Fluoroalkenyl, fluoroalkynyl and fluoroalkyl phosphines

Article abstract of DOI:10.1016/j.jfluchem.2009.08.003

A review of the methods available for the preparation of monodentate P(III) compounds containing fluoroalkenyl, fluoroalkynyl and fluoroalkyl groups is given. The synthesis, properties and coordination chemistry of some fluoroalkenyl- and fluoroalkynyl-co

Full text of DOI:10.1016/j.jfluchem.2009.08.003

Journal of Fluorine Chemistry 130 (2009) 1117–1129
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Fluoroalkenyl, fluoroalkynyl and fluoroalkyl phosphines
*
Kulbinder K. Banger, Alan K. Brisdon , Christopher J. Herbert, Hana Ali Ghaba, Ian S. Tidmarsh
School of Chemistry, The University of Manchester, Manchester M13 9PL, UK
A R T I C L E I N F O
A B S T R A C T
Article history:
A review of the methods available for the preparation of monodentate P(III) compounds containing
fluoroalkenyl, fluoroalkynyl and fluoroalkyl groups is given. The synthesis, properties and coordination
chemistry of some fluoroalkenyl- and fluoroalkynyl-containing phosphines derived from HFC-134a
(CF3CH2F) and HFC-245fa (CF3CH2CH2F) is summarised. The development of the reaction between
trimethylsilyl-containing phosphines and RfI which provides a general method by which bulky
fluoroalkyl groups, such as i-C3F7, t-C4F9, c-C6F11, can be readily introduced into phosphorus(III) centres
is reported. Together these methods provide a way of generating P(III) systems of the type R3ꢀnP(Rf)n
capable of possessing a wide range of steric and electronic properties.
Received 15 May 2009
Received in revised form 13 August 2009
Accepted 13 August 2009
Available online 22 August 2009
Keywords:
Synthesis
Characterisation
Fluoroalkenyl
Fluoroalkynyl
Fluoroalkyl
Phosphines
ß 2009 Elsevier B.V. All rights reserved.
1. Introduction
ligand in the complex [Ni(CO)3P] (where P = phosphine ligand)
˚
assuming an Ni–P bond length of 2.28 A proposed by Tolman [4].
Phosphorus(III) systems of the type PR3 are amongst the most
widely studied ligand systems in transition metal chemistry, and
they, or their derivatives, find applications in a wide range of areas,
including functional materials, such as phase-transfer reagents and
electrolytes, as organocatalysts, and as ligands in a variety of
metal-catalysed C–C bond forming reactions [1]. Part of the reason
for their widespread utility is that by varying the R-groups both the
steric demand and the electronic properties of PR3 compounds may
be systematically modified.
The number of P(III) systems containing one, or more,
perfluorinated substituents (denoted by Rf) is significantly smaller
than for their perprotio analogues and this is often because of
either limitations in synthetic methodology, or available starting
materials. However, where such compounds exist they can show
an unusual combination of steric and electronic properties [2].
Typically, when compared with their non-fluorinated analogue,
fluorine-containing phosphines possess a greater steric demand
(due to the larger size of fluorine compared with hydrogen) whilst
Typical values range from ca. 908, for phosphines with small steric
demand such as PH3, to more than 1808 for P(t-Bu)3 and P(o-tolyl)3.
An alternative measure of the size of a phosphine, called the S40
parameter, is the difference between the sum of the C–P–C angles
between substituents and the sum of the M–P–C angles [5]. This
second measure has the advantage of being easily derived from
crystallographic or computational data, but counter-intuitively
bulky phosphines possess small S04 values.
The electronic properties of phosphines are most frequently
described based on the CO stretching frequency value of a
phosphine-substituted transition metal carbonyl complex. Origin-
ally Tolman’s electronic parameter (TEP) was based on the A10
symmetry
complexes (P = P(III) ligand) [4]. This is a single measure of the
combined and electronic effect of the phosphine towards the
nickel centre [6]; the better the donor the lower the (CO) vibration
as a result of increased back-donation into CO * orbitals, Table 1.
n(CO) stretching frequency of a series of [Ni(CO)3P]
s
p
n
p
However, because of the toxicity and hence difficulty in handling
[Ni(CO)4] a number of alternative metal complexes have been
used, including those of the type [Mo(CO)5P] [7] and [Rh(CO)ClP2]
[8]. Comparison of the values obtained from one complex with
those measured for another show similar trends (Fig. 1 (a) and (b)),
and the values can be empirically inter-related.
electronically the phosphorus centre becomes a poorer
s-donor
and a better -acceptor. Indeed, PF3 and tris(perfluoroalkyl)pho-
p
sphines, have been considered as bulky mimics of the CO ligand [3].
The most frequently used measure of the size of a phosphine
ligand is the solid cone angle that completely encapsulates the
TEP ¼ 1:116
n
ðCOÞMo ꢀ 243 cmꢀ1
ðCOÞRh þ 1621 cmꢀ1
(1.1)
(1.2)
* Corresponding author. Tel.: +44 0 161 306 4459.
TEP ¼ 0:226
n
E-mail address: alan.brisdon@manchester.ac.uk (A.K. Brisdon).
0022-1139/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2009.08.003

1118
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
Table 1
Steric and electronic properties of some monodentate P(III) compounds.a.
Phosphineb
Cone
n
(CO) (cmꢀ1),
n
(CO) (cmꢀ1),
n
(CO) (cmꢀ1),
1J(SeP)
R3P5Se
1J(PtP)
1J(RhP)
angle (8)
Ni(CO)3Pc
Mo(CO)5P
RhCOClP2
PtCl2P2
RhCOClP2
1
2
PH3
87
118
132
160
132
182
145
165
170
194
145
166
122
136
136
140
162
153
145
2083.2
2064.1
2061.7
2059.2
2060.3
2056.1
2068.9
2066.4
2056.4
2066.6
2066.7
2061.9
2065.3
2067.0
2063.7
2066.7
2060.6
2064.8
2068.2
2073
1961
1967
1958
1950
1955
1955
1979
1970
1943
1974
1976
1964
1968
1974
1964
1973
1964
1966
1977
PMe3
2071
2069
684
686
2379
2394
2413
2384
114.7
116.1
3
PEt3
4
Pi-Pr3
5
Pn-Bu3
2070
2073
689
712
731
115.9
126.9
6
Pt-Bu3
7
PPh3
2637
2397
8
PBz3
9
PCy3
673
706
720
805
710
725
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
P(o-Tol)3
P(p-Tol)3
P(NMe2)3
PMe2Ph
PMePh2
PEt2Ph
2066
2070
2071
2073
117.7
122.4
120.4
123.3
PEtPh2
2070
2482
5694
4828
PCy2Ph
PCyPh2
726
726
PPh2(p-Tol)
P(CCMe)Ph2
P(OMe)3
P(OEt)3
2075
2080
2078
107
109
130
128
132
133
139
116
137
124
152
165
169
161
163
2079.5
2076.3
2075.9
2085.3
2072.0
2071.6
2074.6
2074.2
2107.0
2080.9
2069
2014
2016
954
193.9
P(Oi-Pr)3
P(OPh)3
PPh2(OMe)
PPh2(OEt)
PPh2(OPh)
PPh(OEt)2
P(CF3)3
2083
810
796
2104
2086
PMe2CF3
PEt2pfv
1981
1981
1979
2002
1989
730
766
128.4
131.1
130.6
142.3
132.5
Pi-Pr2pfv
PCy2pfv
PPhpfv2
PPh2pfv
PPh2cdfv
PPhcdfv2
Pi-Pr2cdfv
Pi-Pr2tfp
Pi-Prtfp2
Pt-Butfp2
PPh2tfp
2069
2673
2654
3039
2069
2073.1
2084
2077
2078
2082
2074
2077
2085
2079
2077
2081
2078
2078
848
785
2073
2076
2794
2644
2081
173
156
2072
2075
2084
2077
3068
2075
PPhtfp2
2079
P(p-F-Ph)3
P(p-Cl-Ph)3
PPh(C6F5)2
PPh2(C6F5)
P(C6F5)3
P(C6H3CF3-3,5)3
PPh2C2F5
PPh(C2F5)2
PPh2(i-C3F7)
P(C6H4CF3-4)3
PPh2Cl
145
145
171
158
184
160
160
2071.3
2072.8
2082.8
2075.9
2090.9
2073
1982
1984
2002
1982
2008
2000
741
750
136.4
133
2080
3145
2798
2945
4120
2986
155.3
134.7
2078
2080
2089
2081
2088
161
145
138
131
124
104
2079
828
2071
1990
1993
2080.7
2092.1
2102.0
2110.8
867
935
PPhCl2
PCl3
2095
2103
PF3
a
Data taken from Refs. [1,4,7–11].
b
c
pfv: perfluorovinyl (CF5CF2), cdfv: chlorodifluorovinyl (CCl5CF2), tfp: trifluoropropynyl (CBB CCF3).
Values in italics are the average values calculated using Eqs. (1.1) and (1.2).
The TEP, and related parameters, also provide a measure from
complexes have been used previously [10,11], and correlations
between these and TEP exist.
which the electronic contribution due to each R-group of a
phosphine of type PR1R2R3 can be derived, using the formula
Since the early work of Chatt and Wilkins [12] and Tolman [4,9]
rationalising the properties of phosphines there have been many
attempts to model and predict the behaviour of these systems.
These include attempts to extract information about the stereo-
electronic factors that influence the thermodynamics and kinetics
of reactions by quantitative analysis of ligand effects (abbreviated
as QALE) [13] and attempts to generate a phosphine ligand
knowledge database [14]. The approaches used, and their validity
n
(CO) = 2056.1 +
groups, such as alkyl groups have small
(CH3) = 2.6), whilst electron-withdrawing groups have larger
values (e.g. (F) = 18.2, (CF3) = 19.6). Alternatively, coupling
x(R1) +
x(R2) +
x
(R3) cmꢀ1 [9]. Electron-donating
(R) values (e.g.
x
x
x
x
constant information derived from NMR studies may be used as
a gauge of the electronic properties of a P(III) ligand. For example
1J(SeP) from phosphine selenides and 1J(RhP) from Vaska-type

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1119
Fig. 2. A stereoelectronic map for monodentate P(III) ligands.
have the potential to provide P(III) ligands with properties which
are distinct from both perprotio-phosphines and phosphites [18].
To date the majority of known PR3ꢀn(Rf)n (n = 1–3) systems contain
fluoroaryl substituents, an area which has recently been reviewed
[19]. A second family of ligands that have been studied in some
detail are those that possess long fluorinated chains, such as
CH2CH2(CF2)nCF3 (n = 5, 7), for use in fluorous-biphase systems
[20]. These ligands typically include a non-fluorinated –C6H4– or –
CH2CH2– ‘‘spacer’’ group in order to insulate the phosphorus centre
from the electronic effects of the fluorocarbon chain and because of
this they resemble non-fluorinated phosphines in both steric and
electronic respects.
The potential that organofluorine-containing P(III) systems
offer to modify the steric demands and electronic properties of
traditional systems, coupled with the synthetic challenges that
exist in this area of chemistry ensures that this continues to be an
actively pursued research area in academia and industry. The
remainder of this paper describes work associated with mono-
meric P(III) systems containing one or more perfluorinated, non-
aromatic group.
2. Results and discussion
2.1. Fluoroalkenyl-containing phosphines
Fig. 1. Correlation between
n(CO) in [Ni(CO)3P] and (a) [Mo(CO)5P] and (b)
[Rh(CO)ClP2] (P = monodentate P(III) ligand), the number above each point refers to
the entry in Table 1.
Fluoroalkenyl-containing phosphines, such as those containing
a perfluorovinyl group, have been known since 1960 [21]. Initially
these compounds were prepared by the low temperature reaction
of an appropriate chlorophosphine with the Grignard reagent
derived from F2C55CFI [22], or LiCF55CF2 derived from CF255CFBr
and MeLi [23]; these two methods accounted for the synthesis of
perfluorovinyl-containing phosphines of the type P(CF55CF2)3,
PR3ꢀn(CF55CF2)n (n = 1,2; R = Ph, NMe2, NEt2, EtO and BuO). More
recent approaches have utilised CF3CH2F (HFC-134a) as the
perfluorovinyl-synthon, as shown in Scheme 1 [24], and have
provided perfluorovinyl-containing phosphines with R = Cl, Et, i-
Pr, Ph, Cy and the P-chiral compound PPhBu(CF55CF2) [25–27]. In
an analogous fashion 1-chloro-2,2-difluorovinyl phosphines can be
prepared from CF3CH2Cl [28].
Shortly after the first reports of perfluorovinyl phosphines a
number of their spectroscopic properties, including NMR [29], IR
[30] and modes of fragmentation in mass spectrometry experi-
ments [31], were published, along with a limited number of
examples of their reactivity. Cowley and Taylor prepared
PCl3ꢀn(CF55CF2)n from the reaction of P(Me2N)3ꢀn(CF55CF2)n with
HCl, and subsequently converted that to F2P(CF55CF2)n by reaction
with SbF3 [32]. Horn and Kolkmann reported that when preparing
tris(perfluorovinyl)phosphine from PCl3 and an excess of CF2CFLi
nucleophilic attack of the resulting P(CF55CF2)3 occurred to give a
has recently been reviewed [15]. What is clear however, from the
experimental data is that when electronic and steric information is
combined (Fig. 2) a number of interesting features become
apparent.
This, and other similar stereoelectronic maps [16], illustrate the
fact that traditional PR3 phosphine ligands (represented by
squares) whilst spanning a wide range of steric demands have a
much more limited set of electronic properties. When combined
with data for phosphites, P(OR)3, phosphinites PR2(OR0), phos-
phonites, PR(OR0)2 (all represented by diamonds) and halo-
containing phosphines (triangles) it becomes clear that the upper
right part of the stereoelectronic map, which represents sterically
demanding, electron-poor P(III) ligand systems is under repre-
sented. This is unfortunate since in some cases these are exactly
the types of ligands that have demonstrated enhanced reactivity
and catalytic properties [17].
The data that exists for organofluorine-substituted phosphine
ligands (represented by inverted triangles) shows that these
systems offer an opportunity to generate ligands that fill this void,
with electronic properties more akin to those of phosphites, but a
significantly larger steric demand. As such fluorinated phosphines

1120
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
Scheme 1.
product which was formulated as (F2C55CF)2PCF55CF–CF55CF2,
whilst the reaction of P(CF55CF2)3 with methyl lithium gave
(F2C55CF)2PCF55CFMe, and with LiPPh2 (F2C55CF)2PCF55CFPPh2 and
P(CF55CFPPh2)3 were reported [33]. However, until 1999 there was
no coordination chemistry of these P(III) compounds published,
and hence little characterisation of their ligand properties. We
[27,28], and subsequently others [34], have developed the
derivatisation and coordination chemistry of these ligands.
The fluorovinyl phosphines are typically non-malodorous, and
generally stable towards water and aerial oxidation, although they
can be oxidised to their P(V) analogues, P(E)R3ꢀn(CF55CF2)n (E = O,
S, Se) with hydrogen peroxide, elemental sulfur and selenium
respectively [25], or by reaction with Me3SiN3 to quantitatively
give Ph2(CF255CF)P55NSiMe3 [35]. Stereospecific replacement of
the fluorine trans-to the phosphorus centre can also be achieved,
for example hydrogen can be introduced by reaction with LiAlH4 or
LiAlH(OBu)3, whilst R0Li reagents (R0 = alkyl) result in the
incorporation of R0 via an addition elimination reaction [26].
In contrast to the general stability of these phosphines in many
of the reactions investigated we found that in studies of the
quaternisation of these systems many of the compounds showed
poor stability. For example, the reaction of PPh2(CF55CF2) with
methyliodide results in the isolation of a solid product for which
neither the NMR spectra or elemental analysis correspond to the
anticipated product, [Ph2P(CF55CF2)Me]+Iꢀ. The 31P{1H} NMR
of doublet of doublets at ꢀ68.7 ppm and a CFH group as an
overlapping doublet of doublet of quartets at ꢀ214.4 ppm. Taken
together these data suggest the presence of a phosphorus-bound –
CFHCF3 group. In a similar fashion the reaction of benzyl bromide
with Ph2P(CF55CF2) resulted in NMR data which was consistent
with [Ph2P(CFHCF3)(CH2Ph)]Br, the identity of which was subse-
quently confirmed in the solid state by X-ray diffraction, the
resulting structure is shown in Fig. 3.
The crystal structure of [Ph2P(CFHCF3)(CH2Ph)]Br shows the
presence of the two phenyl, one benzyl and the CHFCF3 group
coordinated to the phosphorus centre in an approximately
tetrahedral arrangement. The P–C bond lengths to the non-
˚
fluorinated groups are within the range 1.782(3)–1.806(4) A,
whilst the P–CHFCF3 bond length is significantly longer at
˚
˚
1.850(4) A and d(C1–C2) = 1.504(6) A. There are no X-ray data
for a CFHCF3 group bonded to phosphorus within the CCDC
database [36], although carbon and Pd and Pt-based systems have
been characterised, the latter being formed by the oxidative
addition of CF3CFHI to [M(TMEDA)Me2] (M = Pd, Pt) followed by
reductive elimination of MeI [37,38]. The most similar quaternary
phosphine to have been structurally characterised is
[Ph2MePCF2CF2Br]+Iꢀ. MeI in which the average P–Ph bond length
˚
˚
is 1.774(6) A, the P–CF2CF2Br distance is 1.912(8) A, and d(C–C) of
˚
the halogenated fragment is 1.539(4) A [39].
Since the fluorovinyl phosphines are not water sensitive it is
assumed that the anticipated perfluorovinyl-containing phospho-
nium salt is initially formed, and that this cation is sensitive to
nucleophilic attack by adventitious moisture which results in the
generation of some HF. This subsequently adds across the vinyllic
double bond of another perfluorovinyl phosphonium salt to give
the observed –CFHCF3 containing product, in much the same way
that HBr was observed to add across the C55C bond of
Ph2P(O)CF55CF2 [26]. Further support for this interpretation comes
from studies involving other phosphines and organoiodides. The
slow reaction between Ph2P(CF55CF2) and iodobenzene was
followed by NMR spectroscopy. After a period of 4 weeks the
peaks due to the perfluorovinyl-containing phosphine starting
material were reduced in intensity and new mutually coupling
signals were observed at 19.1 ppm in the 31P{1H} spectrum and in
the fluorine NMR spectrum (ꢀ79.9, ꢀ101.4 and ꢀ186.6 ppm)
which confirm the presence of a new perfluorovinyl-containing
phosphorus species. All the NMR data is consistent with
[Ph3P(CF55CF2)]+Iꢀ, but we cannot be unequivocal of the identity
of this species since attempts to isolate it hastened the decom-
position. After leaving to stand for a further 8 days the signals due
to the starting phosphine were no longer apparent and those of the
intermediate species were being replaced by a new signal in both
the fluorine and phosphorus NMR spectra indicative of decom-
position.
spectrum showed
a doublet of quartets (JPF = 64, 3 Hz) at
22.6 ppm, shifted ca. 50 ppm from the starting phosphine whilst
the 19F NMR spectra show the presence of a CF3 group as a doublet
Fig. 3. An ORTEP representation of the solid state structure of
Interestingly, quaternisation of i-Pr2PCF55CF2 with methylio-
dide, according to NMR evidence, proceeds to the anticipated
quaternary phosphonium species, but in this case the product
appears to be more stable. The 31P{1H} NMR spectrum displays a
[Ph2P(CH2Ph)(CHFCF3)]Br. Selected bond lengths: P1-C1 = 1.850(4), P1-
˚
C3 = 1.782(3), P1-C9 = 1.794(3), P1-C15 = 1.806(4), C1-C2 = 1.504(6) A; selected
angles: C1-P1-C3 = 108.76(18), C1-P1-C9 = 106.87(18), C1-P1-C15 = 105.88(18),
C3-P1-C9 = 108.79(17), C3-P1-C15 = 114.42(17), C9-P1-C15 = 111.79(17)8.

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1121
signal at 42 ppm, which is shifted approximately 50 ppm from the
starting material. Even after a week in solution there was no sign of
decomposition, however, attempts to work-up and isolate the
product was unsuccessful due to decomposition. Furthermore,
attempts to quaternise the P(III) compounds containing more than
one perfluorovinyl group, such as PhP(CF55CF2)2 or i-PrP(CF55CF2)2
were completely unsuccessful. These observations are in line with
the expected trend in nucleophilicity of these compounds.
The coordination chemistry of these ligands has been inves-
tigated in
a number of molybdenum, rhodium, palladium,
platinum and gold systems [25,27,28,35]. Some of these complexes
have been crystallographically characterised, and in those cases it
has been possible to estimate the cone angles of the ligands
involved. Such studies suggest that the cone angles for the mono-
substituted perfluorovinyl phosphines range from ca. 1528 for
PEt2(CF55CF2) to 1698 for PCy2(CF55CF2). Interestingly the cone
angle for PhP(CF55CF2)2 (1618) appears to be somewhat smaller
than that of PhP(C6F5)2 (1718) and this is suggested as the reason
why PhP(CF55CF2)2 will form a complex with [Cp*RhCl(m-Cl)]2 but
that PhP(C6F5)2 does not [40]. It was also shown that intramole-
cular dehydrofluorinative coupling of rhodium bound pentafluor-
omethylcyclopentadienyl and perfluorovinyl phosphine ligands
can occur. In these reactions the fluorine cis- to the phosphorus
centre is lost, as HF, as coupling to one of the CH3 groups of the Cp*
ligand occurs. The reaction is not, however, clean and by-products
containing CF3CHF-phosphines, similar to those observed in the
quaternisation reactions are also detected.
Fig. 4. ORTEP representation of the solid state structure of As(CF5CF2)3. Selected
bond lengths: As1-C1a = 1.922(4), As1-C1b = 1.934(3), As1-C1c = 1.938(4), C1a-
˚
C2a = 1.301(5), C1a-F1a = 1.349(4) C2a-F2a = 1.314(4), C2a-F3a = 1.316(4) A;
selected angles: C1a-As1-C1b = 97.57(17), C1a-As1-C1c = 96.41(15), C1b-As1-
C1c = 96.59(15), As1-C1a-C2a = 122.3(3), C2a-C1a-F1a = 116.5(3), C1a-C2a-
F2a = 126.3, C1a-C2a-F3a = 123.6(3), F2a-C2a-F3a = 110.18.
In order to assess the electronic properties of these ligands two
approaches have been used; the first was based on 1J(PX) coupling
constants in P(Se)Rn(CF55CF2)m compounds (vide supra) or plati-
num or rhodium metal complexes whilst the second was based on
straightforward way of generating a series of P(III) ligands with
electronic properties which may be systematically varied by
changing n and R. Their electronic properties lie between those of
traditional phosphines and phosphites according to measures of
their carbonyl stretching frequencies in metal complexes. It is also
possible to tailor the steric demand of these systems from ca. 150–
1708 by modification of the ancillary R-group. The detailed
parameters are given in Table 1, and a summary of the reactivity
of perfluorovinyl-containing phosphines, exemplified by
Ph2P(CF55CF2), is presented in Scheme 2.
There have been a limited number of studies of more extended
perfluorinated alkenyl P(III) systems, to date these have mainly
been of the E- and Z-isomers of R2P(CF55CFCF3) (R = Me, Et, t-Bu, Ph,
MeO, EtO, Et2N) which have been prepared either from the
Grignard reagent derived from CFI55CFCF3 or by reaction of a
F2C55CFCF3 with R2PH [42]. There are, as far as we are aware, no
published reports of the coordination chemistry of these P(III)
ligands.
the n(CO) stretching frequencies of [Mo(CO)5P] and [RhP2(CO)Cl]
(P = monodentate P(III) ligand) complexes. Comparison of the
carbonyl stretching frequencies obtained from the rhodium
complexes demonstrate that they are poorer donors than non-
fluorinated phosphines, e.g. for the complex of PhP(CF55CF2)2
n
(CO) = 2002 cmꢀ1 which compares with 1979 cmꢀ1 for PPh3,
1996 cmꢀ1 for PhP(C6F5)2 [40], and 2014 cmꢀ1 for P(OMe)3,
Table 1. Thus, whilst the perfluorovinyl group appears to have a
smaller steric demand than –C6F5, it appears to possess a
comparable electron-withdrawing effect, and this is reflected in
the
x
values for these groups both of which are ca. 10, which is
(OEt) = 6.8.
larger than that found for many OR groups, e.g.
x
Whilst the tris-substituted compounds E(CF55CF2)3 (E = P, As,
Sb) have been prepared before [21], none have been structurally
characterised, or assessed as ligands. We were able to prepare all of
these materials in good to moderate yields, but we were
completely unsuccessful in getting them to coordinate to metal
centres. We were equally unsuccessful in obtaining solid state
structures of P(CF55CF2)3 and Sb(CF55CF2)3, both of which froze as
glassy solids. However, As(CF55CF2)3 could be frozen in situ on the
diffractometer within a capillary tube to give a crystalline solid,
and the data collected were solved to give the representation of the
solid state structure shown in Fig. 4. The three As–C distances lie
2.2. Fluoroalkynyl-containing phosphines
The simplest fluoroalkynyl-containing P(III) system should be
those based on –CBB C–F, however, there are no published reports of
these compounds, which is not surprising considering the known
instability of fluoroalkynes. Indeed, the only known perfluoroalk-
ynyl-containing P(III) compounds are those of –CBB C–CF3 and –
CBB C–C6F5 [43], although a greater range of P(V) species of the type
(RO)2P(O)(CBB C–Rf) (R = Me, Et, i-Pr, Ph; Rf = CF3, C2F5, C3F7, CF2Cl
and CF2H) have been prepared and their applications and
conversions, for example to fluoroalkylated vinylphosphonates,
have been studied [44].
A number of synthetic routes to Li(tfp) (tfp = CCCF3) exist based
on a variety of starting materials. In 1952 Henne and Nager
reported the preparation of the Grignard reagent EtMg(tfp),
derived from the metathesis of F3CCCH and EtMgBr [45], whilst
in 1971 Carty et al. obtained the lithium alkynyl reagent by
deprotonation of 3,3,3-trifluoropropyne [46]. More recently,
because of difficulties in obtaining CF3CCH a number of alternative
routes have been proposed. These include the dechlorination of
˚
within the range 1.922(4)–1.938(4) A, whilst the C–F distances
show somewhat greater variation, ranging from 1.313(4) to
˚
1.353(4) A and the average C–As–C bond angle is 96.86(16)8. This
constitutes the first structurally characterised perfluorovinyl-
containing arsenic compound, and the first example of a neutral
homoleptic perfluorovinyl compound of any p-block element.
Moreover, the cone angle of this compound is calculated as 1898,
which suggests that the corresponding upper limit for P(CF55CF2)3
is 1878, based on a previous comparison of values observed for a
series of PR3 and AsR3 compounds [41].
Thus fluoroalkenyl-containing phosphines of the type
PR3ꢀn(CF55CF2)n (n = 1–3) derived from HFC-134a provide
a

1122
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
Scheme 2.
CF3CCl55CCl2 with zinc dust in DMF at 100 8C to give F3CCCZnCl
[47], and the generation of Li(tfp) from F3CCBr55CH2 treated with
two equivalents of LDA at ꢀ78 8C [48]. In 2002 we reported that the
commercially available hydrofluorocarbon CF3CH2CHF2 (HFC-
245fa), on treatment with three equivalents of n-butyl lithium,
results in LiCBB CCF3 (Scheme 3) [49], whilst more recently Shimizu
et al. published the reaction of 1,1,1-trifluoro-3,3-dichloroacetone
with tosyl chloride and triethylamine in DCM to give the enol
tosylate which on treatment with two equivalents of n-BuLi
generates the same lithium reagent [50].
Despite the existence of a number of synthetic routes to Li(tfp)
the only trifluoropropenyl-containing phosphorus(III) compounds
that have been reported in the literature are Ph2P(tfp),
(Et2N)2P(tfp), PhP(tfp)2 and P(tfp)3, none of which have been
structurally characterised either in isolation or when acting as
It is because of their bifunctional nature and the versatility they
offer in synthesis and coordination chemistry that alkynyl-
containing phosphines have attracted attention [55]. The recent
application of such systems includes the generation of hetero-
cycles, proposed catalysts for Heck–Mizoroki carbon–carbon
cross-coupling reactions [56] and the activation of C–H bonds
[57]. There has also been theoretical studies of the electronic
structures of
a series of phosphoralkynyls, including those
containing the P–CBB CCF3 fragment [58].
We were interested to see whether the electron-withdrawing
effect of the remote CF3 group would be transmitted via the CBB C
bond to the phosphorus centre and so offer a relatively small, but
electronically-poor substituent for phosphines.
A number of
trifluoropropynyl-containing phosphines were prepared using the
method outlined in Scheme 3. Subsequent reaction of these ligands
with [Mo(CO)6] afforded complexes of the type [Mo(CO)5P] from
which the CO stretching frequency was obtained. Interestingly, the
data obtained suggests that despite the distance of the CF3 group
from the phosphorus centre the total electronic effects of the ligand
are nearly the same as that derived from the perfluorovinyl group.
simple
s-donor ligands towards metal centres. Indeed when
coordination studies have been undertaken, with Co, Rh, Pd and Pt
metal centres, the phosphines were found to coordinate as
bidentate donors. The phosphine would act as a
s donor to one
metal centre and as a donor to a second metal [51]. Subsequent
p
reactions of the metal-bound P(III) systems with amines [52],
alcohols [53] and secondary phosphines [54] have been investi-
gated, all of which resulted in coupling of the phosphine ligands
with the incoming reagent.
For example, n(CO) of [Mo(CO)5{Ph2PCBB CCF3}] is observed at
2077 cmꢀ1, the same value as that found for the Ph2PCF55CF2
analogue, whilst for [Mo(CO)5{PhP(CBB CCF3)2}] and [Mo(CO)5
{PhP(CF55CF2)2}] the corresponding values are 2085 and
Scheme 3.

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1123
2084 cmꢀ1. However, not surprisingly, the trifluoropropynyl-con-
taining ligand possesses a smaller steric demand, the cone angle of i-
Pr2PCBB CCF3 is estimated to be 1568, bigger than that of i-Pr3P, but 98
smallerthanthatofperfluorovinyl-containinganalogueat1658[59].
The trifluoropropynyl-containing phosphines are found to be
somewhat more reactive than the perfluorovinyl analogues. Thus,
unlike for the perfluorovinyl P(III) systems slow decomposition at
room temperature is observed for the –CBB CCF3 phosphines. One
other area where the greater reactivity of these compounds
becomes obvious is on reaction with alkyl lithium reagents.
Reaction of this unusual P(III) ligand with [PtI2(COD)] resulted
in the formation of a light orange-coloured solid, the phosphorus
NMR spectrum of which exhibits a singlet at
d 24.3 ppm, ca.
30 ppm more positive than that of the uncoordinated phosphine,
with platinum satellites (195Pt, I = 1/2, 33%), 1J(PtP) = 3453 Hz.
Recrystallization from dichloromethane–hexane resulted in crys-
tals suitable for X-ray diffraction, and the solid-state structure of
the product, [PtI2{Pi-Pr2(C3F2t-Bu)}2], is shown in Fig. 5. It
confirms, as expected from the magnitude of the platinum–
phosphorus coupling constant, a trans-arrangement of the ligands
around a square-planar platinum centre, with the metal siting on a
crystallographic centre of symmetry. The bond lengths and angles
observed in this complex are largely unexceptional; the Pt–P bond
Previously we have shown that reaction of
a main-group
trifluoropropynyl compound with t-BuLi results in two possible
products after work-up, one of which corresponds to an unusual
cyclisation process [60]. The reaction of t-BuLi with i-Pr2PCBB CCF3
dissolved in ether was studied at a series of different temperatures.
Two phosphorus-containing products were evident in the 31P{1H}
NMR spectra, the proportions of which changed considerably with
temperature. When the reaction is carried out at ꢀ20 8C, after
work-up, the predominant product is characterised as i-Pr2P(Z-
CH55CF3(t-Bu)), which arises from the syn-addition of t-BuLi across
the triple bond with the direction of addition being consistent with
both the sterics and electronics of the precursor. But on increasing
the temperature the proportion of the second product increased,
until by 40 8C it constituted 90% of the isolated material (Scheme
4). The identity of this second product was initially inferred
from multinuclear NMR data; the fluorine NMR spectrum
displayed a doublet (J(PF) = 4 Hz) at ꢀ103.3 ppm whilst the
13C{1H} data showed triplets due to coupling with two fluorine
nuclei at 106.2 (J(CF) = 273 Hz), 122.7 (J(CF) = 14 Hz) and
149.2 ppm (J(CF) = 10 Hz). These data are consistent with those
previously observed for the gem-difluorocyclopropene unit
obtained in a similar reaction on silicon based systems [60].
Because this cyclisation reaction is not observed for less sterically
demanding RLi reagents we have suggested that steric congestion
which resulted following the addition of t-BuLi across the triple
bond resulted in elimination of LiF and concomitant cyclisation. An
alternative explanation might be that addition of t-BuLi is followed
by LiCF3 elimination which decomposes to generate LiF and
difluorocarbene which then adds back across the triple bond. We
have investigated this reaction in a variety of solvents and over a
range of temperatures in an attempt to provide additional
information on this unusual process. However, even when a good
difluorocarbene acceptor, such as cyclohexene, is used as solvent
we have been unable to detect any product from the addition of
difluorocarbene to the cyclohexene. We therefore have no
evidence against an intramolecular process, but equally we
concede that we cannot distinguish this from a process where
LiCF3 is produced, decomposes and recombines all within the
primary coordination sphere.
˚
˚
length is 2.3201(9) A, which is comparable to of 2.318(2) A found
for trans-[PtI2(PPh3)2] [61]. From the data we can determine a cone
angle value of 1718, which makes this one of the bulkiest
fluorocarbon P(III) ligands we have yet prepared. This is the first
structurally characterised example of any phosphorus ligand
containing a cyclopropene unit, however, Chen and co-workers
have published the structures of two phenyl-substituted gem-
difluorocyclopropene systems [62,63].
2.3. Fluoroalkyl-containing phosphines
The third class of monodentate organofluorine P(III) com-
pounds – those containing phosphorus-bound perfluoroalkyl
groups – is largely restricted to –CF3 and –CF2CF3 containing
compounds. This is because of the limited number of suitable
perfluoroalkyl-delivery reagents; perfluoroalkyl Grignard and
lithium reagents are thermally unstable (e.g. LiC2F5 decomposes
above ꢀ70 8C) or inaccessible (e.g. LiCF3). A variety of specialised
methods have therefore been developed, those that have some
generic capabilities include: direct fluorination of PR3 with F2
(followed by reduction of the resulting P(V) P(Rf)3F2 species) [64]
and electrochemical fluorination of PR3 in anhydrous HF [65]. For
trifluoromethyl-containing phosphines the principal historical
methods have relied on the reaction of perfluoroalkyl transfer
agents, such as Hg(CF3)2, Cd(CF3)2.DME or Te(CF3)2 with PI3 [66] or
heating P4 with RfI [67]. Perfluoroethyl-containing phosphines are
usually prepared from the reaction of an appropriate chloropho-
sphine with LiC2F5 at low temperature [68].
Because of the limitations in Rꢀf delivery reagents a number of
alternative methods and starting materials have been investigated.
One of the simplest and readily available starting materials to use
would be the commercially available perfluoroalkyl iodides.
However, unlike their perprotio analogues, these tend to be
resistant to SN1 and SN2 reactions due to a combination of the
d
d
polarization of the Rf–I (C ꢀ–I +) bond and steric and lone-pair
repulsion from the fluorine substituents. Fortunately, for soft
Scheme 4.

1124
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
Fig. 5. An ORTEP representation of the solid state structure of trans-[PtI2{Pi-Pr2(C2F3t-Bu)}2]. Selected bond lengths: Pt1-I1 = 2.6191(3), Pt1-P1 = 2.3201(9), P1-C1 = 1.799(4),
˚
P1-C8 = 1.848(4), P1-C11 = 1.841(4), C2-C3 = 1.425(5), C2-F1 = 1.365(4), C2-F2 = 1.376(4) A; selected angles: I1-Pt1-P1 = 89.53(2), Pt1-P1-C1 = 115.89(13), Pt1-P1-
C8 = 117.50(13), Pt1-P1-C11 = 113.80(12), C1-P1-C8 = 99.37(17), C1-P1-C11 = 102.14(17), C8-P1-C11 = 106.09(18), 114.42(17), P1-C1-C2 = 147.0(3), C1-C2-C3 = 55.2(2),
C1-C3-C2 = 63.3(3), C1-C2-F1 = 122.7(3), C1-C2-F2 = 122.6(3)8.
nucleophiles, such as those of phosphorus, the SRN1 radical–anion
chain mechanism is possible. Indeed, Vaillard et al. proposed this
for the reaction of Ph2PNa with n-C4F9I or n-C6F13I in HMPA under
photolytic conditions, which yielded, after oxidation, Ph2P(O)(n-
C4F9) and Ph2P(O)(n-C6F13) in 55–85% yields [69].
trifluoromethyl-containing starting materials such as Me3SiCF3.
Whilst the first of these methods provides a generic route to P(Rf)3
compounds it relies on the availability of Me3SiRf reagents, of
which only
a few are commercially available. The second
procedure offers the possibility of generating a family of PR2(CF3)
compounds, but is equally difficult to extend to other fluoroalkyl
groups. One approach that has been successful for the synthesis of
fluoroaryl-containing phosphines involves the reaction of tri-
methylsilyl-containing phosphanes with polyfluoroarenes [72]. In
that reaction, and a similar one reported by Oshima and co-
workers on tertiary organofluorides [73], a C–F bond is replaced by
a C–P bond and Me3SiF is generated. A similar approach has been
employed for the generation of some perprotio-phosphines and,
under transition metal-catalysed conditions, Ph2P(CF2)nBr and
Ph2P(CF2)nPPh2 from Ph2PSiMe3 and Br(CF2)nBr (n = 1 or 2]) [74].
We were surprised to find that this strategy does not appear to
have been applied to the synthesis of perfluoroalkyl-containing
phosphanes.
We have also investigated this approach as a method of
preparing fluoroalkyl-containing phosphines with the emphasis on
sterically demanding Rf groups. The addition of i-C3F7I to a THF
solution of Ph2PLi (prepared from Ph2PCl and lithium wire) results
in an almost immediate loss of the colour associated with Ph2PLi.
Subsequent addition of hexane followed by filtration, to remove
lithium halides, resulted in a low melting point solid, the 31P{1H}
and 19F NMR spectra of which are shown in Fig. 6. The phosphorus
NMR spectrum exhibits an overlapping doublet of septets, and the
fluorine NMR spectrum shows a doublet of doublets and a doublet
of septets in the ratio of 6:1. These spectra are consistent with the
anticipated A6MX spin-system of Ph2P(i-C3F7), and was confirmed
as the identity of the product subsequently by elemental analysis
and finally a single-crystal X-ray structure [75].
The reaction of R2PSiMe3 with i-C3F7I at ꢀ20 8C in hexane, after
work-up, resulted in white-coloured solid material, the NMR
spectra of which showed the same 31P{1H} and 19F signals as those
observed for the product of the reaction between i-C3F7I with
Ph2PLi. Significantly, the reaction is quicker and the product,
Ph2P(i-C3F7), could be isolated in yields of more than 70%.
Extending this work to a wide range of RfI compounds (Rf = CF3,
C2F5, i-C3F7, s-C4F9, t-C4F9, c-C6F11) allowed for the synthesis of a
number of R2PRf analogues, as shown in Table 2 [75]. The
spectroscopic data for Ph2PCF3 [70] and Ph2PCF2CF3 [68] agrees
with those previously reported. We found that the reactions with
primary, short-chain perfluoroalkyl groups were significantly
slower than the reactions with the secondary and tertiary iodides,
whilst reactions with perfluoroaryliodides and perfluorovinylio-
dides were unsuccessful after 5 days. We have subsequently
investigated a number of variations of the procedure, such as the
addition of a fluoride catalyst, replacing the silyl phosphine with
Ph2PH and using perfluoroalkyl chlorides or bromides, instead of
The preparation of lithium phosphides is straightforward either
from the reaction of chloro- or dichloro-phosphines with lithium
wire in THF, or by cleavage of a P–Ar bond of an aryl-phosphine. By
reaction of these phosphides with a number of RfI compounds it is
potentially possible to generate a fairly large library of perfluor-
oalkyl-containing phosphines. The reaction of Ph2PLi and i-Pr2PLi
with a variety of RfI compounds was undertaken, including CF3I,
C2F5I, i-C3F7I, s-C4F9I, t-C4F9I and c-C6F11I; the results of these
reactions are summarised in Table 2. Unfortunately, the yields of
fluoroalkylphosphines prepared in this way tended to be low and
so alternative routes have been investigated (Scheme 5).
In 2005 Caffyn and co-workers reported that P(Rf)3 compounds
can be obtained from the reaction of Me3SiRf with P(OR)3 (R = Ph,
p-C6H4CN) [70]. Subsequently Togni and co-workers synthesised
CF3-containing phosphines from the reaction of R2PH or Ph2PSiMe3
and CF3-containing hypervalent iodine compounds [71], Scheme 5.
Both of these methods benefit from the commercial availability of

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1125
Fig. 6. (a) 31P{1H} and (b) 19F NMR spectra of Ph2P(i-C3F7) in CDCl3 solution, with expansions.
iodides. However, none of these modifications were found to offer
any advantage.
We have undertaken some studies of the reactivity and
coordinationchemistryof theseligandscontainingbulky, fluoroalkyl
groups. Attempts to quaternise some of these P(III) compounds with
benzyl bromide were unsuccessful, presumably because of the
reducednucleophilicityofthephosphoruscentreduetothepresence
oftheelectron-withdrawingi-C3F7 group. However, oxidationtogive
the P(V) compound occurs readily; thus reaction of Ph2P(i-C3F7) with
elemental selenium in toluene generates the selenide, from which
the magnitude of the 1J(SeP) coupling constant (828 Hz) can be
obtained. This value is considerably higher than the values obtained
for non-fluorinated phosphines, which typically lie in the range 670–
730 Hz (Table 1) but is comparable with those for phosphinites (e.g.
Ph2P(OMe) 1J(SeP) = 810 Hz) and between that measured for
Ph2P(CF55CF2) (785 Hz) and PhP(CF55CF2)2 (848 Hz) [76].
Table 2
Selected perfluoroalkyl-containing phosphines prepared from the reaction of RfI
with silyl phosphines and lithium phosphides.
RfI
Product
31P{1H} NMR
Method (yield)
(
d
, ppm)a
CF3I
Ph2PCF3
2.5
Ph2SiMe3 (75%)
C2F5I
Ph2PC2F5
ꢀ1.9
Ph2SiMe3 (80%) Ph2PLi (5%)
We were able to obtain an estimate of the steric demand of
Ph2P(i-C3F7) from its molecular structure which, as expected,
exhibits a pyramidal geometry at the phosphorus centre. The three
i-Pr2P(C2F5)
24.5
i-Pr2SiMe3 (52%)
C8F17
I
Ph2P(C8F17
)
1.1
Ph2SiMe3 (85%)
˚
P–C bond lengths are 1.828(5), 1.831(5) and 1.899(5) A, with the
i-C3F7I
Ph2P(i-C3F7)
i-Pr2P(i-C3F7)
ꢀ0.8
Ph2SiMe3 (75%) Ph2PLi (40%)
i-Pr2PLi (32%)
longer distance being that to the fluorinated group. There are four
other solid state structures of phosphines containing one, or more,
perfluorinated groups, they are PPh2(C2F5) [68], PPh(C2F3)2 [25],
PPh2(C6F5) [77] and P(C6F5)3 [78]. The P–C bond distances are
similar in all five of these phosphines, but Ph2P(i-C3F7) appears to
be the most sterically demanding mono-substituted phosphine
with the sum of the angles around the phosphorus centre being
309.68, which is larger than that observed in PPh2(C2F5) (304.28),
41.7
PhMeP(i-C3F7) ꢀ12.0
PhMeSiMe3 (65%) PhMePLi (36%)
s-C4F9I
c-C6F11
t-C4F9I
Ph2P(s-C4F9)
3.7
ꢀ3.4
15.2
Ph2SiMe3 (56%)
Ph2SiMe3 (48%)
Ph2SiMe3 (30%)
I
Ph2P(c-C6F11)
Ph2P(t-C4F9)
CF25CFI
–
–
C6F5I
a
Data taken from Ref. [75].

1126
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
Scheme 5.
PPh(C2F3)2 (299.68) and PPh2(C6F5) (306.68) and very similar to that
of the tris-substituted P(C6F5)3 (309.98). Furthermore, we have
recently obtained the first crystal structure of a metal complex of
this ligand, which also suggests that the cone angle is slightly
larger than PPh2(C2F5) in a comparable complex [76].
CF3CH2CHF2 by Honeywell. CF3I, C2F5I, i-C3F7I, s-C4F9I, t-C4F9I,
cyclo-C6F11I (all Apollo Scientific), Li (3.2 mm diameter wire), n-
BuLi (10 M or 2.5 M in hexanes), t-BuLi (1.5 M in pentanes), MeLi
(1.6 M in Et2O) (Acros or Aldrich), Ph2PCl, i-Pr2PCl, PPh3, AsCl3,
SbCl3, Me3SiCl and CDCl3 were purchased from commercial
vendors and used as supplied. Perfluorovinyl-containing phos-
phines were prepared as described in the literature [25,27].
Ph2P(SiMe3) and i-Pr2P(SiMe3) were prepared based on literature
methods [79,80]. NMR spectra were recorded on Bruker spectro-
meters operating at 200, 300 or 400 MHz for 1H and referenced
against TMS (1H and 13C), CFCl3 (19F) and 85% H3PO4 (31P). All NMR
data are reported using the high frequency positive convention.
Infrared spectra were recorded using KBr plates on a Nicolet Nexus
FTIR/Raman spectrometer. Elemental analyses were performed by
the departmental microanalytical service.
3. Conclusions
There is considerable interest in the synthesis and potential
applications for bulky, electron-poor phosphines. P(III) systems
containing fluorinated groups offer one solution to this problem,
particularly in systems in which it is possible to vary the identity of
the fluorinated and non-fluorinated groups which allows for the
steric demand and electronic properties of the ligands to be
modified in a systematic fashion. Methods exist by which such
variation can be introduced into fluoroaryl, fluoroalkenyl and
fluoroalkynyl based systems using Grignard or lithium reagents of
suitable precursors. The replacement of CFCs with HFCs has
provided a number of new fluorocarbon building blocks which can
be used to readily generate fluoroalkenyl- and fluoroalkynyl-
containing phosphines. Such systems are shown to possess steric
and electronic properties that place them between traditional
phosphines and phosphites.
The reaction of trimethysilyl-containing phosphines with
perfluoroalkyliodides offers a convenient route by which the a
similar variation may be applied to perfluoroalkyl-containing P(III)
compounds. This general synthetic method has been successfully
used to introduce secondary, tertiary and perfluorocyclohexyl
groups, and as such provides, in many cases for the first time,
access to a new range of sterically demanding, electron-poor
fluorinated P(III) ligands. We are continuing to investigate the
scope of these findings as well as the properties, coordination
chemistry and application of these new ligands.
4.1. X-ray crystallography
X-ray data for [Ph2P(CH2Ph)(CHFCF3)]Br and trans-[PtI2{Pi-
Pr2(c-C2F3t-Bu)}2] were recorded on a Nonius
diffractometer using graphite-monochromated Mo K
˚
(l = 0.71073 A). The data for all structures were solved by direct
k
-CCD 4-circle
radiation
a
methods and subjected to full-matrix least-squares refinement on
F2 using the SHELX-97 program [81]. All non-hydrogen atoms were
refined with anisotropic thermal parameters.
A crystal of
As(CF55CF2)3 was obtained from a sample held in a Pyrex capillary
(o.d. 0.38 mm) mounted on a Sto¨e Stadi-4 diffractometer equipped
with an Oxford Cryosystems low temperature device. Crystal-
lization was achieved by first establishing a stable solid–liquid
equilibrium at 260 K and then cooling the sample at a rate of
10 K hꢀ1. A data set comprising a full sphere of data to 2
collected at 110 K; an absorption correction was applied using
u = 608 was
c
-
scans. The structure was solved by placing the As atom at the origin
and locating the C and F atoms in a subsequent difference synthesis
(CRYSTALS). Data collection and refinement parameters are
summarised in Table 3. Molecular representations shown in the
figures were generated using Pluton [82]. Crystallographic data
(excluding structure factors) for the structures in this paper have
been deposited with the Cambridge Crystallographic Data Centre
as supplementary publication nos. CCDC 732138–732140. Copies
of the data can be obtained, free of charge, on application to CCDC,
12 Union Road, Cambridge CB2 1EZ, UK (fax: +44 1223 336033 or
e-mail: deposit@ccdc.cam.ac.uk).
4. Experimental
Reactions were routinely performed under anaerobic condi-
tions in flame-dried glassware, with moisture-sensitive reagents
being handled under an argon atmosphere in a dry box (Belle
Technologies, UK). Diethyl ether and THF were dried over sodium/
benzophenone for ca. 1 day and then freshly distilled prior to use.
Hexane was dried over sodium wire for ca. 1 day and then freshly
distilled prior to use. CF3CH2F was provided by INEOS Fluor and

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1127
Table 3
Crystal data and structure refinement data.
Compound
[Ph2P(CH2PH)(CHFCF3)]Br
21H18BrF4P
As(CF5CF2)3
trans-[PtI2{Pi-Pr2(c-C2F3t-Bu)}2]
Formula
C
C6AsF9
C26H46F4I2P2Pt
945.45
M
457.23
447.46
Crystal system
Space group
Monoclinic
P21/n (no. 14)
Monoclinic
P21/c
Monoclinic
P21/n (no. 14)
Unit cell parameters
˚
a (A)
10.9179(3)
11.5879(3)
16.7960(5)
90
6.6969(8)
7.2004(8)
19.599(3)
90
9.5442(2)
16.8766(3)
11.1179(2)
90
˚
b (A)
˚
c (A)
˚
a
b
g
(A)
(8)
(8)
104.370(1)
90
99.633(10)
90
113.449(1)
90
3
˚
V (A )
Z
1776.12(10)
4
931.7(2)
4
1642.91(6)
2
D (g cmꢀ3
)
a
1.475
2.267
1.911
m
(Mo K
) (mmꢀ1
)
2.112
3.758
6.283
Crystal size (mm)
0.07 ꢁ 0.15 ꢁ 0.25
29248 (2.60–27.00)
293
0.39 ꢁ 0.27 ꢁ 0.27
3258 (3.02–27.52)
120
0.20 ꢁ 0.20 ꢁ 0.20
16055 (3.10–24.90)
150
No. data collected (
Temperature (K)
F(0 0 0)
u range)
920
600
904
Data/restraints/parameters
4096/0/317
0.0456, 0.1176
0.41, ꢀ0.35
2141/0/146
0.0343, 0.0597
0.567, ꢀ0.325
2877/0/253
0.0204, 0.0466
0.80, ꢀ1.02
R1, wR2½I > 2sðIÞꢂ
ꢀ3
˚
Maximum, minimum residual electron density (e A
)
4.1.1. [Ph2P(CFHCF3)Me]I
(188.3 MHz, CDCl3, CCl3F): ꢀ80.1 (dd, 1F, JFF = 29, 43 Hz, CF55CF2),
ꢀ105.6 (ddd, 1F, JFF = 43, 119 JFP = 9 Hz, CF55CF2), ꢀ189.5 (dd, 1F,
JFF = 30, 119 Hz, CF55CF2). 31P NMR (81.8 MHz, CDCl3, 85% H3PO4)
42.0 (d, JPFb = 59 Hz).
PPh2(CF55CF2) (0.40 g, 1.5 mmol) and iodomethane (0.21 g,
1.5 mmol) were stirred at 46 8C for 7 days in hexane (10 cm3). After
cooling the precipitate which had formed was filtered off to give an
off-white product. (0.095 g 15%). Anal. Calcd. for C15H14PF4I: C
42.1%, H 3.3%, I 29.7%. Found: C 43.1%, H 3.21%, I 30.1%. IR (KBr,
4.1.5. As(CF55CF2)3
neat);
TMS):
n
d
1263, 1197, 1149 (C–F) cmꢀ1
130–137 (C6H5), 113.9 (m, CHF), 8.8 (d, JPC = 62 Hz, CH3).
.
13C NMR (75.5 MHz, CDCl3,
To a three-necked round-bottomed flask, equipped with
nitrogen inlet/outlet and magnetic stirrer, maintained at ꢀ78 8C,
were added CF3CH2F (9.3 ml, 110 mmol) and diethyl ether (400 ml)
followed by n-BuLi (in hexanes, 22.0 ml, 10 M, 220 mmol) at a
sufficiently slow rate to ensure that the temperature of the reaction
mixture did not increase beyond ꢀ70 8C. After 6 h of maintaining
this temperature the solution was cooled to ꢀ85 C and AsCl3
(2.3 ml, 27.4 mmol) dissolved in Et2O (40 ml) was added slowly.
After leaving the reaction mixture to warm to room temperature
overnight hexane (450 ml) was added and the solution was
filtered. Removal of the solvents from the filtrate followed by
distillation (58 8C, 80 mmHg) resulted in As(CF55CF2)3 as a clear
liquid (7.84 g 90%). Anal. Calcd. for C6F9As: C 22.6%, F 53.8%, As
19F NMR (188.3 MHz, CDCl3, CCl3F):
d
ꢀ68.7 (ddd, 3F, JFP = 4,
JFaFb = 16, JFaH = 19 Hz, CHFCF3), ꢀ214.4 (ddq, 1F, JFaFb = 16,
JFbH = 42, JFbP = 61 Hz, CHFCF3). 31P NMR (81.8 MHz, CDCl3, 85%
H3PO4) 22.6 (dd, JPFb = 61, JPFa = 3 Hz).
4.1.2. [Ph2P(CFHCF3)CH2Ph]Br
PPh2(CF55CF2) (1.5 g, 5.6 mmol) and benzyl bromide (0.60 g,
3.5 mmol) were stirred at 46 8C for 7 days in THF (10 cm3). After
cooling the precipitate which had formed was filtered off to give a
white solid. (0.85 g 55%). Anal. Calcd. for C21H18PF4Br: C 55.1, H 4.0,
Br 17.5%. Found: C 55.7%, H 3.9%, Br 18.2%. IR (KBr, neat);
(C55C aromatic), 1197, 1024, 993 (C–F) cmꢀ1 13C NMR (75.5 MHz,
CDCl3, TMS):
125.8–136.8 (C6H5), 28.0 (d, JPC = 41 CH2). 19F NMR
(188.3 MHz, CDCl3, CCl3F):
n 1584
.
23.6%; found C 22.3%, F 53.3%. IR (KBr, neat);
n 1710 (C55C), 1299,
d
1165, 998 (C–F) cmꢀ1 13C NMR (75.5 MHz, neat, external CDCl3):
.
d
ꢀ68.8 (ddd, 3F, JFP = 4, JFaH = 6,
124.1 (dd, J = 306, 65 Hz CF55CF2), 158.1 (ddd, J = 325, 283, 35,
CF55CF2). 19F NMR (188.3 MHz, neat external CCl3F): ꢀ87.1 (dd, 1F,
J = 40, 57 Hz), ꢀ114.2 (dd, 1F, J = 57, 120 Hz), ꢀ180.1 (dd, 1F, J = 41,
120 CF55CF2).
JFaFb = 16 Hz, CHFCF3), ꢀ214.4 (ddq, 1F, JFaFb = 16, JFbH = 41,
JFbP = 57 Hz, CHFCF3). 31P NMR (81.8 MHz, CDCl3, 85% H3PO4) 25.2
(dd, JPFb = 60 Hz, JPFa = 3).
4.1.3. [Ph3P(CF55CF2)]I
4.1.6. Sb(CF55CF2)3
PPh2(CF55CF2) (0.40 g, 1.5 mmol) and iodobenzene (0.40 g,
1.9 mmol) in THF (5 cm3) were stirred at 46 8C for a period of
28 days under a positive pressure of dinitrogen and samples were
taken from the reaction mixture periodically for characterisation
by NMR spectroscopy (see text). 19F NMR (188.3 MHz, CDCl3,
CCl3F): ꢀ79.9 (m, 1F, J = 120 Hz, CF55CFF), ꢀ101.4 (m, 1F, J = 120 Hz
CF55CFF), ꢀ186.6 (m, 1F, JFP = 50 Hz, CF55CF2). 31P NMR (81.8 MHz,
CDCl3, 85% H3PO4) 19.6 (dm, JPFb = 49 Hz).
Using a similar method to that described above using CF3CH2F
(9.3 ml, 110.3 mmol) in diethyl ether (100 ml), n-BuLi (in hexanes,
88 ml, 2.5 M, 220 mmol) and SbCl3 (6.3 g, 27.6 mmol) dissolved in
Et2O (100 ml) resulted, after distillation (42 8C, 10 mmHg) in
Sb(CF55CF2)3 as a clear liquid (6.85 g 70%). Anal. Calcd. for C6F9Sb: C
19.7%, F 46.9%, Sb 33.4%. Found: C 19.9%, F 46.5%. IR (KBr, neat);
1710 (C55C), 1299, 1165, 998 (C–F) cmꢀ1 19F NMR (188.3 MHz,
neat external CCl3F):
n
.
d
ꢀ82.1 (dd, 1F, JFF = 48, 57 Hz), ꢀ114.0 (dd,
1F, JFF = 57, 132 Hz), ꢀ181.0 (dd, 1F, JFF = 48, 132 Hz CF55CF2).
4.1.4. [i-Pr2P(CF55CF2)Me]I
i-Pr2P(CF55CF2) (0.40 g, 1.5 mmol) and iodomethane (0.20 g,
1.4 mmol) in hexane (10 cm3) were stirred at 46 8C for a period of
28 days under a positive pressure of dinitrogen and samples were
taken from the reaction mixture periodically (see text). 19F NMR
4.1.7. i-Pr2PCCCF3
To a three-necked round-bottom flask equipped with magnetic
stirrer and nitrogen inlet and outlet held in a slush bath at ꢀ20 8C
was added diethyl ether (150 ml) and HFC-245fa (3.6 ml,

1128
K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
37 mmol). n-BuLi (42 ml, 2.5 M in hexanes, 105 mmol) was added
slowly over a period of 1 h at a sufficiently slow rate to ensure that
the internal temperature did not rise above ꢀ20 8C. The reaction
mixture was maintained at this temperature for a further 1 h. and
then cooled to ꢀ90 8C. A cold, concentrated solution of chlor-
odiisopropylphosphine (3.8 ml 24 mmol, in 20 ml ether) was
slowly added. The reaction was left to stir overnight and the
temperature then raised to 0 8C. Hexane (150 ml) was added and
the solution was filtered to remove the precipitated lithium salts.
The combined solvents were removed under vacuum and after
distillation (39 8C, 10 mmHg) 2.1 g (40%) of i-Pr2PCCCF3 was
collected as a clear liquid. Anal. Calcd. for C9H14F3P: C 51.4%, H
6.7%, F 27.1%, P 14.7%. Found: C 49.8%, H 6.6%, F 27.3%, P 14.5%. IR
69.6 (dd, 6F, 3J(PF) = 18.5, 3J(FF) = 11.8 Hz, CF3), ꢀ184.9 (dsept, 1F,
2J(PF) = 74.0, 3J(PF) = 11.6, CF) 31P{1H} NMR (CDCl3, 162 MHz, 85%
H3PO4): ꢀ0.7 (dsept, 2J(PF) = 73.0, 3J(PF) = 19.0 Hz) [74].
4.1.11. Typical procedure for R2PSiMe3 reaction with RfI
A dried Schlenk vessel maintained under a positive pressure of
nitrogen was charged with Ph2P(SiMe3) (2.2 ml, 8.5 mmol) and
hexane (20 ml). The solution was cooled to ꢀ30 8C and i-C3F7I
(1.2 ml, 8.5 mmol) was added over ca. 30 min. The solution was
stirred and left to warm to room temperature overnight. MeLi
(1.6 M in Et2O, 5.4 ml, 8.6 mmol) was added to the yellow solution
and stirred for ca. 30 min. The resulting precipitate was filtered off
under an inert atmosphere and the volatiles were removed under
vacuum to yield Ph2P(i-C3F7) as a white solid (2.26 g, 75%).
(KBr, neat);
n .
2960 (CH), 2193 (C55C), 1255, 1140 (C–F) cmꢀ1 1H
NMR (200.2 MHz, CDCl3): 1.27 (m, CH3), 2.03 (s, J = 5 Hz, CH),
13C{1H} NMR (75.5 MHz, CDCl3): 114.1 (q, J = 134 Hz CCCF3), 91.2
(dq, J = 6, 39, CCCF3), 89.4 (dq, J = 34, 4, CCCF3), 24.3 (CH), 19.7
(CH3). 19F NMR (188.3 MHz, CDCl3, CFCl3): ꢀ50.7 (d, J PF = 6 Hz,
CF3), 31P{1H} (81.8 MHz, CDCl3, 85% H3PO4) NMR: ꢀ12.4 (q,
JPF = 6 Hz).
Acknowledgements
We thank the Engineering and Physical Sciences Research
Council (EPSRC) and UMIST for financial support, ICI Klea, INEOS
Fluor and Honeywell for the donation of HFCs, Johnson Matthey for
the loan of precious metal salts and Prof. S Parsons for the structure
of As(CF55CF2)3. AKB would particularly like to thank the many
UMIST and University of Manchester undergraduate and post-
graduate students that have contributed to this work over the past
10 years.
4.1.8. i-Pr2P(c-C2F3t-Bu)
In a round-bottom flask fitted with reflux condenser and
maintained under a positive pressure of nitrogen was placed
diethyl ether (15 ml) and i-Pr2PCCCF3 (0.423 g, 2.01 mmol) and the
solution was brought to reflux. t-BuLi (2.7 ml, 1.5 M in pentanes,
4.05 mmol) was added dropwise over 15 min. The solution was
allowed to cool down and methanol (2 ml) was added to quench
unreacted butyl lithium. Hexane (40 ml) was added and the
solution filtered through celite. Removal of the solvents under
reduced vacuum produced a mixture of i-Pr2P(c-C2F3t-Bu) and i-
Pr2P(CH55CF3t-Bu) as a 91:9 mixture according to fluorine NMR
measurements which was used without further purification. IR
References
[1] See for example in: D.W. Allen, J.C. Tebby (Eds.), Organophosphorus Chem. 35
(2006) 1, The Royal Society of Chemistry, Cambridge, UK.
[2] See for example D.M. Roddick, R.C. Schnabel, ACS Symp. Ser. 555 (1994) 421–437.
[3] (a) J. Apel, J.Z. Grobe, Anorg. Allg. Chem. 453 (1979) 53–67;
(b) A.L. Fernandez, M.R. Wilson, A. Prock, W.P. Giering, Organometallics 20 (2001)
3429–3435.
[4] C.A. Tolman, Chem. Rev. 77 (1977) 313–348.
(KBr, neat);
n
2960 (CH), 1746 (C55C) cmꢀ1
.
1H NMR (200.2 MHz,
[5] B.J. Dunne, R.B. Morris, A.G. Orpen, J. Chem. Soc. Dalton Trans. (1991) 653–661.
[6] G.O. Spessard, G.L. Miessler, Organometallic Chemistry, Prentice-Hall, Englewood
Cliffs, NJ, 1997.
CDCl3): 1.26 (m, CH3), 2.07 (s, J = 5 Hz, CH), 13C{1H} NMR
(75.5 MHz, CDCl3): 149.2 (dt, J = 3, 10 Hz PCCCF2), 122.7 (dt,
J = 62, 14, PCCCF2), 106.2 (dt, J = 5, 273, PCCCF2), 25.0 (CH), 19.9
(CH3). 19F NMR (188.3 MHz, CDCl3, CFCl3): ꢀ103.3 (d, JPF = 4 Hz,
CF2), 31P{1H} (81.8 MHz, CDCl3, 85% H3PO4) NMR: ꢀ5.7 (q,
JPF = 4 Hz).
[7] E.C. Alyea, G. Ferguson, A. Somogyvari, Organometallics 2 (1983) 668–674.
[8] A. Roodt, S. Otto, G. Steyl, Coord. Chem. Rev. 245 (2003) 121–137.
[9] C.A. Tolman, J. Am. Chem. Soc. 92 (1970) 2953–2956.
[10] J. Malito, E.C. Alyea, Phos. Sulfur Silicon 54 (1990) 95–99;
W. McFarlane, D.S. Rycroft, J. Chem. Soc. Dalton Trans. (1973) 2162–2166.
[11] J. Tiburcio, S. Bernes, H. Torrens, Polyhedron 25 (2006) 1549–1554;
B.E. Mann, C. Masters, B.L. Shaw, J. Chem. Soc. A (1971) 1104–1106.
[12] J. Chatt, R.G. Wilkins, J. Chem. Soc. (1952) 273–278.
[13] J. Bartholomew, A.L. Fernandez, B.A. Lorsbach, M.R. Wilson, A. Prock, W.P. Giering,
Organometallics 15 (1996) 295–301.
[14] See for example R.A. Mansson, A.H. Welsh, N. Fey, A.G. Orpen, J. Chem. Inf. Model.
46 (2006) 2591–2600.
[15] O. Kuhl, Coord. Chem. Rev. 249 (2005) 693–704.
[16] K.D. Cooney, T.R. Cundari, N.W. Hoffman, K.A. Pittard, M.D. Temple, Y. Zhao, J. Am.
Chem. Soc. 125 (2003) 4318–4324.
4.1.9. trans-[PtI2{Pi-Pr2(c-C2F3t-Bu)}2]
A round-bottomed flask was charged with dichloromethane
(15 ml), [PtI2(COD)] (0.272 g, 0.05 mmol) and i-Pr2P(c-C2F3t-Bu)
(0.364 g, 0.49 mmol) and the solution was stirred at room
temperature for 24 h. Removal of the solvent under vacuum left
a light orange powder (38 mg, 80%). Recystallisation from a mixed
dichloromethane–hexane solvent system yielded crystals of
suitable quality for X-ray diffraction studies. 1H NMR
(200.2 MHz, CDCl3): 1.20 (m, CH3), 1.87 (s, J = 5 Hz, CH), 19F
NMR (188.3 MHz, CDCl3, CFCl3): ꢀ105.3 (s, JPtF = 14 Hz, CF2),
31P{1H} (81.8 MHz, CDCl3, 85% H3PO4) NMR: 24.3 (s,
JPtP = 3453 Hz).
[17] See for example S. Jeulin, S. Duprat de Paule, V. Ratovelomanana-Vidal, J.P. Genet,
N. Champion, P. Dellis, Angew. Chem. Int. Ed. 43 (2004) 320–325;
R.A. Baber, M.L. Clark, K.M. Heslop, A.C. Marr, A.G. Orpen, P.G. Pringle, A. Wards,
D.E. Zambrano-Williams, Dalton Trans. (2005) 1079–1085;
R.B. DeVashe, J.M. Spruell, D.A. Dixon, G.A. Broker, S.T. Griffin, R.D. Rogers, K.H.
Shaughnessy, Organometallics 24 (2005) 962–971;
R. Wursche, T. Debaerdemaeker, M. Klinga, B. Rieger, Eur. J. Inorg. Chem. (2000)
2063–2070;
E. Rivard, A.D. Sutton, J.C. Fettinger, P.P. Powers, Inorg. Chim. Acta 360 (2007)
1278–1286.
[18] M.L. Clarke, D. Ellis, K.L. Mason, A.G. Orpen, P.G. Pringle, R.L. Wingad, D.A. Zahera,
R.T. Baker, Dalton Trans. (2005) 1294–1300.
[19] C.L. Pollock, G.C. Saunders, E. Smyth, V.I. Sorokin, J. Fluorine Chem. 129 (2008)
142–166.
4.1.10. Typical procedure for reaction of R2PLi with RfI
Under a nitrogen atmosphere a Schlenk flask was charged with
THF (5 ml) and Ph2PCl (0.12 ml, 0.67 mmol) to this was added
lithium wire (0.02 g, 2.9 mmol) and the reaction mixture was
stirred at room temperature for 1 h. i-C3F7I (0.1 ml, 0.67 mmol)
was syringed into the solution and left to stir at room temperature
for 1 h. after which all volatile material was removed under
vacuum to yield a viscous pale yellow oil. The oil was passed down
a short column, the first fraction which eluted from the column
(1:1 hexane:toluene) was collected and the solvent removed to
yield 0.10 g (40%) of Ph2P(i-C3F7). 1H NMR (CDCl3, 400.4 MHz):
7.24–7.45 (m, 6H), 7.69–7.82 (m, 4H); 19F NMR (CDCl3, 376 MHz) –
´
[20] I.T. Horvath, J. Rabai, Science 266 (1994) 72–75;
I.T. Horvath, Acc. Chem. Res. 31 (1998) 641–650, and references therein;
E.G. Hope, R.D.W. Kemmitt, D.R. Paige, A.M. Stuart, J. Fluorine Chem. 99 (1999)
197–200;
´
P. Bhattacharyya, B. Croxtall, J. Fawcett, J. Fawcett, D. Gudmunsen, E.G. Hope,
R.D.W. Kemmitt, D.R. Paige, D.R. Russell, A.M. Stuart, D.R.W. Wood, J. Fluorine
Chem. 101 (2000) 247–255, and references therein.
[21] R.N. Sterlin, R.D. Yatsenko, L.N. Pinkina, I.L. Knunyants, Khim. Nauk. Prom. 4
(1959) 810–811.
[22] A.H. Cowley, M.W. Taylor, J. Amer. Chem. Soc. 91 (1969) 1929–1933.
[23] P. Tarrant, P. Johncock, J. Savory, J. Org. Chem. 28 (1963) 839–843.

K.K. Banger et al. / Journal of Fluorine Chemistry 130 (2009) 1117–1129
1129
[24] K.K. Banger, A.K. Brisdon, A. Gupta, J. Chem. Soc., Chem. Commun. (1997) 139–
140.
[51] A.J. Carty, Inorg. Chem. 15 (1976) 1956–1959;
A.J. Carty, S.E. Jacobson, R.T. Simpson, N.J. Taylor, J. Am. Chem. Soc. 97 (1975)
7254–7262.
[52] A.J. Carty, S.E. Jacobson, N.J. Taylor, P.C. Chieh, J. Chem. Soc., Dalton Trans. (1976)
1375–1380.
[53] S.E. Jacobson, N.J. Taylor, A.J. Carty, Proc. 16th Int. Conf. Coord. Chem., 1974, p. 3.
[54] D.K. Johnson, A.J. Carty, J. Chem. Soc., Chem. Commun. (1977) 903–904.
[55] P.J. Low, J. Cluster Sci. 19 (2008) 5–46;
[25] K.K. Banger, R.P. Banham, A.K. Brisdon, W.I. Cross, G. Damant, S. Parsons, R.G.
Pritchard, A. Sousa-Pedares, J. Chem. Soc., Dalton Trans. (1999) 427–434.
[26] N.A. Barnes, A.K. Brisdon, F.R.W. Brown, W.I. Cross, I.R. Crossley, J.V. Morey, R.G.
Pritchard, L. Sekhri, New J. Chem. 28 (2004) 828–837.
[27] N.A. Barnes, A.K. Brisdon, F.R.W. Brown, W.I. Cross, C.J. Herbert, R.G. Pritchard, G.
Sadiq, Dalton Trans. (2008) 101–114;
N.A. Barnes, A.K. Brisdon, J.G. Fay, R.G. Pritchard, J.E. Warren, Inorg. Chim. Acta 358
(2005) 2543–2548.
J.R. Berenguer, M. Bernechea, J. Fornies, A. Garcıa, E. Lalinde, M.T. Moreno, Inorg.
Chem. 43 (2004) 8185–8198, and references therein.
[28] N.A. Barnes, A.K. Brisdon, M.J. Ellis, W.I. Cross, R.G. Pritchard, J. Fluorine Chem. 112
(2001) 35–45.
[56] A. Jakob, B. Milde, P. Ecorchard, C. Schreiner, H. Lang, J. Organomet. Chem. 693
(2008) 3821–3830.
[29] A.H. Cowley, M.W. Taylor, J. Am. Chem. Soc. 91 (1969) 1026–1027.
[30] R.N. Sterlin, S.S. Dubov, Zh.Vses. Khim, O-va. im. D. I. Mendeleeva 7 (1962) 117–
118.
[31] R.G. Kostyanovsky, A.P. Pleshkova, V.N. Voznesensky, Y.I. Elnatanov, Org. Mass
Spectrom. 11 (1976) 237–257.
[32] A.H. Cowley, M.W. Taylor, J. Am. Chem. Soc. 91 (1969) 1934–1936.
[33] H.G. Horn, F. Kolkmann, Z. Naturforsch. B 33B (1978) 1427–1429.
[34] R.G. Peters, J.D. Palcic, R.G. Baughman, Acta. Cryst. E E59 (2003) m1198–m1200.
[35] H.G. Horn, R. Koentges, F. Kolkmann, H.C. Marsmann, Z. Naturforsch. B 33B (1978)
1422–1426.
[57] See for example M. Bernechea, J.R. Berenguer, E. Lalinde, J. Torroba, Organome-
tallics, 28 (2009) 312–320.
[58] N. Pham-Tran, X.J. Hou, M.T. Nguyen, J. Phys. Org. Chem. 19 (2006) 167–172.
[59] I. Tidmarsh, unpublished results.
[60] A.K. Brisdon, I.R. Crossley, K.R. Flower, R.G. Pritchard, J.E. Warren, Angew. Chem.,
Int. Ed. 42 (2003) 2399–2401.
[61] N.M. Boag, K. Mohan Rao, N.J. Terrill, Acta. Crystallogr., Sect. C Cryst. Struct.
Commun. 47 (1991) 1064–1065.
[62] W. Xu, Q.-Y. Chen, J. Org. Chem. 67 (2002) 9421–9427.
[63] Z.-L. Cheng, Q.-Y. Chen, Synlett (2006) 478–480.
[36] F.H. Allen, Acta. Crystallogr., Sect. B 58 (2002) 380–388.
[37] R.P. Hughes, M.A. Mayer, M.D. Tawa, A.J. Ward, A. Williamson, A.L. Rheingold, L.N.
Zakharov, Inorg. Chem. 43 (2004) 747–756.
[64] J.J. Kampa, J.W. Nail, R.J. Lagow, Angew. Chem., Int. Ed. 34 (1995) 1241–1244.
[65] V.Y. Semenii, V.A. Stepanov, N.V. Ignat’ev, G.G. Furin, L.M. Yagupolskii, Zh. Obshch.
Khim. 55 (1985) 2716–2720;
[38] R.P. Hughes, A.J. Ward, A.L. Rheingold, L.N. Zakharov, Can. J. Chem. 81 (2003)
1270–1279.
[39] P.G. Jones, C. Thone, Private communication to the Cambridge Structural Data-
base, Deposition Number CCDC 241911 (2004).
[40] N.A. Barnes, A.K. Brisdon, M. Niuewenhuyzen, R.G. Pritchard, G.C. Saunders, J.
Fluorine Chem. 128 (2007) 943–951.
N. Ignat’ev, P. Sartori, J. Fluorine Chem. 103 (2000) 57–61.
[66] E.A. Ganja, C.D. Ontiveros, J.A. Morrison, Inorg. Chem. 27 (1988) 4535–4538.
[67] F.W. Bennett, G.R.A. Brandt, H.J. Emeleus, R.N. Haszeldine, Nature 166 (1950) 225;
F.W. Bennett, H.J. Emeleus, R.N. Haszeldine, J. Chem. Soc. (1953) 1565–1571;
H.J. Emeleus, J.D. Smith, J. Chem. Soc. (1959) 375–381;
A.B. Burg, W. Mahler, J. Am. Chem. Soc. 80 (1958) 2334.
[68] R.G. Peters, B.L. Bennett, R.C. Schnabel, D.M. Roddick, Inorg. Chem. 36 (1997)
5962–5965;
J.D. Palcic, P.N. Kapoor, D.M. Roddick, R.G. Peters, Dalton Trans. (2004) 1644–1647.
[69] S.E. Vaillard, A. Postigo, R.A. Rossi, Organometallics 23 (2004) 3003–3007.
[70] M.B. Murphy-Jolly, L.C. Lewis, A.J.M. Caffyn, Chem. Commun. (2005) 4479–4480.
[71] P. Eisenberger, I. Kieltsch, N. Armanino, A. Togni, Chem. Commun. (2008) 1575–
1577.
[72] L.I. Goryunov, J. Grobe, V.D. Shteingarts, B. Krebs, A. Lindemann, E.-U. Wu˝ rthwein,
C. Mu˝ ck-Lichtenfeld, Chem. Eur. J. 6 (2000) 4612–4622;
L.I. Goryunov, V.D. Shteingarts, J. Grobe, B. Krebbs, M.U. Triller, Z. Anorg, Allg.
Chem. 628 (2002) 1770–1779.
[41] N.S. Imyanitov, Koord. Khim. 11 (1985) 1041–1045.
[42] P. Cooper, R. Fields, R.N. Haszeldine, J. Chem. Soc., Perkin I (1975) 702–707;
A. Gentzsch, G.-V. Roschenthaler, Chemiker-Ztg 114 (1990) 6–7;
H. Wessolowski, A. Gentzsch, G.-V. Roschenthaler, G.L. Gard, Heteroatom. Chem. 8
(1997) 467–471.
[43] S.A. Krupoder, G.N. Dolenko, G.G. Furin, O.I. Andreevskaya, G.G. Yakobson, Izves-
tiya Sibirskogo Otdeleniya Akademii Nauk SSSR, 6, Seriya Khimicheskikh Nauk,
1981, pp. 119–127.
[44] Y. Shen, M. Qi, J. Chem. Soc., Perkin Trans. 1 (1993) 2153–2154;
Y. Shen, Y. Lin, Y. Xin, Tetrahedron Lett. 26 (1985) 5137–5138;
G. Jiang, J. Sun, Y. Shen, J. Fluorine Chem. 108 (2001) 207–210;
T. Konno, A. Morigaki, K. Ninomiya, T. Miyabe, T. Ishihara, Synthesis (2008) 564–
572.
[45] A.L. Henne, M. Nager, J. Am. Chem. Soc. 74 (1952) 650–652.
[46] A.J. Carty, N.K. Hota, T.W. Ng, H.A. Patel, T.J. O’Connor, Can. J. Chem. 49 (1971)
2706–2711.
[73] K. Hirano, H. Yorimitsu, K. Oshima, Org. Lett. 6 (2004) 4873–4875.
[74] M. Fild, P.G. Jones, K. Ruhnau, J. Fluorine Chem. 54 (1991) 387;
M.L. Clarke, A.G. Orpen, P.G. Pringle, E. Turley, Dalton Trans. (2003) 4393–4394.
[75] A.K. Brisdon, C.J. Herbert, Chem. Comm., in press, doi:10.1039/b909749j.
[76] C.J. Herbert, personal communication.
[47] J.E. Bunch, C.L. Bumgardner, J. Fluorine Chem. 36 (1987) 313–317;
N. Yoneda, S. Matsuoka, N. Miyaura, T. Fukuhara, A. Suzuki, Bull. Chem. Soc. Jpn.
63 (1990) 2124–2126.
[48] T. Yamazaki, K. Mizutani, T. Kitazume, J. Org. Chem. 60 (1995) 6046–6056;
A.R. Katritzky, M. Qi, A.P. Wells, J. Fluorine Chem. 80 (1996) 145–147.
[49] A.K. Brisdon, I.R. Crossley, Chem. Commun. (2002) 2420–2421.
[50] M. Shimizu, M. Higashi, Y. Takeda, G. Jiang, M. Murai, T. Hiyama, Synlett (2007)
1163–1165.
[77] O.B. Shawkataly, M.-L. Chong, H.-K. Fun, K. Sivakumar, Acta Cryst. Sect. C 52
(1996) 1725–1727.
[78] A. Karipides, C.M. Cosio, Acta Cryst. Sect. C 45 (1989) 1743–1745.
[79] S.E. Tunney, J.K. Stille, J. Org. Chem. 52 (1987) 748–753.
[80] G. Becker, O. Mundt, M. Rossler, E. Schneider, Z. Anorg. Allg. Chem. 443 (1978) 42–
52.
[81] G.M. Sheldrick, SHELXS97, University of Go¨ttingen, Germany.
[82] A.L. Spek, J. Appl. Cryst. 36 (2003) 7–13.