Structural characterization of the head-to-head isomers of the [Pd2(Ph2Ppy)2Cl2] and [PtPd(Ph2Ppy)2I2] complexes (Ph2Ppy = 2-(diphenylphosphino)pyridine)

Article abstract of DOI:10.1016/j.ica.2005.12.021

The molecular structures of the thermodynamically unstable head-to-head isomers, HH-[Pd₂(Ph₂Ppy)₂Cl₂] and HH-[PtPd(Ph₂Ppy)₂I₂], have been determined by single crystal X-ray diffraction. The two complexes have proved to be isostructural. The severe distortions of the bond angles from the ideal square planar geometry around the metal centers ligating the trans phosphorus donor atoms are indicative of a more pronounced internal strain in the HH isomers as compared to the HT counterparts. The enhanced internal strain is thought to be the major driving force responsible for the spontaneous conversion of the head-to-head isomers to their head-to-tail congeners. ¹³C NMR spectra in solution phase as well as solid-state ³¹P MAS NMR spectra have proved to be informative regarding the orientation of the asymmetric Ph₂Ppy ligands.

Full text of DOI:10.1016/j.ica.2005.12.021

Inorganica Chimica Acta 359 (2006) 2933–2941
www.elsevier.com/locate/ica
Structural characterization of the head-to-head isomers of
the [Pd2(Ph2Ppy)2Cl2] and [PtPd(Ph2Ppy)2I2] complexes
(Ph2Ppy = 2-(diphenylphosphino)pyridine)
a
b
a,
*
´
´ ´ ´
´
´
Laszlo Parkanyi , Gabor Szalontai , Gabor Besenyei
a
Chemical Research Center, Hungarian Academy of Sciences, 1525 Budapest, P.O. Box 17, Hungary
b
University of Veszpre´m, NMR Laboratory, 8201 Veszpre´m, P.O. Box 158, Hungary
Received 7 November 2005; accepted 11 December 2005
Available online 9 February 2006
Dedicated to Professor Brian James on the occasion of his 70th birthday.
Abstract
The molecular structures of the thermodynamically unstable head-to-head isomers, HH-[Pd2(Ph2Ppy)2Cl2] and HH-
[PtPd(Ph2Ppy)2I2], have been determined by single crystal X-ray diffraction. The two complexes have proved to be isostructural. The
severe distortions of the bond angles from the ideal square planar geometry around the metal centers ligating the trans phosphorus donor
atoms are indicative of a more pronounced internal strain in the HH isomers as compared to the HT counterparts. The enhanced internal
strain is thought to be the major driving force responsible for the spontaneous conversion of the head-to-head isomers to their head-to-
tail congeners. 13C NMR spectra in solution phase as well as solid-state 31P MAS NMR spectra have proved to be informative regarding
the orientation of the asymmetric Ph2Ppy ligands.
ꢀ 2005 Elsevier B.V. All rights reserved.
Keywords: Head-to-head isomers; Crystal structure; 31P MAS NMR; Dimeric palladium complexes; 2-(Diphenylphosphino)pyridine
1. Introduction
using [Pd2(P–P)2Cl2] complexes as active components (P–
P denotes bis(diphenylphosphino)methane, dppm, or 1,1-
bis(diphenylphosphino)ethane, dppmMe) [2]. The major
achievements of the early years have been summarized in
comprehensive review articles [3]. Recently, the interaction
of [Pd2(dppm)2Cl2] with arenesulfonyl azides, benzoyl
azides and aryl azides has offered a convenient access to
dinuclear nitrene and azide adducts [4].
Our interest in dimeric complexes has been rekindled by
the expectation that they may serve as easily variable model
compounds for studying internal molecular motions in
solution phase or can be used as touchstones in the exten-
sion of MAS NMR spectroscopy to the structural charac-
terization of coordination compounds in the solid state. In
respect to the former subject, we have demonstrated that
the syn and anti isomers of [Pd2(dppmMe)2Cl2] are different
not only by their reactivities [5] but the activation barrier of
the specific intramolecular motion called ‘‘swinging’’ is
Binuclear, metal–metal bonded palladium and platinum
complexes ligating bridging diphosphine ligands have
received special attention since their first syntheses [1].
The ease with which most of these compounds add small
molecules across the metal–metal bond initiated intense
research activities in many laboratories, which explored
the diverse chemical and structural features of the substrate
complexes and the products as well. Outstanding interest
has been devoted to the potential use of dimeric palladium
complexes in gas separation processes, particularly to the
reversible binding of carbon monoxide from coal–gas
streams and to the removal of H2S from gas mixtures by
*
Corresponding author. Tel.: +36 1 438 4141x122/325 7547; fax: +36 1
325 7554.
E-mail address: besenyei@chemres.hu (G. Besenyei).
0020-1693/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2005.12.021

2934
L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
remarkably lower for the sterically more congested syn
dimer [6]. Our first efforts to gain structural information
by the substitution method. The 11 T 31P MAS spectrum
(x0/2p was ꢀ202.5 MHz) was recorded on a Bruker
Avance DRX500 using 4 mm zirkonia rotors. The Haeber-
len convention [10] (|r33 ꢀ riso| > |r11 ꢀ riso| > |r22 ꢀ riso|,
g = r22 ꢀ r11/r33 ꢀ riso) was used to calculate the principal
components of the shielding tensor. The calculations were
done by the Varian’s STRARS program package.
on
a series of seemingly simple palladium dimers
[Pd2(dppm)2X2] (X = Cl, Br, I) have shown that 31P
MAS NMR spectra can truly reflect the non-uniform struc-
tural environments of the phosphorus nuclei, in line with
the X-ray diffraction studies on the same complexes [7].
Although the solid-state spectra of these four-spin systems
were informative regarding certain scalar couplings (e.g.
two-bond trans P–Pd–P coupling constants) which are usu-
ally not available under isotropic conditions, the depen-
dence of the 31P MAS NMR spectra on the rotation rate
could not be treated in terms of the average Hamiltonian
[8]. Afterwards, we turned our attention to simpler models
in which less phosphorus nuclei are involved and which
allow an intentional change of the relative positions of
the coordinated phosphine ligands.
2.2. Crystallography
1-HH and 2 crystallize in the tetragonal space group
I41/a. Data collections were performed on an Enraf-Nonius
CAD4 diffractometer at room temperature, using graphite-
˚
monochromated Mo Ka (k = 0.7107 A) (1-HH) and Cu Ka
˚
radiation (k = 1.5418 A) (2). Empirical absorption correc-
tion (psi-scans) was applied for the intensity data of 2 (the
max. and min. transmissions were 0.3079 and 0.1077). Data
resolution for 2 is rather low (sinh/k = 0.5262) due to the
high absorption. The structures were determined by direct
methods [11] and refined by anisotropic full matrix least-
squares [12] against F2 for the non-hydrogen atoms. The
hydrogen atoms were generated from assumed geometry.
Hydrogen atoms were included in structure factor
calculations and were treated as riding atoms. The
hydrogen isotropic temperature factors were derived from
the carbon atoms they were bonded to. Molecular graphics
were prepared by the program PLATON [13].
We have found promising model compounds satisfying
the above requirements among the known complexes of
the asymmetric 2-(diphenylphosphino)pyridine, Ph2Ppy,
such as [Pd2(Ph2Ppy)2Cl2] [9].
Here we report on the isolation of the hitherto elusive
head-to-head isomer HH-[Pd2(Ph2Ppy)2Cl2] and present
solid-state structural data for the HH-[Pd2(Ph2Ppy)2Cl2]
(1-HH) and HH-[PtPd(Ph2Ppy)2I2] (2) complexes by using
single-crystal X-ray diffraction. Solution phase NMR data
as well as 31P MAS NMR spectra allowing unambiguous
identification of the HH and HT isomers will also be
presented.
2.3. Syntheses
2. Experimental
[Pd(Ph2Ppy)2Cl2] [9a], [Pd2(dba)3 Æ CHCl3] [14] and HH-
[PtPd(Ph2Ppy)2I2] [15] were prepared according to the liter-
ature procedures. Ph2Ppy was purchased from Aldrich and
used as received. Dichloromethane and diethyl ether were
freshly distilled from calcium hydride and Na/benzophe-
none, respectively.
2.1. NMR spectroscopy
Both the liquid-phase and the solid-state NMR experi-
ments were performed on a Varian UNITY 300 NMR
spectrometer equipped with a 5 mm broadband variable
temperature liquid and a room temperature double-bearing
Doty XC5 probe. The solution-phase spectra were
recorded in CD2Cl2 at room temperature. Approximately
20–30 mg samples were used for the solution studies. The
13C and 31P resonance frequencies, x0/2p were ꢀ121.4
and ꢀ75.42 MHz, respectively. The MAS spectra were
recorded with high-power (ꢁ100–120 W) proton decou-
pling. Sample spinning speeds were typically varied
between 3000 and 11000 10 Hz. Stable rotation below
1500 Hz was not possible on this probe. 60–100 mg sample
quantities were used in 5 mm o.d. zirkonia or Si3N4 rotors.
The number of scans varied between 32 and 128. Both
MAS (with recycling delays of 12–15 s) and CP/MAS spec-
tra (with contact times of 0.7–1.2 ms and recycling delays
of 6 s) were recorded and gave practically identical results.
The chemical shifts are referred to external TMS
(dCD2Cl2 = 54 ppm was used for the conversion) and
85% H3PO4. The 31P 90ꢁ pulse width was about 3.6 ls.
For referencing and set-up purposes, polycrystalline PPh3
was applied (diso = ꢀ6 ppm relative to the 85% H3PO4)
2.4. Preparation and isolation of HH-[Pd2(Ph2Ppy)2Cl2]
(1-HH)
To a solution of 257 mg (0.365 mmol) of [Pd(Ph2-
Ppy)2Cl2] in 50 ml of dry dichloromethane was added drop-
wise 190 mg (0.184 mmol) of [Pd2(dba)3 Æ CHCl3] in 50 ml of
dichloromethane at ambient temperature. The progress of
the conproportionation reaction was monitored by using
TLC (Merck, Kieselgel 60 F 254, eluens: CH2Cl2/EtOAc
2:1). After stirring for 90 min, the reaction mixture was fil-
tered to remove any insoluble impurities and the solution
was evaporated to ca. 10 ml under vacuum. To the resulting
dark red solution, 40 ml of diethyl ether was slowly added,
which caused the precipitation of a bright red microcrystal-
line solid. Yield 287 mg (97%). The 31P NMR spectrum of a
sample indicated that the product is a ca. 2:1 mixture of the
HH and HT isomers (the proportion of the HH isomer var-
ied between 64% and 66%).
A mixture of the HH and HT isomers obtained as
described above (532 mg) was dissolved in CH2Cl2 and

L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
2935
filtered through a Celite layer. The volume was reduced to
40 ml under vacuum and 80 ml of methanol was added.
The chlorinated solvent was partly removed under vacuum
causing a sharp drop of the temperature of the solution,
which was accompanied by a partial precipitation of the
dimer (224 mg). The isolated fraction proved to be the HH
isomer of 96% purity, as shown by the integration of the
31P NMR spectrum. 31P NMR (CH2Cl2): ꢀ12.6 ppm (s).
N
N
Pd
P
P
P
Pd
N
Pd
Pd
N
P
HH
HT
2.5. Preparation and isolation of HT-[Pd2(Ph2Ppy)2Cl2]
(1-HT)
isomers
Scheme 1.
105 mg (0.10 mmol) of [Pd2(dba)3 Æ CHCl3] as a solid
was added in small portions to a solution of 142 mg
(0.20 mmol) of [Pd(Ph2Ppy)2Cl2] dissolved in 10 ml of dry
dichloromethane. The reaction mixture was stirred under
N2 at 25 ꢁC for 75 min. After filtration and partial removal
of the solvent, dropwise addition of 25 ml of dry diethyl
ether resulted in a red solid which was dried under vacuum.
Yield 150 mg (93%). 31P NMR spectroscopy indicated that
the product is pure HT isomer. 31P NMR (CH2Cl2):
5.9 ppm (s).
(X = Cl, Br or I). The relatively fast isomerization, espe-
cially when X = Cl, hampered, however, the isolation of
the less stable products.
In agreement with the observation of Fujita and
coworkers, reaction (1) delivered two products in our
hands as shown by the two singlet resonances in the solu-
tion phase 31P NMR spectra. We have found, however, a
simple route for the isolation of the hitherto elusive HH-
[Pd2(Ph2Ppy)2Cl2] dimer. The progress of the conpropor-
tionation reaction was followed by using thin-layer
chromatography allowing to stop the reaction when the
Pd(0) complex could not be detected any more. Full con-
version of the starting dimer was achieved after 60–
90 min at 25 ꢁC at which stage the product proved to be
an approximately 2:1 mixture of the HH and HT isomers.
The close monitoring of the reaction has allowed to shorten
the reaction time suggested earlier (20 h, [16]), which cer-
tainly had a beneficial effect in preserving the relatively
high proportion of the unstable isomer. From this mixture,
the HH-[Pd2(Ph2Ppy)2Cl2] complex could be isolated in
96% purity if the precipitated raw product was redissolved
in dichloromethane, diluted with methanol and the chlori-
nated solvent was partly removed immediately under vac-
uum (for details of the work-up procedure see Section 2).
The compound isolated in this way proved to be sufficiently
pure for liquid phase and solid-state investigations and to
grow single crystals for crystallographic studies.
We note that the concentration of the reaction mixture
has been found to exert a remarkable effect on the che-
moselectivity of reaction (1). While the proportion of the
HH isomer was in the range of 64–66% when the total
Pd-concentration was kept at ca. 8 mM, we could observe
only the HT complex at [Pd] = 40 mM. It was not our aim
to explore the kinetic or mechanistic aspects of reaction (1)
and this observation was not further investigated.
Identification, although not quite straightforward, of
the two isomers was possible by simulations of the AXX0
spin systems (A = 13C, X = X0 = 31P) observed in the
solution phase 13C NMR spectra for all carbon atoms J-
coupled to phosphorus atoms (see Figs. 1 and 2). The split-
ting of the signal A is influenced by the scalar coupling
between the two magnetically non-equivalent P atoms.
The distinction is based on the general observation that
the two-bond scalar coupling values between trans related
3. Results and discussion
3.1. Synthesis and isolation of HH-[Pd2(Ph2Ppy)2Cl2]
(1-HH)
Although the conproportionation reaction of
[Pd2(dba)3 Æ CHCl3] with [Pd(Ph2Ppy)2Cl2] (Eq. (1)) yield-
ing dimeric palladium(I) complexes with bridging Ph2Ppy
ligands has been known for more than twenty years, the
contradictory observations regarding the number and the
stability of the products deserve some attention.
1/2[Pd2(dba)3] + [Pd(Ph2Ppy)2Cl2]
! [Pd2(Ph2Ppy)2Cl2] + 1.5dba
(1)
The first synthesis of the [Pd2(Ph2Ppy)2Cl2] dimer was de-
scribed by Balch and coworkers [9a]. It was envisaged that
the two possible arrangements of the bridging ligands
might result in the formation of head-to-head, HH, and
head-to-tail, HT, isomers (Scheme 1). It was observed that
reaction (1) yields a single product that was assumed to
hold the bridging groups in a HT orientation. This conclu-
sion was based on structural studies carried out on the re-
lated heterodinuclear HT-[PtPd(Ph2Ppy)2Cl2] complex
whose 31P NMR spectra allowed to establish unequivocally
that the phosphorus nuclei are bound to different metal
centers.
The identification of Balch’s product as a HT isomer
was later confirmed by Fujita and coworkers in their com-
parative study on the ligating properties of Me2Ppy and
Ph2Ppy ligands (Me2Ppy stands for 2-(dimethylphosph-
ino)pyridine) [16]. Monitoring the conproportionation
reaction (1) by NMR spectroscopy, they detected the inter-
mediacy of new complexes which were assumed to be the
unstable head-to-head isomers HH-[Pd2(Ph2Ppy)2X2]

2936
L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
172.8
172.7
172.6
172.5
172.4
172.3
[ppm]
[ ppm ]
171.0
171.5
170.5
Fig. 1. 13C{1H} NMR of the HT isomer (expansion of the signal of the
pyridyl carbon atom carrying the Ph2P moiety), bottom: experimental,
top: simulation of the A part of an AXX0 system. Parameters obtained
Fig. 2. 13C{1H} NMR of the 1-HH isomer (expansion of the signal of the
pyridine C2 carbon atom (for numbering see Fig. 5), top: experimental
spectrum, bottom: simulation of the A part of an AXX0 system).
from simulations: JAX = 1J(31P–13C) = 75.3 Hz, JAX ¼ 3Jð31P–13CÞ ¼
0
ꢀ5:8 Hz, JXX ¼ 3Jð31P–31PÞ ¼ 18:0 Hz. The two almost identical central
Parameters: JAX = 1J(31P–13C) = 70 Hz, JAX ¼ 3Jð31P–13CÞ ¼ ꢀ11 Hz,
0
0
JXX ¼ 2Jð31P–31PÞ ¼ 430 Hz. (The splitting does not change any more
0
transitions (indicated by an arrow) are burried by the noise at about
172.1 ppm.
upon increasing J0XX if its value surpasses 200 Hz, the weak outer
transitions are expected to disappear in the noise.)
P nuclei are significantly larger than the similar couplings
between cis related P atoms [17]. Consequently, simulations
of the A part of the AXX0 spin systems can deliver the JPP
coupling constants we are looking for.
dipolar and scalar coupled isolated 31P spin pair of the
HH isomers is expected to show rotational rate-dependent
MAS spectra [18,19]. In the case of HT isomers, due to the
much smaller scalar PP couplings, this phenomenon must
be absent. As revealed by the X-ray data, the relevant inter-
In our case (only resonances of the pyridine a carbon
atoms are shown in the figures), the distance between the
second and fourth lines gives the sum of the one- and
three-bond 31P–13C couplings. The one-bond coupling val-
ues obtained from the simulation are larger than expected
(70–80 Hz), indicating strong s electron involvement in that
particular P–C bond, and are, perhaps, characteristic for
this arrangement. The relevant coupling values between
the ipso-carbon of the phenyl rings and the same P nucleus
are somewhat smaller (Fig. 1).
˚
nuclear d(Pꢂ ꢂ ꢂP) distances are: 1-HH: 4.447 A, 1-HT:
˚
˚
˚
4.417 A, 2: 4.408 A and HT-[Pd2(Ph2Ppy)2I2]: 4.375 A.
Somewhat surprisingly, the Pꢂ ꢂ ꢂP separation is a little smal-
ler in the HT isomer, due to the twist along the Pd–Pd
axis).
Indeed, the spectra confirm these assumptions, in that at
low rotation speeds (<3500 Hz) the spectrum of the HH
isomer (Fig. 3) displays an unusual quartet splitting typical
for the so-called J-recoupling phenomenon [19,20].
We feel important to emphasize that under MAS condi-
tions, J-recoupling may arise even between chemically
identical P atoms unless their chemical shielding tensors
are coincident (i.e. they are related by a center of inver-
sion). Separation of the first and third or the second and
We note, however, that only the lower limit (e.g.
>180 Hz) of the coupling is available from the simulations
2
in the case of large J(P–M–P) coupling (i.e. HH isomer,
Fig. 2), due to the diminishing intensities of the outer
transitions (Table 1).
More convincing evidence could be obtained from
straightforward 31P MAS spectra of the complexes. The
2
fourth lines give the J(P–M–P) coupling value we need.
Table 1
13C NMR chemical shifta (ppm) and splittingb values (in parentheses, Hz) observed for the HH and HT isomers at room temperature in CD2Cl2
Isomer
Pyridine carbons
Phenyl carbons
C(N,P)
C(N)
C(para to N)
C(meta to N)
C(para to P)
ipso
ortho
meta
para
HH
HT
a
170.9 (55.5)b
172.2 (70.6)b
155.9 (10.5)
155.5 (12.1)
138.1 (4.6)
139.7 (5.2)
128.5 (8.2)
131.7 (10.8)
125.4 (s)
125.4 (s)
n.o.
132.3 (52.8)
135.2 (15.4)
135.6 (11.8)
131.4 (s)
132.4 (2.3)
129.3 (10.8)
130.4 (11.2)
Relative to external TMS.
b
Difference between the second and fourth lines; if one-bond PC coupling is involved its value is not identical with this coupling, in other cases the
observed splitting can be considered a good estimate of the relevant nJPC coupling since contributions from 4JPC couplings are generally small. s = singulet,
n.o. = not observed due to overlap.

L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
2937
11800 Hz
11200 Hz
*
impurity
ppm
75
50
25
0
-25
-50
-75
63
50
38
25
13
0
-13
-25
-38
-50
-63
ppm
7800 Hz
4600 Hz
ppm
50
0
-50
4550 Hz
63
50
38
25
13
0
-13
-25
-38
-50
-63
ppm
2940 Hz
ppm
50
0
-50
1800 Hz
63
50
38
25
13
0
-13
-25
-38
-50
-63
ppm
1940 Hz
ppm
50
0
-50
Fig. 4. 31P MAS spectra of the 1-HT isomer. Apart from the disappearing
of the spinning side-bands the spectra do not show rotation rate
dependence. The signals indicated by asterisk are most probably due to
a different polymorph, the solution spectrum did not show any evidence of
impurity.
63
50
38
25
13
0
-13
-25
-38
-50
-63
ppm
Fig. 3. 31P MAS spectra of the 1-HH complex; rotation rate dependence
of the homonuclear spin pair. Separation of the first and third lines of the
low speed (1940 Hz) spectrum is about 430 Hz.
10.4 ppm. The minor difference observed in the relevant
Pd–P bond length of the HT isomer (vide infra) supports
the first alternative. Similar patterns have often been
observed in dimeric palladium(I) complexes [7] and also
in several types of A-frame adducts [21].
At higher rotation speeds this collapses to a singlet indicat-
ing a single chemical environment for the two P atoms in
the crystalline phase. This is in agreement with the C2 sym-
metry of the molecule as shown by the crystallographic
study (see below). The isotropic chemical shift, the shield-
ing anisotropy and the asymmetry parameter of this phos-
phorus site are ꢀ8.6, 59.2 ppm and 0.95, respectively.
On the contrary, the spectrum of 1-HT does not show
this phenomenon (see Fig. 4). The P atoms are separated
by three bonds in this isomer, the expected value of this
coupling is less than the observed linewidths of the MAS
spectrum (150–250 Hz). At the same time, the magnitude
of the dipolar coupling among the P nuclei, a sine qua
non condition for the J-recoupling phenomenon, must be
similar to that of the 1-HH isomer due to the similar inter-
nuclear distances among them (vide supra). The unchanged
AB-like splitting of the isotropic signal observed in the
whole rotation range studied indicates that either the C2
symmetry existing in liquid phase is lost in the solid phase
or there are two different molecules in the asymmetry unit.
The estimated isotropic chemical shifts are 11.6 and
3.2. Crystallographic description of complexes HH-
[Pd2(Ph2Ppy)2Cl2] (1-HH) and HH-[PtPd(Ph2Ppy)2I2]
(2)
Although there are numerous structures reported with 2-
(diphenylphosphino)pyridine, to our knowledge, none of
the complexes with four coordinated transition metal cen-
ters with a HH orientation of the bridging ligands has been
characterized by crystallographic methods.1 The head-to-
head orientation of the Ph2Ppy pyridylphosphine has been
observed, however, and such complexes have been charac-
1
A search of the Cambridge Crytallographic Database (updated August
2005) retrieved 50 dimetallic complexes incorporating Ph2Ppy ligands in a
HH orientation. None of these structures contained two four-coordinated
metal centers.

2938
L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
terized in solution by using NMR spectroscopy [15,16a]. In
order to reveal the structural differences and similarities of
the HH and HT isomers, HT-[Pd2(Ph2Ppy)2Cl2] [16b] and
HT-[Pd2(Ph2Ppy)2I2] [22] were selected as reference com-
pounds for the structural discussion. The crystallographic
data are collected in Table 2, selected bond lengths and
bond angles are presented in Tables 3 and 4.
The asymmetric unit is half a molecule (Z0 = 0.5). The
molecular symmetry is C2, the halogen and metal atoms
sit on a twofold axis.
Compound 1-HH is a typical side-by-side complex
(Fig. 5, structure a) in which the two four-coordinated pal-
ladium centers are connected by a relatively short
Table 4
˚
Selected bond lengths (A) and bond angles (ꢁ) of HH-[PtPd(Ph2Ppy)2I2], 2
and HT-[Pd2(Ph2Ppy)2I2]
HH-[PtPd(Ph2Ppy)2I2]
HT-[Pd2(Ph2Ppy)2I2] [22]
Bond lengths
Pt1–Pd1
Pd1–Pd2
Pt1–P1
Pd–P
Pd–N
Pt1–I1
Pd1–I2
N1–C2
P1–C2
2.571(1)
2.254(2)
2.597(1)
2.217(1); 2.206(1)
2.103(4); 2.121(4)
2.024(8)
2.642(1)
2.703(1)
1.36(1)
Table 2
Crystal data and structure refinement
2.696(1); 2.695(1)
1.344(7); 1.355(6)
1.823(5); 1.827(5)
Empirical formula
Formula weight
Unit cell dimensions
C34H28Cl2N2P2Pd2 C34H28I2N2P2PdPt
810.26
1081.85
1.80(1)
˚
Bond angles
P1–Pt1–P1a
N1–Pd1–N1a
P1–Pd1–N2
P2–Pd2–N1
a (A)
15.483(2)
26.750(3)
6412.6(14)
8
15.694(5)
28.172(5)
6939(3)
8
˚
155.7(1)
174.6(4)
c (A)
3
˚
Volume (A )
169.4(1)
171.2(1)
Z
Density (calculated) (Mg/m3) 1.678
2.071
Absorption coefficient
(lm mꢀ1
F(000)
1.415
26.351
)
a
3216
4048
Crystal colour
red
black
Crystal description
Crystal size (mm)
h-Range for data collection (ꢁ) 2.40 6 h 6 31.94
Index ranges h;k;l
block
0.50 · 0.30 · 0.30
block
0.18 · 0.10 · 0.05
2.53 6 h 6 21.96ꢁ
ꢀ11, 11;ꢀ16,
16;ꢀ29, 29
4761
ꢀ23, 23;ꢀ23,
23;ꢀ39, 39
12147
5474/0.0233
2817
Reflections collected
Independent reflections/Rint
Reflections I > 2r(I)
2109/0.0453
1472
Data/restraints/parameters
Goodness-of-fit on F2
Extinction coefficient
Final R indices
5474/0/192
0.816
2109/0/193
0.984
0.00016(2)
0.0389/0.1015
0.0312/0.0694
[I > 2r(I)], R1/wR2
R indices (all data), R1/wR2
Largest difference peak
0.0883/0.0797
0.557/ꢀ0.640
0.0613/0.1086
1.305/ꢀ1.080
ꢀ3
˚
and hole (e A
)
b
Table 3
˚
Selected bond lengths (A) and bond angles (ꢁ) of HH- and HT-
[Pd2(Ph2Ppy)2Cl2]
HH-[Pd2(Ph2Ppy)2Cl2]
HT-[Pd2(Ph2Ppy)2Cl2] [16b]
a
Bond lengths
Pd1–Pd2
Pd1–P1
Pd2–P2
Pd2–N1
Pd1–N2
Pd1–Cl1
Pd2–Cl2
N1–C2
2.561(1)
2.278(1)
2.594(1)
2.207(2)
2.203(2)
2.106(5)
2.122(5)
2.428(2)
2.411(2)
1.331(8); 1.355(9)
1.842(7); 1.828(7)
a
2.036(2)
2.400(1)
2.411(1)
1.348(3)
1.829(3)
P1–C2
Bond angles
P1–Pd1–P1a
N1–Pd2–N1a
P1–Pd1–N2
P2–Pd2–N1
154.9(1)
177.8(1)
Fig. 5. Molecular diagrams of complexes 1-HH (a) and 2 (b), showing
atomic displacement ellipsoids at the 50% probability level. Hydrogen
atoms are omitted for clarity.
170.2(2)
173.3(1)

L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
2939
˚
(2.561(1) A) metal–metal bond. For comparison, the corre-
of the filled d orbitals that may be especially pronounced in
Pt and Pd complexes with short metal–metal separation.
The most striking difference in the conformations of the
HH and HT isomers is the orientation of the axial phenyl
moieties. The axial phenyls point in opposite directions in
the HH isomers with respect to the best plane containing
the metal, phosphorus, and nitrogen atoms. In the HT iso-
mers, these phenyl rings are on the same side of this plane
(Fig. 6).
˚
sponding distance in 1-HT is 2.594(1) A. The short Pd–Pd
separation is in line with the rigid nature of the Ph2Ppy
ligand that typically allows to span shorter metal–metal
distances as compared to the more flexible dppm. From
the structural point of view, the most interesting part of
the complex is the coordination sphere of the palladium
centers that ligate either two phosphorus or two nitrogen
donor atoms as opposed to the uniform one phosphorus/
one nitrogen coordination in HT isomers. A survey of
the structural data readily reveals the most significant dif-
ferences in the environments of the two Pd centers. It seems
noteworthy that the angles formed by neighboring ligands
on the ‘‘nitrogen side’’ deviate from the ideal right angle
only negligibly, the difference between the larger and smal-
ler angle being only 2.2ꢁ. Severe distortion from the ideal-
ized square-planar geometry can be observed, however, on
the ‘‘phosphorus side’’ where the same angles differ by
25.2ꢁ.
The conformation of the HH isomer leads to a more
compact, close packed arrangement in the crystal lattice
that contains no solvent accessible voids, 68.3% of the unit
cell is filled with atoms (Fig. 7). There are large solvent
accessible areas in the lattice of the HT isomers. The per-
cent filled space for 1-HT and HT-[Pd2(Ph2Ppy)2I2] is
63.1% and 61.4%. The potential solvent areas for these
3
3
˚
˚
two compounds are 820.1 A (11.7%) and 1010.9 A
(13.5%). These voids are in fact filled with disordered sol-
vent molecules.
Undoubtedly, the notable distortions of the bond angles
reflect a substantial amount of internal strain and may be
the major driving force of the spontaneous HH ! HT
isomerization.
Compound 2, HH-[PtPd(Ph2Ppy)2I2], has been prepared
by Balch and coworkers [15] and has been characterized by
multinuclear NMR methods. The thermodynamic instabil-
ity of the dimer was demonstrated unequivocally by its
slow conversion to HT isomer, which could be catalyzed
by free Ph2Ppy. Here we report on the molecular structure
of 2, based on a single crystal X-ray diffraction study.
1-HH and 2 are isostructural. The cell similarity index
[24] P is 0.03195 and the Iv index [25] (that involves all
atoms of the structures) is 81.2%. There is only a relaxed
isostructurality between 1-HT and HT-[Pd2(Ph2Ppy)2I2]
(space group C2/c, P = 0.0147, Iv = 52.2%).
As for the ligands in trans positions, the N1–Pd2–N1a
angle (177.8ꢁ) approaches the linearity as opposed to the
154.9ꢁ of the P–Pd–P linkage. On the contrary, the two
P–Pd–N axes in the HT isomer deviate from the linear
arrangement only by 6.0ꢁ and 9.8ꢁ. The metal and the hal-
ogen atoms sit on a twofold axis, therefore the X–M–M–X
(X = halogen, M = metal) chain is strictly collinear for
both 1-HH and 2.
A remarkable difference of the HT and HH isomers is
reflected by the Pd–P and Pd–N bond lengths. While the
In agreement with the close relationship of the two com-
plexes, the structural features of 2 are very similar to those
of complex 1-HH. The Pt–Pd bond length is 2.571(1) A and
˚
˚
mean dPd–P is 2.205(2) A in HT-[Pd2(Ph2Ppy)2Cl2], it is
˚
longer in the HH counterpart by 0.07 A. The lengthening
adds to the number of structures with short metal–metal
˚
of the Pd–P bond can be rationalized in terms of the strong
trans influence of the phosphine moieties and is also con-
form to the expectation that the bond strength of a strong
p-acceptor ligand is decreased if another ligand of the same
kind occupies the trans position. Not surprisingly, the
opposite trend is observed for the pyridyl parts. While
the head-to-tail arrangement results in a mean Pd–N bond
bonds. Although the Pt–P bonds (2.254(2) A) are shorter
than the corresponding distances in complex 1-HH, they
are remarkably longer of the Pd–P bonds in the reference
˚
compound HT-[Pd2(Ph2Ppy)2I2] (dPdP, av. = 2.210 A). This
feature can be interpreted again in terms of the strong trans
influence of the P nuclei. Accordingly, the Pd–N distances
are notably shorter that those in the HT isomer
˚
˚
length of 2.114 A in the HT-[Pd2(Ph2Ppy)2Cl2] dimer, the
[Pd2(Ph2Ppy)2I2] (2.024(8) vs. dPdN, av = 2.112 A). It seems
˚
Pd–N distance drops to 2.036(2) A in the HH isomer, in
interesting to point out that the metal-iodine bond lengths
agreement with the relatively weak trans influence of the
nitrogen donor atoms.
Other bond lengths and bond angles of the two isomers
do not show characteristic differences. We note, however,
that the dihedral angle of the coordination planes is appre-
ciably larger (46.81ꢁ) in the HH isomer than in the HT con-
gener (37.43ꢁ). It has been demonstrated for a large set of
dimeric palladium and platinum complexes incorporating
bridging diphosphine ligands that increasing metal–metal
distances are accompanied by decreasing dihedral angles
[23]. The structural features of this isomer pair are conform
to this general trend and support the earlier conclusion that
a twist about the metal–metal bond reduces the interaction
HH
HT
Fig. 6. The 1-HH and 1-HT isomers as viewed down the metal–metal axis.

2940
L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
a
b
Fig. 7. Packing diagrams for 1-HT (a) with the solvent accessible voids and 1-HH (b). Hydrogen atoms are omitted for clarity.
˚
in the heterodinuclear complex are different by 0.06 A. This
contrasts with the observation that the Pd-halide bonds are
practically equal in complexes 1-HH (dPdCl = 2.400(1) and
the P-bound platinum center are more distorted than those
formed by the neighboring substituents of palladium as
indicated by the difference of the largest and smallest angles
(24.2ꢁ vs. 5.4ꢁ). Like in the case of the HH and HT
[Pd2(Ph2Ppy)2Cl2] isomer pair, the dihedral angle of the
coordination planes incorporating the platinum and the
palladium centers is larger (48.40ꢁ) than the same feature
of the reference compound HT-[Pd2(Ph2Ppy)2I2] (41.3ꢁ).
All these structural features suggest that it is the HH-
[PtPd(Ph2Ppy)2I2] complex that accumulates more internal
strain and, consequently, is predestinate to isomerize to the
more stable HT counterpart.
˚
˚
2.411(1) A), 1-HT (dPdCl = 2.428(2) and 2.411(2) A),and
˚
HT-[Pd2(Ph2Ppy)2I2] (dPdI = 2.696(1) and 2.695(1) A).
Although the metal–metal bond has a strong trans influ-
ence this should equally affect both the Pd–I and the Pt–I
bonds. We attribute the shorter Pt–I bond length to the
smaller ionic radius of platinum as compared to that of
palladium.
As expected, the deviations from the idealized square
planar geometry around the metal centers are very similar
in complexes 1-HH and 2. While the N1–Pd1–N1a angle in
2 is almost linear (174.6(4)ꢁ), the P1–Pt1–P1a angle is
severely bent (155.7(1)ꢁ). Similarly, other angles around
We find interesting to point out that the structural differ-
ences of the HH and HT isomers are reminding those of the
syn- and anti-[Pd2(dppmMe)2Cl2] complexes. Of the latter

L. Pa´rka´nyi et al. / Inorganica Chimica Acta 359 (2006) 2933–2941
2941
pair of compounds, the syn-[Pd2(dppmMe)2Cl2] dimer
proved to be sterically more congested, and consequently,
the thermodynamically less stable product [23]. Similarly
to the structural differences between the HH and HT com-
plexes, larger dihedral angle, shorter Pd–Pd bond length
and more extensive internal strain were characteristic to
the syn isomer compared to the anti congener. Importantly,
however, the enhanced internal strain of the syn dimer was
accompanied by an increased reactivity. While the anti iso-
mer was reluctant toward all typical reagents but the ele-
mental selenium, the syn congener formed A-frame
adducts with a series of small molecules [5]. Based on the
above considerations, the HH-[Pd2(Ph2Ppy)2Cl2] dimer is
expected to be more reactive than its HT counterpart if
the isomerization can be suppressed efficiently.
1-HH and 2. Supplementary data associated with this arti-
cle can be found, in the online version, at doi:10.1016/
j.ica.2005.12.021.
References
[1] (a) R.G. Holloway, B.R. Penfold, R. Colton, M.J. McCormick, J.
Chem. Soc., Chem. Commun. (1976) 485;
(b) M.P. Brown, R.J. Puddephatt, M. Rashidi, K.R. Seddon, J.
Chem. Soc., Dalton Trans. (1977) 951.
[2] (a) C.L. Lee, Y.P. Yang, S.J. Rettig, B.R. James, D.A. Nelson, M.A.
Lilga, Organometallics 5 (1986) 2220;
(b) G. Besenyei, C.L. Lee, J. Gulinski, S.J. Rettig, B.R. James, D.A.
Nelson, M.A. Lilga, Inorg. Chem. 26 (1987) 3622.
[3] (a) R.J. Puddephatt, Chem. Soc. Rev. 12 (1983) 99;
(b) B. Chaudret, B. Delavaux, R. Poilblanc, Coord. Chem. Rev. 86
(1988) 191.
´
´
´
´
´
[4] (a) I. Foch, L. Parkanyi, G. Besenyei, L.I. Simandi, A. Kalman, J.
4. Conclusions
Chem. Soc., Dalton Trans. (1999) 293;
´
´
´
(b) G. Besenyei, L. Parkanyi, G. Szalontai, S. Holly, I. Papai, G.
Keresztury, A. Nagy, Dalton Trans. (2004) 2041;
The kinetic product of the conproportionation reaction
of [Pd(Ph2Ppy)2Cl2] and [Pd2(dba)3], HH-[Pd2(Ph2-
Ppy)2Cl2], has been isolated and the molecular structure
was determined by using X-ray diffraction. This complex
and the related HH-[PtPd(Ph2Ppy)2I2] dimer are the first
structurally characterized members of the large number
of HH-[M1M2(Ph2Ppy)2X2] complexes (M1,M2 = Pd, Pt;
X = anion). Although a survey of the structural data has
revealed a host of distortions and unfavorable internal
non-bonding interactions in both the HT and HH isomers,
the bond angles around the P-bound metal centers have
proved to be outstandingly distorted in the HH complexes.
This structural feature is in line with the thermodynami-
cally unstable nature of the head-to-head isomers and is
seen as the major driving force of the spontaneous
HH ! HT conversion. The structural differences of the
HT and HH isomers remind those observed for the syn
and anti isomers of [Pd2(dppmMe)2Cl2] but it remains to
be cleared if the potentially enhanced reactivity of the
HH dimers can be observed experimentally. The MAS
31P NMR technique has proved to be a viable experimental
method for recognizing the various arrangements of the
Ph2Ppy ligand in dinuclear complexes.
´
´
´
(c) G. Besenyei, L. Parkanyi, I. Foch, L.I. Simandi, Angew. Chem.
Int., Ed. 39 (2000) 956.
[5] G. Besenyei, C.L. Lee, Y. Xie, B.R. James, Inorg. Chem. 30 (1991)
2446.
[6] G. Szalontai, G. Besenyei, Inorg. Chim. Acta 357 (2004) 4413.
[7] G. Szalontai, G. Besenyei, Monats. Chem. 136 (2005) 663.
[8] G. Wu, R.E. Wasylishen, J. Chem. Phys. 98 (1993) 6138.
[9] (a) First synthesis: A. Maisonnet, J.P. Farr, M.M. Olmstead, C.T.
Hunt, A.L. Balch, Inorg. Chem. 21 (1982) 3961;
(b) , Reviews on the coordination chemistry of Ph2Ppy:Z.Z. Zhang,
H. Cheng, Coord. Chem. Rev. 147 (1996) 1;
(c) P. Espinet, K. Soulantica, Coord. Chem. Rev. 193–195 (1999) 499.
[10] U. HaeberlenHigh Resolution NMR in Solids, Suppl. 1, Academic
Press, New York, 1976.
[11] G.M. Sheldrick, SHELXS-97 Program for Crystal Structure Solution,
University of Go¨ttingen, Germany, 1997.
[12] G.M. Sheldrick, SHELXL-97 Program for Crystal Structure Refine-
ment, University of Go¨ttingen, Germany, 1997.
[13] A.L. Spek, J. Appl. Cryst. 36 (2003) 7.
[14] T. Ukai, H. Kawazura, Y. Ishii, J.J. Bonnett, J.A. Ibers, J.
Organomet. Chem. 65 (1974) 253.
[15] J.P. Farr, F.E. Wood, A.L. Balch, Inorg. Chem. 22 (1983) 3387.
[16] (a) T. Suzuki, J. Fujita, Bull. Chem. Soc. Jpn. 65 (1992) 1016;
(b) T. Suzuki, N. Iitaka, S. Kurachi, M. Kita, K. Kashiwabara, S.
Ohba, J. Fujita, Bull. Chem. Soc. Jpn. 65 (1992) 1817.
[17] P.S. Pregosin, R.W. Kunz, 31P and 13C NMR of Transition Metal
Phosphine Complex, Springer, Heidelberg, 1979.
[18] M.H. Lewitt, D.P. Raleigh, F. Cruzet, R.G. Griffin, J. Chem. Phys.
92 (1990) 6347.
Acknowledgement
[19] K. Eichele, G. Wu, R.E. Wasylishen, J. Magn. Res. A 101 (1993) 157.
[20] P. Tekely, C. Cardinnet, M.J. Potrebowski, A. Sebald, D. Reichert,
Z. Luz, J. Chem. Phys. 116 (2002) 7607.
[21] G. Szalontai, G. Besenyei, unpublished results.
[22] Y. Xie, C.L. Lee, Y. Yang, S.J. Rettig, B.R. James, Can. J. Chem. 70
(1992) 751.
Support from the Hungarian Scientific Research Fund
(OTKA T34355) is acknowledged.
Appendix A. Supplementary material
´
´
´
[23] G. Besenyei, L. Parkanyi, E. Gacs-Baitz, B.R. James, Inorg. Chim.
Acta 327 (2002) 179.
Crystallographic data for the structural analyses have
been deposited with the Cambridge Crystallographic Data
Center, CCDC, Nos. 288659 and 288660 for compounds
´
´
´
´
[24] A. Kalman, L. Parkanyi, Gy. Argay, Acta Crystallogr., Sect. B 49
(1993) 1039.
´
´
´
´
[25] L. Fabian, A. Kalman, Acta Crystallogr., Sect. B 55 (1999) 1099.