4-Membered metallodithiophosphinate rings - Flat or puckered? A comparison of two crystal structures with computational and literature data

Article abstract of DOI:10.1016/j.ica.2005.10.006

Two crystalline forms of (dithiodiphenylphosphinate)(phenyl) (triphenylphosphine)-palladium(II) (C₃₆H₃₀P ₂PdS₂), one without solvent, the other containing THF (C₄H₈O), are obtained after reaction of sodium diphenyldithiophosphinate with (phenyl) (bis-triphenylphosphine) palladium(II) chloride and crystallisation from two different solvent mixtures. The molecular structures, as determined by single crystal X-ray diffraction, differ in the planarity of the 4-membered palladium dithiophosphinate rings. The experimental conformations have been compared to the conformations of four-membered metal-S₂P rings reported in the Cambridge Structural Database. A flat conformation is more common than a puckered one. DFT calculations at the B3LYP level of theory indicate that the flat conformation of a model metallodithiophosphinate ring is very slightly lower in energy (1.2 kcal/mol) than the puckered conformation.

Full text of DOI:10.1016/j.ica.2005.10.006

Inorganica Chimica Acta 359 (2006) 609–616
www.elsevier.com/locate/ica
4-Membered metallodithiophosphinate rings – flat or puckered?
A comparison of two crystal structures with computational
and literature data
a,
a
a
Catharine Esterhuysen *, Gert J. Kruger b,1, Gavin Blewett , Helgard G. Raubenheimer
a
Department of Chemistry, University of Stellenbosch, Private Bag X1, Stellenbosch, Matieland 7602, South Africa
Department of Chemistry and Biochemistry, University of Johannesburg, P.O. Box 524, Auckland Park, 2006, South Africa
b
Received 13 July 2005; accepted 2 October 2005
Available online 22 November 2005
Abstract
Two crystalline forms of (dithiodiphenylphosphinate)(phenyl)(triphenylphosphine)-palladium(II) (C36H30P2PdS2), one without sol-
vent, the other containing THF (C4H8O), are obtained after reaction of sodium diphenyldithiophosphinate with (phenyl) (bis-triphenyl-
phosphine) palladium(II) chloride and crystallisation from two different solvent mixtures. The molecular structures, as determined by
single crystal X-ray diffraction, differ in the planarity of the 4-membered palladium dithiophosphinate rings. The experimental confor-
mations have been compared to the conformations of four-membered metal-S2P rings reported in the Cambridge Structural Database. A
flat conformation is more common than a puckered one. DFT calculations at the B3LYP level of theory indicate that the flat confor-
mation of a model metallodithiophosphinate ring is very slightly lower in energy (1.2 kcal/mol) than the puckered conformation.
ꢀ 2005 Elsevier B.V. All rights reserved.
Keywords: X-ray crystal structures; 4-Membered rings; Palladium complexes
1. Introduction
coordinating sulphur, which is slightly (in b) or signifi-
cantly (in c) longer than the covalent metal-sulphur bond.
The metal-sulphur bond lengths in coordination pattern d
are approximately the same, as are the P–S separations,
due to delocalisation of the negative charge. In both b
and c the P–S double bond is retained, although the double
bond character may be slightly affected by the coordina-
tion. It has been suggested that the coordination mode is
dependent on the metal involved, and the ligand surround-
ing it; with soft metal centres coordination modes a–c are
preferred, while harder metals tend to coordinate in mode
d [2].
This paper describes the preparation and analysis of a
diphenyldithiophosphinate complex of palladium, C36H30-
P2PdS2 (Scheme 2). NMR analysis gave no indication that
the ligand was hemilabile, nevertheless two sets of crystals
exhibiting different crystal morphology were obtained
when the complex was crystallised from two different sol-
vent mixtures. In order to investigate the possibility that
In the course of an ongoing study into improving cata-
lysts for alkene oligomerisation, we have focussed on pre-
paring model palladium complexes of bidentate ligands
that could exhibit hemilabile properties [1]. Our interest
turned to dithiophosphinate compounds as possible candi-
dates, since they can coordinate to virtually all main group
and transition metals in a variety of coordination modes
[2], including monodentate (with only one S atom coordi-
nating to the metal) and bidentate modes (where four-
membered rings are formed), as indicated in Scheme 1.
Coordination patterns b and c are characterised for-
mally by a dative bond between the metal and the second
*
Corresponding author. Tel.: +27 21 8083345; fax: +27 21 8083360.
E-mail addresses: ce@sun.ac.za (C. Esterhuysen), gjk@na.rau.ac.za
(G.J. Kruger).
1
Tel.: +27 11 4892368; fax: +27 11 4892819.
0020-1693/$ - see front matter ꢀ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2005.10.006

610
C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
S
S
S
S
P
P
P
P
M
M
M
M
S
S
S
S
a
b
c
d
Scheme 1.
Ph
S
celite-packed sintered glass filter and evaporated to dry-
ness. The product was purified by crystallisation from
mixed solvents. The 1:1 mixtures of THF/pentane and ben-
zene/pentane, respectively, yielded the distinctly different
solvated and non-solvated batches of crystals.
Pd
PPh2
S
Ph3P
Scheme 2.
Complex C36H30P2PdS2: colourless crystalline material,
1
yield: 0.112 g, 58%. H NMR (300 MHz, CDCl3, 20 ꢁC):
the two could contain the ligand in two different coordina-
tion modes both were analysed by single crystal X-ray dif-
fraction. The product complex exhibited two different
conformations: in the crystal without solvent (hereafter
designated 1), the 4-membered metallo-dithiophosphinate
ring was shown to be distorted from planarity, as opposed
to planar in the crystal that contained THF (hereafter des-
ignated 2). To better understand the reasons for such differ-
ent behaviour, and also establish the occurrence of energy
differences between such conformations we undertook an
analysis of similar compounds in the Cambridge Structural
Database (CSD) [3] and compared the two conformations
in an ab initio computational study.
d = 7.83–7.91 (m, 4H), 7.21–7.44 (m, 21H, PPh3,
[(Ph)2PS2H]), 7.01–7.05 (m, 2H, C6H5), 6.66–6.68 (m,
3H); 13C{1H} NMR (75 MHz, CDCl3, 20 ꢁC): d = 136.0
(d, JC–P = 4.1 Hz, 1H), 134.4 (d, JC–P = 11.7 Hz), 131.2
(d, JC–P = 3.1 Hz), 131.0 (s), 130.6 (s), 130.4 (s), 130.2
(d, JC–P = 2.4 Hz), 130.0 (d, JC–P = 11.7 Hz), 128.3
(d, JC–P = 13.1 Hz), 128.2 (d, JC–P = 10.3 Hz), 128.1
(d, JC–P = 7.9 Hz), 127.4 (s), 122.7 (s); 31P{1H} NMR
(121 MHz, CDCl3, 20 ꢁC): 77.29 (d, 1P) (J = 7.3 Hz),
30.12 (d, 1P) (J = 6.8 Hz).
2.3. X-ray structure determination
2. Experimental
Colourless crystals suitable for X-ray structure determi-
nation were obtained by crystallisation from a solution of
dichloromethane layered in a 1:1 ratio with pentane (1).
Crystallisation from a solution of tetrahydrofuran (THF)
layered in a 1:1 ratio with pentane yielded orange crystals
containing included solvent (2). Data were collected on a
Siemens SMART diffractometer [5] at room temperature
with graphite monochromated Mo Ka radiation and cor-
rected for Lorentz and polarization effects [6]. Empirical
absorption corrections were applied [7]. The structures
were solved by interpretation of a Patterson synthesis
which yielded the positions of the metal atoms. All non-
hydrogen atoms were refined anisotropically by full-matrix
least-squares calculations on F2 using SHELXL-97 [8] within
the X-seed environment [9]. The presence of extra peaks of
electron density on the difference Fourier map indicated the
existence of disorder in the dithiosphosphinate complex in
2, with second atomic positions being located for the Pd,
both S atoms and P(2). The positions of these four atoms
could be refined, with anisotropic displacement parameters
for the Pd, while the site occupancy refined to 5%. This
small contribution by the minor component meant that
the positions of the carbon atoms could not be identified
from the difference Fourier map, and thus the remainder
of the disordered portion was not further modelled. Fur-
ther details of the data collections and structure refine-
ments are listed in Table 1. ORTEP-3 for Windows [10]
was used to generate the figures, with ellipsoid probabilities
at 50%.
2.1. General procedures and instruments
All reactions and manipulations were carried out under
a dry argon atmosphere using standard Schlenk and vac-
uum-line techniques. All solvents were dried and purified
by conventional methods and were freshly distilled under
argon shortly before use. Other reagents were used without
further purification. NMR spectra were measured on a
Varian VXR 300 spectrometer (1H, 300 MHz; 13C,
75.48 MHz; 31P, 127.4 MHz) at 20 ꢁC. Chemical shifts
are reported in ppm relative to signals for residual 1H
and 13C atoms in the deuterated solvents.
2.2. Synthesis
Diphenyldithiophosphinic acid, 0.121 g (4.84 · 10À4 mol),
was dissolved in 10 ml of methanol and 2 ml of THF.
2.42 · 10À4 mol dry sodium carbonate (0.026 g) was added
and the mixture stirred for 1 h. The solution was then
evaporated to dryness to deliver a colourless product that
was placed under high vacuum for extended periods to
ensure all the water formed was removed. 0.073 g (2.69 ·
10À4 mol) of this sodium salt was dissolved in 7 ml of
anhydrous THF and added dropwise to 0.200 g (2.69 ·
10À4 mol) (PPh3)2Pd(Ph)(Cl) [4] and stirred vigorously for
36 h. The reaction mixture turned yellow during this per-
iod. The solution was filtered through an anhydrous

C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
611
Table 1
Crystal data and structure refinement for complexes 1 and 2
1
2
Empirical formula
C36H30P2PdS2
695.06
296(2)
C36H30P2PdS2 Æ C4H8O
767.16
296(2)
Formula weight (g molÀ1
Temperature (K)
)
ꢀ
Crystal system, space group
Unit cell dimensions
monoclinic, P21/n
triclinic, P1
˚
a (A)
12.9199 (6)
16.2919 (8)
15.9092 (8)
90
103.777 (1)
90
3252.4 (3)
4
10.9884 (6)
3.2742 (7)
13.4552 (7)
88.534 (1)
83.825 (1)
71.356 (1)
1848.74 (17)
2
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
Z
Calculated density (Mg mÀ3
Absorption coefficient (l) (mmÀ1
F(000)
)
1.419
0.821
1416
1.378
0.731
788
)
Crystal size
h Range for data collection (ꢁ)
Index ranges
Reflections collected/unique [Rint
Completeness to maximum 2h
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
0.38 · 0.32 · 0.25
1.82–25.00
0.35 · 0.35 · 0.25
1.52–25.00
À15 6 h 6 13,À19 6 k 6 19,À15 6 l 6 18
16679/5661 [0.0437]
98.9%
full-matrix least-squares on F2
5661/0/370
À13 6 h 6 13, À15 6 k 6 15, À15 6 l 6 12
]
9840/6230 [0.0489]
95.9%
full-matrix least-squares on F2
6230/2/458
1.087
1.201
Final R indices [I > 2r(I)]
R indices (all data)
Largest difference in peak and hole (e A
R1 = 0.0466, wR2 = 0.0715
R1 = 0.0733, wR2 = 0.0785
0.426 and À0.329
R1 = 0.0577, wR2 = 0.1210
R1 = 0.0698, wR2 = 0.1287
0.910 and À0.506
À3
˚
)
2.4. CSD analysis and computational details
square-planar configuration, with the dithiophosphinate
ligand coordinating bidentately to form a 4-membered
Pd–S–P–S metallocyclic ring. The overall molecular struc-
ture of 2 is similar, although there are significant differences
in the central core, including the conformation of the
4-membered Pd–S–P–S ring. In order to more clearly
appreciate the differences Fig. 2 shows the spheres of coor-
dination around the Pd atoms.
The February 2005 version of the CSD (338445 entries)
was searched using ConQuest [11] for all structures con-
taining a 4-membered metal-dithiophosphinate ring, M–
S–P–S with M = any metal. The data were analysed using
Vista, as part of the CSD suite of programs [3].
All calculations were performed using GAUSSIAN-98 [15]
at the non-local DFT level of theory with Beckeꢀs 3-param-
eter exchange functional [12] mixed with Lee–Yang–Parrꢀs
correlation functional (B3LYP) [13], using the LANL2DZ
basis set [14]. Simplified model complexes were used, where
the phenyl rings in the structures obtained from the X-ray
diffraction analyses were replaced by methyl groups.
Geometry optimisations of both conformations were per-
formed, with vibrational frequency calculations used to
confirm that minimum energy conformations had been
obtained.
In 1 the deviation from the square-planar configuration
around the Pd is more pronounced than in 2, with a max-
˚
imum deviation from planarity of 0.135 (1) A by C(31) in 1,
˚
compared to 0.038(2) A by C(31) in 2. The two Pd–S sep-
arations are significantly different in both 1 and 2, with
the effect being more pronounced in 2, where the difference
˚
˚
between Pd–S(1) and Pd–S(2) is 0.11 A for the major disor-
der component (although only 0.01 A for the minor com-
ponent), suggesting that the dithiophosphinate ligand is
coordinating according to either pattern b or c in Scheme
1. However, at the same time the difference in S–P bond
lengths in each compound is negligible, thus implying that
coordination mode d should also be considered in both
complexes, despite Pd2+ being a soft metal. The longer
Pd–S(2) distance can instead be attributed to the larger
trans influence of the phenyl group compared to the
phosphine.
3. Results and discussion
3.1. Molecular structures from single-crystal X-ray
diffraction analysis
The molecular structure and numbering scheme of 1 are
shown in Fig. 1. Selected bond lengths and angles are listed
in Table 2. As can be seen from Table 2, the conformation
of the minor disorder component is poorly defined, hence
most of the discussion will focus on the 95% disordered
structure. The ligands are arranged around the Pd in a
In 1 the ring is strongly puckered, as seen in Fig. 2(a),
˚
with P(1) lying 1.188(1) A above the plane containing Pd,
S(1) and S(2) and the puckering angle between the Pd–
S(1)–S(2) and P(1)–S(1)–S(2) planes being 23.2(1)ꢁ,
whereas the ring in 2 (Fig. 2(b) and (c) – major and minor

612
C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
Fig. 1. Molecular structure and numbering scheme of 1. Hydrogens are omitted for clarity. Ellipsoids are shown at the 50% probability level.
disorder components, respectively), is planar, with P(1)
˚
bent only 0.065(3) A away from the Pd–S(1)–S(2) plane
Table 2
Selected bond distances (A) and angles (ꢁ) for 1 and 2
˚
(at puckering angles of 3.0(1)ꢁ and 7(1)ꢁ for the major
and minor disorder components, respectively). The pucker-
ing can also be measured by the Pd–S(1)–P(1)–S(2) torsion
angle, which is 17.45(6)ꢁ in the distorted ring in 1, as
1
2a Pd–S–P–Sb M–S–P–Sc
Bond distances
Pd–C(31)
Pd–P(2)
2.007(4)
2.2534(10)
2.015(6)
2.2628(14)
[2.23(3)]
2.3845(17)
[2.38(4)]
2.493(2)
[2.39(4)]
2.008(2)
[1.99(2)]
2.005(2)
[2.04(2)]
2.990(1)
[3.04(1)]
Pd–S(1)d
Pd–S(2)
S(1)–P(1)
S(2)–P(1)
Pd–P(1)
2.4029(10)
2.4571(11)
2.0142(14)
2.0108(14)
2.956(1)
2.35(3)
2.37(3)
2.01(2)e
2.51(13)
2.57(18)
1.99(3)e
2.99(6)
83.5(9)
3.14(14)
78(4)
Bond angles
S(1)–Pd–S(2)
83.33(3)
83.52(4)
82.20(5)
83.68(6)
[82.2(7)]
85.32(7)
[87.5(11)]
82.56(8)
[86.3(12)]
108.36(9)
[102.1(16)]
Pd–S(1)–P(1)
Pd–S(2)–P(1)
S(1)–P(1)–S(2)
106.78(6)
103(3)
107(3)
Average literature values calculated from CSD data are included for
comparison.
a
Values for minor disorder component given in parenthesis.
Average value calculated from 25 fragments in 19 structures, standard
b
deviation in parenthesis.
c
Average value calculated from 1006 fragments in 499 structures,
standard deviation in parenthesis.
d
Pd–S bond lengths were chosen so that Pd–S(1) < Pd–S(2).
Average for S(1)–P(1) and S(2)–P(1).
Fig. 2. Sphere of coordination around Pd for (a) 1, (b) major disorder
component of 2 and (c) minor disorder component of 2.
e

C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
613
Table 3
opposed to only 2.31(10)ꢁ in 2 (11.4(14) for the minor dis-
˚
p–p interaction distances for 1 and 2 (A)
order component). The puckering observed means that in 1
˚
1
2
the Pd to P(1) distance is shorter than in 2 (2.956(1) A as
˚
opposed to 2.990(2) A or 3.04(1)).
C(123)Á Á ÁH(236)
C(124)Á Á ÁH(224)
H(126)Á Á ÁC(33)
H(123)Á Á ÁC(226)
H(122)Á Á ÁC(214)
H(213)Á Á ÁC(115)
2.87
2.97
2.98
2.99
3.08
3.13
C(124)Á Á ÁH(222)
H(124)Á Á ÁC(213)
C(126)Á Á ÁH(214)
C(122)Á Á ÁH(113)
H(232)Á Á ÁC(213)
2.87
3.06
3.08
3.10
2.87
This puckering in the solid state can be ascribed to the
C(12X) phenyl ring on the diphenyldithiophosphinate
ligand being involved with five different weak off-set face-
to-edge p–p interactions (see Fig. 3); three to each of the
phenyl rings in the triphenylphosphine ligand on a neigh-
bouring molecule (coloured purple), one to a triphenyl-
phosphine group on a second neighbouring molecule
(coloured red) and the fifth to the phenyl ligand on a third
neighbouring molecule (coloured blue) with face-to-edge
to the purple and blue coloured molecules, respectively.
The packing of the molecules relative to each other allows
the formation of these intermolecular interactions without
the concomitant distortion observed for 1. Furthermore, the
longer separations in 2 suggest that the interactions are
weaker and do not supply as much stabilisation to the
structure.
˚
separations of 2.87, 2.99, 3.08, 2.97 and 2.98 A, respectively
(see Table 3). The PÁ Á ÁP distance to the purple coloured
˚
molecule is 7.116(1) A. These distances agree well with val-
ues given for so-called ‘‘phenyl embraces’’ by Dance and
˚
Scudder [16], who suggest CÁ Á ÁH distances of 2.8–3.2 A,
˚
and PÁ Á ÁP distances of 6.4–7.4 A. In order to maximise
3.2. Conformational analysis using CSD
these interactions the diphenyldithiophosphinate ligand
needs to bend away from the molecular plane as observed.
A further off-set face-to-edge interaction is observed for the
second phenyl in the diphenyldithiophosphinate ligand. A
smaller number of similar, although weaker, off-set face-
to-edge p–p interactions (Table 3) are observed in the crys-
tal structure of 2, resulting in an extended 3-dimensional
network of interlocked molecules with the THF solvent
molecules located in narrow channels, as shown in Fig. 4.
The search for 4-membered
metal) yielded 537 structures, with a total of 1074 frag-
ments, that matched the search criteria. 19 of these con-
rings (M = any
M-S-P-S
tained a
ring, with a total of 25 fragments
- - -
Pd S P S
matching the search criteria. The average bond lengths
and angles in the and rings as calculated
Pd-P-S-P
M-S-P-S
from these results are listed in Table 4. These values agree
well with those in 1 and 2, except for the average Pd–S
bond length, which is shorter than all four Pd–S bond
˚
˚
The PÁ Á ÁP distances are longer, 8.244(2) A and 8.611(2) A
Fig. 3. Off-set face-to-edge p–p interactions formed by the C(12X) phenyl ring in 1. The three neighbouring molecules are coloured in purple, red and blue,
respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

614
C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
Fig. 4. Off-set face-to-edge p–p interactions formed by the C(12X) phenyl ring in 2. The three neighbouring molecules are coloured in purple, red and blue,
respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 4
˚
Average bond lengths (A) and angles (ꢁ) calculated from the literature
structures reported in the CSD
Parameters in Pd–S–P–Sa
Parameters in M–S–P–Sb
Pd–S(1)c
Pd–S(2)
S–P
2.35(3)
2.37(3)
2.02(3)
3.00(7)
2.51(13)
2.57(18)
1.99(3)
Scheme 3.
Pd–P
3.14(14)
À
˚
of less than 0.05 A, indicating delocalisation in the PS2
S–P–S
S–M–S
a
103(3)
83.5(9)
107(4)
79(4)
fragment and suggesting coordination mode d.
The two P–S distances can be expected to not differ sig-
nificantly since there are two resonance structures (Scheme
3). This was confirmed by a search for dithiophosphinate
anions, which revealed that 91.5% of the 71 structures
found had a difference in P–S bond lengths of less than
Average value calculated from 25 fragments in 19 structures, standard
deviation in parenthesis.
b
Average value calculated from 1074 fragments in 537 structures,
standard deviation in parenthesis.
c
Pd–S bond lengths were chosen so that Pd–S(1) < Pd–S(2).
˚
0.05 A, just as found in the complexes. The larger differ-
ences observed can be explained by packing effects.
The S–M–S–P torsion angle, used as an indicator of the
puckering in the S–M–S–P ring, varies between 0ꢁ and
22.15ꢁ (absolute values of all torsion angles were taken).
In order to make a comparison between complexes that
can be designated as having a flat conformation and those
having a puckered conformation an arbitrary cutoff value
of 10ꢁ was chosen as indicating a flat geometry. On the
basis of this designation 88.9% of all S–M–S–P fragments
can be considered flat. These results suggest that the puck-
ered conformation exhibited by 1 is probably the strained
conformation, since the flat conformation observed in 2
is more common. In order to test this ab initio calculations
were performed.
lengths in 1 and 2. In fact, the two Pd–S(2) bond lengths in
1 and 2 are the longest Pd–S bond lengths yet observed in
this type of complex. The previous longest Pd–S bond was
˚
2.430 A [17]. In addition, most of the complexes reported in
the literature do not show such a large difference in their
two metal-S bond lengths as observed in 2. In particular,
most Pd complexes have coordination mode d (Scheme
1), with small differences between P–S and Pd–S bond
lengths (the largest previously observed difference between
˚
Pd–S bond lengths is 0.083 A) [18]. Only 11.8% of all frag-
ments have a difference in metal–S bond lengths of greater
˚
than 0.11 A. The majority of these are complexes of dith-
iophosphinate ligands with main group metals (In, Tl,
Ge, Sn, Pb, Sb, Bi, Te) corresponding with coordination
in mode c (Scheme 1). Of the transition metal complexes,
only 6.3% have metal-S bond lengths that differ by more
3.3. Computational results
˚
than the 0.11 A observed for 2. In all of these complexes
The two experimentally obtained molecular structures
were used as a basis for geometry optimisations, although
the difference in P–S bond lengths is relatively small, with
93.7% of structures having a difference in P–S bond lengths

C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
615
Table 6
with the phenyl rings replaced with methyl groups (in the
case of 2 the solvent was omitted in order to directly com-
pare bonding energies). In both cases the geometry optimi-
sation yielded the same minimum energy structure
(designated 3), with a planar conformation of the PdS2P
4-membered ring. The bond lengths and angles of this min-
imum energy structure are listed in Table 5, and the sphere
of coordination around the Pd is shown in Fig. 5(a). In
order to determine the difference in energy between a com-
plex with the ligand in this conformation and one with a
puckered conformation a geometry optimisation was per-
formed where the S(2)–Pd–S(1)–P torsion angle was con-
strained to 17.5ꢁ, as in the crystal structure of 1. The
minimum energy conformation thus obtained is designated
4, and details of the structure are included in Table 5 and
Comparison of calculated absolute energies (a.u.) and energy differences
(kcal/mol) for the optimised model compound, 3, the optimised model
compound with torsion angle S–P–S–Pd fixed at 17.5ꢁ, 4, and the
monodentate model compound, 5
Energies (a.u.)
DE (kcal/mol)
3
4
5
À399.477
À399.474
À399.448
1.191a
17.695b
a
DE = E(4) À E(3).
DE = E(5) À E(3).
b
Fig. 5(b). A further optimisation was performed to deter-
mine whether the dithiophosphinate ligand could coordi-
nate in a monodentate fashion. The corresponding
minimum energy conformation is designated 5, and struc-
tural details are also included in Table 5 and Fig. 5(c).
The absolute energies of these three conformations are
given in Table 6, along with the difference in energies as
compared to the lowest energy conformation, 3.
Table 5
˚
Bond lengths (A) and angles (ꢁ) of the optimised model compound, 3, the
optimised model compound with torsion angle S–P–S–Pd fixed at 17.5ꢁ, 4,
and the monodentate model compound, 5
The results show clearly that the calculated difference in
energy between the flat and puckered conformations is very
small, with the flat conformation being more stable than
the puckered conformation. The complex with the model
ligand in the monodentate coordination mode is much
higher in energy, explaining why it is not found
experimentally.
3
4
5
Pd–S(1)
Pd–S(2)
S(1)–P
S(2)–P
Pd–P
2.510
2.621
2.189
2.181
3.212
2.510
2.627
2.189
2.183
3.143
2.448
4.283
2.242
2.141
3.705
S(1)–P–S(2)
S(1)–Pd–S(2)
S(2)–Pd–S(1)–P
105.4
85.2
0.0
104.3
84.4
17.5
118.6
61.1
26.5
The calculated bond lengths are longer than those
observed experimentally. However, the trends are similar,
˚
with the Pd–S bond lengths differing by 0.11 A, while the
S–P bonds are the same length. The bond lengths in the cal-
culated flat, 3, and puckered, 4, conformations do not differ
from each other, such as was observed experimentally.
Bond angles agree well with experimentally observed val-
ues. The short S(1)–P bond length in the monodentately
coordinated dithiophosphinate ligand in 5 clearly indicates
that it is a double bond, as expected according to Scheme
1a, whereas the S(2)–P bond length indicates a single bond.
4. Conclusion
The results of the CSD analysis and ab initio calcula-
tions show that 4-membered PdS2P rings will take up a flat
conformation in the absence of other influencing factors. In
the case of the two crystalline forms described crystallo-
graphically in this paper, the PdS2P ring in 1 is distorted
from planarity by the dithiophosphinate ligand bending
to enable the formation of weak intermolecular p–p inter-
actions between one of the phenyl rings and phenyl rings
on neighbouring molecules, whereas for 2 intermolecular
p–p interactions can form without distortion and the
PdS2P ring is thus flat.
Acknowledgements
We thank the NRF, the University of Stellenbosch and
the University of Johannesburg for funding.
Fig. 5. Sphere of coordination around Pd for (a) 3, (b) 4, (c) 5.

616
C. Esterhuysen et al. / Inorganica Chimica Acta 359 (2006) 609–616
[5] SMART Data Collection Software, Siemens Analytical X-Ray Instru-
ments Inc., Madison, WI, USA, 1996.
Appendix A. Supplementary material
[6] SAINT Data Reduction Software, Siemens Analytical X-Ray Instru-
ments Inc., Madison, WI, USA, 1996.
[7] G.M. Sheldrick, SADABS, Siemens Analytical X-Ray Instruments Inc.,
Madison, WI, USA, 1996.
[8] G.M. Sheldrick, SHELX-97. Program for Crystal Structure Analysis,
University of Go¨ttingen, Germany, 1997.
[9] L.J. Barbour, J. Supramol. Chem. 1 (2001) 189.
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic Data
Centre, CCDC Nos. 278302 (1) and 278303 (2). Copies of
this information may be obtained free of charge from the
Director, CCDC, 12 Union Road, Cambridge, CB2 1E2,
UK (fax: +44 1223 336 033; e-mail: deposit@ccdc.ca.ac.uk,
www: http://www.ccdc.cam.ac.uk). Supplementary data
associated with this article can be found, in the online ver-
sion, at doi:10.1016/j.ica.2005.10.006.
[10] L.J. Farrugia, J. Appl. Cryst. 30 (1997) 565.
[11] I.J. Bruno, J.C. Cole, P.R. Edgington, M. Kessler, C.F. Macrae, P.
McCabe, J. Pearson, R. Taylor, Acta Cryst. B 58 (2002) 389.
[12] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[13] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[14] (a) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270;
(b) W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284;
(c) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299.
[15] M.J. Frisch et al., GAUSSIAN 98 (Revision A.7), Gaussian, Inc.,
Pittsburgh, PA, 1998.
[16] I. Dance, M. Scudder, J. Chem. Soc., Chem. Commun. (1995) 1039.
[17] J.P. Fackler Jr., L.D. Thompson, I.J.B. Lin, T.A. Stephenson, R.O.
Gould, J.M.C. Alison, A.J.F. Fraser, Inorg. Chem. 21 (1982) 2397.
[18] K. Panneerselvam, T.-H. Lu, S.-F. Tung, S. Narayan, V.K. Jain,
Acta Cryst. C55 (1999) 541.
References
[1] M.W. Esterhuysen, H.G. Raubenheimer, R. Brull, C. Esterhuysen,
¨
G.J. Kruger, J. Organomet. Chem. 619 (2001) 164.
[2] (a) I. Haiduc, D.B. Sowerby, S.-F. Lu, Polyhedron 14 (1995) 3389;
(b) I. Haiduc, D.B. Sowerby, Polyhedron 15 (1996) 2469.
[3] F.H. Allen, Acta Cryst. B 58 (2002) 380.
¨
[4] W.A. Herrmann, C. Broßner, T. Priermeier, K. Ofele, J. Organomet.
Chem. 481 (1994) 97.