Paper
www.rsc.org/dalton | Dalton Transactions
Molecular structure of tris(pentafluoroethyl)phosphane P(C2F5)3
Stuart A. Hayes, Raphael J. F. Berger, Beate Neumann, Norbert W. Mitzel,*
Julia Bader and Berthold Hoge*
5
10
15
Received (in XXX, XXX) Xth XXXXXXXXX 200X, Accepted Xth XXXXXXXXX 200X
First published on the web Xth XXXXXXXXX 200X
DOI: 10.1039/b000000x
The structure of tris(pentafluoroethyl)phosphane, P(C2F 5)3, has been determined in the solid state
by in-situ single crystal growth and subsequent X-ray diffraction and in the gas phase by electron
diffraction. The experimental structure determinations have been supported by and compared with
quantum chemical calculations at different levels of theory. P(C2F 5)3 prefers a conformation of C3
symmetry with the CF 3 groups of the three C2F 5 units oriented to the same side of the molecule as
the phosphorus lone pair of electrons, leading to an estimated Tolman cone angle of 193° (solid)
and 191° (gas), whereas a second stable conformer of C3 symmetry, (only 5 kJ mol-1 higher in
energy) has a smaller Tolman cone angle of 171°. The structures are discussed in context with
related molecules E(CF 3)3 (E = N, P, As, Sb) as well as N(C2F 5 )3 and P(C2H5)3.
Dedicated to Professor David W. H. Rankin on the occasion of his 65 th birthday
Introduction
20
25
30
35
40
45
50
Electron poor ligands such as perfluoroorganylphosphane
ligands are useful tools to tune the redox potential as well as
the Lewis-acidity of corresponding transition metal complexes. 1 One of the most electron poor and therefore strong πacidic perfluoroorganylphosphane ligands is tris(trifluoromethyl)phosphane, P(CF3)3, which is known to form stable
transition metal complexes.2 The synthesis of P(CF3)3 complexes requires special safety precautions because P(CF 3)3 (b.
p. 17°C)3 reacts violently on contact with air,4 while the less
volatile homologue P(C2F 5)3 (b. p. 62°C)5 can be handled by
standard Schlenk techniques. Therefore it is reasonable to
substitute the potentially hazardous P(CF 3)3 ligand by the
P(C2F5)3 ligand which is easy to handle as well as easily
accessible. Tris(pentafluoroethyl)phosphane is accessible on a
large scale via the reduction of P(C2F 5)3F2 - a technical
product of the Merck Company, Darmstadt - with NaBH46 or
by the reaction of Me3SiC2F 5 with P(OPh)3. 5
Based on DFT calculations, the electronic nature of the
P(CF 3)3 and P(C2F 5)3 ligand should be comparable. The
Tolman electronic parameter (TEP) is defined as the A1 COvalence mode of the corresponding [Ni(CO)3PR3] complexes.
For P(CF 3)3 and P(C2F 5)3, the calculated TEPs of 2104.4 and
2102.9 cm-1, respectively, are nearly identical. 7
The reaction of [Mo(CO)4(nbd)] (nbd = norbornadiene) with
P(CF 3)3 leads within hours at room temperature to a nearly
quantitative formation of the literature known complex
[Mo(CO)4{P(CF 3)3}2].8 By contrast, the reaction with
P(C2F5)3 proceeds much slower and results after a prolonged
reaction time mainly in the formation of decomposition products. To demonstrate the different coordination properties of
P(CF 3)3 and P(C2F 5)3, [Mo(CO)4(nbd)] was treated with an
excess of a mixture of P(CF 3)3 and P(C2F 5)3 yielding only the
bis[tris(trifluoromethyl)phosphane] complex [Mo(CO)4{P(CF3)3}2]. Based on these observations, sterical reasons may
be the cause of the different coordination properties of the
This journal is © The Royal Society of Chemistry [year]
55
P(CF 3)3 and P(C2F 5)3 ligand.
Herein we report the molecular structure of tris(pentafluoroethyl)phosphane, P(C2F5)3, in the gas and solid phases. The
structures explain the significantly different coordination
properties of the P(CF 3)3 and P(C2F5)3 ligand.
Results and discussion
60
65
70
Crystal Structure
P(C2F5)3 crystallises in the trigonal space group R, with a
molecular structure belonging to the C3 point group, as shown
in Figure 1 and 2. The lone pairs on phosphorous in adjacent
molecules oppose one another, but the P···P distance is longer
than the sum of van-der-Waals radii. The values of structural
parameters are presented in the following discussion,
alongside those for the theoretical and experimental structures
in the gas phase.
Figure 1 Crystal structure of P(C2F5)3 showing the atom numbering and
the unit cell viewed along the c-axis.
Journal Name, [year], [vol], 00–00 | 1
Figure 2 Crystal structure of P(C2F5)3 showing the atom numbering and
the unit cell viewed along the c-axis.
5
10
15
30
Figure 3 Calculated potential-energy curves for the dihedral angle formed
by the C3 axis, P, C(1) and C(2). The respective solid- and gas-phase
angles refined to 29.8(1)° and 30.8(5)°.
Ab initio and Hybrid Density-Functional Calculations
The electronic structure calculations were performed, primarily, to assist the determination of the experimental gas-phase
structure by GED. In particular, potential energy scans were
used to the search for probable conformers. Geometry optimisations were performed to estimate the relative energies of
local minima on the potential-energy surface (PES) and as
flexible restraints on some geometric parameters in the GED
refinement. Finally, frequency calculations were used to confirm the nature of the stationary points on the PES and, in
combination with the program SHRINK,9 to provide approximate values for the amplitudes of vibration, u, and distance
corrections for curvilinear perpendicular motion, kh1.
35
40
45
Predicted Conformations of P(C2F5)3 in the Gas Phase
20
25
The three -C2F 5 groups allow a large number of possible
conformations, making a systematic search of all conformations by ab initio methods impractical. However, as this molecule adopts C3 symmetry in the solid phase, and does not
possess any inter-atomic contacts shorter than the sum of vander-Waals’ radii, it is likely that the lowest-energy gas-phase
structure also has this symmetry. Potential energy scans were
performed for torsion around the P–C bonds, allowing all
other parameters to relax within the constraint of C3
symmetry, as shown in Figure 3.
50
55
2 | Journal Name, [year], [vol], 00–00
This reveals two low-energy local minima at ca. 30° and 65°,
denoted A and B, with a potential energy barrier of ca. 10 to
20 kJ mol–1 between them. A is close to the crystal structure,
for which the corresponding dihedral angle is 29.8(1)°. Both
A and B were confirmed as minima on the full potentialenergy surface by further geometry optimisations and
frequency analyses, without symmetry constraints.
Such close minima, separated by only 35°, but with an
appreciable energy barrier between them are unusual and
appears to be a result of steric repulsion between fluorine
atoms on adjacent C2F 5 groups. The correlation between the
calculated energy and the shortest non-bonded F···F distance
on neighbouring fluoroethyl groups can be seen by comparing
Figures 3 and 4. This reveals that all three local minima in the
potential energy curve correspond to structures where the
shortest non-bonded F···F distance is longer than the sum of
the van-der-Waals radii of two fluorine atoms.
Figure 4 Shortest F···F distances between neighbouring C2F5 groups for
calculated structures. The dotted line is the sum of van-der-Waals radii for
two fluorine atoms. The y-axis has been inverted to emphasise the
correlation with the calculated energy, shown in Figure 3.
This journal is © The Royal Society of Chemistry [year]
45
A 0 kJ mol–1
B 5 kJ mol–1
50
using the harmonic-oscillator and rigid-rotor approximations,
and are shown alongside the electronic energy in Table 1. The
differences between the values of ∆E and ∆G are almost
entirely due to the thermal contributions to the free energy, as
the calculated zero-point-energy corrections differ by no more
than 0.4 kJ mol–1 (at a given level of theory).
Table 1 Calculated relative electronic and free energies (∆E and ∆G,
respectively) in kJ mol–1, at various levels of theory, and the
corresponding abundances for the MP2 calculation.
Conformer
Sym
A
B
C3
C
C 9 kJ mol–1
D 8 kJ mol–1
D
5
a
Figure 5 Conformers A – D of P(C2F5)3 calculated to be minima on the
potential-energy surface and relative energies (RI-MP2/TZVPP).
10
15
20
A search for asymmetric conformers representing minima on
the potential-energy surface was carried out by identifying
conformers for which repulsive F···F contacts would be
minimised. Potential structures were optimised at the HF/631G* level of theory and subjected to frequency analysis. In
this way, three asymmetric conformers were found, with
energies calculated to be 9.2, 7.2 and 11.6 kJ mol–1 higher
than that of A. These energies are relatively high and, if
correct, these conformers are unlikely to be observable in the
GED experiment, despite the lower symmetry compared to A.
The highest energy conformer was therefore not subjected to
further study, but the geometries of the remaining two
structures (C and D) are shown in Figure 5 alongside those of
the C3 symmetric conformers, A and B.
Relative Energies of conformers A, B, C and D
25
30
35
40
As can be seen from Figure 3, the relative electronic energies
of A and B vary with the level of theory applied. Considering
that an energy difference of 1 kJ mol–1 can significantly
change the population of a species, the discrepancy of ca. 8 kJ
mol–1 between HF and MP2 theory is a large disagreement.
Fluorine is known to be problematic for calculating electron
correlation and the apparent importance of steric F···F interactions in this compound may exacerbate this problem.
However, the 6-31G* basis is not well balanced for MP2 and
this may also be playing a role. The MP2 method was found
to be prohibitively expensive for larger basis sets, but it was
possible to optimise the geometry using the RI-MP2
method 10–13 with the larger TZVPP basis set,14 yielding the
energies in Table 1.
In addition to the problem of calculating the electronic energy, the zero-point energy corrections and thermal contributions to the free energy may be quite different for each conformer. The free energy differences were calculated for all four
conformers at the HF/6-31G* and B3LYP/6-311G* levels
This journal is © The Royal Society of Chemistry [year]
55
60
65
70
C1
HF/6-31G*
B3LYP/6311G*
∆E
∆G
RIMP2/TZVPP
∆E
%a
∆E
∆G
0
0
0
0
0
74.3
7.14
10.88
3.23
4.75
4.75
10.5
9.21
10.33
5.45
4.14
8.62
6.5
7.23
8.56
4.82
2.68
7.88
8.7
Calculated using ∆E at the RI-MP2/TZVPP level of theory.
Theoretical geometries
The convergence of the calculated geometric parameter values
(for conformer A) can be seen from Table 2. The effect of
increasing the basis-set size can be gauged by comparing the
HF/6-31G* and HF/TZVPP parameter values as well as those
at the B3LYP level of theory, where different basis sets have
been used. These effects are small, with bonded distances
changing by little more than 1 pm as the basis set is increased
from double- to triple-zeta quality. In terms of the effect of
the level of theory, the agreement between HF and MP2 is
remarkably good, the only significant disagreement being the
average C–F distance, which HF is expected to underestimate.
The hybrid density functionals predict longer bonded C–P and
C–C distances and, given the good agreement between HF and
MP2, are unreliable. The remaining parameter values are,
however, in good agreement. The calculated (RIMP2/TZVPP) geometries and energies of all four conformers
are provided as supporting information.
Table 2 Values of selected geometric parameters for conformer A at
various levels of theory. Distances are in Å and angles are in degrees. a
75
a
HF
HF
631G*
TZV
PP
RIMP2
TZV
PP
B3L
YP
631G*
B3LY
P
6311G*
rCP
1.900
1.912
1.909
1.924
1.931
B3L
YP
ccpVT
Z
1.937
B3PW
91
ccpVTZ
rCC
1.533
1.545
1.547
1.548
1.554
1.556
1.554
1.337
1.928
rCFav
1.320
1.312
1.337
1.347
1.344
1.342
∠CPC
99.0
99.9
98.2
97.9
98.5
98.8
99.1
∠CCP
113.3
112.6
111.3
111.7
111.6
111.8
111.3
φzPCC
30.6
30.2
30.3
30.7
30.3
30.4
29.8
The TZVPP basis set is similar in size and construction to cc-pVTZ
Journal Name, [year], [vol], 00–00 | 3
dihedral angles, respectively. All parameters were subsequently refined to give the relationship between the R-factor
and the fraction of conformer B shown in Figure 6. From this
plot, the abundance of B is not likely to be more than 3 %.
Gas Electron Diffraction (GED)
5
10
15
20
GED Model. C3 symmetry was assumed in the molecular
model, which reduces the number of required parameters from
60 to 20. In addition, local C3v symmetry was assumed for the
CF 3 group and, on the basis of the ab initio calculations, all
F–C–F angles were assumed to be equal. The geometry was
therefore described in terms of 13 independent parameters, as
shown in Table 3. Parameters p9 to p 11 are deviations of the –
CF 2– group from local C2v symmetry (with respect to the bond
angles) and each of these was defined as a rotation around one
of three orthogonal vectors centred at C(1). The ‘rock’ (p 9)
denotes a rotation of the fluorine atoms around a vector
normal to the P–C–C plane, where a positive value is an
increase in the P–C–F angles at the expense of the C–C–F
angles. The ‘twist’ angle (p 9) is a rotation around the bisector
of the unit vectors in the directions of the C–P and C(1)–C(2)
bonds, a positive value indicating increases in the angles ∠P–
C–F(2) and ∠C–C–F(1). The ‘wag’ is therefore a rotation
around a vector approximately normal to the F–C–F plane,
where a positive value is defined as an increase in the angles
∠P–C–F(2) and ∠C–C–F(2).
Table 3 Refined gas-phase geometric parameters and applied restraints.
Distances are in Å and angles are in degrees.
Parametera
25
GED(rg)
Figure 6 Variation of the R-factor with the addition of a second conformer (the disagreement between the R-factor at x = 0.0 and that given in
Table 3 is due to small differences in data s-limits and weighting points)
50
Figure 7 Experimental (○), theoretical and difference (experimental
minus theoretical) molecular-intensity curves.
RI-MP2(re)b
p1
rP–C
1.904(3)
1.901(3)
1.909
p2
rC–C
1.533(3)
1.524(4)
1.547
p3
rC–F average
1.339(1)
1.338(1)
1.337
p4
rC–F diff. (CF2 – CF3)
0.016(2)
0.017(2)
0.017(2)
p5
rC–F diff. (CF2 group)
0.007(2)
0.006(2)
0.007(2)
p6
∠C–P–C
100.1(2)
99.9(3)
98.2
p7
∠P–C–C
113.9(3)
113.2(3)
111.3
p8
∠F–C–F
108.0(1)
108.0(1)
108.7
p9
∠CF2 rock
3.5(3)
3.9(4)
4.0(5)
p10
∠CF2 twist
0.5(3)
0.4(4)
1.0(5)
4.0(5)
p11
∠CF2 wag
4.3(4)
4.8(4)
p12
φz–P–C–C
φP–C–C–F
30.8(5)
31.0(5)
30.3
182.2(9)
182.1(10)
179.3
RG
4.70%
7.18%
–
p13
a
GED(rh1)
45
For definitions of parameters 9, 10 and 11, see text. The C3 axis is denoted z,
as in the calculations section.
b
Where the theoretical value is followed by
brackets, the GED value was restrained to the RI-MP2/TZVPP value, with an
55
uncertainty given by the number in parentheses.
30
35
40
In addition to the results presented here, a model containing
two conformers, A and B, was also tested. For this purpose,
four additional parameters were introduced: A difference
between the C–P–C angles in A and B, a difference between
the P–C–C angles in A and B, and two new torsional angles
(φz–P–C–C and φP–C–C–F in B). The difference between φz–
P–C–C in A and B and the difference between φP–C–C–F in
A and B were included as dependent parameters. In order to
prevent over-fitting of the data, the difference parameters
were tightly restrained to the RI-MP2/TZVPP values with
uncertainties of 0.1 and 1.0 degrees for the angles and
4 | Journal Name, [year], [vol], 00–00
60
65
GED Refinement. The GED refinement was performed according to the SARACEN method, 15 allowing all thirteen independent geometric parameters to be refined, eight of which
refined unrestrained. The resulting parameter values are
shown in Table 3. In addition, ten groups of amplitudes of
vibration were refined, each corresponding to a single peak in
the radial-distribution curve. These were restrained to the
B3LYP/6-311G* values, with uncertainties of 10 % of the calculated values. A full list of inter-nuclear distances, camerato-nozzle distances, electron wavelengths, data cut-offs and
weighting points; the least-squares correlation matrix can be
found in supplementary information (Tables S1, S2 and S3).
The single-conformer model gave a good fit to the
experimental data, as shown by the molecular-intensity and
radial-distribution curves (Figures 7 and 8, respectively).
This journal is © The Royal Society of Chemistry [year]
Whilst the refined parameter values are similar for the rg and
rh1 refinements (Table 3), the quality of fit is substantially
better for the rh1 structure type, reducing the R-factor from
7.18 to 4.70 %.
45
Table 4 Comparison of parameter values for P(C2F5)3 as determined by
single-crystal x-ray diffraction, ab initio MP2 theory and gas-phase
electron diffraction
XRD
5
Figure 8 Experimental and difference (experimental minus theoretical)
radial-distribution curves. Prior to Fourier transformation, the molecularscattering intensities were multiplied by s⋅exp[(–0.00003s2)/(ZC – fC)(ZF –
fF)].
10
15
20
25
30
35
40
Discussion
The molecular-structure parameters determined by the three
methods used (single crystal XRD, ab initio calculations and
GED) are compared in Table 4. These three methods are in
good agreement, indicating that the effects of crystal packing
are small. Even the torsional angle around the P–C bond (φ z–
P–C–C) is virtually the same in the solid phase 29.8(1)° and
in the gas phase 30.8(5)°. A couple of small differences can
also be seen between the calculated equilibrium geometry and
the experimental GED structure, where the XRD structure is
most similar to the calculated geometry, namely, the C–C
distance and the P–C–C angle. As can be seen from Table 3,
the C–C distance is one of the few parameters that is
dependent on the structure type and increases from 1.524(3) to
1.533(3) with the inclusion of curvilinear distance corrections
in the structure refinement. As the distance corrections are
fixed, the uncertainty on this value is larger than the standard
deviation obtained from the least-squares fit. The P–C–C
angle is not as dependent on the structure type, but may still
be influenced by thermal motion that cannot be extrapolated
from the minimum of the potential-energy surface. However,
there also appears to be some flexibility in this angle as the
calculated (RI-MP2/TZVPP) values of the P–C–C angles in
conformers B to D range from 109° to 124°.
A glance back at Table 2 reveals that the hybrid DFT methods
do not reproduce the experimental geometry to a high degree
of accuracy and for most parameter values perform less well
than HF. In particular, the P–C bond length is overestimated
by ca. 3 pm by B3LYP and ca. 2 pm by B3PW91 (with the
relatively large cc-pVTZ basis set). In general, however, the
parameter values predicted by B3PW91 are slightly better
than those predicted by B3LYP (where the solid- and gasphase structures are in good agreement).
This journal is © The Royal Society of Chemistry [year]
a
50
RI-MP2/TZVPP
GED (rh1)
rP–C
1.908(1)
1.909
1.904(3)
rC–C
1.543(2)
1.547
1.533(3)
rC–F(1)
1.351(1)
1.351
1.351(2)
rC–F(2)
1.348(1)
1.344
1.345(2)
rC–F(3) a
1.330(2)
1.332
1.333(1)
rC–F(4) a
1.323(1)
1.330
1.333(1)
rC–F(5) a
1.328(2)
1.326
1.333(1)
∠C–P–C
99.24(5)
98.2
100.0(2)
∠P–C–C
110.78(8)
111.3
113.9(3)
∠P–C–F(1)
108.69(8)
108.4
108.0(3)
∠P–C–F(2)
114.57(8)
114.3
112.8(3)
∠C–C–F(1)
105.96(9)
105.7
105.2(4)
∠C–C–F(2)
108.49(9)
108.4
108.4(4)
φ z–P–C–C
29.8(1)
30.3
30.8(5)
Assumed to be equal in the GED model.
Table 5 Comparison of structure parameters for different tris(perfluoroalkyl)pnictogenes and P(C2H5)3
compound
method
d(E-C)a
d(C-F)
N(CF3)316
GED
1.426(6)
1.323(4)
P(CF3)317
GED
1.937(17)
1.342(13)
<C-E-C a
117.9(4)
compound
method
d(E-C)a
N(C2F5)319
GED (3σ)
1.482(7)
d(C-F)
1.330(2)
∠C-E-C
a
a
119.3(4)
∠E-C-C a
118.2(26)
φ z–E–C–C a
147.0(5)
E = N, P, As, Sb;
b
Sb(CF3)317
GED
2.202(16)
1.326(9)
99.6(25)
As(CF3)318
GED
2.007(2)
1.337(3)
1.326(3)
95.5(3)
P(C2F5)3b
GED
1.904(3)
P(C2F5)3b
XRD
1.908(1)
P(C2H5)320
XRD
1.842(2)
1.841(2)
1.844(2)
-
1.348av (CF2) 1.350av (CF2)
1.333(1) (CF3) 1.327av (CF3)
100.0(2)
99.24(5)
this work;
c
113.9(3)
110.78(8)
30.8(5)
29.8(1)
fixed, i.e. not refined;
d
100.0(35)
100.2(1)
98.8(1)
99.4(1)
113.0(2)
113.5(2)
112.5(2)
48 - 50
not reported, calcd
from given coordinates; av = average
55
60
65
Data for comparison of the structure of P(C2F 5)3 with related
compounds are presented in Table 5. Such include the
trifluoromethylpnictogenes N(CF 3)3, P(CF3)3, As(CF 3)3 and
Sb(CF 3)3, the pentafluoroethyl nitrogen analogue N(C2F 5)3
and the non-fluorinated analogue P(C2H5)3.
The value for the P-C bond length for P(CF 3)3 at 1.937(17) Å
seems somewhat larger than that in P(C2F 5)3 at 1.904(3) (gas
phase), but is equal considering the large standard deviation
for P(CF 3)3 (and Sb(CF3)3) due to the structure determinations
have been done at that time without least squares fitting and
rotating sector device in the diffractometer (we have recently
reinvestigated the structure of As(CF 3)3 with modern instruJournal Name, [year], [vol], 00–00 | 5
5
10
15
20
25
30
35
40
45
50
55
mentation and advanced data analysis procedures and could
drastically improve the accuracy). 17
Expectedly one observes for the trifluoromethylpnictogenes
E(CF 3)3 (E = N, P, As, Sb) a drastical decrease of the C-E-C
angle from E = N to P. There is a slight widening of this C-NC angle from N(CF 3)3 to N(C2F 5)3, but both are unusually
large, even for amines, reflecting the unusual steric and
electronic situation in such almost planarised nitrogen
compounds. 21
However, the C-P-C angles in P(CF3)3 and P(C2F5)3 (solid and
gas) are remarkably similar, showing the very comparable
demand for space directly at the phosphorus atom (but again,
an exact comparison is hampered by the low accuracy of the
old P(CF 3)3 structure). Interestingly this angle is not affected
by fluorination of the ethyl groups, as P(C2H5)3 has almost the
same value for this structural parameter.
The most interesting parameter to look at is the torsion of the
ethyl groups, as this defines the orientation of the latter and
ultimately the accessibility of the pnictogen atom expressed
by the Tolman angle for the purpose of describing their binding properties as Lewis-basic ligands towards metal atoms.
This torsion, defined relative to the lone pair of electrons
position, i.e. relative to the axis of symmetry (where applicable). In this respect, there is a marked difference between
the nitrogen and phosphorus compounds, as for N(C2F 5)3 with
its almost planar nitrogen coordination geometry the torsion
angle is 147.0(5)°, whereas for the phosphorus analogue it is
30.8(5)°, both in the gas phase. As pointed out earlier, the
conformation of P(C2F 5)3 seems to be determined by intergroup F⋅⋅⋅F contacts and thus a difference should be expected
in this structural feature between P(C2F5)3 and the nonfluorinated P(C2H5)3. Indeed is there a larger torsional angle
for P(C2H5)3 at 48 – 50° for the three crystallographically
different ethyl groups.
60
65
70
Experimental
75
Crystal Structure Determination
A single crystal of tris-pentafluoroethyl phosphane (PFEP),
P(C2F5)3, was grown in a sealed glass capillary, in situ on a
Nonius Kappa CCD diffractometer at ca. 200 K. The crystal
was then cooled to 100(2) K for data acquisition.
80
85
90
Conclusion
The Tolman cone angle is one of the most established models
to describe the steric bulk of phosphane ligands. 22 The cone
angle of the conformer A of P(C2F 5)3 (comp. Figure 5) was
calculated to be 193° in the solid state (XRD) and 191° in the
gas phase (GED) (calculated for a hypothetical P-M distance
of 2.28 Å and a F van-der-Waals radius of 1.50 Å). Based on
DFT calculations, the P(C2F 5)3 ligand will favour a B type
conformer (conf. Figure 5) in the coordination sphere of a
Ni(CO)3 moiety, 7 which reduces the cone angle to about 171°
(based on the calculations on conformer B described above).
These cone angles exceed the cone angle - the steric bulk - of
the P(CF 3)3 ligand significantly and classify the P(C2F 5)3
ligand as a sterically demanding ligand. Tolman published a
cone angle of 137° for the P(CF 3)3 ligand,22 which will be
reduced to 128°23 in the coordination sphere of the iron
complex [Fe(CO)2{P(CF 3)3}3]. 24 These considerably different
steric bulks might be one reason for the distinctly different
reactivity of P(CF3)3 and P(C2F 5)3 towards [Mo(CO)4(nbd)].
In addition, this fact might also be the reason why no example
of a P(C2F 5)3 transition metal complex is known so far.
At first view the different coordination properties of the
P(CF 3)3 and P(C2F 5)3 ligands may be a surprise because the
6 | Journal Name, [year], [vol], 00–00
substitution of only one perfluoroorganyl group by a less sterically demanding non fluorinated alkyl group leads to ligands
of nearly comparable cone angles. The estimated cone angles
of the most commonly used bidentate bis(perfluoroorganyl)phosphane ligands (CF 3)2PCH2CH2P(CF 3)2 and (C2F 5)2PCH2CH2P(C2F 5)2 are estimated to be 120 and 129°, respectively. 25
Further investigations are in progress to obtain evidence about
the complexation ability of P(C2F5)3 and to answer the question which of the conformers A and B (or even C and E, see
Figure 5) will be presumably favoured by P(C2F 5)3 in the role
of a ligand.
95
100
105
110
Crystal data. C6F 15P, M = 388.03, trigonal, space group R ,
a = 13.4590 (4), c = 10.8110 (3), V = 1695.98 (9) Å3, T = 100
(2) K, Z = 6, µ(Mo-Kα) = 0.443 mm-1, 6601 reflections
measured, 1086 unique (Rint = 0.028). Structure solutions and
refinements were undertaken with the programs SHELXS-97 26
and SHELXL-97.27 Final R1, wR2 values on all data 0.0309,
0.0781. R1, wR2 values on [Io > 2σ(Io), 1009 reflections] data
0.0287, 0.0765. CCDC reference number XXX. For crystallographic data in CIF or other electronic format see DOI:
XXX/bXXXXx.
Ab initio and Hybrid DFT Calculations
The majority of calculations were performed using the
Gaussian 03 suite of programs,28 the exceptions being the RIMP2/TZVPP 10–14 calculations, which were performed using
the Turbomole29 program. All calculations were restricted and
MP2 calculations were performed with frozen cores. The
B3LYP and B3PW91 make use of Becke’s three parameter
hybrid functional,30 where the non-local correlation energy is
provided by the LYP expression 31 in B3LYP and by the
Perdew-Wang expression 32 in B3PW91.
GED
Electron scattering intensities for P(C2F 5)3 were recorded at
room temperature on combination of reusable Fuji and Kodak
imaging plates using a Balzers KD-G233 Gas-Eldigraph
(formerly operated in Tübingen by H. Oberhammer34), equipped with a new electron source (STAIB Instruments), operating at ca. 60 kV. The image plates that were exposed at the
shorter nozzle-to-plate distance (252.78 mm) were scanned
using a Fuji FLA 3000 scanner, whilst those recorded at the
longer distance (501.88 mm) were scanned using a Fuji BAS
1800 scanner. In both cases the scanning software yielded
digital 16-bit grey-scale image data. These data were reduced
to total intensities using T. G. Strand’s program PIMAG35
(version 040827) in connection with a sector curve, which was
based on experimental xenon scattering data and tabulated
scattering factors of xenon. Further data reduction (yielding
This journal is © The Royal Society of Chemistry [year]
5
10
20
molecular-intensity curves), the molecular structure refinement, and the electron wavelength determination (from benzene data) were performed using version 2.4 of the ed@ed program. 36 The scattering factors employed were those of Ross et
al. 37 Further details about the Bielefeld GED apparatus and
methods are to be published elsewhere.38 Table S1 (see
supplementary information) gives the data analysis parameters
for each data set: R-factors (RD and RG), scale factors, correlation parameter values, data ranges, weighting points, nozzleto-plate distances and electron wavelengths. Amplitudes of
vibration, u, and distance corrections for curvilinear
Notes and references
perpendicular motion, kh1, were calculated using the program
SHRINK,9 making use of frequency calculations at the
B3LYP/6-311G* level of theory.
15
Acknowledgements
This work was supported by Deutsche Forschungsgemeinschaft. The Merck Company, Darmstadt, is acknowledged for
financial support. We want to thank Dr. N. Ignatiev (Merck
AG) for helpful discussions.
75
*
25
Fakultät für Chemie, Universität Bielefeld, Universitätsstraße 25,
33615, Bielefeld, Germany. E-mail: [email protected]; Fax:
+49 (0) 521 106 6026
† Electronic
Supplementary
Information
(ESI)
available:
XXXXXXXXXXXXXXXXXXXXXX see DOI: 10.1039/b000000x/
1.
30
2.
3.
4.
5.
35
6.
7.
40
8.
45
9.
10.
11.
12.
13.
14.
50
15.
55
16.
60
17.
18.
19.
65
70
For example: E. J. Velazco, A. J. M. Caffyn, X. F. Le Goff and L.
Ricard, Organometallics 2008, 27, 2402 and literature cited
therein.
J. Apel, J. Grobe, Z. Anorg. Allg. Chem. 1979, 453, 53.
F. W. Bennett, H. J. Emeléus and R. N. Haszeldine, J. Chem. Soc.
1953, 1565.
W. Mahler, Inorg. Chem. 1979, 18, 352.
M. B. Murphy-Jolly, L. C. Lewis and A. J. M. Caffyn, Chem.
Commun. 2005, 35, 4479.
U. Welz-Biermann, N. Ignatyev, M. Weiden, M. Schmidt, U.
Heider, A. Miller, H. Willner and P. Sartori, WO 03/087113A1,
Merck Patent GmbH, Darmstadt.
D. G. Gusev, Organometallics 2009, 28, 763; and personal
communications.
J. Apel, R. Bacher, J. Grobe and D. Le Van, Z. Anorg. Allg. Chem.
1979, 453, 39.
V. A. Sipachev, J. Mol. Struct. (THEOCHEM) 1985, 121, 143.
R. Ahlrichs, Phys. Chem. Chem. Phys. 2004, 6, 5119.
F. Weigend and M. Häser, Theor. Chem. Acc. 1997, 97, 331.
F. Weigend, M. Häser, H. Patzelt and R. Ahlrichs, Chem. Phys.
Lett. 1998, 294, 143.
F. Haase and R. Ahlrichs, J. Comp. Chem. 1993, 14, 907.
A. Schäfer, C. Huber and R. Ahlrichs, J. Chem. Phys. 1994, 100,
5829.
a) N. W. Mitzel, D. W. H. Rankin, J. Chem. Soc., Dalton Trans.
2003, 3650; b) A. J. Blake, P. T. Brain, H. McNab, J. Miller, C. A.
Morrison, S. Parsons, D. W. H. Rankin, H. E. Robertson and B. A.
Smart, J. Phys. Chem. 1996, 100, 12280. c) N. W. Mitzel, H.
Schmidbaur, D. W. H. Rankin, B. A. Smart, M. Hofmann and P. v.
R. Schleyer, Inorg. Chem. 1997, 36, 4360; d) N. W. Mitzel, B. A.
Smart, A. J. Blake, H. E. Robertson, D. W. H. Rankin, J. Phys.
Chem. 1996, 100, 9339.
H. Bürger, H. Niepel, G. Pawelke and H. Oberhammer, J. Mol.
Struct. 1979, 54, 159.
C. J. Marsden and L. S. Bartell, Inorg. Chem. 1976, 15, 2713.
H. J. M. Bowen, Trans. Faraday Soc. 1954, 50, 463.
M. Gaensslen, U. Groß, H. Oberhammer and S. Rüdiger, Angew.
Chem. 1992, 104, 1525.
22.
23.
24.
25.
80
26.
27.
85
28.
29.
90
30.
31.
32.
95
33.
34.
35.
100
36.
37.
105
38.
Soc., 1996, 118, 2664; f) N. W. Mitzel and H. Oberhammer, Inorg.
Chem. 1998, 37, 3593.
For example: C. A. Tolman, Chem. Rev. 1977, 77, 313.
B. J. Dunne, R. B. Morris and A. G. Orpen, J. Chem. Soc., Dalton
Trans. 1991, 653.
A. B. Burg, Inorg. Chem. 1986, 25, 4751.
E. J. Velazco and A. J. M. Caffyn, Organometallics, 2008, 27,
2402.
G. M. Sheldrick, SHELXS-97, Acta Crystallogr., Sect. A 1990, 46,
467.
G. M. Sheldrick, SHELXL-97, Program for refinement of crystal
structures, Universität Göttingen, 1997.
Gaussian 03, M. J. Frisch et al., Gaussian, Inc., Wallingford CT,
2004.
R. Ahlrichs, M. Bär, M. Häser, H. Horn and C. Kölmel, Chem.
Phys. Lett. 1989, 162, 165.
A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
a) C. Lee, W. Wang and R. G. Parr, Phys. Rev. B, 1988, 33, 8822.
(b) B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys.
Lett. 1989, 157, 200.
J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B 1998, 54,
16533.
W. Zeil, J. Haase and L. Wegmann, Z. Instr. 1966, 74, 84.
H. Oberhammer in Molecular Structure by Diffraction Methods,
Vol. 4 (Eds: G. A. Sim, L. E. Sutton), The Chemical Society,
London, 1976, p. 24.
S. Gundersen, S. Samdal, T. G. Strand and H. V. Volden, J. Mol.
Struct. 2007, 832, 164.
S. L. Hinchley, H. E. Robertson, K. B. Borisenko, A. R. Turner, B.
F. Johnston, D. W. H. Rankin, M. Ahmadian, J. N. Jones and A. H.
Cowley, Dalton Trans. 2004, 2469.
A. W. Ross, M. Fink, R. Hilderbrandt in International Tables for
Crystallography, Vol. C (Ed.: A. J. C. Wilson), Kluwer Academic
Publishers, Dordrecht, 1992, p. 245.
R. Berger, M. Hoffmann, S. A. Hayes, N. W. Mitzel, Z.
Naturforsch. 2009, 64b, 1259.
20. J. Bruckmann and C. Krüger, Acta Crystallogr. Sect. C 1995, 51, 1155.
21. a) H. Bock, I. Goebel, Z. Havlas, S. Liedle and H. Oberhammer,
Angew. Chem. Int. Ed. 1991, 30, 187; b) N. W. Mitzel, A. Schier
and H. Schmidbaur, Chem. Ber. 1992, 125, 2711; c) N. W. Mitzel,
K. Angermaier and H. Schmidbaur, Chem. Ber. 1994, 127, 841; d)
N. W. Mitzel, J. Riede, A. Schier and H. Schmidbaur, Acta Cryst.
(C), 1995, 51, 756; e) N. W. Mitzel, E. Breuning, A. J. Blake, H.
E. Robertson, D. W. H. Rankin and B. A. Smart, J. Am. Chem.
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 7
Table of Contents Entry
5
Molecular Structure of Tris(pentafluoroethyl)phosphane P(C2F5)
S. A. Hayes, R. J. F. Berger, B. Neumann, N. W.
Mitzel,* J. Bader and B. Hoge*
10
15
The structures of tris(pentafluoroethyl)phosphane, P(C2F 5)3,
determined in the solid and gaseous phases, show it to have
a preferred conformation with a sterically shielded binding
site; this is expressed by a Tolman angle of ca. 191° (gas)
and explains the greater difficulty to act as ligand towards
metal atoms as compared with the slimmer P(CF3)3.
25
20
8 | Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]