Synthesis and characterisation of PCsp3P phosphine and phosphinite platinum(ii) complexes. Cyclometallation and simple coordination

Article abstract of DOI:10.1039/b712660c

New and improved preparative routes to the previously known PCP ligands cis-1,3-bis(di-isopropylphosphinito)cyclohexane and cis-1,3-bis[(di-tert- butylphosphino)methyl]cyclohexane are reported. They react with 1 equivalent of dichloro(1,5-cyclooctadiene)p

Full text of DOI:10.1039/b712660c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
3
Synthesis and characterisation of PCsp P phosphine and phosphinite
platinum(II) complexes. Cyclometallation and simple coordination†
Daniel Olsson, Athimoolam Arunachalampillai and Ola F. Wendt*
Received 16th August 2007, Accepted 12th September 2007
First published as an Advance Article on the web 28th September 2007
DOI: 10.1039/b712660c
New and improved preparative routes to the previously known PCP ligands cis-1,3-bis(di-
isopropylphosphinito)cyclohexane and cis-1,3-bis[(di-tert-butylphosphino)methyl]cyclohexane are
reported. They react with 1 equivalent of dichloro(1,5-cyclooctadiene)platinum(II) [(COD)PtCl2] to give
the cis coordinated complex cis-[PtCl2{cis-1,3-bis(di-isopropylphosphinito)}cyclohexane] and the
3
Csp -H activated complex trans-[PtCl{cis-1,3-bis(di-tert-butylphosphino)}cyclohexane]. The new PCP
ligand cis-1,3-bis(di-tert-butylphosphinito)cyclohexane was synthesised and reacts with [(COD)PtCl2]
giving the di-nuclear trans-[PtCl2{cis-1,3-bis(di-tert-butylphosphinito)cyclohexane}]2, which is highly
insoluble. All metal complexes were characterised with X-ray crystallography. DFT calculations
indicate that the inability of the phosphinite ligands to cyclometallate is due to a kinetic barrier,
possibly involving an axial–equatorial conformational change necessary for the C–H activation
process.
Introduction
Experimental
Transition metal complexes with PCP pincer ligands display a
rich chemistry with many applications in homogeneous catalysis.1
The strongly chelating nature of PCP ligands usually gives
high thermal and oxidative stabilities to their transition metal
complexes although it has been shown that there is decomposition
at high temperatures giving highly active metal(0) species.2 Most
General procedures and materials
All experiments with metal complexes and the phos-
phine/phosphinite ligands were carried out under an atmosphere
of nitrogen in a glovebox or using standard Schlenk or high
vacuum-line techniques. Caution! Several reactions are performed
in closed vessels at temperatures well above the boiling point of the
solvent. This is a potential explosion hazard and should only be
performed behind a blast shield using specially designed reaction
vessels. We use glass (3.2 mm thick and tested for pressures up
to 20 bar, giving a safe working pressure of ca. 10 bar) Strauss
flasks equipped with 8 mm Kontes Teflon valves as described
elsewhere.7 All non-deuterated solvents were vacuum-transferred
from sodium benzophenone ketyl directly to the reaction vessel.
Anhydrous CuCl2 was obtained by keeping CuCl2·2H2O at 140 ◦C
over night. Water used for washing in purification steps was
degassed with N2 for 2 h prior to use. KH in mineral oil was
obtained from Aldrich and washed with hexane three times prior
to use. All other commercially available reagents were purchased
from Sigma Aldrich and used as received. Commercial cis,trans-
1,3-cyclohexanediol consists of 55% cis and 45% trans according
to 1H NMR spectroscopy. The pure cis-isomer was obtained
2
investigated are the complexes of PCsp P type but pioneering work
by Shaw et al. describes the synthesis of numerous platinum–metal
3
3
complexes of PCsp P type. We have also reported on the synthesis
and catalytic applications of cyclohexyl based PCP ligands cis-1,3-
bis[(di-tert-butylphosphino)methyl]cyclohexane (1) and cis-1,3-
bis(di-isopropylphosphinito)cyclohexane (2).4 In particular ligand
2 behaved differently than the corresponding aromatic ligand and
did not give cyclometallation with palladium(II). A number of
complexes with aliphatic backbones have been described recently
in both nickel5 and rhodium6 chemistry.
In this paper we further extend the family of PCP ligands
with aliphatic cyclohexane backbones. The synthesis of the
phosphinite ligand cis-1,3-bis(di-tert-butylphosphinito)cyclo-
hexane 3 and desulfurisation of cis-1,3-bis[(di-tert-butylthio-
phosphoryl)methyl]cyclohexane 4 are described. Additionally, the
coordination chemistry involving cycloocta-1,5-diene platinum
dichloride [(COD)PtCl2] and cyclohexyl based PCP ligands is
presented. We also report on DFT calculations on these systems
and provide a possible explanation for the different reactivity of
phosphine and phosphinite based ligands.
using a modified literature method,8 and gave a satisfactory H
1
NMR spectrum.9 [(COD)PtCl2] was prepared according to the
literature.10 1H, 13C and 31P NMR spectra were recorded on a
Varian Unity INOVA 500 spectrometer working at 499.77 MHz
(1H). Chemical shifts are given in ppm downfield from TMS using
residual solvent peaks (1H and 13C NMR) or H3PO4 as reference.
NMR multiplicities are abbreviated as follows: s = singlet, d =
doublet, t = triplet, q = quartet, m = multiplet, b = broad,
v = virtual. Coupling constants for virtual triplets in 1H and 13
C
Organic Chemistry, Department of Chemistry, Lund University, P. O. Box
124, S-221 00 Lund, Sweden. E-mail: ola.wendt@organic.lu.se
† CCDC reference numbers 657410–657412. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b712660c
NMR spectra are reported according to Cohen and Sheppard.11
Elemental analyses were performed by H. Kolbe, Mu¨lheim an der
Ruhr, Germany.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 5427–5433 | 5427
©

View Article Online
Preparations
Method B. To a THF solution (30 cm3) of cis-1,3-cyclo-
hexanediol (0.58 g, 5.0 mmol), KH (0.40 g, 10.0 mmol) was slowly
added and the solution was stirred for 14 h. Then a solution of
di-t-butylchlorophosphine (1.9 cm3, 10.0 mmol) in THF (10 cm3)
was added to the reaction mixture and it was stirred for another
17 h. The solvent was evaporated, the mixture was dissolved in
ether (75 cm3) and filtered. Removal of ether afforded a colourless
viscous oil which crystallised slowly upon standing overnight.
Yield: 1.80 g (89%). Anal. Calc. for C22H46O2P2: C, 65.32; H, 11.46;
P, 15.31. Found: C, 65.18; H, 11.40; P, 15.26. 1H NMR (C6D6) d
ppm: 0.81 (tq, 1H, JH–H = 14.0 and JP–H = 3.5 Hz, Cy), 0.95–1.15
(m, 2H, Cy), 1.08 and 1.10 (d, 36H, 3JP–H = 5.0 Hz, t-Bu), 1.35–
1.45 (m, 2H, Cy), 1.95 (d, 2H, JH–H = 12.5 Hz, Cy), 2.80 (d, 1H,
JH–H = 12.5 Hz, Cy) and 3.50 (m, 2H, Cy). 13C NMR (C6D6) d
ppm: 20.84 (s, CH2, Cy), 27.73 and 27.61 (d, CH3, 2JP–C = 2.9 Hz,
t-Bu), 33.82 (d, CH2, 3JP–C = 4.4 Hz, Cy), 34.83 and 34.63 (d, CH2,
3JP–C = 6.5 Hz, Cy), 43.35 (vt, P–C(CH3)3, 1JP–C = 3.8 Hz, t-Bu),
cis-1,3-Bis[(di-tert-butylphosphino)methyl]cyclohexane (1).
A
thick-walled Strauss-flask containing cis-1,3-bis[(di-t-butylthio-
phosphoryl)methyl]cyclohexane 4 (0.31 g, 0.67 mmol) and LiAlH4
(1.05 g, 27.7 mmol) was evacuated and THF (40 cm3) was vacuum-
transferred into the mixture. The flask was placed in an oil bath
at 150 ◦C for 12 h before cooling to room temperature, releasing
positive pressure of H2S, purging with N2 and withdrawing 0.6 cm3
for NMR analysis. Thus, five NMR samples were taken during the
3 d reaction time and after each test three freeze–pump–thaw cycles
were carried out. The grey suspension was quenched by removal of
THF and H2S on the Schlenk line followed by addition of EtOAc
(5.5 cm3) and saturated Na2SO4 (aq) (5.5 cm3). THF (10 cm3) was
added for extraction and the clear, colour free organic solution
˚
was separated and dried over 4 A molecular sieves. Removal of
the solvent and drying the white sticky residue under vacuum gave
pure 1. Yield 0.24 g (90%). 1H, 13C{ H} and 31P{ H} NMR spectra
1
1
1
78.91 (d, P–O–CH, 2JP–C = 18.2 Hz, Cy). 31P{ H} NMR (C6D6) d
were consistent with previous reports.4a
1
ppm: 153.9 (s). 31P{ H} NMR (THF-d8): d 153.5 (s).
cis-[PtCl2{cis-1,3-bis(di-isopropylphosphinito)cyclohexane}] (5).
In a glovebox, 2 (0.48 g, 1.37 mmol) was added to a pale grey
suspension of [(COD)PtCl2] (0.52 g, 1.38 mmol) in THF (40 cm3)
at room temperature. Within minutes the reaction mixture turned
to a pale yellow solution and a 31P NMR spectrum confirmed that
no starting material was present in the solution. Evaporation of
the solvent and recrystallisation from THF–heptane (1 : 1) gave
crystals suitable for X-ray diffraction. Yield 0.61 g (72%).
cis-1,3-Bis(di-isopropylphosphinito)cyclohexane (2). Method A.
The literature procedure used for the corresponding cis/trans
mixture was applied.4b The crude reaction mixture was extracted
with Et2O and filtration followed by evaporation resulted in a
yellow oil in almost quantitative yield according to 31P NMR
spectroscopy.
Method B. To a THF solution (30 cm3) of cis-1,3-cyclo-
hexanediol (0.58 g, 5.0 mmol), KH (0.40 g, 10.0 mmol) was slowly
added and the solution was stirred for 14 h. Then a solution of
ClPiPr2 (1.6 cm3, 10.0 mmol) in THF (10 cm3) was added to
the reaction mixture and it was stirred for another 17 h. The
solvent was evaporated and the mixture was dissolved in ether
and filtered. Removal of the ether afforded a colourless viscous
oil which crystallised slowly upon standing. Yield: 1.25 g (72%).
NMR data were consistent with those reported in the literature.4b
Anal. Calc. for C18H38Cl2O2P2Pt: C, 35.19; H, 6.23. Found: C,
1
35.27; H, 6.19. H NMR (THF-d8): d = 0.90 (m, 1H, Cy), 0.96,
1.23, 1.35, 1.63, (dd, 24H, 3JP–H = 7.0 Hz and 3JH–H = 16.7 Hz, i-Pr
Me), 1.00 (m, 2H, Cy), 1.25 (m, 2H, Cy), 1.45 (m, 2H, Cy), 1.70 (m,
1
2H, Cy), 2.45 (m, 2H, i-PrCH), 3.25 (m, 4H, Cy and i-Pr). 13C{ H}
NMR (THF-d8): d = 16.78 (s, CyCH2), 18.42, 19.76, 21.53 (s, i-Pr
CH3), 29.49 and 32.98 (d, CH2, 3JP–C = 4.9 Hz, Cy), 32.02 (vt, CH,
1
3JP–C = 3.3 Hz, i-Pr), 71.50 (d, 2JP–C = 20.1 Hz, P–O–CH). 31P{ H}
NMR (THF-d8): d = 111.60 (1JP–Pt = 4028 Hz).
trans-[PtCl2{cis-1,3-bis(di-t-butylphosphinito)cyclohexane}]2 (6).
To a colourless solution of 3 (0.25 g, 0.62 mmol) in THF (15 cm3),
[(COD)PtCl2] (0.23 g, 0.62 mmol) was added and the mixture was
heated at 150◦ C overnight without stirring. Yellow crystals were
formed, and the crystals were filtered off and washed with hexane
(5 cm3), methanol (10 cm3) and ether (5 cm3) and dried. Yield:
0.36 g (86%). Anal. Calc for C45H96Cl4O5P4Pt2: C, 39.36; H, 7.05;
P, 9.02. Found: C, 39.25; H, 7.08; P, 9.18
cis-1,3-Bis(di-tert-butylphosphinito)cyclohexane (3). Method A.
A hexane solution of BuLi (1.6 M, 3.5 cm3, 5.6 mmol) was
carefully added to an ice-cold THF (15 cm3) solution of cis-1,3-
cyclohexanediol (0.291 g, 2.51 mmol). The ice-bath was removed
and the white suspension was allowed to stir for 2 h at room
temperature, before being cooled down to 0 ◦C. A THF (15 cm3)
solution of ClPtBu2 (1.00 cm3, 5.26 mmol) was added and the
suspension was aged for 2 h at room temperature. The white
suspension was heated at 160 ◦C overnight, which resulted in a
pale yellow solution with a white solid after sedimentation. The
organic phase was carefully separated using a cannula and the
white solid was washed with a new portion of THF (10 cm3). The
combined organic phases were evaporated, producing a semi-solid
white residue. The residue was distributed between H2O (10 cm3)
and Et2O (20 cm3), giving a pale yellow, transparent upper organic
phase and a whitish somewhat cloudy water phase. The Et2O was
separated and the water phase extracted with another portion
of Et2O (10 cm3). The combined organic phases were dried with
trans-[PtCl{cis-1,3-bis(di-tert-butylphosphino)cyclohexyl}] (7).
In a glovebox, 1 (0.31 g, 0.78 mmol) was added to a pale grey
suspension of [(COD)PtCl2] (0.29 g, 0.77 mmol) in THF (40 cm3)
in a Strauss-flask. The flask was put in an oil-bath and at 140 ◦C
a pale yellow solution resulted. After 30 min at 140 ◦C the
reaction solution was allowed to cool down to room temperature.
Evaporation of the solvent and recrystallisation from pentane gave
crystals suitable for X-ray diffraction. Yield 0.30 g (63%). Anal.
Calc for C24H49ClP2Pt: C, 45.75; H, 7.84; P, 9.83. Found: C, 45.83;
1
H, 7.79; P, 9.76. H NMR (THF-d8): d 0.90–2.10 (m region, CH
˚
4 A molecular sieves over night. Filtration and evaporation of the
& CH2), 1.38 (vt, 3JPH = 13.2 Hz, 18H, C(CH3)3), 1.42 (vt, 3JPH
=
1
4
solvent produced a waxy white residue. 31P NMR spectroscopy
13.3 Hz, 18H, t-Bu). 13C{ H} NMR (C6D6): d 27.8 (t, JPC
=
showed the product to be almost pure.
1.4 Hz), 29.7 (vt, 2JPC = 5.5 Hz, C(CH3)3), 30.2 (vt, 2JPC = 5.3 Hz,
5428 | Dalton Trans., 2007, 5427–5433
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 1 Crystal data for 5, 6 and 7
5
6
7
Formula
Fw
C18H38Cl2O2P2Pt
614.41
Pbca
12.8734(8)
14.8662(9)
24.0444(13)
90
4601.6(5)
8
1.774
6.480
2.26–32.89
43 390
8030
3388
C45H96Cl4O5P4Pt2
1373.08
C2/c
28.8263(11)
8.7302(2)
44.3883(10)
96.065(3)
11108.2(6)
8
1.642
5.379
2.21–32.77
40 857
17 354
C24H49ClP2Pt
630.11
P21/n
12.0176(8)
14.8027(9)
15.7279(9)
100.786(5)
2748.5(3)
4
1.523
5.327
2.21–32.72
24 215
9275
Space group
˚
a/A
˚
b/A
˚
c/A
b/◦
3
˚
V/A
Z
Dcalcd/g cm−3
l/mm−1
h range/◦
No. reflns collected
No. of unique reflns
No. of reflns I > 2r(I)
R(F) (I > 2r(I))a
wR2(F2) (all data)b
Sc
4742
0.0586
0.1724
0.735
4711
0.0347
0.0878
0.984
0.0607
0.1481
0.925
Rint
0.1426
0.0721
0.0410
a R = R (|Fo| − |Fc|)/R|Fo| b wR2 = [Rw(|Fo| − |Fc|)2/R|Fo|2]1/2 c S = [Rw(|Fo| − |Fc|)2/(m − n)]1/2
.
C(CH3)3), 33.8 (vt, JPC = 24.5 Hz), 34.3 (vt, 2JPC = 20.2 Hz), 36.9
with Raney-nickel or trichlorosilane have been unsuccessful16 and
we decided to use LiAlH4 in THF. The reaction of 4 with LiAlH4
in THF at 150 ◦C over 3 d gave 1 as a single product with H2S as
a by-product, cf. Scheme 1.
3
3
1
(vt, JPC = 20.9 Hz), 36.8 (vt, JPC = 16.9 Hz), 48.8 (s, JC–Pt
=
1
730 Hz, HC–Pt), 50.8 (vt, 2JPC = 14.7 Hz). 31P{ H} NMR (C6D6):
d 63.9 (s, JP–Pt = 3127 Hz).
Crystallography†
Intensity data were collected with an Oxford Diffraction Xcalibur
12
˚
3 system using x-scans and Mo-Ka (k = 0.71073 A) radiation.
Intensity data were extracted and integrated using Crysalis RED.13
The structures were solved by direct methods and refined by full-
matrix least-squares calculations on F2 using SHELXTL 5.1.14
Non-H atoms were refined with anisotropic displacement. All
hydrogen atoms were constrained to parent sites, using a riding
model. Crystal data and details about data collection are given in
Table 1.
Scheme 1
CCDC reference numbers 657410–657412.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b712660c
The reaction was conveniently monitored using 31P NMR
spectroscopy, and at the same time the excess hydrogen sul-
fide was ventilated from the reaction mixture. In accordance
with previous work4a the starting material and product dis-
play singlets at d = 76.3 and 20.7 ppm, respectively, in 31P
NMR spectra with THF-d8 as the solvent. Monitoring the
reaction mixture also revealed the presence of the mono-
sulfurated intermediate cis-(1-di-tert-butylthiophosphorylmethyl-
3-di-tert-butylphosphinomethyl)cyclohexane (8) seen as two peaks
in a 1 : 1 ratio at d = 76.6 and 20.4 ppm. The reaction progress
shows typical intermediate behaviour of 8, where it accumulates
to approximately 60% and then disappears again. After 100 h the
reaction is complete and compound 1 was obtained in excellent
yield and high purity through a simple work-up procedure.
DFT calculations
The geometries of the model compounds were optimised using
B3LYP DFT implemented by the programme Jaguar15 with the
lav3p basis set and electronic energies were calculated. Starting
geometries were obtained from the crystal structures of 5 and 7.
Results
Reduction of
1,3-cis-bis[(di-tert-butylthiophosphoryl)methyl]cyclohexane (4)
In previous syntheses of ligand 1 the purification and identification
of the product was hampered by the oxygen sensitivity of this
diphosphine and hence it has been converted to the corresponding
phosphine sulfide, which was isolated and fully characterised.4a
Previous attempts to cleanly remove the sulfur on the PCP ligand
Synthesis of phosphinite ligands 2 and 3
The cis isomer of 1,3-cyclohexanediol was separated from the
commercially available cis/trans mixture according to the protocol
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 5427–5433 | 5429
©

View Article Online
of Lehtonen et al.,8 involving complexation with anhydrous
CuCl2 followed by liberation of the cis-diol by treatment with
gaseous NH3. To generate cis-1,3-bis(di-isopropylphosphinito)-
cyclohexane (2) and cis-1,3-bis(di-tert-butylphosphinito)cyclo-
hexane (3) the synthetic strategy of Sjo¨vall et al. was initially
used.4b
Treating cis-1,3-cyclohexanediol with n-BuLi gave the corre-
sponding alkoxides that readily reacted with ClPiPr2 and ClPtBu2,
respectively, affording the phosphinite PCP ligands 2 and 3 in
sufficient purity according to 31P NMR spectroscopy, cf. Scheme 2.
A remarkable difference in reactivity was observed between these
two chlorophosphines. When ClPiPr2 was mixed with the alkoxide
at 0 ◦C the reaction started instantaneously as indicated by the
fact that the reaction suspension slowly turned into a solution as
the reaction mixture reached room temperature. Contrary to this,
the reaction of ClPtBu2 with the alkoxide required an elevated
temperature to start and was only driven to completeness at
160 ◦C, as indicated by 31P NMR spectroscopy. When using n-BuLi
as a deprotonation agent both phosphinite products were tedious
to purify and the product mixtures never reached analytical purity.
Continued purification also resulted in a large decrease in yield
and was always stopped when the 31P NMR spectrum showed
satisfactory purity. Using KH as the deprotonation agent giving
the potassium salts of the diol instead of the lithium salts was
a superior method. The potassium chloride formed during the
reaction was easily separated by dissolving the reaction mixture
in ether and filtering. Thus, ligands 2 and 3 were isolated in a
good yield and fully characterised, with the 31P NMR spectrum of
ligand 2 consistent with that reported in the literature.4b
Scheme 3
Fig. 1 DIAMOND drawing of 5 at the 30% probability level. Most
hydrogens and the methyl carbons on the tert-b◦utyl groups are omitted for
˚
clarity. Selected bond lengths (A) and angles ( ) with estimated standard
deviations: Pt–Cl1 2.374(2), Pt–Cl2 2.370(2), Pt–P1 2.238(2), Pt–P2
2.234(2), Pt · · · C1 3.922(9), Cl1–Pt–Cl2 86.96(8), P1–Pt–P2 100.59(8),
Cl1–Pt–P1 88.01(8), Cl2–Pt–P2 84.45(7).
The structure exhibits a distorted cis square-planar geometry.
Due to the rigid structure of the cyclohexane ring the chelate
causes slight deformation of the complex giving a P1–Pt–P2
angle of 100.59(8)◦. This is somewhat larger than the earlier
reported palladium iodine analogue4b and deviation is in general
expected in an eight-membered ring with a naturally large bite
angle. The diaxial substitution pattern on the cyclohexane ring
causes the ring to be positioned well above the plane of the
complex and the cyclohexane ring adopts the chair conformation,
where the bond angles compare fairly well to those in free
cyclohexane (111.5◦).17
Scheme 2
Reaction of ligands 2 and 3 with [(COD)PtCl2]
When a THF solution of 2 was added to a THF suspension
of [(COD)PtCl2] at room temperature the platinum precursor
dissolved within minutes, cf. Scheme 3. This resulted in a pale
yellow solution, which was analysed by 31P NMR spectroscopy
5 min after mixing. The spectrum revealed total consumption of 2
and evaporation of the solvent and recrystallisation from boiling
THF–heptane 1 : 1 yielded complex 5 as air-stable, colourless
Intramolecular hydrogen bonding is present between three
of the isopropyl hydrogens and two of the chlorines (Cl2–
˚
˚
˚
H9 2.726(2) A; Cl1–H8 2.724(2) A; H10–Cl2 2.961(2) A). Our
initial aim to activate the C1–H bond had not been achieved.
˚
It can be noted that the Pt–C1 distance is 3.92 A, which is far
crystals. The 31P{ H} NMR spectrum in THF-d8 gives rise to a
from an agostic interaction. Attempts to induce C–H activation
by using elevated temperature, treatment with silver triflate or
triethyl amine were not fruitful. Addition of silver triflate gave
immediate precipitation of silver chloride but the 31P NMR
spectrum indicated formation of several products none of which
was the desired cyclometallated product. Filtration through Celite
and heating did not enrich any possible product. Treating 5 with
1
singlet at d = 111.6 ppm with a 1JPtP of 4028 Hz, indicating a cis
geometry. The cis coordination mode has earlier been observed for
the iodo-palladium analogue.4b X-Ray diffraction of the crystals
of 5 confirmed the cis-structure with a diaxial arrangement of the
cyclohexyl ring. A perspective view of the molecular structure is
shown in Fig. 1 including selected bond lengths and angles.
5430 | Dalton Trans., 2007, 5427–5433
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
triethyl amine in an effort to deprotonate C1 had absolut◦ely no
effect nor had a heating of the complex overnight at 125 C. In
conclusion, trying to convert cis coordinated complex 5 to a trans,
C–H activated complex seems unfeasible. Ligand 2 apparently
prefers a diaxial conformation whereas C–H activation seems to
demand a diequatorial conformation.
Reaction of ligand 1 with [(COD)PtCl2]
When a THF solution of phosphine ligand 1 was added to a THF
suspension of [(COD)PtCl2] at room temperature no dissolution
of the platinum precursor was observed. To obtain a reaction
demanded a temperature of 140 ◦C. This differs substantially from
the phosphinite complexation. By increasing the temperature in
small steps an intermediate was observed when monitoring the
reaction by 31P NMR spectroscopy. The intermediate was observed
as a singlet (d = 37.4) with satellites (JPtP = 4282 Hz). It was not
isolated and disappeared upon further heating but we strongly
suspect it to be the cis coordinated complex 9, cf. Scheme 3.
Continued heating of the THF mixture of 1 and [(COD)PtCl2]
at 140 ◦C for 30 min resulted in a pale yellow solution containing
only compound 7. No base was needed to induce cyclometallation
Reaction of [COD]PtCl2 with the t-butyl analogue 3 in THF at
150◦ C afforded a material, which was not soluble in any of the
common organic solvents, precluding an NMR characterisation.
When the reaction was carried out without stirring, air-stable,
yellow crystals, which were filtered off and washed with methanol
and ether, were obtained. These were also insoluble but the
quality of the crystals enabled a single crystal X-ray diffraction
experiment and they were also characterised by elemental analysis.
A perspective view of the molecular structure is shown in Fig. 2.
The structure shows that there is no cyclometallation on the C1
carbon of the ligand and the complex obtained is a dimer in which
the platinum coordinates to two trans oriented phosphorus atoms
◦
and, interestingly, 7 is robust in a THF solution at 140 C with
one equivalent of HCl. Compound 7 gives rise to a 31P NMR
resonance at 63.9 ppm with a 1JPt–P of 3127 Hz indicating a trans
geometry. The cyclometallation is also substantiated in the 13C
NMR spectrum where the C1 resonance comes at 48.8 ppm with
platinum satellites (1JPt–C = 730 Hz). Our conclusions from the
NMR investigation were unambiguously confirmed by an X-ray
crystallographic analysis. A perspective view of the molecular
structure is shown in Fig. 3 including selected bond lengths and
angles.
˚
from two different ligands. The average Pt–P distance is 2.324 (3) A
and the average P–Pt–P angle is 176.6 (9)◦. Also the two chlorine
atoms bonded to the platinum are trans to each other. The Pt–Cl
˚
bond distances vary dramatically from 2.298(4) to 2.382(4) A and
the average Cl–Pt–Cl angle is 158.8 (3)◦, strongly deviating from
the expected 180◦. The overall geometry around the metal centre
is distorted square-planar. A similar type of complex has been
observed in the case of palladium using the isopropyl ligand 2.4b
We believe that the formation of a dinuclear complex instead of a
mononuclear one as in the case of the isopropyl analogue is due
to steric constraints. By adopting the structure shown in Fig. 2
the two bulky phosphine substituents can be trans oriented and
equatorial although no cyclometallation is obtained. Together the
16-membered ring forms a cage like structure with a methanol
(from the washing) as a guest. This methanol shows a disorder as
indicated by the unreasonably long C–O distance; any attempt to
resolve it failed. From a packing diagram it can be seen that the
cages form a tunnel-like structure along the b axis. However, from
a space-filling plot it is clear that the cage is accessible only from
one side, viz. the top side in the representation in Fig. 2.
Fig. 3 DIAMOND drawing of compound 7 at the 30% probability level.
˚
Hydrogen atoms are omitted for clarity. Selected bond lengths (A) and
angles (◦) with estimated standard deviations: Pt1–C1 2.065(4), Pt1–Cl1
2.434(2), Pt1–P1 2.303(3), Pt1–P2 2.308(5), P1–Pt1–P2 167.61(4).
The structure displays a distorted square-planar geometry with
the phosphorus atoms located nearly trans to each other. The
P1–Pt1–P2 angle of 167.61(4)◦ illustrates the distortion caused
by the chelate, forcing the two phosphines to pinch back, away
˚
from the chlorine. The Pt–Cl bond is unusually long at 2.434(2) A,
illustrating the high trans influence of the cyclohexyl ring. The
cyclohexane ring adopts the chair conformation and is aligned
with the plane of the complex, with the activated C–H bond in
an equatorial position. Similar to complex 5 there is seemingly
very little strain in the ring backbone. Confirming the observation
in solution the molecules has Cs-symmetry with a mirror plane
Fig. 2 DIAMOND drawing of compound 6 at the 30% probability level.
Hydrogen atoms and methyl carbons on the t-butyl groups are omitted
1
involving Cl1, Pt1, C1 and C4. Thus the 13C{ H} NMR spectrum
contains nine unique resonances with the tert-butyl groups above
and below the coordination plane being magnetically inequivalent.
This is observed in both the 1H (D 19.0 Hz) and 13C{H} (D 68.9 Hz)
NMR spectra.
for clarity. Note the disordered methanol in the cavity formed by the
◦
˚
16-membered ring. Selected bond lengths (A) and angles ( ) with estimated
standard deviations: Pt1–Cl1 2.382(4), Pt1–Cl2 2.298(4), Pt2–Cl3 2.301(7),
Pt2–Cl4 2.335(5), Cl1–Pt–Cl2 158.25(19), Cl3-Pt2–Cl4 159.3(3).
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 5427–5433 | 5431
©

View Article Online
DFT calculations
phosphinite ligand which has been reported to easily give pincer
complexes with Ni(II).5b This leads us to the conclusion that it is the
axial–equatorial flip in the present system and the corresponding
palladium system4b that gives rise to an unusually large barrier in
the phosphinites. Due to the strain of the cyclohexyl ring the
conformational change in 9 is probably on the verge of what
is possible and the shorter bond distances in the phosphinites
completely inhibit the ring flip.21
To shed some light on the inability of ligands 2 and 3 to
cyclometallate we decided to study some model reactions using
DFT calculations. There are always errors in trying to calculate
enthalpies or free energies but by looking at the energy difference
between two similar reactions, those of ligands 1 and 3, the errors
are minimised. The reactions studied are shown with their energies
in Scheme 4.
Acknowledgements
Financial support from the Swedish Research Council and the
Knut and Alice Wallenberg Foundation is gratefully acknowl-
edged. We also thank Mr Leo Kirsten for help with the calcu-
lations, Mr Magnus Johnson for experimental assistance and
Mr Anders Sundin and Prof. Ulf Berg for valuable discussions.
References
Scheme 4
1 M. Ohff, A. Ohff, M. E. van der Boom and D. Milstein, J. Am. Chem.
Soc., 1997, 119, 11687. For reviews see: J. T. Singleton, Tetrahedron,
2003, 59, 1837; M. E. van der Boom and D. Milstein, Chem. Rev., 2003,
103, 1759.
By calculating the electronic energies for the six compounds two
DDE values were obtained and in both cases the reaction is 2 kcal
mol−1 more exothermic in the phosphine case as compared to the
phosphinite case. This indicates, as experimentally observed, that
there is a small thermodynamic preference for the cyclometallation
of 1. However, the differences both in coordination of the ligands
and in the actual cyclometallation are small and the inability of 2
and 3 to cyclometallate is probably based on things other than an
unfavourable equilibrium.
2 See for example: (a) M. R. Eberhard, Org. Lett., 2004, 6, 2125; (b) W. J.
Sommer, K. Yu, J. S. Sears, Y. Ji, X. Zheng, R. J. Davis, C. D. Sherrill,
C. W. Jones and M. Weck, Organometallics, 2005, 24, 4351; (c) D.
Olsson, P. Nilsson, M. El Masnaouy and O. F. Wendt, Dalton Trans.,
2005, 1924; (d) P. Nilsson and O. F. Wendt, J. Organomet. Chem., 2005,
690, 4197.
3 N. A. Al-Salem, H. D. Empsall, R. Markham, B. L. Shaw and B.
Weeks, J. Chem. Soc., Dalton Trans., 1979, 1972; N. A. Al-Salem, W. S.
McDonald, R. Markham, M. C. Norton and B. L. Shaw, J. Chem. Soc.,
Dalton Trans., 1980, 59; C. Crocker, R. J. Errington, R. Markham, C. J.
Moulton, K. J. Odell and B. L. Shaw, J. Am. Chem. Soc., 1980, 102,
4373; C. Crocker, R. J. Errington, R. Markham, C. J. Moulton and
B. L. Shaw, J. Chem. Soc., Dalton Trans., 1982, 387; C. Crocker, H. D.
Empsall, R. J. Errington, E. M. Hyde, W. S. McDonald, R. Markham,
M. C. Norton, B. L. Shaw and B. Weeks, J. Chem. Soc., Dalton Trans.,
1982, 1217; J. R. Briggs, A. G. Constable, W. S. McDonald and B. L.
Shaw, J. Chem. Soc., Dalton Trans., 1982, 1225; R. J. Errington, W. S.
McDonald and B. L. Shaw, J. Chem. Soc., Dalton Trans., 1982, 1829;
H. D. Empsall, E. M. Hyde, R. Markham, W. S. McDonald, M. C.
Norton, B. L. Shaw and B. Weeks, J. Chem. Soc., Chem. Commun.,
1977, 589; R. J. Errington and B. L. Shaw, J. Organomet. Chem., 1982,
238, 319; B. L. Shaw, J. Organomet. Chem., 1980, 200, 307.
4 (a) S. Sjo¨vall, O. F. Wendt and C. Andersson, J. Chem. Soc., Dalton
Trans., 2002, 1396; (b) S. Sjo¨vall, C. Andersson and O. F. Wendt, Inorg.
Chim. Acta, 2001, 325, 182.
Discussion
The present cyclometallation reactions are C–H activations with a
concerted mechanism (either electrophilic or oxidative addition)18
and it comes as no surprise that the C(sp3)–H bond with its higher
p-character is more difficult to activate than the corresponding
C(sp2)–H bond.19 It is plausible that ligand 1 is coordinated in
a cis diaxial manner in intermediate 820 and that activation then
takes place at an equatorial C–H bond with both phosphine arms
in an equatorial conformation.
5 (a) A. Castonguay, C. Sui-Seng, D. Zargarian and A. L. Beauchamp,
Organometallics, 2006, 25, 602; (b) V. Pandarus and D. Zargarian,
Chem. Commun., 2007, 978.
6 V. F. Kuznetsov, A. J. Lough and D. G. Gusev, Inorg. Chim. Acta, 2006,
359, 2806.
7 B. J. Burger and J. E. Bercaw, in New Developments in the Synthesis,
Manipulation and Characterization of Organometallic Compounds, ed.
A. Wayda and M. Y. Darensbourg, American Chemical Society,
Washington, DC, USA, 1987, vol. 357, ch. 4.
8 A. Lehtonen, R. Kiveka¨s and R. Sillanpa¨a¨, Polyhedron, 2002, 21, 1133.
9 R. J. Abraham, E. J. Chambers and W. A. Thomas, J. Chem. Soc.,
Perkin Trans. 2, 1993, 1061.
This conformational change brings the C–H bond into the
vicinity of the metal centre as shown in Scheme 5. The calculations
indicate that there is a kinetic barrier towards cyclometallation of
the phosphinite ligands. However, both inter- and intramolecular
C–H activations are usually favoured by electrophilic metal
centres, which facilitate coordination and in this respect we
would have expected the phosphinites to favour cyclometallation.
Interestingly, this is also observed for a straight chain aliphatic
10 H. C. Clark and L. E. Manzer, J. Organomet. Chem., 1973, 59, 411.
11 A. D. Cohen and N. Sheppard, Proc. R. Soc. London, Ser. A, 1959,
A252, 488.
12 Crysalis CCD, Oxford Diffraction Ltd., Abingdon, Oxfordshire, UK,
2005.
13 Crysalis RED, Oxford Diffraction Ltd., Abingdon, Oxfordshire, UK,
2005.
Scheme 5
5432 | Dalton Trans., 2007, 5427–5433
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
14 G. M. Sheldrick, SHELXTL, 5.1, Program for Structure Solution
and Least Squares Refinement, University of Go¨ttingen, Go¨ttingen,
Germany, 1998.
15 Maestro, v. 7.5, Schro¨dinger, LLC, New York, NY, USA, 2005.
16 S. Sjo¨vall, M. H. Johansson and C. Andersson, Eur. J. Inorg. Chem.,
2001, 2907.
19 For such reactivity trends see for example: M. E. Thompson, S. M.
Baxter, A. R. Bulls, B. J. Burger, M. C. Nolan, B. D. Santarsiero, W. P.
Schaefer and J. E. Bercaw, J. Am. Chem. Soc., 1987, 109, 203.
20 All cis coordinated complexes of both phosphinites and phosphines of
this type are diaxial, cf. references 4b and 16.
˚
21 Measured bond distances (in A) in the calculated cis-structures are:
17 M. Davis and O. Hassel, Acta Chem. Scand., 1963, 17, 1181.
18 A. D. Ryabov, Chem. Rev., 1990, 90, 403.
C–C 1.55, C–P 1.92 and P–Pt 2.57 in the phosphine and C–O 1.44,
O–P 1.73, P–Pt 2.51 in the phosphinite.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 5427–5433 | 5433
©