Remote C-F-metal interactions in late-transition-metal complexes

Article abstract of DOI:10.1021/om100261e

A series of phosphorus ligands containing the 5,6,7,8- tetrafluoronaphthalen-1-yl group (NpF₄) has been prepared, starting from 1-bromo-5,6,7,8-tetrafluoronaphthalene. The coordination chemistry of diphenyl(5,6,7,8-tetrafluoronaphthalen-1-yl)phosphine (1) with Ir(III), Pd(II), Pt(II), Rh(I), Ir(I), and Au(I) has been investigated, focusing on a possible remote interaction of the fluorine atom at position 8 of the naphthyl group (F¹) with the metal center. The Ir(III) complex [IrCp*Cl ₂(1)] (12) displays a weak J(P,F¹) coupling constant of 8 Hz and no Ir-F¹ interaction. A cationic derivative, the formally coordinatively unsaturated [IrCp*Cl(1)]SbF₆ (13), is generated upon chloride abstraction. For this complex J(P,F¹) is 80 Hz and the Ir-F¹ distance is 2.956(2) A. [MCl(η³-allyl)(1)] complexes (14, M = Pt; 15, M = Pd) display what is interpreted as a weaker M-F¹ interaction despite corresponding J(P,F¹) values of 67 and 98 Hz and M-F¹ distances of 3.069(2) and 2.995(2) A, respectively. Similar considerations apply to the analogous Rh(I) and Ir(I) complexes [IrCl(COD)(1)] (18) and [RhCl(COD)(1)] (19) (J(P,F¹) = 62 and 75 Hz and M-F¹ = 3.0744(14) and 3.0997(8) A, respectively). With the chiral enantiopure ligand (11bS)-4-(5,6,7,8-tetrafluoronaphthalen-1- yl)dinaphtho[2,1-d:1′,2′-f][1,3,2]dioxaphosphepine (10) analogous complexes were prepared. Ligand 10 is formally a derivative of 1 where the two phenyl groups have been replaced by a BINOL unit. In the case of [PtCl(η³-allyl)(10)] (16) and [RhCl(COD)(10)] (21) the coupling between F¹ and the NMR-active nuclei ¹⁹⁵Pt and ¹⁰³Rh, respectively, was detected in 2D NMR heteronuclear correlation experiments.

Full text of DOI:10.1021/om100261e

2540 Organometallics 2010, 29, 2540–2546
DOI: 10.1021/om100261e
Remote C-F-Metal Interactions in Late-Transition-Metal Complexes
†
Kyrill Stanek, Barbara Czarniecki, Raphael Aardoom, Heinz Ruegger, and
Antonio Togni*
€
Department of Chemistry and Applied Biosciences, Swiss Federal Institute of Technology, ETH Zurich,
CH-8093 Zu€rich, Switzerland. †Deceased June 28, 2009.
Received April 2, 2010
A series of phosphorus ligands containing the 5,6,7,8-tetrafluoronaphthalen-1-yl group (NpF4)
has been prepared, starting from 1-bromo-5,6,7,8-tetrafluoronaphthalene. The coordination chemi-
stry of diphenyl(5,6,7,8-tetrafluoronaphthalen-1-yl)phosphine (1) with Ir(III), Pd(II), Pt(II), Rh(I),
Ir(I), and Au(I) has been investigated, focusing on a possible remote interaction of the fluorine atom
at position 8 of the naphthyl group (F1) with the metal center. The Ir(III) complex [IrCp*Cl2(1)] (12)
displays a weak J(P,F1) coupling constant of 8 Hz and no Ir-F1 interaction. A cationic derivative, the
formally coordinatively unsaturated [IrCp*Cl(1)]SbF6 (13), is generated upon chloride abstraction.
1
1
3
˚
For this complex J(P,F ) is 80 Hz and the Ir-F distance is 2.956(2) A. [MCl(η -allyl)(1)] complexes
(14, M = Pt; 15, M = Pd) display what is interpreted as a weaker M-F1 interaction despite
corresponding J(P,F ) values of 67 and 98 Hz and M-F distances of 3.069(2) and 2.995(2) A,
respectively. Similar considerations apply to the analogous Rh(I) and Ir(I) complexes [IrCl(COD)(1)]
(18) and [RhCl(COD)(1)] (19) (J(P,F )=62 and 75 Hz and M-F =3.0744(14) and 3.0997(8) A,
respectively). With the chiral enantiopure ligand (11bS)-4-(5,6,7,8-tetrafluoronaphthalen-1-
yl)dinaphtho[2,1-d:10,20-f][1,3,2]dioxaphosphepine (10) analogous complexes were prepared. Ligand
10 is formally a derivative of 1 where the two phenyl groups have been replaced by a BINOL unit. In
the case of [PtCl(η3-allyl)(10)] (16) and [RhCl(COD)(10)] (21) the coupling between F1 and the NMR-
active nuclei 195Pt and 103Rh, respectively, was detected in 2D NMR heteronuclear correlation
experiments.
1
1
˚
1
1
˚
Introduction
were carried out by Plenio3-6 and Takemura,7-9 who used
X-ray analysis and stability constant measurements as well as
NMR spectroscopy (mainly the 1J(C,F) coupling constant, the
19F chemical shift, and 6,7Li/107,109Ag/133Cs/205Tl-19F direct
coupling) as means for the detection of such interactions in the
case of monofluorinated cryptands and crown ethers. In 1997,
Glusker’s first detailed and systematic search in the Cam-
bridge Structural Database (CSD) for exceptionally short
CF metal distances (for group 1 and 2 metals) in 1983 led
to the conclusion that “the CF bond is capable of significant,
if not prominent, interactions with both alkaline metal
cations and proton donors”.1 Though the relevance of this
topic was put forward by Glusker, only scattered reports
appeared initially in the literature and the investigation of
3 3 3
Plenio reviewed the complete work on CF metal interactions
3 3 3
and defined for the first time a set of criteria for a CF
M
3 3 3
interaction to be accepted as such.3 For instance, it was pointed
out that a minimum of four bonds must separate the metal from
the remote fluorine, thus ruling out complexes of fluorinated
ligands such as -C6F5, η6-C6F6, -CF3, dCF2, and F2CdCF2.
Plenio argued that “It is very possible that the presence of fluorine
in the complexes of such ligands is stabilizing, but it should also be
considered that such contacts are the unavoidable result of the
existence of such metal complexes.”3 Furthermore, it was stated
such “remote” CF metal interactions mainly focused on
3 3 3
group 1 and 2 metal cations, since it was reasoned that the
hard fluorine atom (categorized by the HSAB principle2)
would preferably interact with hard metal cations. Thus,
systematic investigations mainly of group 1 and 2 metals
that a comparison of the investigated M F distance with the
3 3 3
*To whom correspondence should be addressed. E-mail: atogni@
ethz.ch. Fax: (þ41) 44-632-1310.
sum of the corresponding van der Waals radii appears to be too
loose a criterion for the judgment of an interaction. Instead,
Plenio suggested the use of the ionic van der Waals radii for
group 1 and 2 metals (for typical coordination numbers). Thus,
(1) Murray-Rust, P.; Stallings, W. C.; Monti, C. T.; Preston, R. K.;
Glusker, J. P. J. Am. Chem. Soc. 1983, 105, 3206–3214.
(2) Pearson, R. G. In Chemical Hardness; Sen, K. D., Ed.; Springer-
Verlag: Berlin, New York, 1993; Vol. 80.
for a M F distance to be considered as a significant interac-
3 3 3
(3) (a) Plenio, H. Chem. Rev. 1997, 97, 3363–3384. For specific
examples see: (b) Plenio, H.; Diodone, R. Angew. Chem., Int. Ed. Engl.
1994, 33, 2175–2177. (c) Plenio, H.; Burth, D. J. Chem. Soc., Chem.
Commun. 1994, 2297–2298.
˚
tion, an overall threshold of e2.7 A was set for first-row
transition metals, a value smaller than the sum of the van der
(4) Plenio, H. ChemBioChem 2004, 5, 650–655.
(5) Plenio, H.; Diodone, R. J. Am. Chem. Soc. 1996, 118, 356–367.
(6) Plenio, H.; Diodone, R. Chem. Ber. 1996, 129, 1211–1217.
(7) Takemura, H.; Kon, N.; Yasutake, M.; Kariyazono, H.; Shinmyozu,
T.; Inazu, T. Angew. Chem., Int. Ed. 1999, 38, 959–961.
(8) Takemura, H.; Nakashima, S.; Kon, N.; Yasutake, M.; Shinmyozu,
T.; Inazu, T. J. Am. Chem. Soc. 2001, 123, 9293–9298.
(9) Takemura, H. Curr. Org. Chem. 2005, 9, 521–533.
r
pubs.acs.org/Organometallics
Published on Web 05/19/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 11, 2010 2541
Figure 1. Collection of reported metal complexes showing
CF M interactions.
Figure 2. Examples of late second- and third-row transition
metals, where the CF M interaction has been proven by
X-ray diffraction analysis and/or NMR spectroscopy.
3 3 3
3 3 3
Waals radii. Applying the same principle to second- and third-
˚
row transition metals results in a threshold of e3.0 A.
A brief overview of the most relevant substructures of
recurring ligands and metals that have been used for the
study of the coordination chemistry of the CF unit is shown
in Figure 1.
the P and F atoms undergoing an interaction with the metal
center are separated by four bonds in a rigid, planar frame-
work. Some methodologies for the detection of the CF
M
3 3 3
interactions have also been adapted to this very purpose. In
particular, to the best of our knowledge, rare 2D fluorine-
metal NMR correlation spectra have been used for the first
time in this field. It will also become clear that the nature of
It turns out that late second- and third-row transition
metals are underrepresented, and four examples of well-
documented complexes are depicted in Figure 2. Ruthenium
has been repeatedly reported to interact with CF units.
Perera and Shaw10 ascribed the large J(P,F) coupling con-
stant of 68 Hz observed for complex A in Figure 2 to a Ru-F
interaction. Grubbs11 attributed the enhanced performance
of a ring-closing metathesis catalyst (B) to the existence of a
Ru-F interaction and found Ru-F distances as short as
the CF metal interaction very much depends on the metal,
3 3 3
its oxidation state, and its ligand environment.
Results and Discussion
During our efforts to prepare new phosphorus ligands
containing the NpF4 group to be used in transition-metal-
catalyzed reactions, we observed an interesting NMR spectro-
˚
2.5203(12) A in certain cases. Another short Ru-F distance
of 2.489(6) A was reported by Cruz-Garritz.12,13 Crabtree14
˚
scopic property of such compounds. Indeed, the 31P andthe19
F
found an exceptionally short Ir-F distance and used NMR
data, i.e. the large J(F,H) coupling constant, to corroborate
the hypothesis of a M-F interaction (C). Already in 1987,
Cavell15 explained the unexpectedly high J(Pt,F) coupling
constant in complex D by the spatial proximity of fluorine to
platinum, which “allows facile van der Waals penetration”
and leads to “through space spin-spin coupling”.15
NMR spectra show an unexpected, very strong coupling
between phosphorus and the fluorine atom at position 8
of the naphthyl fragment, here indicated for simplicity as F1.
This behavior is commonly referred to as “through-space
coupling”,16-18 indicating that it does not primarily take place
via the bonds of the connecting carbon framework. A better
term would be “through-lone-pair coupling”, since this arises
from the fact that the phosphorus lone pair has a nonvanishing
spatial probability density overlapping with the fluorine lone
pair or even encompassing the 19F nucleus.19 Thus, a series of
different P-NpF4 derivatives containing various substituents
at phosphorus have been prepared by standard procedures,
starting from 1-bromo-5,6,7,8-tetrafluoronaphthalene.20 These
compounds, together with their corresponding J(P,F) values,
are shown in Table 1. Depending on the two remaining
substituents on phosphorus, the J(P,F) coupling constant
can be as high as 324 Hz for Cl2P-NpF4 (9). Note for
comparison that the J(P,F) value for a P-CF3 group, for
Although this collection of examples is not complete, it
already covers the most important work done in the field of
CF M interactions for late transition metals, thus also
revealing the paucity of data in this area.
3 3 3
In the present work, we expanded the scope of transition
metals possibly undergoing a remote M-F interaction by
utilizing new phosphorus ligands containing the 5,6,7,8-
tetrafluoronaphthalen-1-yl group (NpF4). In these ligands
(10) Perera, S. D.; Shaw, B. L. Inorg. Chim. Acta 1995, 228, 127–131.
(11) (a) Ritter, T.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc.
2006, 128, 11768–11769. (b) Ragone, F.; Poater, A.; Cavallo, L. J. Am.
Chem. Soc. 2010, 132, 4249–4258.
(12) Catala, R. M.; Cruzgarritz, D.; Hills, A.; Hughes, D. L.;
Richards, R. L.; Sosa, P.; Torrens, H. J. Chem. Soc., Chem. Commun.
1987, 261–262.
(13) Catala, R. M.; Cruzgarritz, D.; Sosa, P.; Terreros, P.; Torrens,
H.; Hills, A.; Hughes, D. L.; Richards, R. L. J. Organomet. Chem. 1989,
359, 219–232.
(16) Berger, S.; Braun, S.; Kalinowski, H. O. In NMR-Spektroskopie
von Nichtmetallen; Thieme: Stuttgart, Germany, 1994; Vol. 4.
(17) McFarlan., W J. Chem. Soc., Chem. Commun. 1971, 609–610.
(18) Karcar, A.; Freytag, M.; Jones, P. G.; Bartsch, R.; Schmutzler,
R. Z. Anorg. Allg. Chem. 2002, 628, 533–544.
(14) Kulawiec, R. J.; Holt, E. M.; Lavin, M.; Crabtree, R. H. Inorg.
Chem. 1987, 26, 2559–2561.
(15) Phillips, I. G.; Ball, R. G.; Cavell, R. G. Inorg. Chem. 1987, 26,
4074–4079.
(19) Ng, S.; Sederholm, C. H. J. Chem. Phys. 1964, 40, 2090–2094.
(20) See: Gribble, G. W.; Lehoullier, C. S.; Sibi, M. P.; Allen, R. W.
J. Org. Chem. 1985, 50, 1611–1616. An improved procedure is provided in
the Supporting Information.

2542 Organometallics, Vol. 29, No. 11, 2010
Stanek et al.
Table 1. Through-Lone-Pair Coupling Constants of Various
Tetrafluoronaphthalenyl-Containing Phosphorus Ligands
(Measured at Room Temperature)
compd
R1
R2
J(P,F) (Hz)
1
2
3
Ph
Ph
Ph
Ph
Ph
198
200
213
219
232
242
246
282
324
250
235
N(Et)2
NpF4
CF3
NpF4
O-(1R)-menthyl
Ph
NpF4
Cl
4 (rac)
5a
NpF4
Ph
O-(1R)-menthyl
Cl
Cl
6 (S)b
7 (R)b
8a
9a
10
11
O,O’-(S)-BINOL
O,O’-(R,R)-Ph-TADDOL
a These compounds were observed in a NMR experiment on mixing
1 equiv of NpF4-Li with 1 equiv of PCl3. b The CIP descriptor defines the
phosphorus stereocenter; the diastereoisomers were separated by pre-
parative HPLC.
1
example, is only around ca. 80 Hz,21 whereas the J(P,F)
value for PF3 is 1400 Hz.22
The measured J(P,F) coupling constants nicely reflect the s
character of the phosphorus lone pair, which increases when
electron-withdrawing substituents are introduced, according
to Bent’s rule.22 Thus, this parameter may be viewed as
equivalent to J(Se,P) for phosphine selenides used for the
quantification of σ-donor properties of phosphorus
ligands.24-28
Figure 3. 31P and 19F NMR spectra corresponding to phos-
phine 1 (top) and the neutral as well as the monocationic Ir(III)
complexes 12 (middle) and 13 (bottom), respectively, at 298 K in
CD2Cl2.
1
Since this type of coupling involves the phosphorus lone
pair, the J(P,F) value should decrease upon coordination to a
metal. We reacted phosphine 1 with [Ir2Cp*2Cl2(μ-Cl)2], and
indeed, J(P,F) decreased to only 8 Hz upon formation of
complex 12 (Scheme 1, Figure 3). The resulting Ir(III)
complex was then treated with a chloride scavenger
(AgSbF6 or AgBF4) and J(P,F) increased again to 80 Hz,
clearly indicating that the electron-deficient 16-electron
complex 13 saturates its coordination sphere by interacting
with the remote F1 of the NpF4 unit. The 19F and 31P NMR
spectra corresponding to compounds 1, 12, and 13, respec-
tively, are shown in Figure 3.
Scheme 1. Synthesis of an Ir(III) Complex with Phosphine 1
and Subsequent Chloride Abstraction Resulting in a CF Ir
Interaction
3 3 3
It is easily recognized that the Ir(III) center is stereogenic
in complex 13. Therefore, the two phenyl substituents on
phosphorus are diastereotopic and should give rise to
different sets of resonances in NMR spectra. However, as
shown in Figure 3, the 19F and 31P resonances of compound
13 are rather broad, thus indicating a dynamic behavior at
room temperature. When samples were cooled to 193 K,
resolved signals for the two diastereotopic phenyl groups
were detected and, at the same time, the 31P resonance
became much sharper, showing a slightly larger coupling
constant to F1 of 87 Hz (Figure 4). These findings corro-
borate the assumption of a stable Ir-F bond at low
temperature, however, being involved in a dynamic process
corresponding to an epimerization of the Ir center at room
temperature.
(21) See, e.g.: Eisenberger, P.; Kieltsch, I.; Armanino, N.; Togni, A.
Chem. Commun. 2008, 1575–1577.
(22) Jameson, C. J.; Jameson, A. K.; Wille, S. J. Chem. Phys. 1979, 83,
3372–3378.
(23) Bent, H. A. Chem. Rev. 1961, 61, 275–311.
(24) Duddeck, H. Prog. Nucl. Magn. Reson. Spectrosc. 1995, 27,
1–323.
(25) Muller, A.; Otto, S.; Roodt, A. Dalton Trans. 2008, 650–657.
(26) Chevykalova, M. N.; Manzhukova, L. F.; Artemova, N. V.;
Luzikov, Y. N.; Nifant’ev, I. E.; Nifant’ev, E. E. Russ. Chem. Bull. 2003,
52, 78–84.
The same sequence of iridium(III) complex formation/
chloride abstraction was carried out with the BINOL-
derived chiral phosphepine ligand 10. The presence of a chiral
BINOL backbone on phosphorus should lead to two
(27) Allen, D. W.; Taylor, B. F. J. Chem. Soc., Dalton Trans. 1982,
51–54.
(28) Andersen, N. G.; Keay, B. A. Chem. Rev. 2001, 101, 997–1030.

Article
Organometallics, Vol. 29, No. 11, 2010 2543
Figure 4. Low-temperature (193 K) 13C and 31P NMR spectra of complex 13. Diastereotopic ortho and meta C-Ph signals corroborate
the Ir-F interaction.
˚
diastereomeric complexes upon chloride abstraction. In-
deed, two expected products were observed in a 1.7:1 ratio,
showing again augmented J(P,F) coupling constants of
40 Hz (major) and 51 Hz (minor). Unfortunately, these pro-
ducts were accompanied by two minor, unidentified side
products and could not be characterized properly (see the
Supporting Information for further details).
0.02 A. Though this displacement is barely significant, it is
interesting to note that it occurs toward the same side of the
plane as the fluorine atom F1, whose distance from the plane
˚
is 2.666 A. The observation of such a trigonal-pyramidal
geometry led to the question whether common (pseudo)
square-planar complexes could be expanded to square-
pyramidal complexes with the remote fluorine atom occupying
the apical position.
We obtained X-ray-quality single crystals of complexes 12
and 13, and the corresponding crystal structures were deter-
mined. ORTEP views are shown in Figure 5. The Ir-F1
We therefore prepared chiral and nonchiral complexes of
Pd(II) and Pt(II), respectively, as shown in Scheme 2.
All targeted complexes were obtained in fair to good yields.
In the case where the chiral phosphepine ligand 10 was used,
two different isomers were observed for each of the two
metals, due to the two possible orientations of the allyl ligand
with respect to the main coordination plane. Pleasingly, all six
complexes showed clearly augmented J(P,F) coupling con-
stants as shown in Table 2, again strongly supporting a
metal-fluorine interaction. Furthermore, a coupling amount-
ing to ca. 200 Hz between F1 and 195Pt was detected by the
characteristic 195Pt satellites in the 19F NMR spectrum of the
˚
˚
distance of 2.956(2) A in compound 13 is just below the 3.0 A
threshold for a third-row transition-metal-fluorine interac-
tion. However, since the interaction is strongly supported by
our NMR experiments and because of the significant con-
figurational change occurring on going from complex 12 to
derivative 13, there should be no doubt about the existence of
a stable interaction between iridium and fluorine in this case.
Complexes 12 and 13 display a significantly different
geometry around the respective Ir(III) centers. Whereas
complex 12 is pseudo-tetrahedral, as expected, 13 is distorted
trigonal pyramidal with the P, Cl, and Ir atoms and the Cp*
centroid almost perfectly lying in one common plane, while
the remote fluorine occupies the apical position of the
pyramid. In fact, the distance of the Ir atom from the plane
defined by the Cp* centroid and the P and Cl atoms is just
1
Pt complexes. This lies at the lower limit of J(Pt,F) values
reported for Pt-F bonds in the range of 200-2000 Hz29 but is
(29) Mason, J. Multinuclear NMR; Plenum Press: New York, 1987;
p 552.

2544 Organometallics, Vol. 29, No. 11, 2010
Stanek et al.
˚
Figure 5. ORTEP view (30% probability ellipsoids) of complex 12 (left) and the cation 13 (right). The Ir-F distance in 13 is 2.956(2) A.
˚
Important differences between 12 and 13: the Ir-P distance increases from 2.3172(15) to 2.3523(10) A; the F(1)-C(8) distance increases
˚
from 1.325(7) to 1.346(4) A; the Ir-P-C(1) angle decreases from 115.53(19) to 113.66(11)°.
Scheme 2. Synthesis of Pseudo-Square-Planar Palladium and
Platinum Complexes Containing Achiral or Chiral
NpF4 Ligands
Table 2. Selected Coupling Constants Supporting a Metal-
Fluorine Interaction in Complexes 14-17 (Measured at Room
Temperature)
complex
J(P,F) (Hz)
J(Pt,F) (Hz)
1J(Pt,P) (Hz)
14
67
61
52
204
173
241
4551
6498
6423
16a
16b
15
17a
17b
98
96a
99a
a Measured at 223 K.
nonfluorinated analogues of these compounds. However, a
structurally somewhat related complex containing diphenyl-
(1-naphthyl)phosphine is [Pt(N3)2(SMe2)PPh2Np].31 In this com-
pound the distance between Pt and the hydrogen atom attached
to C(8) of the naphthyl unit, corresponding to F1 in complex 14,
˚
is 3.37 A: i.e., longer than the Pt-F distance in our example.
In order to obtain more data for comparison, four related
d8 square-planar complexes were prepared, i.e. the Rh(I) and
Ir(I) complexes 18-21 (Scheme 3). The isolated yields of
these compounds were rather low, due to their high solubility
even in pentane, which was used for their precipitation from
the reaction mixtures, as well as to their reactivity and hence
lower stability. The chiral complexes 20 and 21 were not
isolated as solids but only investigated in solution by NMR.
These related compounds showed a J(P,F) coupling con-
stant in the expected range. The corresponding values are
collected in Table 3.
still much higher than in complex D, shown in Figure 2, for
which a J(Pt,F) value of 11.5 Hz was interpreted as indicating
an interaction.15
In a 2D NMR 195Pt-19F correlation experiment performed
on complexes 14 and 16a,b it was possible to directly detect the
interactions of interest. The spectrum of complexes 16a,b is
depicted in Figure 6, the cross peaks between F1 and Pt nicely
confirming the interaction (the corresponding spectrum of
complex 14 can be found in the Supporting Information, p
S47). To the best of our knowledge, it appears that this spectrum
represents the first published example of a direct correlation
experiment involving 195Pt and 19F, although similar experi-
ments have beenreported for NMR-active metals such as109Ag,
183W, and 207Pb.30
Unfortunately, the Rh-F1 coupling in 19 and 21 is not
directly observable, this not being surprising, since it is known
that 1J(Rh,F) values lie around 50 Hz29 and the coupling in
this case is expected to be weaker than 1J(Rh,F). Because F1
also couples to other fluorine atoms of the NpF4 group, the
extent of the coupling being around 20 Hz, the interesting
J(Rh,F) parameter becomes hardly discernible. In principle, a
stable CF-metal bonding interaction should lead to two
different diastereoisomers in the case of complexes 20 and
21. However, in our cases only one isomer was detected below
room temperature. This observation indicates a very weak
metal-fluorine interaction, with the NpF4 unit still freely
rotating. These additional complications demanded the use of
2D NMR 103Rh-19F correlation spectroscopy for the unam-
biguous identification of the interaction. The correlation spec-
trum obtained for complex 21 is shown in Figure 7.
Although the metal-fluorine interaction is strongly sup-
ported by NMR spectroscopy, it was necessary to corroborate
the results with geometric parameters of the corresponding
derivatives in the solid state. We were able to obtain single
crystals of the achiral complexes 14 and 15, which were subjected
to an X-ray crystal structure study (see the Supporting Informa-
tion for full details). As expected, in these two compounds the
NpF4 unit points toward the metal center and the crucial dis-
tances Pt-F1 for complex 14 and Pd-F1 for 15 are 3.0690(17)
˚
and 2.995(2) A, respectively. There have been no reports of
(30) Battiste, J.; Newmark, R. A. Prog. Nucl. Magn. Reson. Spec-
trosc. 2006, 48, 1–23.
(31) Huh, H. S.; Lee, Y. K.; Lee, S. W. J. Mol. Struct. 2006, 789, 209–
219.

Article
Organometallics, Vol. 29, No. 11, 2010 2545
Figure 6. 195Pt, 19F correlation spectrum of diastereomeric complexes 16, indicating the Pt-F interaction.
Scheme 3. Synthesis of Ir(I) and Rh(I) Complexes, Showing
Metal-Fluorine Interactions
Table 3. Selected Coupling Constants for Complexes 18-21
complex
J(P,F) (Hz)
J(Rh,F) (Hz)
1J(Rh,P) (Hz)
18
19
20
21
62
75
50
66
Figure 7. HMQC 103Rh, 19F correlation spectrum of complex 21.
n.d.
150
234
<10
criterion and put it into a broader perspective. It is clear that
for any other metal center its formal oxidation state, its
ligand environment, and its possible geometrical constraints
will change its ability to undergo such an interaction. Our
analysis of complexes 18 and 19 indicates that such an
interaction is operative, albeit rather weakly, at a M-F
Although J(Rh,F) is not directly observable in 1D NMR
spectra, the cross peaks in the 2D correlation spectrum
directly identify the Rh-F interaction in complex 21. Since
the J(P,F) values for the iridium analogues are also strongly
increased, we assume a similar metal-fluorine interaction in
the iridium(I) complexes 18 and 20. To substantiate this
assumption, single crystals of compounds 18 and 19 were
analyzed by X-ray diffraction. The corresponding ORTEP
views are shown in Figure 8.
The metal-fluorine distances in these cases lie above
the threshold of 3.0 A set by Plenio. However, since the
interaction has been identified by means of 2D NMR techni-
ques and a fairly large J(P,F) coupling constant has been
detected, it may be necessary to relativize the simple distance
˚
distance beyond 3.0 A.
To enhance the interaction between the metal and the
fluorine in the series of complexes 14-21, the obvious next
step was the generation of coordinatively unsaturated
species by chloride abstraction. Unfortunately, all experi-
ments commonly using AgSbF6 as scavenger resulted in
complex and highly unstable product mixtures, from which
no single product could be isolated. In some rare cases,
though, it has been possible to detect species in solution
3
˚

2546 Organometallics, Vol. 29, No. 11, 2010
Stanek et al.
Ir(III) complex 13, for which the bonding interaction is
further supported by stereochemical considerations, and its
precursor 12, having a much lower J(P,F) coupling constant.
However, the interaction of a metal center with the remote
fluorine seems to be weaker for the d8 and d10 centers Ir(I),
Rh(I), Pd(II), Pt(II), and Au(I). Overall, the very nature of
the observed interactions remains speculative and is prob-
ably not identical in all cases. In this respect, the difference
between complex 13 and the examined d8 systems is remi-
niscent of the distinction adopted for interactions of CH
units with coordinatively unsaturated transition-metal cen-
ters. Thus, remote CH-M interactions are divided into
agostic (three-center-two-electron bonds) and anagostic
(electrostatic) interactions.32 It is not surprising that there
is no such discrimination known for remote CF-M inter-
actions, because of the comparatively small selection of
complexes displaying CF-M contacts and because of the
general tendency of CF units to undergo electrostatic (i.e.,
“CF-anagostic”) interactions. Thereby, such an electrostatic
interaction could be seen as one between the metal and the
fluorine atom or as a metal-dipole (CF) interaction, as
already described by Takemura.8 Since not only the CF
bond but also the whole NpF4 unit possesses significant
dipole moments (2.83 D for 1-methyl-5,6,7,8-tetrafluoro-
naphthalene, as calculated by DFT B3LYP 6-31G*, and
1.83 D for the CF bond in fluoromethane33), the alignment
of the NpF4 dipole antiparallel to the one of the Pd-P bonds
may also be a strong stabilizing electrostatic effect. There are
many examples of so-called “closed-shell interactions”,34 i.e.
stabilizing interactions between d8-d10-s2 systems. These
interactions are generally weaker than covalent or ionic bonds
but stronger than other van der Waals bonds. Whether or not
the interaction of fluorine with Pd(II), Pt(II), Ir(I), Rh(I), and
Au(I) could be interpreted as such a closed-shell interaction
remains an open question. We are still pursuing the study of
further organometallic systems possibly displaying these kinds
of subtle interactions.
Figure 8. ORTEP view (30% probability ellipsoids) of complexes
1
˚
18 and 19, respectively. Selected distances (A): for 18, Ir-F =
3.0744(14), Ir-P=2.3203(7), Ir-Cl=2.3614(6); for 19, Rh-F1=
3.0997(8), Rh-P=2.3381(3), Rh-Cl=2.3699(3).
displaying large J(P,F) coupling constants up to 150 Hz.
However, the complex mixtures rapidly degraded, giving
elemental metals within a few hours. Because of these
difficulties, further investigations were abandoned. Simi-
lar considerations apply to the derivatives [RuCp*Cl2-
(PPh2NpF4)], [Ru(p-cymene)Cl2(PPh2NpF4)], and [RhCp*Cl2-
(PPh2NpF4)].
The scope of transition metals was further extended to
gold(I). In particular, the simple linear complexes [AuCl(1)]
(22) and [AuCl(10)] (23) were prepared. In both cases, and by
analogy to the previously discussed complexes, the observed
J(P,F) coupling constants of 75 and 66 Hz, respectively, seem
to indicate a M-F interaction. For complex 22, which was
characterized by X-ray crystallography, the fluorine-gold
˚
distance is 3.1771(13) A. This is in fact longer than the sum of
˚
the individual van der Waals radii of 3.13 A, but we also
assign this as remote, weak metal-fluorine interactions due
to the increased J(P,F) coupling constants.
Conclusions
Acknowledgment. This work is dedicated to the memory
€
of one of the coauthors, Dr. Heinz Ruegger, deceased June
Utilizing the new phosphorus ligands 1 and 10, containing
the 5,6,7,8-tetrafluoronaphthalen-1-yl fragment, a series of
late-transition-metal complexes have been prepared and
characterized by multinuclear NMR spectroscopy and
X-ray crystal structural analyses, in view of identifying a
possible bonding interaction of the fluorine atom at position
8 of the naphthyl group with the metal center. The main
argument for claiming such an interaction is the relati-
vely large J(P,F) coupling constant in the approximate
range 40-100 Hz observed for most of the derivatives. The
argument is mainly based on the comparison between the
28, 2009. Financial support from the ETH Zurich and the
Swiss National Science Foundation is gratefully acknow-
ledged (Ph.D. grants to K.S. and R.A., respectively).
Supporting Information Available: Text and figures giving full
experimental details for the preparation of all new compounds
described in this paper, including copies of 1H, 19F, 31P and 13
C
NMR spectra, and CIF files giving crystallographic data and
refinement parameters and bond lengths and angles of com-
pounds 1, 12-15, 18, 19, and 22. This material is available free of
charge via the Internet at http://pubs.acs.org.
(32) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Natl. Acad.
Sci. U.S.A. 2007, 104, 6908–6914.
(33) O’Hagan, D. Chem. Soc. Rev. 2008, 37, 308–319.
(34) Pykko, P. Chem. Rev. 1997, 97, 597–636.
€