A molecular mechanics approach to mapping the conformational space of diaryl and triarylphosphines

Article abstract of DOI:10.1039/b610123b

A molecular mechanics force field has been developed which accurately reproduces experimental solid state structures and conformer interconversion barriers for a series of sterically congested diaryl and triaryl phosphines and some of their chalcogenide and Cr(CO)₅ derivatives. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b610123b

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
A molecular mechanics approach to mapping the conformational space of
diaryl and triarylphosphines†
Natalie Fey,a James A. S. Howell,*b Jonathan D. Lovatt,b Paul C. Yates,b Desmond Cunningham,c
Patrick McArdle,c Hugo E. Gottliebd and Simon J. Colese
Received 14th July 2006, Accepted 28th September 2006
First published as an Advance Article on the web 13th October 2006
DOI: 10.1039/b610123b
A molecular mechanics force field has been developed which accurately reproduces experimental solid
state structures and conformer interconversion barriers for a series of sterically congested diaryl and
triaryl phosphines and some of their chalcogenide and Cr(CO)5 derivatives.
to correlate the steric and electronic properties of phosphine
Introduction
ligands.10a–c
Phosphines and phosphites are ubiquitous ligands in the homo-
We report here the testing of an adapted MM force field for
geneous catalysis of organic reactions using transition metal com-
triaryl and alkyldiaryl phosphines and their chalcogenide and
plexes, primarily because structural variation allows not only a fine
Cr(CO)5 derivatives which can be incorporated into standard
tuning of activity and selectivity (including enantioselectivity)1a–g
computational packages as an aid in the design of phosphines
but also because it can be used to alter catalyst properties such that
and phosphine complexes for use in catalysed organic synthesis.
they may be used in water,2a–d fluorous solvents,3a,b ionic liquids4a–c
or supercritical CO2
The force field has been tested against experimental structural
5a,b
or bound to solid supports.6a–c Promi-
and dynamic data obtained by X-ray crystallography and variable
temperature NMR spectroscopy for a series of sterically congested
2-RC6H4 substituted derivatives. As previous calculations11a,b on
PPh3 and HCPh3 indicate an almost flat energy surface in the
torsion angle range 20 to 50◦ (A–B–Cipso–Ca; A = lone pair,
B = P; A = H, B = C), we have chosen to use experimental
data from more sterically congested ortho-substituted derivatives
where conformational energy minima are more clearly defined and
both conformational equilibrium constants and interconversion
barriers are measurable by variable temperature NMR techniques.
nent among these ligands are triarylphosphines and chelating
diphosphines terminated by a diarylphosphine motif. Chelating
phosphines can be particularly effective in promoting asymmetric
synthesis, and the interaction of the substrate with the array of
phenyl rings presented by the catalyst is an important if not
determining factor in enantioselection.
Although thorough quantum mechanics‡ calculations of ‘real’,
conformationally flexible metal–phosphine complexes are still
not realistically possible in terms of computational time, many
recent studies have used quantum mechanics/molecular mechan-
ics (QM/MM) methods in which QM and MM are applied
either sequentially7 or (more often) simultaneously.8a,b The MM
component has been applied using a variety of common MM
force fields which have not been ideally parameterised to deal
with arylphosphines, though a parameterisation for alkyl phos-
phines has recently been reported.9 Other recent approaches have
identified various metal–ligand descriptors which may be used
Results and discussion
Force field parameterisation
Equilibrium values for bond lengths and angles involving the
phosphorus atom were derived from a survey of 55 relevant crystal
structures retrieved from the Cambridge Crystallographic Data
Base.12 A slight underestimation of ideal bond lengths and angles
proved more effective than the use of excessive force constants
in controlling the reproduction of observed geometries. The
lone pair–phosphorus distance was derived from an electron de-
formation density refinement of bis(diphenyphosphino)butane.13
The corresponding force constant was chosen to reproduce this
distance. Stretching force constants for XP–PX terms (X = O, S,
aSchool of Chemistry, Cantock’s Close, Clifton, Bristol, UK BS8 1TS.
E-mail: natalie.fey@bristol.ac.uk; Fax: +44 (0)117 9251295; Tel: +44
(0)117 9546398
bSchool of Physical and Geographical Sciences, Lennard Jones Lab-
oratories, Keele University, Keele, Staffordshire, UK ST5 5BG.
E-mail: j.a.s.howell@keele.ac.uk; Fax: +44 (0)1782 712378; Tel: +44
(0)1782 583041
cChemistry Department, National University of Ireland, Galway, Uni-
versity Road, Galway, Eire. E-mail: des.cunningham@nuigalway.ie;
p.mcardle@ucg.ie.; Fax: +353 91 525700; Tel: +353 91 492460
=
Se) were derived by extrapolation from the existing C O term,
taking experimentally determined stretching force constants into
account.14 Unique labelling as described by Brown et al.15a,b was
implemented for the Cr(CO)5 complexes. Brown’s parameters
for Cr–C bonds were adjusted until good agreement with the
dChemistry Department, Bar Ilan University, Ramat Gan, 52900, Israel.
E-mail: gottlieb@mail.biu.ac.il; Fax: +972 3 5351250; Tel: +972 3 5318828
eSchool of Chemistry, University of Southampton, Highfield, Southampton,
UK SO17 1BJ. E-mail: sjc5@soton.ac.uk; Fax: +44 (0)23 80593781;
Tel: +44 (0)23 80596722
16
structure of (2-MeC6H4)3PCr(CO)5 was achieved. Any torsion
terms involving the chromium atom in Cr(CO)5 complexes have
been set to zero.
Phenylphosphine and its oxide and sulfide were used for
parameterisation of the XP–PX–C3–C3 term. Diphenylphosphine
† Electronic supplementary information (ESI) available: NMR spectral
data; force field parameterisation; variable temperature NMR lineshape
analysis. See DOI: 10.1039/b610123b
‡ QM here is taken to include density functional theory calculations.
5464 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
and its oxide and sulfide were used for parameterisation of the
C3–PX–C3–C3 term. The required torsion to one ring was driven
whilst the corresponding torsion to the other ring was allowed to
optimise. Methylphosphine and its oxide and sulfide were used
for parameterisation of the H–C4–PX–XP term. Butphosphine
and its oxide and sulfide were used for parameterisation of the
C4–C4–PX–XP term. Methylphenylphosphine and its oxide and
sulfide were used for parameterisation of the C3–C3–PX–C4 term.
Torsional parameters were classified as either phosphorus
dependent or independent. Independent parameters are unlikely
to be influenced significantly by the phosphorus substituents and
should be fully transferable between different substituents. In con-
trast, dependent torsional parameters are potentially influenced
by the extent of p-conjugation and are thus not transferable
between different phosphine derivatives. For triarylphosphines,
the torsions XP–PX–C3–C3 and C3–PX–C3-C3 were regarded as
phosphorus dependent. Additionally for diarylphosphines and the
corresponding oxides and sulfides the C4–PX–C3–C3, H–C4–PX–
XP and C4–C4–PX–XP parameters were regarded as phosphorus
dependent. Since these terms are potentially interdependent, they
were parameterised using a literature procedure.17a,b Using model
compounds, an ab initio Hartree–Fock (HF) rotational energy
profile was generated, whilst the same profile was produced at
the MM level with the torsional constant of interest set to zero. The
torsional energy potential function was then fitted to the difference
plot to determine the magnitude of the torsional parameters.
The MM torsional potential function Et = 0.5 V1 [1 + cos(φ)] +
0.5 V2 [1 − cos(2φ)] + 0.5 V3 [1 + cos(3φ)] was fitted analytically
to the HF-MM difference to determine the magnitude of V1, V2
and V3.
and Cr(CO)5 derivatives were prepared by established methods
(see Experimental section).
The structure and stereodynamics of tris(o-alkyl)phosphines and
their derivatives
Triarylphosphines often adopt a chiral, helical conformation.
Differing ring substituents may introduce central chirality at
the phosphorus atom, and molecules may also differ in the
arrangement of ring substituents with respect to a reference plane
defined by the three ipso carbon atoms.
Correlated phosphorus-ring rotation has previously been ob-
served for the 2-Me and 2-CF3 derivatives19,20 and is a general
feature of molecular propellers of the types Ar3Z and Ar3ZX.21a,b
Crystal structures of (2-MeC6H4)3PX compounds22 show either
an exo3 or an exo2 conformation of the aryl rings (see Scheme 1
for conformer nomenclature). On the assumption that only these
two conformers are populated, P–C rotational isomerism for
an Ar3PX compound via 1-, 2- and 3-ring flip mechanisms
may be represented as in Scheme 1. The 1-, 2- and 3-ring
flip intermediates of idealised structures I to V are required
for conformer interconversion along edges and diagonals. Four
conformers [A,B] and their helical enantiomeric equivalents [Aꢀ,
Bꢀ] are possible.
Using the adapted force field, a complete MM analysis for
(2-MeC6H4)3PO 1b was performed using a systematic driver
calculation to fully explore conformational space with respect
to the three ring torsions. The methyl substituents adjust easily
to differing ring conformations, and were not analysed as an
additional variable. The resulting conformational energy surface
is thus four-dimensional, combining three ring torsions with
the calculated steric energy. The minimum energy in each data
series may be identified, thus permitting correlation with the
corresponding angle values to generate a map of energy minima
suitable for the identification of conformers and lowest energy
interconversion pathways. Fig. 1 displays two contour maps
correlating two of these torsional angles with energy in the most
important regions of the energy surface.
These maps show energy minima for the exo3, exo2 and exo1
conformers. The map identifies the transition state III to lie along
the lowest energy path of exo2/exo3 exchange, a process which
is accompanied by ring exchange and a reversal of helicity. The
transition state IV lies along the path which represents helix
inversion in the exo3 isomer. Both 1-ring flip pathways which
invert helicity and exchange rings in the exo2 conformer may be
identified, with that occurring via transition state II representing
the lower energy alternative. However, a sequence of 2- and 3-ring
flips appears to be the most likely process for exo2 helix inversion
on this conformational energy surface. The exo1 conformer may
also be identified as a local energy minimum, whereas the exo0
conformer is higher in energy and obscured by part of the exo1
surface. The geometries of calculated transition states are close to
those of the idealised structures.
To permit a detailed sampling of conformational space at
reasonable computational cost, geometries were first optimised
semi-empirically (MNDO/d)in allcases, followed byasingle point
calculation at the HF level (HF/STO-3G*). Calculation of relative
ground state and transition state energies of some of the simpler
compounds by DFT (B3LYP) using a higher precision basis set (6-
311G**) gave virtually identical values (see ESI†). As MNDO/d is
not parameterised for selenium compounds (in HYPERCHEM),
the selenium parameters were obtained by scaling the equivalent
sulfur values according to the difference in atomic radius between
sulfur and selenium.
A list of all parameters which have been modified or added to
the MM+ force field in HYPERCHEM18 are given in the ESI.†
Triaryl (2-RC6H4)3PX derivatives
Members of the 2-MeC6H4 and 2-CF3C6H4 triaryl series 1a–e and
2a,b have previously been characterised both in the solid state and
in solution (throughout, a denotes phosphine, b, c and d denote
phosphine oxide, sulfide and selenide respectively and e denotes
the Cr(CO)5 complex).19,20 (2-EtC6H4)3P 3a and (2-iPrC6H4)3P
4a were prepared by reaction of PCl3 with 2-EtC6H4Li and 2-
iPrC6H4MgBr respectively. Only a single 2-tBuC6H4 substituent
could be introduced through the reaction of (2-MeC6H4)2PCl
with 2-tBuC6H4MgI to give (2-MeC6H4)2(2-tBuC6H4)P 5a. The
As the detailed analysis of 1b accurately reproduced the ground
state solid state structure and the conformer distribution and P–
C rotational barriers in solution, a less intensive computational
approach was adopted for other derivatives in which ground state
conformations were explored by 1000 iteration conformational
t
diaryl derivatives (2-MeC6H4)2MeP 6a and (2-MeC6H4)2 BuP 7a
were prepared by reaction of (2-MeC6H4)2PCl with MeLi and
tBuPCl2 with 2-MeC6H4Li respectively. Oxide, sulfide, selenide
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5465
©

View Online
Scheme 1 Ring flip mechanism for triarylphosphine.
searches and ring flip barriers were calculated using the geometries
of the idealised transition states (shown in Scheme 1).
The crystal structures of 1a, 2a13,14 and 3a to 5a (Fig. 2) also
show exo3 conformations of the aryl rings. In 5a, the 2-tBuC6H4
ring is much closer to collinearity with the P-lone pair axis, whilst
the remaining 2-MeC6H4 rings are flattened to some degree to
compensate. As with 1a and 2a, 31P NMR spectra of 3a to 5a
are temperature invariant down to 163 K, indicating an exclusive
5466 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
resonances of 5a; these become diastereoisotopic in the presence
of helix inversion which is slow on the NMR timescale.
Except for 5b and 1e, experimental barriers and conformer pop-
ulations for the exo2/exo3 exchange process for the oxides, sulfides
and selenides may be obtained from the variable temperature 31
P
NMR spectra which show two resonances at low temperature
assignable to the exo3 and exo2 conformers.¶ Barriers for helix
inversion may be obtained from line shape analyses of the higher
temperature 1H spectra where possible. The 31P spectra of 5b and 1e
are temperature invariant, implying a single conformer in solution.
The crystal structure of 5b (Fig. 3) shows that this may be described
as exo3, albeit highly distorted in terms of the ring torsion angles,
whereas that of 1e (Fig. 3) is clearly exo2.
Agreement between calculated and observed P-ring torsional
angles in the solid state is excellent (Table 1). Given the substantial
variation of conformational equilibrium constant with tempera-
ture for many of the compounds, and given the usual difficulties
of comparing calculated relative stabilities in the gas phase at
zero Kelvin with solution values at non-zero temperatures, the
agreement between calculated and experimental thermodynamic
parameters is good (Table 1). The standard deviation of the relative
percent difference between calculated and experimental rotational
barriers for the eleven barriers shown in Table 1 ( 10%) is the same
as the approximate precision of the measurement of the enthalpy
of activation from the NMR spectra.
The exo3 conformation becomes increasingly less favourable
relative to exo2 as the size of the phosphorus substituent X is
increased; exo2 is calculated to be the most stable conformer for
X = Se, Cr(CO)5, though at no stage do exo1 or exo0 become
the global minimum. The idealised 2-ring flip transition state
III represents the lowest energy exchange process for exo2/exo3
exchange. This transition state, and that for 3-ring flip helix
inversion in the exo3 conformer (IV), have two or three rings
respectively which are exo and perpendicular to the reference
plane; the energies of these transition states increase as the size of
the X substituent increases. In contrast, the energies of the 1-ring
flip transition states, which contain only a single perpendicular exo
ring, are calculated to remain constant or decrease slightly from
the oxide to the selenide. For 1e, there is also good agreement
regarding the orientation of the phosphine relative to the Cr(CO)4
plane (Fig. 2), and the calculated magnitude of the Cr–P rotational
barrier (26 kJ mol−1) is close to that observed experimentally
from line broadening of the CO resonance (≤32 kJ mol−1). The
lowest energy process for ring exchange for 1e is calculated to be a
sequential series of 1-ring flips which exchanges rings in the exo2
conformer without conformer interconversion.
Fig. 1 3-Dimensional energy graphs for 1b.
Ring-substituent rotational profiles may also be investigated
by MM calculation (Table 2). For example, the energy profile for
ring-tBu rotation in the exo3 isomer of 5b shows a 3-fold symmetry
Fig. 2 Solid state structures of 3a, 4a and 5a (hydrogens omitted).
=
with three equivalent global minima in which the P O bond
bisects the angle between two methyl groups. This is observed in
the solid state structures of 5a,b (Fig. 2, 3). Probably as a result of
steric destabilisation of the ground state, the predicted rotational
barriers for 5a and 5b are surprisingly low (9 and 14 kJ mol−1), and
population of the exo3 conformer.§ The higher energy process
which results in helix inversion is NMR silent for 1a and 2a, but
may be observed and the barrier quantified through line shape
analysis in the variable temperature behaviour of the 1H CH2CH3
resonances of 3a, the CHMe2 resonances of 4a and the o-Me
¶ The major conformer may be assigned on the basis of the multiplicity
of the resonance for the major conformer observed in the 1H or 19F
subspectrum for the o-Me or o-CF3 substituent (one for exo3, three for
exo2).
§ The characteristic11a high field resonance of H6 (6.7–6.8 ppm) is consistent
only with exo3.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5467
©

View Online
Fig. 3 Solid state structures of 1b–e and 5b (hydrogens omitted).
in agreement, no broadening of the tBu resonance is observed in
the 1H NMR of either 5a, b down to 163 K. As predicted, the solid
state structures of the ethyl and isopropyl compounds 3a and 4a
place the methyl substituents pointing away from the P-lone pair
axis.
The structure and stereodynamics of alkyldiarylphosphines and
their derivatives
Conformational isomerism in ring substituted alkyldiarylphos-
phines Ar2PR and their derivatives may be analysed in the same
topological manner (Scheme 2). Thus, where Ar = 2-MeC6H4
6a–e, the four possible conformers exo2, exo1a,b and exo0 (and
their corresponding enantiomers of opposite helicity) may be
identified. For alkyldiaryl derivatives, systematic variation of the
two ring torsion angles permits generation of a three-dimensional
energy surface, thus facilitating a direct analysis of conformational
preferences and interconversion mechanisms. This is fortunate,
since for these sterically more flexible compounds, calculations
of rotational barriers using idealised transition states do not
accurately reproduce experimental data. Ground state confor-
mational preferences were explored in detail by 1000 iteration
conformational searches, allowing free adjustment of the Me or
tBu group. For the P–Me series 6a–e, agreement between predicted
and observed solid state structures for 6b, c, e is also good (Table 3,
Fig. 4). Three-dimensional plots are shown in Fig. 5 and 6 for the
phosphines and for their Cr(CO)5 complexes, whilst key calculated
structural, kinetic and thermodynamic data are summarized in
Table 3.
Fig. 4 Solid state structures of 6b, c, e and 7b.
plane is also excellent. Only the Cr(CO)5 derivative 6e exhibits
1
low temperature limiting H NMR spectra. This may be taken
as indicating restricted ring exchange via slowing of P–C rotation
in exclusively the exo1a isomer, since 31P spectra are independent
of temperature over the same range. The calculated barrier for
direct helix exchange in the exo1a isomer is much higher than
that observed experimentally, and a more likely scenario involves
interconversion of exo1a with exo2 accompanied by helix inversion
in the exo2 conformer.
For the more sterically demanding tBu series 7a–e, low temper-
ature limiting spectra can be obtained for all but the phosphine
itself 7a. For 7b–e, 31P spectra are resolved into two resonances
assignable to exo2 and exo1a with the relative amount of exo2
decreasing across the series. This is in agreement with the
calculation of relative stability for both the tBu and Me series. As
intuitively expected on steric grounds, the exo2 isomer is stabilised
Though the lowest energy structure for 6b is calculated to
be exo2, the compound crystallises as the exo1 conformer. The
calculated energy difference, however, is sufficiently small to be due
to crystal packing forces. For 6e, agreement between the predicted
and observed orientation of the ligand relative to the Cr(CO)4
5468 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
Table 1 Comparison of calculated and experimental structural, thermodynamic and kinetic data for triarylphosphines and their derivatives
a
b
c
d
e
1
ꢀ
a
39 3 (43 3)
exo3
0 (exo3 only)
9.0
24.6
42.3
122.1
51.3
11.7
40 3 (46 4)
exo3
0 (0)
6.4 (2.4, 163 K)
19.8
33.5
128.4
53.6
19.1 (30)
−161, 51, 56 (−163, 57, 57) −164, 54, 56 (−168, 60, 54) −180, 73, 56 (−179, 70, 60)
exo2
exo2
exo2
b
exo3
exo2
exo1
exo0
0 (4.3, 173 K)
1.0 (0)
8.9
0.2 (4.6, 163 K)
0 (0)
7.0
17.5
25.1
0 (exo2 only)
0.8
12.1
19.1
1-ringa (I)
1-ringb (II)
2-ring (I)
106.4
44.6
33.7 (38)
103.1
41.9
38.7 (40)
104.1
44.3 (39)
90.8
3ringa (III)
29.6
38.7
80.0
82.7
112.5
98.4
123.1
220.0
201.7
3-ringb (IV) 74.3
2
ꢀ
a
39 1(35 2)
exo3
0 (exo3 only)
9.0
24.6
42.3
122.1
37 1 (42 7)
exo3
0 (0)
6.4 (4.0, 163 K)
19.8
33.5
128.4
53.6
19.1 (34)
b
exo3
exo2
exo1
exo0
1-ringa (II)
1-ringb (III) 51.3
2-ring (I) 11.7
3-ringa (IV) 29.6
38.7
80.0
3-ringb (V)
74.3
3
ꢀ
a
39 0 (41 6)
exo3
0 (exo3 only)
11.0
29.6
45, 41, 36
b
exo3
0 (1.4, 173 K)
4.8 (0)
23.5
40.6
151.3
69.8
27.2 (33/36)
49.0
90.1
exo2
exo1
exo0
57.2
133.7
1-ringa (II)
1-ringb (III) 70.1
2-ring (I)
19.0
3-ringa (IV) 33.9 (46)
3-ringb (V)
81.2
4
ꢀ
a
38 16 (43 3, 40 9)c 47
exo3
0 (exo3 only)
7.2
29.5
51.0
143.7
2
−173, 58, 59
exo2
−174, 59, 60
exo3
exo2
b
exo3
0 (3.7, 173 K)
2.1 (0)
25.0
37.1
194.8
4.1
exo2
0 (exo2 only)
16.1
26.0
134.7
58.1 (55)
57.9 (50)
133.3
exo1
exo0
1-ringa (II)
1-ringb (III) 58.8
61.1
2-ring (I)
15.8
32.3 (39)
65.4 (53)
103.8
3-ringa (IV) 34.1 (47)
3-ringb (V)
84.2
155.7
5
ꢀ
a
23, 47, 47 (38, 35, 48)
exo3
0 (exo3 only)
14.2
16.3
48.9
29.5
57.1
69.6
38, 35, 48 (39, 39, 54)
exo3
0 (exo3 only)
10.5
9.5
33.6
24.4
42.3
52.7
exo3
exo2a
exo2b
exo2c
exo1a
exo1b
exo1c
exo0
97.4
70.5
1-ringa
1-ringa
1-ringb
1-ringb
157.2
121.4
98.4
157.1
129.6
89.1
ꢀ
ꢀ
63.1
65.8
2-ringa
2-ringb
2-ringc
3-ringa
3-ringb
39.6
21.6
21.1
47.8 (54)
142.5
86.9
38.2
29.7
30.9
57.4 (73)
146.8
94.6
ꢀ
3-ringb
a P-ring torsional angles (◦); observed solid state values are in parentheses. b Relative energies (kJ mol−1); experimental values from variable temperature
NMR are given in parentheses. c Two molecules in unit cell.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5469
©

View Online
Table 2 Calculated and observed structural and kinetic data for ring-substituent orientation
◦
av. |ssubst|a/
Barrier to substituent rotation/kJ mol−1
b
b
exo3
exo2a
exo2b
exo2c
exo3
exo2a
exo2b
exo2c
3ac
3bc
4ac
4bc
4cc
5a
84 (121)
97
exo2: 100
exo2: 103
exo2: 12
exo2: 10
8
23
32
32
43
54
9
exo: 26,d endo: 56
exo: 34,d endo: 56
exo: 40, d endo: 67
exo: 47,d endo: 66
exo: 56,d endo: 68
7
8 (5, 4)e
10
9
179 (177)
166 (168)
179
166
176
163
164
164
9
14
24
24
5b
14
9
a Ring-substituent torsional angles. Observed solid state values are given in parentheses. b Substituent endo. c exo2 conformations indistinguishable. d Lower
barrier. e Two molecules in unit cell.
Scheme 2 Ring flip mechanism for alkyldiarylphosphine.
t
relative to exo1a more in the Bu series than in the methyl series.
Thus, whereas 6b has an exo1a solid state structure, that of 7b is
exo2 (Fig. 4).
In contrast, barriers for P–tBu rotation in 7b–d are substantially
higher, and calculated barriers are in good agreement with those
measured by variable temperature NMR. The crystal structure
All the calculated barriers for P–CH3 rotation in the lowest
energy conformers of 6a–d lie below the limit of NMR measure-
ment, and no decoalescence of the 1H P–Me resonance is observed.
of the oxide 7b (Fig. 4) confirms the predicted lowest energy
t
=
conformation of the Bu substituent in which the P O bond bisects
a pair of methyl groups.
5470 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
Fig. 5 3-Dimensional energy graphs for 6a (left), e (right).
NMR spectra were performed by computer simulation. For the
case of two exchanging nuclei, an in-house program based on the
equation of Sutherland23 was used. For greater numbers of nuclei,
a program written by R. E. D. McClung (University of Alberta,
Canada) was adapted for use. Full data obtained from the line
shape analyses can be found in the ESI.†
Conclusions
A force field has been developed which can accurately reproduce
relative conformer stabilities and rotational barriers for triaryl-
and diarylalkyl phosphines and their chalcogenide and Cr(CO)5
derivatives. Further work is in progress on the application of this
force field to the modelling of bidentate phosphines and their
chalcogenide and Cr(CO)4 complexes.
Syntheses
1. Synthesis of 2-iPrC6H4Br24. Aqueous HBr (65 ml, 48%)
was added rapidly to a mechanically stirred refluxing solution of 2-
iPrC6H4NH2 (17.2 g, 128 mmol) in water (160 ml). After refluxing
for 30 minutes, the solution was cooled to 0 ◦C and a solution of
NaNO2 (8.8 g, 127 mmol) in water (45 ml) was added, maintaining
the temperature between 0 and 5 ◦C. The cold solution of the
diazonium salt was added rapidly to a stirred solution of CuBr
(21.0 g, 147 mmol), aqueous HBr (43 ml, 48%) and water (120 ml).
After stirring at room temperature for 20 minutes, the mixture was
heated on a steam bath for 1 hour and then steam distilled. The
distillate was extracted with diethyl ether (300 ml), washed with
Experimental
Reactions involving lithium and magnesium reagents were con-
ducted under argon. Except for the preparation of phosphine
oxides, reactions were conducted under dinitrogen. NMR spec-
tra were recorded using a Bruker DPX300 spectrometer (1H,
300 MHz; 13C, 75 MHz; 31P, 121 MHz). Full NMR data for all
compounds are provided in the ESI.†
Temperatures were measured using the in-built copper–
constantan thermocouple which had previously been calibrated
with a platinum resistance thermometer. Line shape analyses of
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5471
©

View Online
Fig. 6 3-Dimensional energy graphs for 7a (left), e (right).
water and dried over MgSO4. After removal of solvent, the residue
was purifie◦d by column chromatography [silica, petroleum ether
(bp 40–60 C)] to give the product (15.3 g, 60%) as a colourless
liquid from the fastest moving fraction.
27.0 mmol) to a solution of 2-EtC6H4Br (5.0 g, 27.0 mmol) in dry
diethyl ether (30 ml) at −5 ◦C. After stirring for 1 hour, PCl3 (1.13 g,
◦
8.19 mmol) in diethyl ether (15 ml) was added dropwise at 0 C
over 15 minutes. The solution was warmed to room temperature
and stirred for a further hour. After filtration and removal of
solvent, the residue was purified by column chromatography
[silica, petroleum ether (bp 60–80 ◦C)] and recrystallised from
petroleum ether (bp 60–80 ◦C) to give 3a as white crystals (1.7 g,
57%); mp 98–99 ◦C (Found: C, 83.2; H, 7.79%. C24H27P requires
C, 83.2; H, 7.80%).
2. Synthesis of 2-tBuC6H4I25. 2-tBuC6H4NH2 (22 g,
147 mmol) was added to 3.4 M H2SO4 (20 ml) and cooled
to −10 ◦C. A saturated aqueous solution of NaNO2 (10.5 g,
152 mmol) was added with vigorous stirring over 5 minutes to
give a light brown slurry. After stirring for a further hour at −10
to 0 ◦C, the slurry of the diazonium salt was added rapidly to
a concentrated ice-cold solution of KI (75 g, 0.45 mol). After
stirring at 0 ◦C for 2 hours, the suspension was extracted with
diethyl ether (150 ml). After removal of solvent, purification by
column chromatography [silica, petroleum ether (bp 40–60 ◦C)]
gave the product (22.8 g, 60%) as a colourless liquid from the
fastest moving fraction.
4. Synthesis of (2-iPrC6H4)3P (4a).
A solution of 2-
iPrC6H4MgBr was prepared by addition of 2-iPrC6H4Br (5.0 g,
22.2 mmol) to magnesium turnings (0.54 g, 22.3 mmol) in
tetrahydrofuran (10 ml) using 1,2-dibromoethane as initiator. The
rate of reaction was controlled by the addition of a further 40 ml
of tetrahydrofuran. After refluxing for 2 hours and cooling to
0
◦C, PCl3 (0.78 g, 5.69 mmol) in tetrahydrofuran (10 ml) was
3. Synthesis of (2-EtC6H4)3P (3a). 2-EtC6H4Li was prepared
added dropwise. After stirring at room temperature for 2 hours,
by addition of nBuLi (16.9 ml of 1.6 M solution in hexane,
the solution was heated to 50 ◦C for a further 2 hours. After
5472 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
Table 3 Calculated and experimental structural, thermodynamic and kinetic data for diarylalkylphosphines and their derivatives
a
b
c
d
e
6
ꢀ
a
60, 13
exo2
0
62, −162 (54, −174)
67, −163 (66, −174)
70, −164
exo1a
2.9
0
4.6
21.1
16.2
99.1
35.2
12.3
18.7
22.3
12.9
81, −163 (81, −152)
exo1a
exo1a
1.4
exo1a
b
exo2
0
4.8
8.7
30.5
6.5
133.0
43.1
15.8
15.6
0 (exo1a only)
26.1
12.8
49.7 (47)
125.8
66.7
37.7
35.6
exo1a
6.5
0
4.2
exo1b
33.9
33.9
5.8
129.3
44.3
18.1
18.4
20.4
11.3
exo0
22.3
16.9
99.7
36.9
12.0
18.0
22.6
13.0
exo2 (helix inversion)
exo1a (helix inversion)
exo1b (helix inversion)
exo2/exo1a
exo2/exo1b
exo1a/exo1b
CH3 rotational barrier
18.9
21.9
41.8
9.7
14.9 (exo2), 11.7 (exo1a)
7
ꢀ
a
22, 15
exo2
0
13.0
41.6
60.4
5.3
131.9
132.4
20.6
13.3
35, 16 (51, 18)
exo2
0 (0)
8.7 (2.6, 173 K)
40.2
56.2
10.9
134.0
212.2
18.1 (33)
75, 22
exo2
0 (2.9, 173 K)
3.2 (0)
45.8
54.1
26.5 (55)
97.6
97.5
77, 23
exo2
0 (4.3, 203 K)
2.2 (0)
46.4
53.5
36.0 (64)
95.8
91.9
81, −159
exo1a
b
exo2
7.2 (3.0, 313 K)
exo1a
0 (0)
68.9
56.6
105.0
111.1
188.4
exo1b
exo0
exo2 (helix inversion)
exo1a (helix inversion)
exo1b (helix inversion)
exo2/exo1a
19.8 (41)
46.3 (42)
21.7 (42)
46.9 (42)
54.0 (68)
29.8 (35, exo1a; 38, exo2)
tBu rotational barrier
21.2 (36)
a P-ring torsional angle (◦); observed solid state values are in parentheses. b Relative energies (kJ mol−1); experimental values measured by variable
temperature NMR are in parentheses.
cooling, water (2 ml) was added carefully. Removal of solvent
and recrystallisation from ethanol gave 4a as white crystals (1.3 g,
59%); mp 108–110 ◦C (Found: C, 83.5; H, 8.50%. C27H33P requires
C, 83.5; H, 8.57%).
stirred solution of 2-MeC6H4Br at −50 ◦C. After warming to
room temperature, the ◦solution was stirred for a further 3 hours.
t
After cooling to −10 C, BuPCl2 (7.0 g, 44 mmol) in diethyl
ether (50 ml) was added over 30 minutes. The solution was stirred
overnight and finally refluxed for 30 minutes. After addition of
water (2 ml) and removal of solvent, the residue was purified by
column chromatography [silica, petroleum ether (bp 40–60 ◦C)]
and recrystallised from methanol to give 7a as white crystals (9.8 g,
83%); mp 86–88 ◦C (Found: C, 79.8; H, 8.87%. C18H23P requires
C, 80.0; H, 8.59%).
5. Synthesis of (2-MeC6H4)2(2-tBuC6H4)P (5a). 2-tBuC6H4I
(2.26 g, 8.67 mmol) in tetrahydrofuran (40 ml) was added to
magnesium turnings (0.22 g, 9.0 mmol). After stirring at reflux for
7 hours, only a small amount of magnesium remained unreacted.
After cooling to 0 ◦C, (2-MeC6H4)2PCl (2.10 g, 8.67 mmol)
in tetrahydrofuran (30 ml) was added slowly over 30 minutes.
The reaction was stirred overnight at room temperature and
water (1 ml) was added carefully. After removal of solvent, the
residue was purified by column chromatography [silica, 9 : 1
petroleum ether (bp 40–60 ◦C)–dichloromethane] followed by
recrystallisation from petroleum ether (bp 60–80 ◦C) to give 5a
8. Synthesis of (2-MeC6H4)2MePO (6b). A solution of m-
chloroperbenzoic acid (0.3 g, 1.97 mmol) in toluene (15 ml) was
added dropwise to a stirred solution of (2-MeC6H4)2MeP (0.3 g,
1.3 mmol) in toluene (10 ml). After stirring for 12 hours, the
solution was washed with 10% Na2CO3 solution (30 ml). After
drying with MgSO4 and removal of solvent, the residue was
recrystallised from petroleum ether (bp 60–80 ◦C) to give 6b as
a white solid (0.3 g, 97%). Other phosphine oxides were prepared
in a similar manner.
◦
as white crystals (0.56 g, 17%); mp 150–153 C (Found: C, 83.3;
H, 7.87%. C24H27P requires C, 83.2; H, 7.80%).
6. Synthesis of (2-MeC6H4)2MeP (6a). MeLi (26.5 ml of a
1.4 M solution in hexane, 37.1 mmol) was added to a stirred
solution of (2-MeC6H4)2PCl (9.12 g, 36.7 mmol) in diethyl ether
(50 ml) at −40 ◦C. After stirring for 1 hour, the solution was
warmed to room temperature and stirred overnight. After addition
of saturated NH4Cl solution (5 ml), the product was extracted with
diethyl ether (60 ml). After removal of solvent, the crude product
was purified by vacuum distillation (100 ◦C, 0.01 mmHg) to give
6a as a white solid (3.9 g, 85%); mp 53–55 ◦C (Found: C, 79.3; H,
7.38%. C15H17P requires C, 79.0; H, 7.52%).
(2-MeC6H4)2MePO. Mp 126–127 ◦C (Found: C, 73.8; H,
7.20%. C15H17OP requires C, 73.9; H, 7.02%).
(2-MeC6H4)2 BuPO. Mp 134–136 ◦C (Found: C, 75.7; H,
t
8.27%. C18H23OP requires C, 75.5; H, 8.11%).
(2-EtC6H4)3PO. Mp 164–166 ◦C (Found: C, 79.1; H, 7.43%.
C24 H27OP requires C, 79.6; H, 7.46%).
◦
(2-iPrC6H4)3PO. Mp 198–202 C (Found: C, 80.4; H, 8.79%.
C27H33OP requires C, 80.3; H, 8.24%).
7. Synthesis of (2-MeC6H4)2 BuP (7a). nBuLi (59.9 ml of a
(2-MeC6H4)2(2-tBuC6H4)PO. Mp 186–188 ◦C (Found: C,
t
2.5 M solution in hexane, 150 mmol) was added dropwise to a
79.8; H, 7.62%. C24H27OP requires C, 79.6; H, 7.46%).
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5473
©

View Online
9. Synthesis of (2-MeC6H4)2MePS (6c). Sulfur (0.1 g,
3.2 mmol) and (2-MeC6H4)2MeP (0.3 g, 1.3 mmol) were refluxed
in CS2 (15 ml) for 4 hours. Removal of solvent, followed by
purification by preparative tlc (silica, CH2Cl2) and recrystallisation
from ethanol gave 6c as white crystals (0.2 g, 44%). Other
phosphine sulfides were prepared in a similar manner.
Crystallographic studies
Data were collected on either an Enraf-Nonius CAD4f diffrac-
tometer (3a, 4a, 5a, b and 6b, c, e) or an Enraf-Nonius Kappa
CCDD detector (7b). Structures were solved by direct methods26
and refined by full matrix least squares.27 The structure of 7b has
already been deposited in the Cambridge Crysytallographic Data
Base (reference BAKTIR). See Table 4 for crystallographic data.
CCDC reference numbers 614969–6144975 and 6208454.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b610123b
(2-MeC6H4)2MePS. Mp 146–148 ◦C (Found: C, 69.4; H,
7.01%. C15H17PS requires C, 69.2; H, 6.54%).
(2-MeC6H4)2 BuPS. Mp 137–138 ◦C (Found: C, 71.4; H,
t
7.79%. C18H23PS requires C, 71.5; H, 7.68%).
◦
(2-iPrC6H4)3PS. Mp 120–122 C (Found: C, 77.5; H, 7.99%.
C27H33PS requires C, 77.2; H, 7.92%).
Computational studies
10. Synthesis of (2-MeC6H4)2MePSe (6d). A solution of
(2-MeC6H4)2MeP (0.5 g, 2.19 mmol) in CHCl3 (10 ml) was
refluxed overnight with selenium powder (0.5 g, 6.64 mmol). The
suspension was filtered through Celite and the solvent removed.
Recrystallisation from ethanol gave white crystals of 6d (0.5 g,
74%). Other phosphine selenides were prepared in a similar
manner.
All calculations were performed in vacuo using the Polak-Ribiere
conjugate gradient algorithm for MM, semi-empirical and ab
initio calculations. Investigations of conformational space for tri-
arylphosphines combined several calculation approaches. Ground
state conformations were sampled using the conformational search
facility in ChemPlus. Conformations were generated by random
rotation of named torsional angles employing a usage directed
scheme which cycles through previously found conformations
in order of increasing energy. Structures were optimised to a
(2-MeC6H4)2MePSe. Mp 158–160 ◦C (Found: C, 58.5; H,
5.85%. C15H17PSe requires C, 58.7; H, 5.57%).
maximum of 750 or 1000 cycles depending on the structural
(2-MeC6H4)2 BuPSe. Mp 164–166 ◦C (Found: C, 61.8; H,
t
−1
complexity, with a convergence criterion of 0.075 kcal A mol−1.
˚
6.81%. C18H23PSe requires C, 61.9; H, 6.65%).
Post-optimisation, structures were rejected if they were in excess
of 20 kcal mol−1 higher in energy than the lowest energy accepted
structure or if two consecutive structures had torsional angles
within 5◦ or energies within 0.2 kcal mol−1 of each other. Ring
flip transition states were investigated by restraining ring torsional
angles to ideal values (combinations of 0, 90 and 180◦). High
11. Synthesis of (2-MeC6H4)2MePCr(CO)5 (6e). Dinitrogen
was bubbled through a stirred degassed solution of Cr(CO)6 (0.5 g,
2.19 mmol) and (2-MeC6H4)MeP (0.5 g, 2.19 mmol) in 1 : 1 THF–
heptane (200 ml). The solution was irradiated for one hour (90 W
mercury arc), filtered through Celite and evaporated to dryness.
The crude product was purified by preparative TLC [silica, 1 :
9 CHCl3–petroleum ether (bp 40–60 ◦C)] and recrystallised from
petroleum ether (bp 60–80 ◦C) to give 6e (0.6 g, 66%). 7e was
prepared in a similar manner.
force constants (500 kcal mol−1 degree−2) were applied and the
−1
geometry was optimised to a gradient of 0.075 kcal A mol−1 or
˚
a maximum of 750 cycles. The restraints were then removed and a
single point energy calculated. For 1b, all 26011 combinations of
ꢀ
ꢀ
ꢀ
the three torsional angles (
,
= −180 to +180◦,
= −180
1
3
2
(2-MeC6H4)2MePCr(CO)5. Mp 78–80 ◦C (Found: C, 57.4; H,
4.11%. C20H17CrO5P requires C, 57.1; H, 4.08%).
to 0◦, 10◦ steps) were generated using a script of commands and
geometries were optimised to a gradient of 0.1 kcal A mol−1 or
−1
˚
(2-MeC6H4)2 BuPCr(CO)5. Mp 130–32 ◦C (Found: C, 59.1;
a maximum of 1000 cycles, subject to restraints of 500 kcal mol−1
degree−2 on these angles. Restraints were then removed and single
t
H, 4.99%. C23H23CrO5P requires C, 59.7%; H, 5.02%).
Table 4 Crystallographic data
3a
4a
5a
5b
6b
6c
6e
7b
Formula
Habit
C24H27P
Monoclinic Triclinic
10.737(2)
12.291(2)
15.286(2)
C27H33P
C24H27P
C24H27OP
C15H17OP
C15H17PS
C20H17CrO5P C18H23OP
Monoclinic Monoclinic Monoclinic Monoclinic Monoclinic
Orthorhombic
7.524(2)
13.097(3)
˚
a/A
11.005(2)
12.990(3)
18.795(5)
75.89(2)
88.14(2)
69.87(2)
0.71069
P1
7.4490(10)
17.291(2)
15.649(2)
14.398(2)
8.274(2)
17.451(4)
8.2401(9)
10.7830(9)
15.5965(10) 15.790(4)
13.974(3)
6.7730(10)
10.763(1)
13.300(2)
14.269(1)
˚
b/A
˚
c/A
16.665(3)
a/◦
b/◦
c /◦
90.39(2)
99.290(10)
101.973(16) 102.280(1)
111.380(1)
96.62(1)
90
˚
k/A
0.71069
P21/n
0.139
0.71069
P21/c
0.141
0.71069
P21/a
0.145
0.71069
P21/c
0.185
0.71069
P21/n
0.323
0.71069
P21/n
0.669
0.71073
P212121
0.162
Space group
l/mm−1
0.121
Z
4
4
4
4
4
4
4
4
h Range/◦
2.13–24.98
3874
3549
2.12–21.97
6526
5976
2.36–24.98
5567
3495
2.39–31.99
5244
4853
2.31–27.96
3572
3249
2.44–29.96
4018
3742
2.10–29.97
6371
5899
1.98–27.38
10051
3602
Measured reflections
Independent reflections
R
Rw
0.0472
0.1393
0.0517
0.1568
0.0375
0.1077
0.0444
0.1189
0.0468
0.1289
0.0618
0.1811
0.0443
0.1342
0.0420
0.1093
5474 | Dalton Trans., 2006, 5464–5475
This journal is
The Royal Society of Chemistry 2006
©

View Online
point energies calculated. Due to the high symmetry of the energy
surface, the remaining data points could be generated as a mirror
image. Ring-substituent conformational space was explored using
4 For recent reviews, see: (a) T. Welton, Coord. Chem. Rev., 2004, 248,
2459; (b) J. Dupont, R. F. de Souza and P. A. Z. Suarez, Chem. Rev.,
2002, 102, 3667; (c) R. Sheldon, Chem. Commun., 2001, 2399.
5 For recent reviews, see: (a) W. Leitner, Acc. Chem. Res, 2002, 35, 746;
(b) E. J. Beckman, J. Supercrit. Fluids, 2004, 28, 121.
6 For recent reviews, see: (a) D. E. Bergbreiter, Top. Curr. Chem., 2004,
242, 113; (b) C. Muller, M. G. Nijkamp and D. Vogt, Eur. J. Inorg.
Chem., 2005, 4011; (c) N. End and K. U. Schoning, Top. Curr. Chem.,
2004, 242, 241.
7 A. M. Gillespie, G. R. Morello and D. P. White, Organometallics, 2002,
21, 3913 and references therein.
8 (a) G. Ujaque and F. Maseras, Struct. Bonding, 2004, 112, 117; (b) D.
Balcells, G. Drudis-Sole, M. Besora, N. Dolker, G. Ujaque, F. Maseras
and A. Lledos, Faraday Discuss., 2003, 124, 429.
9 P. M. Todebush, G. Liang and J. P. Owen, Chirality, 2002, 13, 220.
10 (a) N. Fey, A. C. Tsipsis, S. E. Harris, J. N. Harvey, A. G. Orpen and
R. A. Mansson, Chem.–Eur. J., 2006, 12, 291; (b) K. D. Cooney, T. R.
Cundari, N. W. Hoffman, K. A. Pittard, M. D. Temple and Y. Zhao,
J. Am. Chem. Soc., 2003, 125, 4318; (c) H. R. Bjorsvik, U. M. Hansen
and R. Carlson, Acta Chem. Scand., 1997, 51, 733 and references
therein.
11 (a) C. P. Brock and J. A. Ibers, Acta Crystallogr., Sect. B, 1973, 29,
2426; (b) D. C. Levendis and I. Bernal, Struct. Chem., 1997, 8, 263.
12 D. A. Fletcher, R. F. McMeeking and D. Parkin, J. Chem. Inf. Comput.
Sci., 1996, 36, 746.
a script of driver calculations on the torsional angle s = Cring–Cring
–
Calkyl–Calkyl. For both ground and transition states, the selected
torsional angle was driven in 10◦ increments from −180 to +180◦.
−1
The geometry was optimised to a gradient of 0.1 kcal A mol−1
˚
or a maximum of 700 cycles using a large restraint (500 kcal mol−1
degree−2) and after removal of restraints, a single point energy
was calculated. If more than one ring bore substituents, driver
calculations were performed successively on each substituent,
allowing free adjustment of the remaining geometry. Using the
same script, conformational preferences with respect to Cr(CO)5
orientation were explored.
In the case of diarylphosphines, all 1369 possible combinations
of ring torsional angles (−180 to +180◦ on both rings, 10◦ steps)
−1
were generated and optimised to a gradient of 0.1 kcal A mol−1
˚
or a maximum of 1000 cycles, applying a high torsional restraint
(500 kcal mol−1 degree−2). Removal of restraints was followed by
a single point energy calculation. The barrier to rotation of the
alkyl carbon–phosphorus bond was obtained similarly.
13 J. Bruckmann and C. Krueger, J. Organomet. Chem., 1997, 536, 465–
537.
14 K. A. Jensen and P. H. Nielsen, Acta Chem. Scand., 1963, 17, 1875.
15 (a) M. L. Caffery and T. L. Brown, Inorg. Chem., 1991, 30, 3907; (b) K. J.
Lee and T. L. Brown, Inorg. Chem., 1992, 31, 289.
Acknowledgements
16 J. A. S. Howell, P. C. Yates, M. G. Palin, P. McArdle, D. Cunningham,
Z. Goldschmidt, H. E. Gottlieb and D. Hezroni-Langerman, J. Chem.
Soc., Dalton Trans., 1993, 2775.
17 (a) A. G. Hopfinger and R. A. Pearlstein, J. Comput. Chem., 1984, 5,
486; (b) P. Ugliengo, J. Ahmed, D. Viterbo and M. Ceruti, Gazz. Chim.
Ital., 1989, 119, 493.
18 HyperChem 5.11, Hypercube Inc., Gainesville, FL, 2002.
19 2-MeC6H4 series: (a) J. A. S. Howell, M. G. Palin, P. C. Yates,
P. McArdle, D. Cunningham, Z. Goldschmidt, H. E. Gottlieb and
D. Hezroni-Langerman, J. Chem. Soc., Perkin Trans. 2, 1992, 1769;
(b) T. S. Cameron and B. Dahlen, J. Chem. Soc., Perkin Trans. 2, 1975,
1737.
20 2-CF3C6H4 series: J. A. S. Howell, N. Fey, J. D. Lovatt, P. C. Yates, P.
McArdle, D. Cunningham, E. Sadeh, H. E. Gottlieb, Z. Goldschmidt,
M. B. Hursthouse and M. E. Light, J. Chem. Soc., Dalton Trans., 1999,
3015.
We wish to acknowledge the use of the EPSRC’s Chemical
Database Service at Daresbury.13 One of us (N.F.) also wishes
to acknowledge financial support from Keele University and
the Society for Chemical Industry for the award of a Messel
Scholarship.
References
1 For recent reviews of specific areas of catalysed asymmetric synthesis,
see: (a) F. Speiser, P. Braunstein and W. Saussine, Acc. Chem. Res.,
2005, 38, 784; (b) X. H. Cui and K. Burgess, Chem. Rev., 2005, 105,
3272; (c) S. Castillon, C. Claver and Y. Diaz, Chem. Soc. Rev., 2005,
34, 702; (d) A. M. Carroll, T. P. O’Sullivan and P. J. Guiry, Adv. Synth.
Catal., 2005, 347, 609; (e) R. C. J. Atkinson, V. C. Gibson and N. J.
Long, Chem. Soc. Rev., 2004, 33, 313; (f) T. P. Clark and C. R. Landis,
Tetrahedron: Asymmetry, 2004, 15, 2123; (g) K. V. L. Crepy and T.
Imamoto, Top. Curr. Chem., 2003, 229, 1.
2 For recent reviews, see: (a) D. Sinou, Adv. Synth. Catal., 2002, 344,
221; (b) N. Pinault and D. W. Bruce, Coord. Chem. Rev., 2003, 241, 1;
(c) K. H. Shaughnessy and R. B. De Vasher, Curr. Org. Chem., 2005,
9, 585; (d) A. D. Phillips, L. Gonsalvi, A. Rornerosa, F. Vizza and M.
Perruzzini, Coord. Chem. Rev., 2004, 248, 955.
21 (a) K. Mislow, Chemtracts: Org. Chem., 1989, 2, 151; (b) K. Mislow, D.
Gust, R. J. Finocchiaro and R. J. Boettcher, Top. Curr. Chem., 1974,
47, 1.
22 See references 23, 24 in reference 20.
23 I. O. Sutherland, Annu. Rep. NMR Spectrosc., 1971, 4, 80.
24 For an alternative synthesis, see: G. M. Whitesides, M. Eisenhut and
W. M. Bunting, J. Am. Chem. Soc., 1974, 96, 5368.
˚
25 B. Akermark and A. Ljungqvist, J. Organomet. Chem., 1978, 149, 97.
3 For recent reviews, see: (a) A. P. Dobbs and M. R. Kimberley, J. Fluorine
Chem., 2002, 118, 3; (b) G. Pozzi and I. Shepperson, Coord. Chem. Rev.,
2003, 242, 115.
26 G. M. Sheldrick, SHELXS-97, Acta Crystallogr., Sect. A, 1990, 46, 467.
27 G. M. Sheldrick, SHELXL97, University of Go¨ttingen, Germany,
1997.
This journal is
The Royal Society of Chemistry 2006
Dalton Trans., 2006, 5464–5475 | 5475
©