two phenyl rings
accounts for the 10 equivalent protons observed in the spectrum
(Figure 2d). We also observed similar intramolecular bis-arene
complexes for other members of the Cr(I) bis(phosphine)
complexes (see the Supporting Information, Figure S3).
This ligand-based [Cr(1-bis-η6-arene)]+ assignment and the
role of the ligand phenyl groups in its formation were further
confirmed by synthesizing the cyclohexyl derivative bearing no
phenyl groups (i.e., [Cr(CO)4bis(dicyclohexylphosphine)]+,
[Cr(CO)4(2)]+ in Scheme 1). [Cr(CO)4(2)]+ was dissolved
in dichloromethane and reacted with Et3Al/hexane; in this case
no [Cr(bis-arene)]+ signal was observed (see the Supporting
Information, Figure S5) in the absence of any arene ring to
coordinate the Cr(I). However, when this [Cr(CO)4(2)]+
complex was dissolved in dichloromethane/toluene (50/50)
and activated using Et3Al/hexane, a strong [Cr(bis-arene)]+
signal was subsequently detected by EPR (see the Supporting
Information, Figure S5); in this case the bis-arene can only arise
by reaction of Cr(I) ions with the toluene in solution: i.e.,
producing the solvent-based [Cr(bis-tolyl)]+ species I.
P
7.6 MHz, implying a maximum experimental aiso of only 3.78
MHz (assuming a triplet from two equivalent 31P nuclei).12 We
attempted to resolve the 31P using CW-ENDOR (see the
Supporting Information, Figure S7). Unfortunately, the spectra
were dominated by large anisotropic hfc’s to the arene ring
protons; thus, the 31P hfc’s could not be detected.
The above results indicate that the nature of the observed
[Cr(bis-arene)]+ complex depends on which solvent is used
(solvent-based [Cr(bis-tolyl)]+ in aromatic solvents or ligand-
based [Cr(1-bis-η6-arene)]+ in aliphatic solvents; Scheme 2). To
investigate which one is preferentially formed under competitive
conditions, a further experiment was performed in which
[Cr(CO)4(1)]+ was dissolved in dichloromethane/perdeuter-
ated-toluene (50/50) and activated with Et3Al/hexane. The
spectral resolution disappeared completely, producing a broad
unresolved line (as expected, 10 equivalent 2H compared to 1H,
with giso = 1.987 and 2Haiso ≈ 2.8 MHz) due to the deuterated
bis-tolyl [Cr(η6-CD3C6D5)2]+ (see the Supporting Informa-
tion, Figure S5). In other words, the solvent-based [Cr(η6-
CD3C6D5)2]+ complex forms preferentially over the ligand-
based [Cr(1-bis-η6-arene)]+.
In conclusion, we have shown that intramolecular coordina-
tion of the Cr(I) center to the ligand phenyl groups occurred
when [Cr(CO)4(1)]+ was activated using Et3Al in the absence of
aromatic solvents. The resulting [Cr(1-bis-η6-arene)]+ complex
was characterized by EPR and DFT. However, in aromatic
solvents (such as toluene), solvent-based [Cr(bis-tolyl)]+ com-
plexes are preferentially formed. These results are fully consistent
with the enhancement in catalytic performance observed when
Cr(I)bis(phosphine) complexes are used in aliphatic solvents,
since studies have shown deactivation occurs more rapidly in
aromatic solvents compared to the longer catalyst lifetimes
observed in alkane solvents.3 While [Cr(bis-tolyl)]+ is catalyti-
cally inactive, the facile formation of [Cr(1-bis-η6-arene)]+
simply by CO loss implies reversible phosphine binding could
The paramagnetic [Cr(1-bis-η6-arene)]+ complex represents
a small component of the total [Cr(CO)4(1)]+ content,11
indicating that a large fraction of the Cr(I) centers were
converted to EPR-silent or diamagnetic species following Et3Al
addition. This mixture of chromium centers hampered any
4507
dx.doi.org/10.1021/om2006062 |Organometallics 2011, 30, 4505–4508

Organometallics
COMMUNICATION
occur under high pressure catalytic conditions, suggesting an
indirect role for this complex in the catalysis.
(12) The Fermi contact term (Paiso) for 31P is very large: 13 291
MHz. Hence, the small difference between the calculated Paiso value of
4.594 MHz vs the estimated experimental value of Paiso = 3.78 MHz is
likely due to the limitations of the calculations.
’ ASSOCIATED CONTENT
S
Supporting Information. Text, figures, and a table giving
b
details of the synthesis and characterization data for 1 and 2 and
full details for the EPR/ENDOR experiments and computational
DFT studies. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ ACKNOWLEDGMENT
L.M. acknowledges funding from Sasol. EPSRC funding
(EP/H023879) is also acknowledged.
’ REFERENCES
(1) (a) Dixon, J. T.; Green, M. J.; Hess, F. M.; Morgan, D. H.
J. Organomet. Chem. 2004, 689, 3641. (b) McGuinness, D. S. Chem. Rev.
2011, 111, 2321. (c) McGuinness, D. S. Chem. Rev. 2011, 111,
2321–2341.
(2) (a) Rucklidge, A. J.; McGuinness, D. S.; Tooze, R. P.; Slawin,
A. M. Z.; Pelletier, J. D. A.; Hanton, M. J.; Webb, P. B. Organometallics
2007, 26, 2782. (b) McGuinness, D. S.; Wasserscheid, P.; Keim, W.;
Morgan, D.; Dixon, J. T.; Bollmann, A.; Maumela, H.; Hess, F.; Englert,
U. J. Am. Chem. Soc. 2003, 125, 5272. (c) Overett, M. J.; Blann, K.;
Bollmann, A.; Dixon, J. T.; Hess, F. M.; Killian, E.; Maumela, H.;
Morgan, D. H.; Neveling, A.; Otto, S. Chem. Commun. 2005, 622.
(3) Blann, K.; Bollmann, A.; de Bod, H.; Dixon, J. T.; Killian, E.;
Nongodlwana, P.; Maumela, M. C.; Maumela, H.; McConnell, A. E.;
Morgan, D. H.; Overett, M. J.; Pretorius, M.; Kuhlmann, S.; Wasserscheid,
P. J. Catal. 2007, 249, 244.
(4) Kohn, R. D.; Haufe, M.; Kociok-Kohn, G.; Wasserscheid, P.;
Keim, W. Angew. Chem., Int. Ed. 2000, 39, 4337.
(5) (a) Crewdson, J. C.; Gambarotta, S.; Kokobkov, I.; Duchateau,
R. Organometallics 2006, 25, 715. (b) Agapie, T.; Labinger, J. A.; Bercaw,
J. E. J. Am. Chem. Soc. 2007, 129, 14281. (c) Schofer, S. J.; Day, M. W.;
Henling, M.; Labinger, J. A.; Bercaw, J. E. Organometallics 2006,
25, 2743. (d) Deckers, P. J. W.; Hessen, B.; Teuben, J. H. Organometallics
2002, 21, 5122.
(6) Kundig, E. P.; Pache, S. H. In Organometallics: Compounds of
Group 7-3; Imamoto, T., Ed.; Thieme: Stuttgart, Germany, 2003; Vol. 2,
pp 155À228.
(7) Bruckner, A.; Jabor, J. K.; McConnell, A. E. C.; Webb, P. B.
Organometallics 2008, 27, 3849.
(8) (a) Rieger, A. L.; Rieger, P. H. Organometallics 2002, 21, 5868.
(b) Li, T. T.; Kung, W.; Ward, D. L.; McCulloch, B.; Brubaker, C. H., Jr.
Organometallics 1982, 1, 1229. (c) Prins, R.; Reinders, F. J. Chem. Phys.
Lett. 1969, 3, 45.
(9) (a) Song, L.-C.; Yu, G.-A.; Hu, Q.-M.; Che, C.-M.; Zhu, N.;
Huang, J.-S. J. Organomet. Chem. 2006, 691, 787. (b) Nesmeyanov,
A. N.; Yur’eva, L. P.; Zaitseva, N. N.; Domrachev, G. A.; Zinov’ev, V. D.;
Shevelev, Yu. A.; Vasyukova, N. I. J. Organomet. Chem. 1976, 121, C52.
(c) Elschenbroich, C.; Heck, J. Angew. Chem. 1977, 16, 479. (d)
Elschenbroich, C.; Kroker, J.; Massa, W.; Wunsch, M.; Ashe, A. J. Angew.
Chem. 1986, 25, 571. (e) Elschenbroich, C.; Heikenfeld, G.; Wunsch,
M.; Massa, W.; Baum, G. Angew. Chem. 1988, 27, 414. (f) Elschenbroich,
C.; Sebbach, J.; Metz, B.; Heikenfeld, G. J. Organomet. Chem. 1992,
426, 173.
(10) McDyre, L. E.; Hamilton, T.; Murphy, D. M.; Cavell, K. J.;
Gabrielli, W. F.; Hanton, M. J.; Smith, D. M. Dalton Trans. 2010,
39, 7792.
(11) Upon activation ca. 80% of the original [Cr(CO)4(1)]+ para-
magnetic EPR signal is lost. However, there was no evidence of any
Cr(III) ions in the EPR spectra.
4508
dx.doi.org/10.1021/om2006062 |Organometallics 2011, 30, 4505–4508