Metal Complexes of Mesitylphosphine: Synthesis, Structure, and Spectroscopy

Article abstract of DOI:10.1021/ic960529q

A series of primary phosphine homoleptic complexes [ML₄]ⁿ⁺X_n (1, M = Ni, n = 0; 2, M = Pd, n = 2, X = BF₄; 3, M = Cu, n = 1, X = PF₆; 4, M = Ag, n = 1, X = BF₄; L = PH₂Mes, Mes = 2,4,6-Me₃,C₆H₂] was prepared from mesitylphosphine and Ni(COD)₂, [Pd(NCMe)₄][BF₄]₂, [Cu(NCMe)₄]PF₆, and AgBF₄, respectively. Reactions of 1-4 with MeC(CH₂PPh₂)₃ (triphos) or [P(CH₂CH₂PPh₂)₃] (tetraphos) afforded the derivatives [M(L')L]ⁿ⁺ X_n (L' = triphos; 6, M = Ni, n = 0; 7, M = Cu, n = 1, X = PF₆; 8, M = Ag, n = 1, X = BF₄; L' = tetraphos; 9, M = Pd, n = 2, X = BF₄). Addition of NOBF₄ to 1 yielded the nitrosyl compound [NiL₃(NO)]BF₄, 5. The solution structure and dynamics of 1-9 were studied by ³¹P NMR spectroscopy (including the first reported analyses of a 12-spin system for 1-2). Complexes 1,3,6, and 7·solvent were characterized crystallographically. The structural and spectroscopic studies suggest that the coordination properties of L are dominated by its relatively small cone angle and that the basicity of L is comparable to that of more commonly used tertiary phosphines.

Full text of DOI:10.1021/ic960529q

6708
Inorg. Chem. 1996, 35, 6708-6716
Metal Complexes of Mesitylphosphine: Synthesis, Structure, and Spectroscopy
Igor V. Kourkine, Stanislav V. Maslennikov, Robert Ditchfield, and David S. Glueck*
6128 Burke Laboratory, Department of Chemistry, Dartmouth College, Hanover, New Hampshire 03755
Glenn P. A. Yap, Louise M. Liable-Sands, and Arnold L. Rheingold
Department of Chemistry, University of Delaware, Newark, Delaware 19716
ReceiVed May 16, 1996X
A series of primary phosphine homoleptic complexes [ML4]n+Xn (1, M ) Ni, n ) 0; 2, M ) Pd, n ) 2, X ) BF4;
3, M ) Cu, n ) 1, X ) PF6; 4, M ) Ag, n ) 1, X ) BF4; L ) PH2Mes, Mes ) 2,4,6-Me3C6H2] was prepared
from mesitylphosphine and Ni(COD)2, [Pd(NCMe)4][BF4]2, [Cu(NCMe)4]PF6, and AgBF4, respectively. Reactions
of 1-4 with MeC(CH2PPh2)3 (triphos) or [P(CH2CH2PPh2)3] (tetraphos) afforded the derivatives [M(L′)L]n+Xn
(L′ ) triphos; 6, M ) Ni, n ) 0; 7, M ) Cu, n ) 1, X ) PF6; 8, M ) Ag, n ) 1, X ) BF4; L′ ) tetraphos; 9,
M ) Pd, n ) 2, X ) BF4). Addition of NOBF4 to 1 yielded the nitrosyl compound [NiL3(NO)]BF4, 5. The
solution structure and dynamics of 1-9 were studied by 31P NMR spectroscopy (including the first reported
analyses of a 12-spin system for 1-2). Complexes 1, 3, 6, and 7‚solvent were characterized crystallographically.
The structural and spectroscopic studies suggest that the coordination properties of L are dominated by its relatively
small cone angle and that the basicity of L is comparable to that of more commonly used tertiary phosphines.
Introduction
makes this ligand sufficiently bulky (cone angle 110°)3 that a
comparison of its coordination chemistry with that of tertiary
phosphines should not be completely dominated by steric effects
(for comparison, the cone angle of PMe3 is 118°).3b Moreover,
Although the coordination chemistry of tertiary and secondary
phosphines has been studied in significant detail, little has been
reported about the transition-metal chemistry of primary phos-
phines.1 The properties of these ligands are expected to differ
from those of tertiary phosphines for both steric and electronic
reasons. Replacing R in PR3 by H reduces both the cone angle
of the ligand and, because of the relative inductive effect of R
and H, the basicity of the phosphorus lone pair. Moreover, in
a recent study of the donor strength of a variety of phosphine
ligands, Drago found that PH3 and primary phosphines PH2R
were weaker donors than would be expected if the inductive
effect of the hydrogen substituent were the only influence on
base strength. The decreased basicity was suggested to be due
to the reduced bond angles in primary phosphines (∼90°) in
comparison to those in tertiary ones (∼108°); the accompanying
change in hybridization gives increased s character in the P lone-
pair orbital.2
1
the Me groups provide a useful handle for H NMR spectros-
copy, and the crystal structure of the free ligand has been
reported.4
Systematic investigations of the coordination behavior of
primary phosphines have been hampered by the susceptibility
of P-H bonds to metal-mediated cleavage to form phosphido
(PHR) or phosphinidene (PR) ligands.1 The complexes de-
scribed below do not undergo these reactionsssince they do
not contain halide coligands, the common dehydrohalogenation5
reaction is avoided, and they are not sufficiently reactive to add
the P-H bond oxidatively.6 The structural and spectroscopic
studies reported here both establish the ligand properties of PH2-
Mes at a variety of metal centers and allow direct comparisons
of them with those of a more commonly used tertiary phosphine.
We report here the synthesis of a series of homoleptic
complexes of mesitylphosphine (L ) PH2Mes, Mes ) 2,4,6-
Me3C6H2) as well as some derivatives containing tertiary phos-
phine ligands. We chose to study mesitylphosphine because
in comparison to most other primary phosphines it is signifi-
cantly less air-sensitive and malodorous. The mesityl group
Results and Discussion
The homoleptic complexes NiL4 (1), [PdL4][BF4]2 (2), [CuL4]-
PF6 (3), and [AgL4]BF4 (4) (Scheme 1) were prepared by
displacement of the weakly coordinating ligands cyclooctadiene,
acetonitrile, and tetrafluoroborate by mesitylphosphine; complex
X Abstract published in AdVance ACS Abstracts, October 15, 1996.
(1) Review: Levason, W. In The Chemistry of Organophosphorus
Compounds; Hartley, F. R., Ed.; John Wiley and Sons: Chichester,
England, 1990; Vol. 1; pp 567-641. For selected recent references,
see: (a) Ho, J.; Hou, Z.; Drake, R. J.; Stephan, D. W. Organometallics
1993, 12, 3145-3157. (b) Goerlich, J. R.; Fischer, A.; Jones, P. G.;
Schmutzler, R. Polyhedron 1993, 12, 2279-2289. (c) Goerlich, J. R.;
Schmutzler, R. Z. Anorg. Allg. Chem. 1994, 620, 898-907. (d) Bleeke,
J. R.; Rohde, A. M.; Robinson, K. D. Organometallics 1994, 13, 401-
403. (e) Ho, J.; Rousseau, R.; Stephan, D. W. Organometallics 1994,
13, 1918-1926. (f) Blake, A. J.; Champness, N. R.; Forder, R. J.;
Frampton, C. S.; Frost, C. A.; Reid, G.; Simpson, R. H. J. Chem.
Soc., Dalton Trans. 1994, 3377-3382.
(3) Determined by Tolman’s method for unsymmetrical phosphines, see:
(a) Tolman, C. A.; Seidel, W. C.; Gosser, L. W. J. Am. Chem. Soc.
1974, 96, 53-60. (b) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.
(4) Bartlett, R. A.; Olmstead, M. M.; Power, P. P.; Sigel, G. A. Inorg.
Chem. 1987, 26, 1941-1946.
(5) For some examples, see: (a) Goldwhite, H.; Hirschon, A. S. Transition
Met. Chem. 1977, 2, 144-149. (b) Issleib, K.; Roloff, H.-R. Z. Anorg.
Allg. Chem. 1963, 324, 250-258. (c) Issleib, K.; Wilde, G. Z. Anorg.
Allg. Chem. 1961, 312, 287-298. (d) Kourkine, I. V.; Chapman, M.
B.; Glueck, D. S.; Eichele, K.; Wasylishen, R. E.; Yap, G. P. A.;
Liable-Sands, L. M.; Rheingold, A. L. Inorg. Chem. 1996, 35, 1478-
1485.
(6) (a) Powell, J.; Fuchs, E.; Gregg, M. R.; Phillips, J.; Stainer, M. V. R.
Organometallics 1990, 9, 387-393. (b) Ebsworth, E. A. V.; Mayo,
R. A. J. Chem. Soc., Dalton Trans. 1988, 477-484. (c) Bonanno, J.
B.; Wolczanski, P. T.; Lobkovsky, E. B. J. Am. Chem. Soc. 1994,
116, 11159-11160. (d) Ref 1a.
(2) Drago, R. S. Organometallics 1995, 14, 3408-3417. For a more
general discussion of such hybridization effects (Bent’s rules), see:
(a) Bent, H. A. J. Chem. Educ. 1960, 37, 616-624. (b) Bent, H. A.
Chem. ReV. 1961, 61, 275-311.
S0020-1669(96)00529-0 CCC: $12.00 © 1996 American Chemical Society

Metal Complexes of Mesitylphosphine
Inorganic Chemistry, Vol. 35, No. 23, 1996 6709
Table 1. Selected NMR and IR Data for the Mesitylphosphine
Scheme 1
Complexesa
complex
PH2Mes (L)
NiL4 (1)
[PdL4][BF4]2 (2)
[CuL4]PF6 (3)
[AgL4]BF4 (4)
[NiL3(NO)]BF4 (5)
Ni(triphos)L (6)
[Cu(triphos)L]PF6 (7)
[Ag(triphos)L]BF4 (8)
[Pd(tetraphos)L][BF4]2 (9)
BH3‚L (11)
δ
31P
δ 1H 1JPH
νPH
2302
2278
2301, 2368
2357, 2329
2372
2359, 2336
2269
2346
2347
2301
2246
-153.0
-86.5
-73.2
-121.0
-131.5
-89.7
-96.2
-122.0
-139.0
-94.5
-68.8
3.60
4.83
5.09
4.70
4.62
5.27
6.23
5.69
4.70
5.64
4.81
204
283
422
317
297
350
281
317
280
381
370
Scheme 2
a NMR data (at 22 °C) in C6D6 for L, 1, and 6, in CD2Cl2 for 2-5
and 7 and in CD3NO2 for 8 and 9. Chemical shifts are reported in
ppm and coupling constants in hertz. IR data (in cm-1) for KBr pellets
for 1-11 and for a solution of L in CH2Cl2.
1 can also be made by reduction of [Ni(NCMe)6][BF4]2‚0.5NCMe
with an excess of L.7 The complexes are isolated as white to
yellow solids by recrystallization. Compounds 2-4 are stable
to air and water as solids and in solution, but 1 is air-sensitive.
Complex 1 reacts with NOBF4 (Scheme 2) to yield the purple
nitrosyl complex [NiL3(NO)]BF4 (5), which is formulated as a
tetrahedral d10 Ni(0) species containing linear NO+. The NO
stretching frequency for 5 (1751 cm-1, KBr) is similar to those
for the previously reported8 red [Ni(triphos)(NO)]+ and [Ni-
(tep)NO]+ [triphos ) MeC(CH2PPh2)3, tep ) MeC(CH2PEt2)3],
which range from 1750 to 1770 cm-1 (Nujol mulls), depending
on the counterion. In principle, the NO stretching frequencies
should provide information on the relative donor properties of
L and these multidentate tertiary phosphines.9 However, the
data suggest that this parameter is insensitive to the electronic
nature of the phosphine ligands.
nitromethane, compound 8 is obtained as shown by 31P NMR
spectroscopy.
Structure assignments for the mesitylphosphine complexes
are consistent with elemental analyses and spectroscopic data.
The IR spectra for the complexes show νPH bands between 2269
and 2372 cm-1 13
NMR spectroscopy demonstrates changes
.
1
in the 31P chemical shifts and increases in the JPH-coupling
constant, consistent with primary phosphine coordination.14 In
some cases the 31P and 1H NMR spectra are not first order, and
spectral simulation is required to extract the P-H and other
coupling constants, as discussed in more detail below. The
NMR and IR data characterizing the P-H bonds in 1-9, in
comparison to free L and the borane complex L-BH3 (11,
prepared from BH3‚THF and isolated as white needles, see the
Experimental Section for details), are summarized in Table 1.
1
For complexes 1-9, increase of the JPH for PH2Mes on
Complexes 1-4 react with triphos or tetraphos [P(CH2CH2-
PPh2)3] to yield the derivatives Ni(triphos)L (6), [Cu(triphos)L]-
PF6 (7), [Ag(triphos)L]BF4 (8), and [Pd(tetraphos)L][BF4]2 (9).
These substitutions proceed quickly at room temperature, except
for the synthesis of 6, which requires heating at 55 °C for 3 d.
Compounds 6 and 8-9 can also be prepared by treating the
metal precursors with a 1:1 mixture of triphos or tetraphos and
L, as shown in Scheme 2. Complexes 5-7 and 9 are stable to
air and water as solids and in solution.
coordination to the metal centers is consistent with the change
in the hybridization of P to sp3 leading to more s character in
the P-H bond.14 Coordination to BH3 in 11 gives a similar
result. Literature data on a variety of phosphorus compounds
show that electronegative substituents X on P increase P-H
couplings, an effect rationalized by increased p character in the
P-X bonds, which leads to increased s character in the P-H
bonds.15 Our data on the new metal complexes are consistent
with these reports. In particular, the largest 1JPH couplings are
found for dicationic 2 and 9, in which the metal center should
be the most electron-withdrawing. Comparison of the formally
isoelectronic Ni complexes 1 and 5 also fits this trend; the
complex with the π-acceptor nitrosyl group exhibits a larger
Compound 8 is stable as a solid but slowly decomposes in
CH2Cl2 solution by loss of L. The resulting precipitate,
identified as Ag(triphos)BF4 (10), could be isolated from the
reaction of AgBF4 and triphos in THF/nitromethane. Its 31P{1H}
NMR spectrum shows two doublets centered at δ 0.3 due to
107Ag-31P and 109Ag-31P coupling (J ) 511 and 591 Hz,
1
1JPH. The JPH values for the Cu(I) and Ag(I) complexes are
similar, as expected from the electronegativities of these ions.16
If, as expected, the tertiary phosphine ligands in 6-9 are
better σ donors than the primary phosphines in 1-4, then the
respectively). The observed 1J
P/1J
107Ag31
109Ag31
P ratio of 0.865 is
in good agreement with that calculated from the gyromagnetic
ratios of the Ag nuclei γ(107Ag)/γ(109Ag) ) 0.870.10 These
Ag-P coupling constants are similar in magnitude to those
reported for Ag(triphos)X (X ) NO3, ClO4).11 We assume that
10 contains coordinated BF4, as in structurally characterized
Cu(PPh3)3BF4.12 When complex 10 is treated with L in
+
ML3 fragment should be less electronegative in 6-9 and the
P-H coupling should be less than that in 1-4. However, the
observed P-H couplings for homoleptic 1-4 are very similar
to those for the triphos and tetraphos complexes 6-9. These
(13) Nakamoto, K. Infrared Spectra of Inorganic and Coordination
Compounds, 4th ed.; John Wiley and Sons: New York, 1986.
(14) Verkade, J. G.; Mosbo, J. A. In Phosphorus-31 NMR Spectroscopy in
Stereochemical Analysis; Verkade, J. G., Quin, L. D., Eds.; VCH:
Deerfield Beach, FL, 1987; pp 425-463.
(15) For some examples, see: (a) Vande Griend, L. J.; Verkade, J. G. J.
Am. Chem. Soc. 1975, 97, 5958-5960. (b) McFarlane, W.; White, R.
F. M. J. Chem. Soc., Chem. Commun. 1969, 744. (c) Anderson, D.
W. W.; Ebsworth, E. A. V.; Meikle, G. D.; Rankin, D. W. H. Mol.
Phys. 1973, 25, 381-385. See also ref 2 and references therein.
(16) Huheey, J. E.; Keiter, E. A.; Keiter, R. L. Inorganic Chemistry:
Principles of Structure and ReactiVity, 4th ed.; HarperCollins: New
York, 1993; pp 182-199.
(7) The homoleptic secondary phosphine complex Ni(PHPh2)4 was also
prepared by reduction of Ni(II) with excess phosphine. See: Weston,
C. W.; Bailey, G. W.; Nelson, J. H.; Jonassen, H. B. J. Inorg. Nucl.
Chem. 1972, 34, 1752-1755.
(8) Berglund, D.; Meek, D. W. Inorg. Chem. 1972, 11, 1493-1496.
(9) Richter-Addo, G. B.; Legzdins, P. Metal Nitrosyls; Oxford University
Press: Oxford, 1992.
(10) Wu, G.; Wasylishen, R. E.; Pan, H.; Liu, C. W.; Fackler, J. P., Jr.;
Shang, M. Magn. Reson. Chem. 1995, 734-738.
(11) Camalli, M.; Caruso, F. Inorg. Chim. Acta 1990, 169, 189-194.
(12) Gaughan, A. P., Jr.; Dori, Z.; Ibers, J. A. Inorg. Chem. 1974, 13, 1657-
1667.

6710 Inorganic Chemistry, Vol. 35, No. 23, 1996
Kourkine et al.
Figure 1. Experimental (top) and simulated (bottom) 31P NMR spectra
(22 °C) of Ni(PH2Mes)4 (1) in C6D6.
Figure 4. Experimental (top) and simulated (bottom) 1H NMR spectra
(22 °C) of [Pd(PH2Mes)4][BF4]2 (2) in CD2Cl2. The peak at 5.32 ppm
is due to residual protons in the solvent.
fit with the experimental spectra.) The higher symmetry of
1
complex 1 leads to only three coupling constants: JPH ) 283,
3
2JPP ) 33, and JPH ) 18 Hz. We assume that, in both cases,
4JHH is zero, and simulations with 4JHH ) 1-2 Hz did not affect
the spectra.
2
As expected, the JPP coupling in these complexes depends
Figure 2. Experimental (top) and simulated (bottom) 1H NMR spectra
(22 °C) of Ni(PH2Mes)4 (1) in C6D6.
on the P-M-P angle and is greater in 2 than in 1. Although
many homoleptic tertiary phosphine complexes [ML4]n+ are
2
2
known, JPP can rarely be determined. Some JPP values,
however, have been reported for homoleptic nickel(0)-fluoro-
phosphine complexes18a-c and for [PtL4]2+ (L ) PMe3,
P(OMe)3);18d these are similar in magnitude to those found for
1 and 2, respectively.
The 31P{1H} NMR spectrum of 6 (AX3 spin system) shows
a doublet at δ 9.8 (triphos) and a quartet at δ -96.2 (L) with
a small cis coupling constant (2JPP ) 11 Hz), which appear as
a broad singlet and a broad triplet, respectively, in the 31P
1
NMR spectrum (1JPH ) 281 Hz). In the H NMR spectrum
the PH protons appear at δ 6.23 (C6D6) with an additional 3JPH
of 15 Hz.
Figure 3. Experimental (top) and simulated (bottom) 31P NMR spectra
(22 °C) of [Pd(PH2Mes)4][BF4]2 (2) in CD2Cl2.
The 31P{1H} NMR spectrum of 9 (AM3X spin system)
exhibits a doublet at δ 160.2 (tetraphos apical P, 2JPPtrans ) 323
observations could be explained in two different ways: either
the primary phosphines and tertiary phosphines have similar
σ-donor ability or the P-H coupling is insensitive to fine points
like the nature of the phosphine ligand and is affected only by
large changes in charge at the metal center.
2
Hz), a doublet at δ 47.3 (tetraphos equatorial P, JPPcis ) 38
Hz) and a doublet of quartets at δ -94.5 (L). The coupling
between the two nonequivalent phosphorus species of the
tetraphos is either too small to be observed or, as suggested by
a referee, 3JPP (through-backbone) and 2JPP (through-metal) may
have equal magnitudes and opposite signs.19 The PH protons
in the 1H NMR spectrum appear as a doublet of multiplets with
additional unresolved coupling to the tetraphos P nuclei. For
comparison, the 31P{1H} NMR spectrum of the related complex
[Pd(tetraphos)P(OMe)3][BF4]2 exhibits two doublets [tetraphos
apical P and P(OMe)3, the coupling constant was not reported]
and a singlet (tetraphos equatorial P).20
31P NMR spectroscopy provides information on the solution
structure and dynamic behavior (intramolecular exchange) of
the homoleptic complexes and their derivatives. For the Ni(0)
and Pd(II) complexes 1-2, 6, and 9, the spectra are consistent
either with slow intramolecular exchange of L (on the 31P NMR
time scale) or with the complete absence of such an exchange
process. The 31P NMR spectra of 1-2 show very complex
multiplets (AA′A′′A′′′X2X′2X′′2X′′′2 spin systems). Using a
version of the program Laocoon17 modified to handle up to 12
spin-1/2 nuclei, we have successfully simulated these spectra.
In contrast, the 31P NMR spectra of the copper and silver
complexes 3-4 and 7-8, and nickel nitrosyl complex 5 do not
show P-P and long-range P-H couplings like those seen in
the nickel and palladium complexes.
The 31P{1H} NMR spectrum of 3 shows a broad peak at δ
-121.0 with no observable 63Cu-31P or 65Cu-31P coupling and
a sharp septet due to PF6. Improved resolution for the broad
Cu-L signal could not be obtained at -60 °C, but at 60 °C (in
1
These simulations also reproduce the experimental H NMR
resonances due to the PH protons of 1 and 2. These results
allow deduction of the 31P-1H and the 31P-31P coupling
constants in 1-2.
The experimental and simulated 31P and 1H NMR spectra of
1 and 2 are shown in Figures 1-4. The results are consistent
with the expected geometries: tetrahedral for the Ni(0) complex
1 and square planar for the Pd(II) complex 2. For 2, five
(18) (a) Crocker, C.; Goodfellow, R. J. J. Chem. Soc., Dalton Trans. 1977,
1687-1689. (b) Lynden-Bell, R. M.; Nixon, J. F.; Schmutzler, R. J.
Chem. Soc. A 1970, 565-567. (c) Kruck, T.; Hofler, M.; Jung, H.
Chem. Ber. 1974, 107, 2133-2144. (d) Goodfellow, R. J.; Taylor, B.
F. J. Chem. Soc., Dalton Trans. 1974, 1676-1684.
1
2
coupling constants are required: JPH ) 422, JPPtrans ) 394,
2JPPcis ) -30, 3JPHtrans ) -3, and 3JPHcis ) 6 Hz. (The cis and
trans 3JPH values were assigned arbitrarily and refined to a good
(19) Crumbliss, A. L.; Topping, R. J. In Phosphorus-31 NMR Spectroscopy
in Stereochemical Analysis; Verkade, J. G., Quin, L. D., Eds.; VCH:
Deerfield Beach, FL, 1987; pp 531-557.
(17) Castellano, S.; Bothner-By, A. A. J. Chem. Phys. 1964, 41, 3863-
3869.
(20) Aizawa, S.; Funahashi, S. Chem Lett. 1994, 2023-2026.

Metal Complexes of Mesitylphosphine
Inorganic Chemistry, Vol. 35, No. 23, 1996 6711
THF) a singlet, which appeared as a triplet (1JPH ) 317 Hz) in
the 1H-coupled spectrum, was observed. The chemical shift of
this signal did not change with temperature. These results can
be explained by rapid intramolecular exchange due to ligand
dissociation/reassociation and/or by quadrupolar broadening
induced by the 63/65Cu nuclei in an unsymmetrical environment.
Despite such broadening, 63/65Cu-31P coupling in homoleptic
phosphine complexes has been observed in high-symmetry
Scheme 3
In several cases, only one peak due to mesitylphosphine, with
an average chemical shift between that of free and coordinated
L, was observed. A solution containing 2 and a trace of L
showed only one signal at δ -90, which broadened considerably
on addition of more L, consistent with rapid ligand exchange.
Likewise, a mixture of 3 and L shows a broad peak at δ -130.0
and a sharp peak for the PF6 anion (δ -141.5), and similar
results are obtained for 4 (δ -145.9, sharp) and 8 [δ -13.5 (d,
1JAgP ) 197 Hz, triphos); δ -150.1 (L)]. For 7 + L, the
observed peaks both for the complex (δ -17.1 and -125.0)
and for L (δ -154.9) broaden and change slightly in chemical
shift (the signals due to free and coordinated L move toward
each other) while the PF6 signal (δ -141.5) remains sharp.
These observations are consistent with ligand exchange occur-
ring at an intermediate rate on the NMR time scale.
1
31
environments.21 For example, [Cu(PH2Ph)4]+ shows J63
Cu
P
) 800 Hz at 0 °C.22 We assume that the bulkier mesityl group
both increases the rate of dissociation of L and reduces the
effective symmetry at Cu in 3.
Likewise, no 63Cu-31P or 65Cu-31P coupling was observed
in the 31P{1H} NMR spectrum of 7, which exhibits two broad
signals at δ -18.0 (triphos) and -122.0 (L) and a sharp septet
(PF6) between -60 and 40 °C. The chemical shift of these
signals did not change with temperature. However, the PH
1
resonances in the H NMR spectrum of 7 show an additional
3JPH ) 9 Hz, which suggests that quadrupolar broadening, not
rapid exchange, is responsible for the lack of observed P-P
coupling.
In contrast, a mixture of 5 and 5 equiv of L exhibits a sharp
peak at δ -87.3 (1JPH ) 350 Hz, 5) and a broad peak at δ
-102.6. Use of 10 equiv of L changes the chemical shift of 5
Similarly, no 107,109Ag-31P coupling was observed in 4,
whose 31P{1H} spectrum shows a sharp singlet from 20 to -70
°C. In this case, the chemical shift slowly moved downfield
(δ 3.4 total) as the temperature was lowered to -70 °C. The
31P{1H} NMR spectrum of 8 exhibits a doublet at δ -15.3
(triphos, 1JAgP ) 197 Hz) and a broad peak at δ -139.0 (L) at
ambient temperature. These observations are consistent with
intramolecular exchange of L via dissociation/reassociation
equilibria fast on the NMR time scale, as reported for homoleptic
Ag(I) complexes of the more sterically demanding ligands P(p-
Tol)3 and P(OEt)3, which show a temperature-dependent chemi-
cal shift and Ag-P coupling only at low temperature.23 The
Ag-P coupling observed for 8 is similar to that (224-230 Hz,
depending on the counterion) reported in the tetraphosphine-
Ag(I) complex [Ag(P(p-Tol)3)4]+ at -80 °C.23
1
slightly (to δ -90.3, t, JPH ) 350 Hz) and that of the broad
peak more (to δ -149.3, with further broadening.) In addition
1
to peaks due to the Mes group, the H NMR spectrum of this
1
mixture shows a PH peak due to 5 at δ 5.30 (dm, JPH ) 350
Hz). At -70 °C, an additional PH resonance (δ 3.82, broad d,
1JPH ) 228 Hz) can be observed; this corresponds to a broad
triplet in the 31P NMR spectrum [δ -142.0 (1JPH ) 228 Hz)].
At this temperature, the 31P NMR signal due to 5 is observed
1
at δ -85.6 (t, JPH ) 350 Hz). These results are consistent
with intermolecular exchange of 5 with L at an intermediate
rate on the NMR time scale; the mechanism of exchange is
considered in more detail below.
These results are consistent with the observations on intramo-
lecular exchange and with the expected mechanisms for
intermolecular ligand substitution in these complexes. The
simplest cases are 1, 6, and 9, where the dissociative step
expected for substitution is slow on the NMR time scale, and
3, 4, and 8, where it is fast. The intermediate rate of exchange
observed for 7 is also consistent with its spectra in the absence
of added L. The rapid intermolecular substitution observed for
square planar 2 is consistent with the associative mechanism
expected for this reaction.
Compound 5 shows a triplet in the 31P NMR spectrum from
2
20 to -70 °C with no observable JPP. The chemical shift of
this signal did not change with temperature. This is strikingly
different from the behavior of the neutral Ni complexes 1 and
6. This observation can be explained by the strong labilizing
effect of the nitrosyl ligand. Alternatively, reversible isomer-
ization of 5 to a square planar complex with an anionic bent
nitrosyl could produce the observed spectra without requiring
dissociation of L.
The results for 5 are somewhat more complicated and may
be explained by the previously discussed reversible isomeriza-
tion of 5 to a square planar species with a bent nitrosyl ligand
(5b). Intermolecular exchange could then occur by an associa-
tive mechanism via the five-coordinate bent nitrosyl complex
[NiL4(NO)][BF4] (5c, Scheme 3). The broad peak observed
on addition of L to 5 could then be due to rapid exchange of L
with 5c, while rapid intramolecular isomerization of 5 and 5b
accounts for the lack of observed P-P and long-range P-H
coupling in 5. The IR spectrum of 5 in CH2Cl2 shows a peak
at 1818 cm-1 with a shoulder at 1788 cm-1; this signal, like
the one at 1751 cm-1 in KBr, is assigned to a linear NO. A
similar solution containing 10 equiv of added L produces an
additional peak at 1444 cm-1, which may be assigned to the
proposed bent NO ligand in 5c.9 The previously reported
equilibrium between [NiH(PEt3)3]+ and [NiH(PEt3)4]+, which
could also be observed by NMR, provides a precedent for this
proposed mechanism.24
Intermolecular ligand exchange in complexes 1-9 was
studied by 31P NMR spectroscopy at ambient temperature.
Addition of L (ca. 4 equiv) to solutions of complexes 1, 6, and
9 causes slow ligand exchange on the 31P NMR time scale. The
spectrum of a mixture of 1 and L shows two sharp peaks at δ
-86.8 (1) and -156.8 (L). The spectrum of a mixture of 6
and L shows three sharp peaks at δ 5.2 and -101.0 (coordinated
triphos and L in 6) and at δ -156.8 (L). For 9, the 31P{1H}
NMR spectrum shows signals at δ 157.5, 44.5, and -94.0 due
to 9 and at δ -157.3 (L). Ligand substitution does occur in 1,
6, and 9 as shown by the rapid (minutes) exchange of bound
PH2Mes on addition of PD2Mes; these reactions could also be
conveniently monitored by 31P NMR.
Faster ligand exchange on the 31P NMR time scale was
observed in similar experiments with complexes 2-5 and 7-8.
(21) King, R. W.; Huttemann, T. J.; Verkade, J. G. J. Chem. Soc., Chem.
Commun. 1965, 561.
(22) Black, J. R.; Levason, W.; Spicer, M. D.; Webster, M. J. Chem. Soc.,
Dalton Trans. 1993, 3129-3136.
(23) Muetterties, E. L.; Alegranti, C. W. J. Am. Chem. Soc. 1972, 94, 6386-
6391.
(24) English, A. D.; Meakin, P.; Jesson, J. P. J. Am. Chem. Soc. 1976, 98,
422-436.

6712 Inorganic Chemistry, Vol. 35, No. 23, 1996
Table 2. Crystallographic Data for 3, 1, 7‚(solvent), and 6
Kourkine et al.
3
1
7‚solvent
6
formula
fw
space group P1h
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
z
C36H52CuF6P5 C36H52NiP4 C50H52CuF6P5 C50H52P4Ni
817.2
667.4
P21/n
1009.3
P21/c
835.51
P1h
11.842(3)
18.584(4) 19.780(3)
11.506(3) 13.853(2)
18.965(4) 18.743(4)
9.587(1)
11.528(1)
22.194(2)
102.314(7)
90.921(9)
114.15(1)
2172.0(4)
2
11.891(2)
16.486(4)
72.62(2)
79.17(2)
68.14(2)
2048.3(9)
2
115.84(5) 90.20(1)
3650(2)
4
5136(1)
4
cryst color colorless
yellow-
green
1.215
colorless
yellow-
orange
1.278
D(calc), g
1.325
1.305
6.36
298
cm-3
µ(Mo KR), 7.80
7.30
6.28
293
cm-1
T, K
Figure 5. ORTEP drawing of [Ni(PH2Mes)4] (1), with 35% thermal
ellipsoids. Phenyl ring carbons were refined anisotropically, and
hydrogens have been removed for clarity.
296
252
radiation
R(F), %
R(wF), %
Mo KR (λ )0.710 73 Å)
5.94a
7.59
9.00a
10.64
8.94b
4.14b
14.96c
10.75c
a Quantity minimized ) ∑w∆2; R ) ∑∆/∑(Fo); R(w) ) ∑∆w1/2
/
2
∑(Fo w1/2), ∆ ) |(Fo - Fc)|. b Quantity minimized ) R(wF2) ) ∑[w(Fo
2
- Fc2)2]/∑[(wFo )2]1/2; R ) ∑∆/∑(Fo), ∆ ) |Fo - Fc)|. c R(wF2), %.
These NMR results on the dynamic behavior and ligand
substitution reactions of the mesitylphosphine complexes cor-
relate well with those reported in the literature for phosphine
complexes of Ni(0), Cu(I), and Ag(I) and reflect the small size
of L. As shown by Tolman,3a phosphine dissociation in Ni-
(PR3)4 complexes is controlled by sterics; larger ligands
dissociate more readily. The behavior of complexes 1 and 6 is
similar to that of Ni(PMe3)4, in which slow ligand dissociation
on the NMR time scale is demonstrated by the observation of
61Ni-31P coupling.25 Similar results, such as the slow intramo-
lecular ligand exchange at low temperature in [Cu(PMe3)4]+
and [Cu(PH2Ph)4]+ were observed in earlier studies of Cu(I)
complexes.22 The rapid intra- and intermolecular phosphine
exchange in silver complexes 4 and 8 is similar to that reported
by Muetterties in [AgL4]+ [L ) P(OEt)3, P(p-Tol)3].23 Although
homoleptic Pd(II) phosphine complexes, such as [Pd(PPh3)4]2+
and the triphenylphosphite analog, have been reported,26
information on their ligand substitution reactions is lacking.
Complementary information on the coordination properties
of L was obtained by structural characterization of compounds
1, 3, 6, and 7‚solvent by X-ray crystallography. Crystal, data
collection, and refinement parameters are given in Table 2.
ORTEP diagrams are shown in Figures 5-8. Tables 3 and 4
contain selected bond lengths and angles. Further details of
the structure determinations are given in the Experimental
Section and the Supporting Information.
Figure 6. ORTEP drawing of [Cu(PH2Mes)4]PF6 (3), with 35%
probability thermal ellipsoids. The anion is not shown, and hydrogens
have been removed for clarity.
To our knowledge, the only primary phosphine homoleptic
complexes reported previously are [Cu(PH2Ph)4]+22 and {Cu-
[1,2(PH2)2C6H4]2}+;27 the isoelectronic Ni(0) and Cu(I) com-
plexes 1 and 3 are the first ones to be structurally characterized.
Despite the large number of homoleptic nickel-phosphine
complexes NiL4, surprisingly few crystal structures are known.28
Consistent with the solution NMR data, complex 1 is tetrahedral
in the solid state, with an average Ni-P bond length of 2.149-
Figure 7. ORTEP drawing of Ni(triphos)PH2Mes (6), with 35%
probability thermal ellipsoids. Hydrogens have been removed for clarity.
(8) Å and Ni-P-Ni angles ranging from 106.9(3)° to 111.3-
(2)°. Complex 3 shows distorted tetrahedral geometry with two
sets of Cu-P distances [2.269(2) and 2.271(2) Å, 2.298(3) and
2.295(2) Å] and P-Cu-P angles ranging from 104.0(1)° to
116.5(1)°. The solution NMR data described above suggest that
the four phosphine ligands are equivalent in solution, so these
distortions may be a consequence of crystal packing forces.
(25) Avent, A. G.; Cloke, F. G. N.; Day, J. P.; Seddon, E. A.; Seddon, K.
R.; Smedley, S. M. J. Organomet. Chem. 1988, 341, 535-541.
(26) Yamazaki, S. Inorg. Chem. 1982, 21, 1638-1640.
(27) (a) Wild, S. B. Pure Appl. Chem. 1990, 62, 1139-1144. (b) Kang, Y.
B.; Pabel, M.; Pathak, D. D.; Willis, A. C.; Wild, S. B. Main Group
Chem. 1995, 1, 89-98.
(28) Hursthouse, M. B.; Izod, K. J.; Motevalli, M.; Thorton, P. Polyhedron
1994, 13, 151-153.

Metal Complexes of Mesitylphosphine
Inorganic Chemistry, Vol. 35, No. 23, 1996 6713
is more meaningful for the PH2Mes and PEt3 complexes, since
the short Ni-P bond distance in Ni(PF3)4 is likely a result of
the π-acceptor properties of PF3. Similarly, the average Cu-P
bond length in 3 of 2.28 Å is similar to that of the good σ-donor
ligand PMe3 in [Cu(PMe3)4]+ (av ) 2.26 Å, PMe3 cone angle
118°)30 and smaller than that in the more hindered [Cu(PPh3)4]+
[av ) 2.56 Å,31 PPh3 cone angle 145°].32
On the basis of effective nuclear charge of the d10 species
Cu(I) and Ni(0), the Cu-P bonds in 3 would be expected to be
shorter than the Ni-P ones in 1. Observation of the opposite
result suggests that the Ni-L interaction is stronger than the
Cu-L one; this may be attributed to π-back-bonding interactions
in the Ni, but not in the Cu complex.33 Orpen reported that
such interactions in PPh3 complexes led to increases in P-C
bond length and attributed this to occupation of P-C antibond-
ing orbitals.34 The P-C bond lengths in 1 range from 1.803-
(13) to 1.840(10) Åsnot significantly different from the P-C
bond length in the free phosphine [1.807(5) Å].4 However, a
similar trend is observed in the triphos complexes 6 and 7. The
Ni-P(L) bond length in 6 of 2.145(1) Å is significantly shorter
than the Cu-P(L) bond length in 7 (2.241(3) Å). In this case,
moreover, the P-C bond length in Ni-coordinated mesitylphos-
phine [1.857(2) Å] is significantly longer than that in the free
phosphine, which is similar to that in the copper complexes 3
[av ) 1.805(9) Å] and 7 [1.812(6) Å].
Figure 8. ORTEP drawing of [Cu(triphos)PH2Mes]PF6 (7‚solvent),
with 35% probability thermal ellipsoids. The anion and apparent solvent
are not shown, and hydrogens have been removed for clarity.
Table 3. Selected Bond Lengths (Å) and Angles (deg) for
Homoleptic Mesitylphosphine Complexes
complex
NiL4 (1)
[CuL4]PF6 (3)
M-P
2.142(7)
2.156(7)
2.158(7)
2.140(8)
1.803(13)
1.837(11)
1.840(10)
1.819(13)
111.2(3)
111.3(2)
109.5(2)
108.6(2)
109.2(2)
106.9(3)
2.271(2)
2.295(2)
2.298(3)
2.269(2)
1.795(9)
1.799(8)
1.820(6)
1.806(7)
113.5(1)
104.6(1)
106.5(1)
116.5(1)
104.0(1)
111.5(1)
A similar trend is apparent in the triphos complexes 6 and 7.
The Ni-P (triphos) bond length [av ) 2.1496(11) Å] is
significantly shorter than the Cu-P one [av ) 2.277(3) Å], and
the P-C(Ar) bonds in the triphos ligand are elongated in 6 [av
) 1.852(2) Å] but not in 7 [av ) 1.822(6) Å]. These results
suggest that a Ni-triphos back-bonding interaction in 6 is also
important.
P-C
P-M-P
Comparison of the M-L bond lengths in the structures of
the homoleptic and the triphos complexes suggest that bonding
of L to the three-coordinate fragments [ML3]n+ and [M-
(triphos)]n+ is similar, despite the steric and electronic differ-
ences of the supporting ligands. For example, the Ni-PH2-
Mes distance in 6 [2.1448(12) Å] and the Cu-PH2Mes distance
in 7 [2.241(3) Å] are similar to those in the homoleptic 1 (av
) 2.149(8) Å) and 3 [av ) 2.283(3) Å], respectively. These
results are consistent with the observations of very similar P-H
and P-P couplings in 1 and 6 observed by the NMR studies
described above.
Table 4. Selected Bond Lengths (Å) and Angles (deg) for Mixed
Triphos Mesitylphosphine Complexes
[Cu(triphos)L]PF6
complex
Ni(triphos)L (6)
(7‚solvent)
M-P (L)
2.1448(12)
2.1361(9)
2.145(1)
2.168(1)
1.857(2)
1.839(2)
1.867(2)
1.848(2)
1.864(2)
1.846(2)
1.850(2)
1.873(4)
1.870(4)
1.866(3)
117.01(4)
117.03(4)
125.72(4)
101.08(4)
94.82(4)
96.16(4)
2.241(3)
2.279(3)
2.289(3)
2.262(3)
1.812(6)
1.811(5)
1.818(6)
1.817(5)
1.839(5)
1.820(5)
1.828(5)
1.868(8)
1.859(8)
1.858(8)
119.5(1)
118.6(1)
125.8(1)
94.6(1)
M-P (triphos)
P-C (L)
P-C (Ph-triphos)
The geometries of the triphos ligands in 6 and 7 are normal.
The Ni-P distance and the P-Ni-P angle for the triphos in 6
(2.1496(11) Å and 97.35(4)° average) do not differ much from
those in the complex Ni(triphos)SO2 [2.205(3) Å and 94.9(1)°
average].35 Likewise, the Cu-P distance and the P-Cu-P
angle for the triphos in 7 [2.2767(3) Å and 95.34(10)° average]
P-C (CH2-triphos)
P-M-P (L)
(29) (a) Almenningen, A.; Andersen, B.; Astrup, E. E. Acta Chem. Scand.
1970, 24, 1579. (b) Marriot, J. C.; Salthouse, J. A.; Ware, M. J.;
Freeman, J. M. J. Chem. Soc., Chem. Commun. 1970, 595.
(30) Dempsey, D. F.; Girolami, G. S. Organometallics 1988, 7, 1208-
1213.
(31) Engelhardt, L. M.; Pakawatchai, C.; White, A. H.; Healy, P. C. J.
Chem. Soc., Dalton Trans. 1985, 125-133.
P-M-P (triphos)
97.2(1)
94.13(9)
(32) (a) McAuliffe, C. A. in ComprehensiVe Coordination Chemistry;
Wilkinson, G., Gillard, R. D., McCleverty, J. A., Eds.; Pergamon:
Oxford, 1987; Vol. 2; pp 1012-1029. (b) Tolman, C. A. Chem. ReV.
1977, 77, 313-348.
Comparison of these structures with those of other homoleptic
phosphine complexes suggests that the M-P bond lengths are
controlled by the steric bulk of the phosphine ligand, as well
as its donor properties. The Ni-P bond lengths in 1 [av )
2.149(8) Å] and the related compounds Ni(PF3)4 [Ni-P )
2.099(3) Å]29 and Ni(PEt3)4 [av Ni-P ) 2.2125(10) Å]28
correlate with the phosphine cone angles (PH2Mes, 110°; PF3,
104°; PEt3, 132°).3 As pointed out by a referee, this comparison
(33) Sakaki, S.; Kitaura, K.; Morokuma, K. Inorg. Chem. 1982, 21, 760-
765.
(34) (a) Orpen, A. G.; Connelly, N. G. Organometallics 1990, 9, 1206-
1210. (b) Dunne, B. J.; Morris, R. B.; Orpen, A. G. J. Chem. Soc.,
Dalton Trans. 1991, 653-661.
(35) Dapporto, P.; Midollini, S.; Orlandini, A.; Sacconi, L. Inorg. Chem.
1976, 15, 2768-2774.

6714 Inorganic Chemistry, Vol. 35, No. 23, 1996
Kourkine et al.
(24H, o-CH3), 2.12 (12H, p-CH3). 13C{1H} NMR (C6D6): δ 139.2
(Ar), 136.4 (Ar), 129.1 (p Ar), 21.8 (CH3), 21.4 (CH3). The expected
fourth Ar peak was not observed for complexes 1-3 and 5. 31P{1H}
NMR (C6D6): δ -86.5. 31P NMR (C6D6): δ -86.5 (m).
Alternatively, in an NMR scale experiment, to a solution of [Ni-
(NCMe)6][BF4]2‚0.5NCMe (20 mg, 0.04 mmol) in CH3NO2 (1 mL)
was added PH2Mes (40 mg, 0.25 mmol) to afford a yellow solution.
31P NMR spectroscopy showed that complex 1 was formed.
are comparable to those in Cu(triphos)Ph [2.304(2) Å and
92.0(1)°]36 and Cu(triphos)(BH4) [2.303(3) Å and 94.0(1)°
average].37
Conclusion. We have prepared and characterized a series
of metal complexes of mesitylphosphine and used X-ray
crystallography and NMR spectroscopy to examine its ligand
properties in the absence of P-H bond cleavage reactions. As
expected, the small cone angle of L reduces the rate of ligand
dissociation in the complexes in comparison to larger phos-
[Pd(PH2Mes)4][BF4]2 (2). To a solution of [Pd(NCMe)4][BF4]2 (150
mg, 0.34 mmol) in CH3NO2 (3 mL) was added PH2Mes (260 mg, 1.69
mmol) dissolved in CH3NO2 (1 mL), and the mixture was stirred
overnight. During this time a pale-yellow solid precipitated. The CH3-
NO2 solution was decanted, and the solid was recrystallized from CH2-
Cl2/Et2O at -20 °C to give 120 mg (40% yield) of the pale-yellow
product. Anal. Calcd for C36H52B2F8P4Pd‚0.54CH2Cl2: C, 46.79; H,
5.72. Found: C, 46.67; H, 5.83. The amount of cocrystallized CH2-
Cl2 was confirmed by integration of the 1H NMR spectrum. IR: 3050,
2969, 2915, 2368, 2301, 1603 (s), 1560, 1452 (s), 1378, 1055 (s, broad),
1
phines, and the JPH coupling constants can be used to gauge
the electronegativity of the metal center. The X-ray structural
studies are consistent with a Ni-L π-back-bonding interaction.
The data do not offer conclusive information on the relative
donor abilities of L and those of the more commonly used
tertiary phosphines. Comparison of the NMR properties and
structural parameters in homoleptic 1 and 3 and triphos
complexes 6 and 7 suggests that the basicity of triphos is similar
to that of three mesitylphosphines. However, it is not clear if
changes in parameters like JPP, JPH, and M-P bond distance
are a good measure of ligand donor ability. More structural
and spectroscopic studies including other primary phosphines
and metal centers will be required to see if these results are
general and to further develop the coordination chemistry of
primary phosphines.
889 (s), 851 (s) cm-1 1H NMR (CD2Cl2): δ 7.06 (8H, Ar), 5.09 (8H,
.
m, PH2), 2.38 (24H, o-CH3), 2.32 (12H, p-CH3). 13C{1H} NMR (CD2-
Cl2): δ 144.9 (Ar), 143.1 (Ar), 130.9 (Ar), 22.9 (CH3), 21.6 (CH3).
31P{1H} NMR (CD2Cl2): δ -73.2. 31P NMR (CD2Cl2): δ -73.2 (m).
[Cu(PH2Mes)4]PF6 (3). To an off-white slurry of [Cu(NCMe)4]-
[PF6] (500 mg, 1.34 mmol) in THF (3 mL) was added PH2Mes (860
mg, 5.63 mmol), and the resulting mixture was stirred overnight, during
which time a pink solid precipitated. The THF solution was decanted,
and the solid was washed with petroleum ether (2 × 5 mL). The
compound was recrystallized from THF/petroleum ether at -20 °C to
give 450 mg (45% yield) of pale-pink needles. Anal. Calcd for
C36H52CuF6P5: C, 52.90; H, 6.43. Found: C, 53.28; H, 6.41. IR:
3050, 2915 (s), 2736, 2357 (s), 2329 (s), 1602 (s), 1557(m), 1377 (s),
Experimental Section
General Experimental Details. Unless otherwise noted, all reac-
tions and manipulations were performed in dry glassware under a
nitrogen atmosphere at 20°C in a drybox or using standard Schlenk
techniques. Petroleum ether (bp 38-53 °C), ether, THF, and toluene
were dried over and distilled from Na/benzophenone before use.
CH2Cl2 was distilled from CaH2. Nitromethane (spectrochemical grade,
Aldrich) was used as received.
1294, 1083 (s), 1026 (s), 856 (s, broad), 702 (s) cm-1
.
1H NMR
1
(CDCl3): δ 6.78 (8H, Ar), 4.70 (8H, d, JPH ) 317 Hz, PH2), 2.45
(12H, p-CH3), 2.03 (24H, o-CH3). 13C{1H} NMR (CD2Cl2): δ 140.3
(Ar), 140.0 (Ar), 129.8 (Ar), 22.6 (CH3), 21.3 (CH3). 31P{1H} NMR
1
(CD2Cl2): δ -121.0 (broad), -141.5 (septet, JPF ) 711 Hz).
NMR spectra were recorded on a Varian 300 MHz spectrometer.
1H or 13C NMR chemical shifts are reported vs Me4Si and were
[Ag(PH2Mes)4]BF4 (4). To AgBF4 (50 mg, 0.26 mmol) was added
a solution of PH2Mes (164 mg, 1.08 mmol) in THF (4 mL), and the
resulting mixture was protected from light and stirred for 4 h. The
solvent was removed, and the off-white residue was washed with
petroleum ether (3 × 1 mL) and dried in vacuo to give 145 mg (70%
yield) of the white product, which slowly decomposes in solution. An
analytical sample was obtained by recrystallization from CH2Cl2/Et2O.
Anal. Calcd for C36H52AgBF4P4: C, 53.82; H, 6.53. Found: C, 53.33;
H, 6.53. Repeated analyses gave consistently low results for carbon.
We believe this is due to decomposition as described above. IR: 2961,
1
determined by reference to the residual H or 13C solvent peaks. 31P
NMR chemical shifts are reported vs H3PO4 (85%) used as an external
reference. Coupling constants are reported in hertz. Unless otherwise
noted, peaks in NMR spectra are singlets. IR (KBr) spectra were
recorded on a Perkin-Elmer 1600 series FTIR machine and are reported
in cm-1. Elemental analyses were provided by Schwarzkopf Microana-
lytical Laboratory.
The NMR simulations were performed on an IBM RS/6000-580
workstation using a version of Laocoon 3 which has been modified to
handle up to 12 spin-1/2 nuclei. Full details will be reported separately.
Unless otherwise noted, reagents were from commercial suppliers.
The following compounds were made by the literature procedures: PH2-
Mes,38 [Pd(NCMe)4][BF4]2,39 [Cu(NCMe)4][PF6],40 and [Ni(NCMe)6]-
[BF4]2‚0.5NCMe.41 PD2Mes was prepared from PCl2Mes and LiAlD4.5d
Ni(PH2Mes)4 (1). To a yellow solution of Ni(COD)2 (150 mg, 0.55
mmol) in THF (4 mL) was added a solution of PH2Mes (415 mg, 2.73
mmol) in THF (2 mL), and the resulting mixture was stirred overnight.
The volume of the solvent was reduced to 2 mL, petroleum ether was
layered on top, and the cooling of the solution to -20 °C gave 240 mg
(66% yield) of air-sensitive pale-green needles.
1
2372, 1732, 1603 (s), 1559, 1457 (s), 1375, 1294, 1170, 1035 (s). H
NMR (CD2Cl2): δ 6.85 (d, 4JPH ) 3, Hz, 8H, Ar), 4.62 (d, 1JPH ) 297
Hz, 8H, PH2), 2.28 (24H, o-CH3), 2.11 (12H, p-CH3). 13C{1H} NMR
(CD2Cl2): δ 140.1(d, JPC ) 9 Hz, Ar), 139.7 (Ar), 129.5 (d, JPC ) 6
Hz, Ar), 118.8 (d, JPC ) 23 Hz, Ar), 22.5 (d, JPC )11 Hz, CH3), 21.1
(CH3). 31P{1H} NMR (CD2Cl2): δ -131.5. 31P NMR (CD2Cl2): δ
1
-131.5 (t, JPH ) 297 Hz).
[Ni(PH2Mes)3(NO)]BF4 (5). To a pale-green solution of Ni(PH2-
Mes)4 prepared in situ by mixing Ni(COD)2 (30 mg, 0.1 mmol) and
PH2Mes (68 mg, 0.45 mmol) in THF (2 mL) was added NOBF4 (15
mg, 0.13 mmol) to afford a purple solution. The solvent was removed,
and the residue was washed with petroleum ether (2 × 5 mL) and dried
to give 56 mg (80% yield) of the product. The complex can be
recrystallized from a dilute solution in CH2Cl2/diethyl ether. Anal.
Calcd for C27H39BF4NNiOP3: C, 51.30; H, 6.23; N, 2.21. Found: C,
51.14; H, 6.52; N, 1.90. IR: 2936 (vs), 2854 (vs), 2359, 2336, 1751
(broad, s), 1605 (s), 1558 (s), 1456 (broad, s), 1374 (s), 1113 (s), 847
Anal. Calcd for C36H52NiP4: C, 64.78; H, 7.87. Found: C, 64.64;
H, 8.07. IR: 2959 (s), 2914 (s), 2278 (s), 1603 (s), 1560, 1463 (s),
1371, 1148, 1059 (s), 1025 (s), 861 (vs), 840 (vs), 702, 627, 612, 570,
546 cm-1 1H NMR (C6D6): δ 6.69 (8H, Ar), 4.83 (8H, m, PH2), 2.17
.
(36) Gambarotta, S.; Strologo, S.; Floriani, C.; Chiesi-Villa, A.; Guastini,
C. Organometallics 1984, 3, 1444-1445.
(s), 702, 620 cm-1
.
1H NMR (CD2Cl2): δ 6.93 (6H, Ar), 5.27 (6H,
1
dm, JPH ) 350 Hz, PH2), 2.30 (9H, p-CH3), 2.19 (18H, o-CH3).
13C{1H} NMR (CD2Cl2): δ 141.4 (Ar), 140.6 (Ar), 130.0 (Ar), 22.2
(CH3), 21.2 (CH3). 31P{1H} NMR (CD2Cl2) δ -89.7. 31P NMR (CD2-
(37) Ghilardi, C. A.; Midolini, S.; Orlandini, A. Inorg. Chem. 1982, 21,
4096-4098.
(38) (a) Oshikawa, T.; Yamashita, M. Chem. Ind. (London) 1985, 126-
127. (b) Reference 4.
1
Cl2) δ -89.7 (t, JPH ) 350 Hz).
(39) Thomas, R. R.; Sen, A. Inorg. Synth. 1990, 28, 63-67.
(40) Kubas, G. J. Inorg. Synth. 1979, 14, 68-71.
(41) Hathaway, B. J.; Holah, D. G.; Underhill, A. E. J. Chem. Soc. 1962,
2444-2448.
Ni(triphos)PH2Mes (6). A green mixture of Ni(PH2Mes)4 (200 mg,
0.3 mmol) and triphos (187 mg, 0.3 mmol) was heated in THF (10
mL) at 55 °C for 3 d. The solvent was removed, and the product was

Metal Complexes of Mesitylphosphine
Inorganic Chemistry, Vol. 35, No. 23, 1996 6715
recrystallized from petroleum ether (10 mL) to afford 160 mg (72%
yield) of a yellow solid.
The solution was filtered, and an orange voluminous solid was
precipitated by addition of diethyl ether (15 mL). The solvent was
decanted, and the solid was washed with diethyl ether (15 mL) and
dried in vacuo to give 260 mg (70% yield) of analytically pure 9.
Anal. Calcd for C51H55B2F8P5Pd: C, 55.50; H, 5.04. Found: C,
55.39; H, 5.03. IR: 3052, 2917, 2301, 1604, 1571, 1482, 1435 (vs),
1306, 1280, 1061 (broad, vs), 998 (s), 888, 852, 828, 806, 744 (vs),
Alternatively, a mixture of Ni(COD)2 (100 mg, 0.36 mmol), triphos
(227 mg, 0.36 mmol), and PH2Mes (72 mg, 0.47 mmol) in THF (20
mL) was heated at 60 °C for 4 d. The solvent was removed, and the
residue was recrystallized from petroleum ether to give 200 mg (66%
yield) of the product.
Anal. Calcd for C50H52NiP4: C, 71.87; H, 6.29. Found: C, 70.45;
H, 6.50. Repeated analyses, even on crystals from the batch used for
X-ray crystallography, gave consistently low results for carbon. IR:
3050, 2918, 2269, 1480, 1434 (vs), 1183, 1118, 1096 (s), 1026, 998,
724 (s), 698 (vs), 520 (vs), 484 (s) cm-1
.
1H NMR (CD3NO2): δ 7.62
(5H, m, Ar), 7.41 (12H, m, Ar), 7.21 (13H, m, Ar), 6.9 (2H, d, JPH
)
4 Hz, Ar), 5.64 (dm, 1JPH ) 381 Hz, PH2), 3.10 (4H, br, CH2), 3.00-
2.8 (8H, br, CH2), 2.31 (3H, p-CH3), 1.74 (6H, o-CH3). 13C{1H} NMR
(CD3NO2): δ 144.7 (Ar), 143.1 (d, JPC ) 9 Hz, Ar), 135.4 (m, Ar),
134.2 (m, Ar), 133.6 (m, Ar), 133.4 (Ar), 132.1-131.6 (m, Ar), 131.1
(m, Ar), 31.1 (m, CH2), 27.6 (m, CH2), 23.5 (d, JPC ) 10 Hz, o-CH3),
882, 847, 738 (s), 695 (vs), 566, 515 (vs) cm-1
.
1H NMR (C6D6): δ
1
7.22 (12H, m, Ar), 6.81 (20H, m, Ar), 6.23 (2H, d quart, JPH ) 281
3
4
Hz, JPH ) 15 Hz, PH2), 2.60 (6H, CH2), 2.25 (6H, d, JPH ) 6 Hz,
o-CH3), 2.18 (3H, p-CH3), 1.21 (3H, CH3). 13C{1H} NMR (CD2Cl2):
δ 142.9 (m, Ar), 139.0 (d, JPC ) 8 Hz, Ar), 136.3 (Ar), 131.9 (m, Ar),
129.1 (d, JPC ) 5 Hz, Ar), 128.7 (Ar), 127.4 (Ar), 127.3 (Ar), 38.8
(CH2), 38.0 (MeC), 36.9 (CH3), 22.5 (d, JPC ) 10 Hz, o-CH3), 21.0
(p-CH3). 31P{1H} NMR (C6D6): δ 9.8 (d, 2JPP ) 11 Hz), -96.2 (quart,
2
21.3 (p-CH3). 31P{1H} NMR (CD3NO2): δ 160.2 (d, JPPtrans ) 323
Hz), 47.3 (d, 2JPPcis ) 38 Hz), -94.5 (d quart, 2JPPtrans ) 323 Hz, 2JPPcis
) 38 Hz).
Ag(triphos)BF4 (10). To a slurry of AgBF4 (40 mg, 0.2 mmol) in
THF (4 mL) was added a solution of triphos (128 mg, 0.2 mmol) in 4
mL of 1:1 THF/CH3NO2 to afford a pale-yellow solution. The solvent
was removed, and the residue was washed with Et2O (2 × 10 mL) to
give 140 mg (83% yield) of a pale-yellow solid, which was recrystal-
lized from CH3NO2/Et2O for analysis. Anal. Calcd for C41H39-
AgBF4P3: C, 60.09; H, 4.81. Found: C, 59.86; H, 4.65. IR: 3054
(s), 2957, 1586, 1554 (s), 1483 (s), 1435 (vs), 1400, 1376, 1308, 1281,
1189, 1060 (broad, vs), 998 (vs), 850, 740 (vs), 694 (vs), 655, 512
2JPP ) 11 Hz). 31P NMR (C6D6): δ 9.8 (broad), -96.2 (broad t, JPH
1
) 281 Hz).
[Cu(triphos)PH2Mes]PF6 (7). To a pink solution of [Cu(PH2Mes)4]-
PF6 (40 mg, 0.05 mmol) in THF (2 mL) was added a solution of triphos
(31 mg, 0.05 mmol) in THF (2 mL), and the resulting mixture was
stirred for 30 min. Addition of petroleum ether (8 mL) precipitated
an off-white solid. The solvent was decanted, and the solid was dried
in vacuo to give 33 mg (67% yield) of the product. Anal. Calcd for
C50H52CuF6P5: C, 60.94; H, 5.33. Found: C, 61.14; H, 5.53. IR:
3053, 2920, 2346, 1604, 1482, 1435 (vs), 1381, 1095 (s), 1026, 1000,
(vs), 476 cm-1
.
1H NMR (CD3NO2): δ 7.53-6.80 (30H, m, Ar), 2.78
(3H, CH3), 2.08 (6H, CH2). 31P{1H} NMR (CD3NO2): δ -0.3 [d,
1
1
109
107
J
) 511 Hz, J
) 591 Hz].
AgP
AgP
839 (vs), 738 (s), 695 (vs), 557 (s), 515 (vs) cm-1
.
1H NMR (CD2-
H3B‚PH2Mes (11). To PH2Mes (500 mg, 3.3 mmol) was added
1
Cl2): δ 7.23 (5H, m, Ar), 7.04 (27H, m, Ar), 5.69 (2H, d quart, JPH
) 317 Hz, 3JPH ) 9 Hz, PH2), 2.54 (6H, broad, CH2), 2.43 (6H, o-CH3),
2.38 (3H, p-CH3), 1.69 (3H, broad m, CH3). 13C{1H} NMR (CD3-
NO2): δ 141.0 (Ar), 140.9 (Ar), 135.0 (m, Ar), 132.6 (m, Ar), 130.9
(Ar), 130.5 (d, JPC ) 6 Hz, Ar), 129.6 (m, Ar), 38.3 (m, CH3), 36.6
(MeC), 36.1 (broad, CH2), 23.0 (d, JPC ) 10 Hz, o-CH3), 20.8 (p-CH3).
The expected eighth Ar peak was not observed for complexes 7 and 8.
31P{1H} NMR (CD2Cl2) (40° to -30 °C): δ -18.0 (broad), -122.0
BH3‚THF (4 mL of a 1 M THF solution, 4 mmol). After 1 h, the
solvent was pumped off and the white residue was recrystallized from
THF/petroleum ether to give a white solid, 456 mg (83% yield). An
analytical sample (colorless needles, mp 83 °C) was obtained by another
recrystallization from ether/petroleum ether. Anal. Calcd for C9H16-
BP: C, 65.10; H, 9.73. Found: C, 65.32; H, 9.95. IR: 2978, 2937,
2918, 2866, 2739, 2556, 2392 (vs and broad), 2246, 1786, 1746, 1603,
1560, 1448, 1416, 1381, 1296, 1161, 1128, 1095, 1055, 1030, 956,
1
(broad), -141.1 (septet, JPF ) 711 Hz).
944, 901, 855, 742, 685, 630, 574, 553, 542, 530 cm-1 1H NMR
.
(C6D6): δ 6.51 (d, J ) 3 Hz, 2H, Ar); 4.81 (d quartets, JPH ) 370 Hz,
JBH ) 7.8 Hz, 2H, PH2); 2.07 (6H, o-Me); 1.97 (3H, p-Me). 31P{1H}
NMR (C6D6): δ -68.8 (broad d, J ) 32 Hz). 31P NMR (C6D6): δ
-68.8 (broad t, JPH ) 370 Hz). 13C{1H} NMR (C6D6): δ 141.3 (d, J
) 7.7 Hz, quaternary Ar), 137.5 (d, J ) 108.5 Hz, quaternary Ar),
129.9 (d, J ) 8.3 Hz, m-Ar), 118.1 (d, J ) 57.3 Hz, quaternary Ar),
21.8 (d, J ) 8.8 Hz, o-Me), 21.4 (p-Me).
Crystallographic Studies. Crystal, data collection, and refinement
parameters are given in Table 2. The systematic absences in the data
for 1 and 7‚solvent are uniquely consistent for the reported space
groups. No evidence of symmetry higher than triclinic was observed
in either the photographic or diffraction data for 3 and 6. Solution in
P1h yielded chemically reasonable and computationally stable results.
For 1, all inspected crystals were either twinned or multiple. Ap-
proximately 5% of the data were rejected because of asymmetrical
profiles caused by diffraction contributions from a minor twin or satellite
crystal.
The structures were solved by direct methods, completed by
subsequent difference Fourier syntheses, and refined by full-matrix least-
squares procedures. No absorption corrections were necessary because
of the <10% variation of the azimuthal scans. To conserve data in 1,
phenyl carbon atoms were isotropically refined as part of rigid bodies.
Attempts to model two residual peaks in the difference map of 7‚-
solvent, which were located near an inversion center and away from
the compound molecule, as a chemically recognizable solvent were
unsuccessful; they were assigned carbon atom identities and refined
isotropically. All other non-hydrogen atoms were refined with aniso-
tropic displacement coefficients. Hydrogen atoms were treated as
idealized contributions except those for the apparent solvent molecule
in 7‚solvent which were ignored.
[Ag(triphos)PH2Mes]BF4 (8). To a pale-yellow solution of [Ag(PH2-
Mes)4]BF4 (100 mg, 0.12 mmol) in THF (5 mL) was added a solution
of triphos (76 mg, 0.12 mmol) in THF (4 mL). A white solid was
precipitated by addition of petroleum ether to give 90 mg (75% yield)
of the spectroscopically pure product. The complex decomposes in
polar solvents (CH2Cl2, CH3NO2) to Ag(triphos)BF4 (10) and L over a
period of days and therefore was not obtained analytically pure by
recrystallization. This decomposition occasionally caused difficulties
in the NMR analysis, because the liberated PH2Mes exchanged with
the coordinated L, causing changes in the chemical shift and the
1
JPH of the PH protons in the H NMR spectrum and of coordinated L
in the 31P NMR spectrum. Anal. Calcd for C50H52AgBF4P4: C,
61.81; H, 5.41. Found: C, 62.65; H, 5.42. IR: 3062, 2958, 2347,
1604, 1482, 1435 (vs), 1378, 1058 (broad, vs), 999, 826 (s), 743
(vs), 696 (vs), 514 (vs) cm-1 1H NMR (CD2Cl2): δ 7.24 (5H, m,
.
Ar), 7.1 (25H, m, Ar), 6.96 (4H, Ar), 4.70 (4H, d, broad, 1JP-H ) 280
Hz, PH2), 2.48 (6H, CH2), 2.37 (12H, o-CH3), 2.31 (6H, p-CH3), 1.55
(3H, CH3). 13C{1H} NMR (CD2Cl2): δ 140.5 (Ar), 140.4 (Ar), 139.1
(Ar), 133.9 (m, Ar), 132.3 (m, Ar), 130.6 (Ar), 129.3 (broad, Ar), 38.7
(MeC), 37.2 (CH2), 36.8 (CH3), 22.9 (d, JPC ) 11 Hz, o-CH3), 21.1
1
(p-CH3). 31P{1H} NMR (CD2Cl2): δ -15.3 (d, JAgP ) 197 Hz),
1
-133.7 (broad). 31P NMR (CD2Cl2): δ -15.3 (d, JAgP ) 197 Hz),
-133.7 (broad t).
[Pd(tetraphos)PH2Mes][BF4]2 (9). To a pale-yellow solution of
[Pd(PH2Mes)4][BF4]2 (100 mg, 0.011 mmol) in CH2Cl2 (3 mL) was
added a solution of tetraphos (76 mg, 0.011 mmol) in CH2Cl2 (3 mL)
to afford a red solution, from which an orange solid was precipitated
by addition of diethyl ether (15 mL). The solid was filtered and dried
in vacuo to give 75 mg (30% yield) of the pure product.
Alternatively, a solution of tetraphos (226 mg, 0.34 mmol) in CH2-
Cl2 (3 mL) was mixed with a gray slurry of [Pd(NCMe)4][BF4]2 (150
mg, 0.34 mmol) in CH2Cl2 (4 mL) to afford a pale yellow-green slurry,
to which PH2Mes (51 mg, 0.34 mmol) was added to give a red solution.
All software and sources of the scattering factors are contained in
either the SHELXTL (5.3) or the SHELXTL PLUS (4.2) program
libraries (G. Sheldrick, Siemens XRD, Madison, WI).

6716 Inorganic Chemistry, Vol. 35, No. 23, 1996
Kourkine et al.
Addition of PH2Mes to Complexes 1-9. The general procedure
follows that described here for 2. PH2Mes (14 mg, 0.092 mmol) was
added to a solution of 2 (20 mg, 0.023 mmol) in CH2Cl2 (1.5 mL), and
the reaction was monitored by 31P{1H} NMR. Exchange in Pd
complexes 2 and 9 was also monitored similarly in nitromethane.
Aesar/Alfa for a loan of PdCl2 and Professor Russell Drago for
a helpful discussion.
Supporting Information Available: Tables of bond lengths, bond
angles, atomic coordinates and equivalent isotropic displacement coef-
ficients, and H-atom coordinates and isotropic displacement coefficients
for 1, 3, 6, and 7‚solvent (28 pages). Four X-ray crystallographic files,
in CIF format, are available. Access and ordering information is given
on any current masthead page.
Acknowledgments
We thank Dartmouth College and the Petroleum Research
Fund, administered by the American Chemical Society, for
partial support. S.V.M. thanks NECUSE for a summer under-
graduate research fellowship. We also thank Johnson-Matthey/
IC960529Q