Reactions of {[Pd(μ-SC6F5)(μ-dppm)Pd](μ-SC6F 5)}4·2O(C2H5)2. Crystal Structures of the Complexes [(Ph3P)Pd(μ-SC6F5)(μ-dppm)Pd(SC 6F5)]·1.4CH2Cl2

Article abstract of DOI:10.1021/ic960818e

{[Pd(μ-SC₆F₅)(μ-dppm)Pd](μ-SC ₆F₅)}₄ reacts 1:4 with neutral ligands L to give [LPd(μ-SC₆F₅)(μ-dppm)Pd(SC₆F ₅)] or 1:8 to form [LPd(μ-SC₆F₅)(μ-dppm)PdL]⁺ (dppm = bis(diphenylphosphino)methane). These binuclear complexes retain the palladium-palladium bond and the two dissimilar bridging ligands, as demonstrated by the X-ray structural determinations carried out on [(Ph₃P)Pd(μ-SC₆F₅)(μ-dppm)Pd(SC ₆F₅)]·1.4CH₂Cl₂ and [(Ph₃P)Pd(μ-SC₆F₅)(μ-dppm)Pd(PPh ₃)]SO₃CF₃·2CH₂Cl₂. Ab initio calculations on the model systems [(H₃P)Pd(μ-H₂PCH₂-PH ₂)(μ-SH)Pd(PH₃)]⁺ and [(H₃P)Pd(μ-H₂PCH₂PH₂)Pd(PH ₃)]²⁺ show that the metal-metal bond arises mainly from interactions between palladium sp orbitals, which also play a predominant role in the binding with the sulfur bridge.

Full text of DOI:10.1021/ic960818e

1912
Inorg. Chem. 1997, 36, 1912-1922
Reactions of {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2. Crystal Structures of the
Complexes [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]‚1.4CH2Cl2 and
[(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)]SO3CF3‚2CH2Cl2 and ab Initio MO Calculations on
the Model Systems [(H3P)Pd(µ-H2PCH2PH2)(µ-SH)Pd(PH3)]+ and
[(H3P)Pd(µ-H2PCH2PH2)Pd(PH3)]2+
Rafael Uso´n,*,1a Juan Fornie´s,1a Javier Ferna´ndez Sanz,*,1b Miguel A. Uso´n,1a
Isabel Uso´n,1c and Santiago Herrero1a
Departamento de Qu´ımica Inorga´nica, Instituto de Ciencia de Materiales de Arago´n, Universidad de
Zaragoza-CSIC, E-50009 Zaragoza, Spain, Departamento de Qu´ımica F´ısica, Facultad de Qu´ımicas,
Universidad de Sevilla, E-41012 Sevilla, Spain, and Institut fu¨r Anorganische Chemie, Universita¨t
Go¨ttingen, Tammanstrasse 4, D-3400 Go¨ttingen, Germany
ReceiVed July 11, 1996X
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 reacts 1:4 with neutral ligands L to give [LPd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]
or 1:8 to form [LPd(µ-SC6F5)(µ-dppm)PdL]+ (dppm ) bis(diphenylphosphino)methane). These binuclear
complexes retain the palladium-palladium bond and the two dissimilar bridging ligands, as demonstrated by the
X-ray structural determinations carried out on [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]‚1.4CH2Cl2 and [(Ph3P)Pd(µ-
SC6F5)(µ-dppm)Pd(PPh3)]SO3CF3‚2CH2Cl2. Ab initio calculations on the model systems [(H3P)Pd(µ-H2PCH2-
PH2)(µ-SH)Pd(PH3)]+ and [(H3P)Pd(µ-H2PCH2PH2)Pd(PH3)]2+ show that the metal-metal bond arises mainly
from interactions between palladium sp orbitals, which also play a predominant role in the binding with the
sulfur bridge.
Introduction
The role of bis(diphenylphospine)methane (dppm) as a
constraining ligand that holds metal centers at short distances,
favoring unusual formal oxidation states, has been long known.
Thus, many binuclear complexes containing the palladium-
palladium-bonded “Pd2(dppm)2” moiety have been synthesized;2
similar compounds with two different bridging ligands are
scarce.
We recently described3 the synthesis of an octanuclear
palladium(I) compound (1) that contains four equivalent dipal-
4[Pd(dppm-P,P′)(SC6F5)2] + 2Pd2(dba)3‚CHCl3 f
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4
+ 6dba + 2CHCl3
1
(1)
dba ) 1,5-diphenyl-2,4-pentanedien-3-one
ladium(I) units asymmetrically bridged by bis(diphenylphos-
phino)methane and a pentafluorobenzenethiolate anion. As
shown from X-ray diffraction data (see Figure 1), the most
outstanding features of these moieties arise from the doubly
bridged PdI-PdI structure. In it, two palladium(I) ions are
simultaneously bridged through dppm and a sulfur center, giving
Figure 1. View of compound 1. Phenyl groups are omitted and only
the ipso carbons of the thiolates are shown for the sake of clarity.
rise to an unusual bicyclic structure. The noticeable structure
of this compound prompted us to examine both the reactivity
and the electronic structure of the bonds involved. The various
reaction possibilities (the thiolato bridges, the presumably
strained Pd2S cycles, the metal-metal bonds) of this species
make it an interesting subject for reactivity studies, especially
if the reaction target can be chosen.
From a theoretical point of view, electronic properties of
metal-metal bonds have been the subject of considerable high-
level theoretical work. However, as far as the PdI-PdI bond is
concerned, only a few semiempirical calculations have been
reported.4-7 The large number of PdI-PdI binuclear complexes
known makes it worthwhile to undertake a more reliable, high-
X Abstract published in AdVance ACS Abstracts, March 15, 1997.
(1) (a) Instituto de Ciencia de Materiales de Arago´n. (b) Universidad de
Sevilla. (c) Universita¨t Go¨ttingen.
(2) (a) Chaudret, B.; Delavaux, B.; Poilblanc, R. Coord. Chem. ReV. 1988,
86, 191-243. (b) Russell, M. J. H.; Barnard, C. F. J. In ComprehensiVe
Coordination Chemistry; Wilkinson, G., Gillard, R. D., McCleverty,
J. A., Eds.; Pergamon Press Ltd.: Oxford, England, 1987; Vol. 5, pp
1103-1111. (c) Canty, A. J. In ComprehensiVe Organometallic
Chemistry; Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.;
Pergamon Press Ltd.: Oxford, England, 1995; Vol. 9, pp 225-227.
(3) Uso´n, R.; Fornie´s, J.; Falvello, L. R.; Uso´n, M. A.; Uso´n, I.; Herrero,
S. Inorg. Chem. 1993, 32, 1066-1067.
S0020-1669(96)00818-X CCC: $14.00 © 1997 American Chemical Society

dppm-Bridged Dipalladium(I) Complexes
Inorganic Chemistry, Vol. 36, No. 9, 1997 1913
Table 1. Elemental Analyses and Molar Conductivities of Complexes 2-8
found (calcd)
a
no.
complex
% C
% H
% S
ΛM
2
3
4
5b
5c
6
7
8
[Pd2(SC6F5)2dppm(PPh3)]
[Pd2(SC6F5)2dppm(PCy3]
[Pd2(SC6F5)2dppm(CO)]‚CH2Cl2
[Pd2(SC6F5)2dppm(PPh3)2]ClO4
[Pd2(SC6F5)2dppm(PPh3)2]SO3CF3
[Pd2(SC6F5)2dppm(PMePh2)2]SO3CF3
52.73 (52.52)
51.82 (51.78)
42.29 (42.26)
56.31 (56.66)
55.35 (55.56)
51.45 (51.76)
50.65 (50.45)
49.09 (49.45)
2.98 (2.97)
4.16 (4.34)
2.05 (2.18)
3.71 (3.69)
3.49 (3.57)
3.70 (3.59)
3.33 (3.31)
3.08 (3.17)
5.05 (5.10)
5.12 (5.03)
6.11 (5.79)
2.42 (2.26)
4.50 (4.36)
4.96 (4.77)
4.14 (3.90)
4.09 (3.88)
128.8
126.5
112.2
114.6
110.9
.
[Pd2(SC6F5)2dppm(AsPh3)2]SO3CF3 CH2Cl2
[Pd2(SC6F5)2dppm(SbPh3)2]SO3CF3
a Ω-1 cm2 mol-1
.
- dppm - Ph + PPh3, 6%), 797 (M - SC6F5 - PPh3, 12%), 707 (M
- C6F5 - dppm, 12%), 689 (M - Pd - SC6F5 - PPh3, 12%), 339
(PPh4, 100%).
level theoretical analysis of this bond, first, and to then study
how it is perturbed when a thiolato group bridges the PdI-PdI
unit.
Synthesis of [(Cy3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)] (3). To a
solution of 0.0713 g (0.2 mmol) of Cy3PCS2 in 15 mL of dichlo-
romethane was added 0.2065 g (0.05 mmol) of {[Pd(µ-SC6F5)(µ-dppm)-
Pd](µ-SC6F5)}4‚2O(C2H5)2. After 1 h of stirring, the solution was
concentrated to 3 mL, and 40 mL of pentane was added. The
precipitated solid was filtered off, washed with 5 mL of pentane, and
suction-dried. Prolonged drying in Vacuo was necessary for the total
removal of CS2. Yield: 74%. IR (cm-1): 1631 w, 1589 w, 1576 w,
1512 s, 1312 vw, 1268 vw, 1192 w, 1175 vw, 1163 w, 1130 m, 1101
m, 1084 s, 1050 s, 1027 w, 1001 s, 978 s, 918 w, 888 w, 851 s, 777
w, 694 s, 676 w, 624 w, 583 w, 523 m, 503 m, 490 w, 340 w. 19F
NMR: δ -131.0 (d, Jom ) 25 Hz, Fo), -133.4 (d, Jom ) 23 Hz, Fo),
-163.7 (m, Fm), -157.3 (t, Jmp ) 20 Hz), -166.5 (m, Fm + Fp). 31P
NMR: δ 27.1 (dd, Jab ) 61 Hz, Jac ) 51 Hz, Pa), 4.8 (dd, Jbc ) 34 Hz,
Pb), -2.8 (dd, Pc). MS(FAB+), m/z: 2354 (2M - SC6F5, 37%), 2072
(2M - SC6F5 - PCy3, 21%), 1791 (2M - SC6F5 - 2PCy3, 26%),
1276 (M, 62%), 1077 (M - SC6F5, 68%), 689 (M - Pd - SC6F5 -
PCy3, 100%). MS(FAB-), m/z: 1195 (M - PCy3 + SC6F5, 30%), 996
(M - PCy3, 19%), 703 (M - Pd - dppm - PCy3 + SC6F5, 83%),
504 (M - Pd - dppm - PCy3, 100%).
Synthesis of [(SC6F5)Pd(µ-CO)(µ-dppm)Pd(SC6F5)] (4). A 0.1033
g sample (0.025 mmol) of {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚2O-
(C2H5)2 was suspended in 40 mL of dichloromethane, and CO was
bubbled through the suspension for 1 h. The resulting dark solution
was gravity-filtered and evaporated to 4 mL at 0 °C. After the addition
of hexane (40 mL), the solid was filtered off, washed with hexane (2
mL), and dried in a stream of carbon monoxide. Yield: 34%. IR
(cm-1): 1897 s, br, 1511 vs, 1439 vs, 1098 s, 1081 s, 972 s, 857 s,
768 m, br, 691 s, 517 m, 503 m, 478 m. 19F NMR: δ -131.6 (d, Jom
) 27 Hz, Fo), -165.0 (m, Fm), -162.0 (t, Jmp ) 21 Hz). 31P NMR:
δ -6.5 (s).
In this paper, we report the reactivity of compound 1 and
the theoretical calculations on models based on PdI2 and the
PdI2S cycle. The paper is arranged as follows: First, some
reactions of compound 1 that preserve the “Pd(µ-SC6F5)(µ-
dppm)Pd” unit are described. Characterizations of the obtained
compounds by both spectroscopic (NMR, IR, MS) and X-ray
diffraction techniques are presented. Then a theoretical analysis
based on ab initio Hartree-Fock calculations on PdI2 and PdI2S
models is reported. In this analysis, optimized geometries and
vibrational frequencies are provided and compared with ex-
perimental results. Finally, the main conclusions are outlined.
Experimental Section
Unless otherwise stated, all reactions were routinely carried out under
dry nitrogen, at room temperature, in solvents purified by standard
procedures. The compound {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚
2O(C2H5)2 was prepared as previously reported.3
C, H, and S analyses were performed with a Perkin-Elmer 2400
microanalyzer. IR spectra were recorded (over the range 4000-250
cm-1) on Perkin-Elmer 833 and 1730 FT spectrophotometers, using
Nujol mulls between polyethylene sheets.8 19F and 31P NMR spectra
of CDCl3 solutions of the compounds were run on a Varian XL-200 or
UNITY 300 spectrometer, and chemical shifts are relative to CFCl3
and external 85% H3PO4, respectively. Conductivities of ∼5 × 10-4
M acetone solutions were measured with a Philips PW9509 apparatus,
using a PW9550/60 cell. Mass spectrometric data were obtained using
FAB+ and FAB- techniques on a VG Autospec apparatus. The matrix
used was 3-nitrobenzyl alcohol, and the samples were dissolved in CH2-
Cl2.
Elemental analyses and molar conductivities are given in Table 1.
Synthesis of [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)] (2). To a
dichloromethane (25 mL) solution containing 0.0525 g (0.2 mmol) of
PPh3 was added 0.2065 g (0.05 mmol) of {[Pd(µ-SC6F5)(µ-dppm)Pd]-
(µ-SC6F5)}4‚2O(C2H5)2. After 3 h of stirring, the solvent was evapo-
rated and the residue was treated with a mixture of diethyl ether (3
mL) and hexane (5 mL). The solid obtained was filtered off, washed
with hexane (2 × 3 mL), and dried in Vacuo. Yield: 89%. IR
(cm-1): 1633 w, 1619 w, 1587 w, 1573 w, 1512 vs, 1437 vs, 1309 w,
1274 vw, 1184 m, 1160 w, 1098 vs, 1083 vs, 1027 m, 999 m, 977 vs,
Synthesis of [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)]ClO4 (5b).
Caution! Perchlorate salts of metal complexes with organic ligands
are potentially explosiVe. Only small amounts of these materials should
be prepared, and they should be handled with great care.
(a) A 0.1991 g sample (0.048 mmol) of {[Pd(µ-SC6F5)(µ-dppm)-
Pd](µ-SC6F5)}4‚2O(C2H5)2 was added to a dichloromethane (10 mL)
solution of 0.0415 g (0.2 mmol) of Ag(ClO4) and 0.1049 g (0.4 mmol)
of PPh3. After 17 h of stirring, the Ag(SC6F5) precipitate was filtered
off and the solvent was removed. The crude orange solid was
recrystallized from dichloromethane (1 mL)/hexane (10 mL), washed
with hexane (15 mL), and dried in Vacuo over P2O5. Yield: 89%.
918 w, 855 vs, 781 s, 739 s, 692 vs, 519 vs, 506 s, 489 m, 456 m. 19
F
NMR: δ -131.8 (d, Jom ) 25 Hz, Fo), -133.4 (d, Jom ) 22 Hz, Fo),
-164.1 (m, Fm), -158.3 (t, Jmp ) 21 Hz, Fp), -166.3 (m, Fm + Fp).
31P NMR: δ 10.1 (dd, Jab ) 63 Hz, Jac ) 51 Hz, Pa), 3.9 (dd, Jbc ) 32
Hz, Pb), 0.0 (dd, Pc). MS(FAB+), m/z: 2317 (2M - SC6F5, 4%), 2054
(2M - SC6F5 - PPh3, 2%), 1791 (2M - SC6F5 - 2PPh3, 7%), 1321
(M - SC6F5 + PPh3, 10%), 1059 (M - SC6F5, 11%), 860 (M - SC6F5
(b) To a mixture of 0.2623 g (0.1 mmol) of PPh3 and 0.2073 g (0.1
mmol) of Ag(ClO4) in dichloromethane (20 mL) was added 0.1258 g
(0.1 mmol) of [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]. After 150 min
of stirring, the Ag(SC6F5) precipitate was filtered off and the reddish
filtrate was evaporated to dryness. The solid so obtained was
recrystallized from dichloromethane (1 mL)/hexane (10 mL). Yield:
90%. IR (cm-1): 1639 vw, 1587 vw, 1573 vw, 1511 s, 1435 vs, 1310
w, 1185 vw, 1143 vw, 1094 vs, 1027 w, 999 w, 978 m, 853 m, 784 w,
(4) Goldberg, S. Z.; Eisenberg, R. Inorg. Chem. 1976, 15, 535-541.
(5) Harvey, P. D.; Murtaza, Z. Inorg. Chem. 1993, 32, 4721-4729.
(6) Tanase, T.; Nomura, T.; Fukushima, T.; Yamamoto, Y.; Kobayashi,
K. Inorg. Chem. 1993, 32, 4578-4584.
(7) Harvey, P. D.; Hubig, S. M.; Ziegler, T. Inorg. Chem. 1994, 33, 3700-
3710.
741 s, 693 vs, 623 m, 520 vs, 507 vw. 19F NMR: δ -131.0 (d, Jom
)
25 Hz, Fo), -162.6 (m, Fm), -156.5 (t, Jmp ) 21Hz, Fp). 31P NMR:
δ 16.7 (m, Pa), 2.3 (m, Pb). MS(FAB+), m/z: 1321 (M, 100%), 1059
(8) Uso´n, M. A. J. Chem. Educ. 1990, 66, 412.

1914 Inorganic Chemistry, Vol. 36, No. 9, 1997
Uso´n et al.
(M - PPh3, 39%), 860 (M - dppm - Ph, 23%), 797 (M - 2PPh3,
17%), 707 (M - dppm - 3Ph + H, 19%), 689 (M - 2PPh3 - Pd,
26%).
1792 (2M - 4SbPh3 + SC6F5, 17%), 1503 (M, 44%), 1149 (M -
SbPh3, 27%), 950 (M - SC6F5 - SbPh3, 24%), 873 (M - SC6F5 -
SbPh3 - Ph, 33%), 797 (M - 2SbPh3, 100%), 521 (M - SC6F5 -
2SbPh3 - Ph, 33%).
Synthesis of [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)]SO3CF3 (5c).
To a solution of 0.0514 g (0.2 mmol) of Ag(SO3CF3) and 0.1049 g
(0.4 mmol) of PPh3 in dichloromethane (30 mL) was added 0.2065 g
(0.05 mmol) of {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2, and
the mixture was stirred for 105 min. The Ag(SC6F5) that formed was
filtered off and washed with dichloromethane (3 × 3 mL). The
collected filtrate and washings were evaporated to dryness under
reduced pressure, the crude solid was just dissolved in dichloromethane
(ca. 2 mL), and the solution was treated with hexane (12 mL). The
tomato red solid thus obtained was filtered off, washed with hexane (2
× 2 mL), and dried in Vacuo over P2O5. Yield: 91%. IR (cm-1):
1640 vw, 1587 vw, 1573 vw, 1512 vs, 1436 vs, 1309 vw, 1185 w,
1097 s, 1085 s, 1030 vs, 1269 vs, 1223 w, 1151 s, 999 w, 978 s, 858
m, 788 m, 743 vs, 693 vs, 521 vs, 506 m, 494 m, 481 m, 638 vs, 366
w. 19F NMR: δ -78.4 (s, SO3CF3), -131.1 (d, Jom ) 26 Hz, Fo),
-162.5 (m, Fm), -156.3 (t, Jmp ) 20 Hz, Fp). 31P NMR: δ 16.7 (m,
Pa), 2.3 (m, Pb). MS(FAB+), m/z: 1321 (M, 100%), 1059 (M - PPh3,
53%), 860 (M - dppm - Ph, 26%), 797 (M - 2PPh3, 30%), 707 (M
- dppm - 3Ph + H, 30%), 689 (M - 2PPh3 - Pd, 28%).
Synthesis of [(Ph2MeP)Pd(µ-SC6F5)(µ-dppm)Pd(PMePh2)]SO3CF3
(6). To a suspension of 0.0514 g (0.2 mmol) of Ag(SO3CF3) in 10
mL of dichloromethane were added first 79 µL (0.42 mmol) of PMePh2
and then 0.2065 g (0.05 mmol) of {[Pd2(µ-SC6F5)(µ-dppm)](µ-
SC6F5)}4‚2O(C2H5)2 and a further 10 mL of the same solvent. After
16 h of stirring, the Ag(SC6F5) precipitate was filtered off and the
solvent was removed. The crude residue was dissolved in 2 mL of
dichloromethane, and addition of 20 mL of heptane gave an oil, which
was stirred for 8 h. The resulting orange solid was filtered off, washed
with 5 mL of pentane, and suction-dried. Yield: 93%. IR (cm-1):
1642 vw, 1587 vw, 1576 vw, 1510 s, 1438 vs, 1307 vw, 1280 vs, 1265
vs, 1225w, 1188 vw, 1156 m, 1132 s, 1103 m, 1080w, 1000 vw, 976
s, 854 m, 789 m, 745 m, 692 m, 637 s, 526 w, 509 s, 490 w, 342 w.
19F NMR: δ -78.4 (s, SO3CF3), -131.7 (d, Jom ) 26 Hz, Fo), -162.9
(m, Fm), -156.8 (t, Jmp ) 21 Hz, Fp). 31P NMR: δ 2.4 (m, Pa), -2.7
(m, Pb). MS(FAB+), m/z: 1196 (M, 100%), 996 (M - PPh2Me, 87%),
889 (M - Pd - PPh2Me, 16%), 797 (M - 2PPh2Me, 60%), 689 (M
- Pd - 2PPh2Me, 43%).
Synthesis of [(Ph3As)Pd(µ-SC6F5)(µ-dppm)Pd(AsPh3)]SO3CF3
(7). To a solution of 0.1225 g (0.4 mmol) of AsPh3 in 30 mL of
dichloromethane were added first 0.2065 g (0.05 mmol) of {[Pd(µ-
SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2 and then 0.0514 g (0.2
mmol) of Ag(SO3CF3), resulting in a dark amber solution. After 5 h
of stirring, the precipitated Ag(SC6F5) was filtered off, the solvent was
partially evaporated (to ca. 2 mL), and 5 mL of pentane was added.
The solid thus obtained was filtered off, washed with a further 5 mL
of pentane, and suction-dried. Yield: 84%. IR (cm-1): 1640 vw, 1582
w, 1513 s, 1436 vs, 1307 vw, 1272 vs, 1223 w, 1186 vw, 1150 s, 1099
m, 1084 m, 1031 s, 999 w, 978 m, 853 m, 788 w, 739 vs, 692 vs, 637
s, 524 m, 505 vw, 479 s, 329 m. 19F NMR: δ -78.4 (s, SO3CF3),
-131.3 (d, Jom ) 23 Hz, Fo), -161.9 (m, Fm), -155.5 (t, Jmp ) 21 Hz,
Fp). 31P NMR: δ 1.2 (s). MS(FAB+), m/z: 1410 (M, 26%), 1102 (M
- AsPh3, 34%), 904 (M - SC6F5 - AsPh3, 20%), 797 (M - 2AsPh3,
100%), 689 (M - Pd - 2AsPh3, 17%).
Structure Determinations. Single-crystal X-ray structure deter-
minations were carried out using Stoe-Siemens AED2 diffractometers
with graphite-monochromated Mo KR radiation (λ ) 0.710 73 Å). The
crystal data and data collection and refinement parameters are sum-
marized in Table 2.
(a) [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]‚1.4CH2Cl2. Slow dif-
fusion (at -28 °C) of hexane into a dichloromethane solution of
compound 2 gave four huge parallelepipedic red/green (dichroic)
crystals. A fragment was mounted on top of a glass fiber in a drop of
HMP lithium grease and rapidly cooled. Heavy-atom methods9 were
used to locate the palladium atoms, while subsequent cycles of least-
squares refinements and difference Fourier maps were used to locate
the remaining non-hydrogen atoms;10 hydrogen atoms were placed at
geometrically calculated positions and refined “riding” on the corre-
sponding carbon atom, with dependent isotropic displacement param-
eters.
(b) [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)]SO3CF3‚2CH2Cl2. Slow
diffusion of diethyl ether into a dichloromethane solution of 5c gave
red plates, one of which was mounted on a glass fiber in a drop of
rapidly cooled perfluoropolyether.11 Data vere collected at constant
speed, by following the learned profile method.12 Heavy-atom methods9
were used to locate the palladium, sulfur, and phosphorus atoms, while
subsequent cycles of least-squares refinements and difference Fourier
maps were used to locate the remaining non-hydrogen atoms;10
hydrogen atoms were placed at geometrically calculated positions, and
their positional coordinates were constrained to ride on those of the
corresponding carbon atoms, with dependent isotropic displacement
parameters. An extended version of SHELXL-9310 was used in the
last least-squares refinement to accommodate the 1757 parameters.
Results and Discussion
Although the octanuclear palladium(I) compound {[Pd(µ-
SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2 is stable in the solid
state and soluble in most common organic solvents, it rapidly
decomposes when taken into solution. On the other hand, in
diethyl ether or hexane, where the compound is insoluble,
reactions do not take place. Thus, only reactions faster than
the decay processes yield clean products, which makes it
necessary to add the solid substrate to the already dissolved
reactants.
Addition (1:4) of {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4‚
2O(C2H5)2 to a dichloromethane solution of a tertiary phosphine
L leads (eq 2) to the formation of the corresponding neutral
dipalladium(I) complex.
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 + 4L f
4[LPd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]
(2)
L ) PPh3(2), PCy3(3)
As expected, the 19F NMR spectra of these compounds show,
in the region assigned to ortho-fluorine nuclei of pentafluo-
robenzenethiolato groups (-125 to -135 ppm),13 two doublets,
the signals at higher fields being somewhat broad. The 31P
NMR spectra are characteristic of AMX spin systems, which
requires the phosphorus nuclei of the diphosphine to be
chemically inequivalent. The positions of the signals exclude
chelate behavior of the dppm ligand.14
Synthesis of [(Ph3Sb)Pd(µ-SC6F5)(µ-dppm)Pd(SbPh3)]SO3CF3
(8). To a solution of 0.1412 g (0.4 mmol) of SbPh3 in 30 mL of
dichloromethane was added 0.2065 g (0.05 mmol) of {[Pd(µ-SC6F5)-
(µ-dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2. After 15 min of stirring, 0.0514
g (0.2 mmol) of Ag(SO3CF3) was added to give a dark red-orange
solution. The mixture was stirred for a further 3 h in the absence of
light, and the precipitated Ag(SC6F5) was filtered off. The solution
was evaporated to dryness, the residue was extracted with 2 mL of
dichloromethane, and 40 mL of pentane was added. The resulting
precipitate was filtered off, washed with 5 mL of pentane, and suction-
dried. Yield: 81%. IR (cm-1): 1639 vw, 1578 w, 1511 vs, 1435 vs,
1305 vw, 1263 vs, 1223 w, 1187 vw, 1150 s, 1100 m, 1084 m, 1067
w, 1031 vs, 998 m, 978 m, 851 m, 785 w, 693 vs, 524 m, 505 vw, 637
vs, 454 m, 350 vw, 277 m. 19F NMR: δ -78.4 (s, SO3CF3), -131.2
(9) SHELXTL PLUS, release 4.0; Siemens Analytical X-Ray Instruments,
Inc.: Madison, WI, 1990.
(10) Sheldrick, G. M. SHELXL-93: A Program for Crystal Structure
Refinement. Universita¨t Go¨ttingen, Germany, 1993.
(11) Kottke, T.; Stalke, D. J. Appl. Crystallogr. 1993, 26, 615-619.
(12) Clegg, W. Acta Crystallogr. 1981, A37, 22-28.
(13) Uso´n, R.; Fornie´s, J.; Uso´n, M. A.; Apaolaza, J. A. Inorg. Chim. Acta
1991, 187, 175-180.
(d, Jom ) 24 Hz, Fo), -161.2 (m, Fm), -154.6 (t, Jmp ) 21 Hz, Fp). 31
NMR: δ 0.2 (s). MS(FAB+), m/z: 2498 (2M - 2SbPh3 + SC6F5, 3%),
P

dppm-Bridged Dipalladium(I) Complexes
Inorganic Chemistry, Vol. 36, No. 9, 1997 1915
Table 2. Crystal Data for Compounds 2, 5b, and 5c
2‚1.4CH2Cl2
5b
5c‚2CH2Cl2
formula
fw
crystal system
space group
a (Å)
b (Å)
c (Å)
C56.4H39.8Cl2.8F10P3Pd2S2
1376.57
“C67H52ClF5O4P4Pd2S”
1420.28
C70H56Cl4F8O3P4Pd2S2
1639.75
monoclinic
P21/n (No. 14)
14.054(2)
26.528(5)
16.532(2)
90
112.034(12)
90
5713.3(14)
4
orthorhombic
Pca21 (No. 29)
28.198(6)
17.668(4)
28.197(6)
90
90
90
14048(5)
8
1.343
orthorhombic
Pca21 (No. 29)
27.804(2)
17.659(2)
27.795(3)
90
90
90
13647(2)
8
1.596
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
d
calcd (g cm-3
)
1.600
cryst size (mm3)
0.57 × 0.53 × 0.49
0.57 × 0.57 × 0.34
0.75 × 0.60 × 0.10
µ (mm-1
)
0.987
0.727
0.908
transm factors: max, min
data collecn instrument
radiation
0.630, 0.502
0.987, 0.901
Stoe-Siemens AED2
0.9449, 0.6486
Mo KR (λ ) 0.71073 Å) (graphite monochromated)
temp (K)
200(2)
173(2)
153(2)
scan method
θ collecn range (deg)
no. of unique data
ω/θ
ω/2θ
ω
2.03-22.51
1.54-23.46
3.02-25.10
7457
10 583
15 768
no. of obsd reflns
7456
721
0.0559 (5769)
0.1591
0.0763, 28.6693
1.097
10 558
875
0.0984 (5685)
0.2412
0.1554, 50.8462
1.044
15 768
1757
0.0555 (12 873)
0.1341
no. of params refined
R1a [I > 2σ(I)] (no. of data)
wR2b(all data)
weighting parameters: g1, g2
quality of fitd (on F2)
largest shift/esd, final cycle
largest peak, hole (e/Å3)
c
0.0625, 43.3824
1.005
0.000
1.42, -0.72
0.002
1.325, -0.675
20.588e
0.633, -0.694
2
2
2
a R1 ) ∑||Fo| - |Fc||/∑|Fo|. b wR2 ) [∑w(Fo2 - Fc2)2/∑w(Fo )2]0.5
.
c w ) [σ2(Fo ) + (g1P)2 + g2P]-1; P ) [max(Fo ;0) + 2Fc2]/3. d Quality of
2
fit ) [∑w(Fo - Fc2)2/(Nobservns - Nparams)]0.5
.
e For U23(Cl(3′)) of the disordered solvent; maximum shift/esd for the atoms of the cations 0.000.
Table 3. Atomic Coordinates (×104) and Equivalent Isotropic
Displacement Parameters (Å2 × 103) for 2a
x
y
z
U(eq)a
Pd(1)
Pd(2)
P(1)
P(2)
P(3)
S(1)
S(2)
C(1)
797(1)
-515(1)
1101(2)
-153(2)
-1557(2)
1934(2)
-228(2)
123(7)
1940(1)
1770(1)
2767(1)
2539(1)
1350(1)
1785(1)
1235(1)
3019(3)
1899(1)
2640(1)
2199(1)
3271(1)
3248(1)
1174(2)
1645(1)
2577(5)
23(1)
21(1)
24(1)
23(1)
28(1)
36(1)
26(1)
28(2)
a U(eq) is defined as one-third of the trace of the orthogonalized Uij
tensor.
Table 4. Selected Bond Lengths (Å) and Angles (deg) for 2
Pd(1)-P(1)
Pd(1)-S(2)
Pd(1)-S(1)
Pd(1)-Pd(2)
Pd(2)-P(2)
2.255(2)
2.301(2)
2.366(2)
2.6068(9)
2.260(2)
Pd(2)-S(2)
Pd(2)-P(3)
P(1)-C(1)
P(2)-C(1)
2.321(2)
2.346(2)
1.836(8)
1.850(8)
P(1)-Pd(1)-S(2)
P(1)-Pd(1)-S(1)
S(2)-Pd(1)-S(1)
P(1)-Pd(1)-Pd(2)
S(2)-Pd(1)-Pd(2)
152.27(8) S(2)-Pd(2)-P(3)
99.90(8) P(2)-Pd(2)-Pd(1)
105.20(8) S(2)-Pd(2)-Pd(1)
108.37(8)
89.18(6)
55.30(5)
Figure 2. The molecule of compound 2 in the crystal (ellipsoids at
50% electron probability level; other atoms as circles of arbitrary radii;
H atoms omitted).
99.98(6) P(3)-Pd(2)-Pd(1) 161.53(6)
56.03(5) C(1)-P(1)-Pd(1)
108.9(3)
113.4(3)
68.67(6)
108.6(4)
S(1)-Pd(1)-Pd(2) 159.97(7) C(1)-P(2)-Pd(2)
The structure of compound 2 has been established by a single-
crystal X-ray diffraction study (see Figure 2 for structure and
numbering scheme). Table 3 lists selected atomic positional
parameters, and relevant bond distances and angles are given
in Table 4. The binuclear complex contains dichloromethane
as crystallization solvent, in two different sites (one of them
with partial occupation; the other modeled with two alternative
positions); bond distance and angle restraints were applied to
these molecules.
P(2)-Pd(2)-S(2)
P(2)-Pd(2)-P(3)
144.21(8) Pd(1)-S(2)-Pd(2)
107.41(8) P(1)-C(1)-P(2)
If the starting material is considered3 to be formed by four
Pd(µ-SC6F5)(µ-dppm)Pd units bridged through four additional
pentafluorobenzenethiolato groups to form a 12-membered
cycle, compound 2 can be described as the result of cleavage
of these bridges by PPh3; i.e., it contains the same SPd2P2C
moiety present in the octanuclear precursor 1.
The five-membered Pd2P2C ring is puckered15 (84% twist
with axis through Pd(1) and Pd(2) and 16% envelope with flap
(14) Fornie´s, J.; Navarro, R.; Urriolabeitia, E. P. J. Organomet. Chem. 1990,
390, 257-265.

1916 Inorganic Chemistry, Vol. 36, No. 9, 1997
Uso´n et al.
at P(2)) in contrast to the statistically preferred envelope
conformation found by Orpen16,17 for such systems; accordingly,
the P(1)-Pd(1)-Pd(2)-P(2) torsion angle (-10.5°) is far from
zero. For the six-membered SPd2P2C ring, P(2) lies 0.66 Å
above and S(2) 0.78 Å under the best plane, described by the
carbon, the remaining phosphorus, and both palladium atoms
(mean deviation 0.06 Å).
On the other hand, reaction (1:8) between {[Pd(µ-SC6F5)(µ-
dppm)Pd](µ-SC6F5)}4‚2O(C2H5)2 and a dichloromethane solu-
tion of triphenylphosphine leads to a cationic binuclear complex
(eq 4).
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 + 8PPh3 f
4[(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)]SC6F5
(4)
As compared to that of complex 1, the palladium-palladium
bond length (2.607(1) Å) increases slightly, as do the pal-
ladium-phosphorus distances to the donor atoms of the bridging
diphosphine ligand (Pd(1)-P(1) ) 2.255(2) and Pd(2)-P(2)
) 2.260(2) Å), which are however shorter than the Pd(2)-P(3)
distance of 2.346(2) Å. The terminal ligands do not lie collinear
with the palladium-palladium bond (Pd(2)-Pd(1)-S(1) )
160.0 and Pd(1)-Pd(2)-P(3) ) 161.5°; torsion angle P(3)-
Pd(2)-Pd(1)-S(1) ) -8.0°), as has already been found even
for symmetrically doubly-bridged M2(dppm)2X2 metal-metal-
bonded species.18
As in the starting octanuclear compound, the perfluorophenyl
rings of the thiolato groups are essentially parallel with (dihedral
angle 3.45°) and close to each other (mean of the centroid to
plane distances 3.34 Å) and the rings are offset (1.05 Å) as a
consequence of the polar/π interaction.19 The aromatic ring of
the bridging pentafluorobenzenethiolato group is also roughly
parallel with (8.6°) and close to (3.43 Å) a phenyl ring of the
triphenylphosphine ligand. The rings are tilted about the
S-Cipso bond, forming dihedral angles of 62.6 and 65.9° with
the mean SPd2P2C plane; the nonbonding distances to the nearest
ortho-fluorine atoms are however longer (Pd(1)-F(86) ) 3.275
and Pd(2)-F(92) ) 3.117 Å) than those in the parent compound.
A different type of complex is obtained when the octanuclear
complex is treated in dichloromethane with CO (eq 3); although
5a
As shown by NMR spectroscopy, the reaction proceeds, in a
first fast step, to the formation of the neutral compound 2, which
is converted in a second, slower step to the cationic complex
5a.
The 19F NMR spectrum of complex 5a shows in the ortho-
fluorine region13 two doublets (1:1); that at higher fields is
assignable (Vide infra) to the free pentafluorobenzenethiolate
anion (δ ) -132.2 ppm). The 31P NMR spectrum consists of
two groups of signals; i.e., in solution and on the NMR time
scale both PPh3 ligands are chemically equivalent, as are both
phosphorus nuclei of the diphosphine ligand, which implies an
average Cs symmetry of the species or C2 symmetry along with
rapid inversion about the sulfur atom. The spectrum belongs
to an AA′XX′ system, but the multiplets are too compressed to
allow all the P-P couplings to be ascertained. Chelate
coordination of the diphosphine can be discarded from the shift
of the signals.14
Since this compound could not be isolated as a pure solid,
the reaction was repeated in the presence of silver salts of
noncoordinating anions in order to exchange the anion
(eq 5). The low solubility of Ag(SC6F5) not only forces the
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 + 8L + 4AgX f
4[LPd(µ-SC F )(µ-dppm)PdL]X
+ 4Ag(SC6F5) (5)
6
5
{[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 + CO(exc) f
5b,c, 6-8
4[(SC6F5)Pd(µ-CO)(µ-dppm)Pd(SC6F5)]
(3)
4
X ) ClO4, L ) PPh3 (5b)
this is the only species formed, it could not be isolated as a
pure solid because it is unstable if not under a carbon monoxide
atmosphere.
X ) SO3CF3, L ) PPh3 (5c), PMePh2 (6), AsPh3 (7),
SbPh3 (8)
The 19F NMR spectrum of the reaction solution shows (down
to -90 °C) only a doublet in the region assigned to ortho-
fluorine nuclei of pentafluorobenzenethiolato groups,13 showing
the thiolato groups to be chemically equivalent. The 31P NMR
spectrum contains only one peak, which requires the phosphorus
nuclei of the diphosphine to be chemically equivalent. The
position of the signal excludes chelate behavior of the dppm
ligand.14 Finally, the infrared spectrum of the freshly obtained
solid shows a strong broad band at 1897 cm-1, assignable to a
CO ligand bridging a PdI-PdI unit.20 Starting decomposition
can be detected in the infrared spectrum, and a 2-week-old
sample shows no bands in the ν(CO) region.
exchange of the counterion but also greatly shortens the
completion time (from 5 days to less than 3 h).
The 19F NMR spectra of complexes 5b,c show only one
doublet in the ortho-fluorine region, corresponding to that at
higher frequencies for compound 5a, whereas the same two
multiplets as for compound 5a are seen in their 31P NMR
spectra. The NMR spectra for compounds 6-8 are as expected
for the above formulation.
Since compound 2 reacts with equimolecular quantities of
Ag(ClO4) and PPh3 to give compound 5b, the analogous reaction
with 1 mol of PPh2Me was attempted in order to obtain a mixed-
phosphine complex. However, redistribution of the neutral
ligands readily occurred and formation of the bis(triphenylphos-
phine), the bis(diphenylmethylphosphine), and the mixed com-
pound was observed in solution by 31P and 19F NMR spectros-
copy and in the mass spectrum of the solid precipitated on
addition of hexane. In a similar way, reaction of the octanuclear
compound {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 with (NBu4)-
Br, in an attempt to synthesize the corresponding anionic
dipalladium(I) complex, led to redistribution of the ligands trans
to the metal-metal bond, and a ternary mixture was observed,
which could not be resolved.
Attempts to grow single-crystals of compound 4, even using
CO-saturated solvents under a carbon monoxide atmosphere,
were unsuccessful.
(15) Gould, R. O.; Taylor, P.; Thorpe, M. C. PUCKER. University of
Edinburgh, U.K., 1995.
(16) Morton, D. A. V.; Orpen, A. G. J. Chem. Soc., Dalton Trans. 1992,
641-653.
(17) Orpen, A. G. Chem. Soc. ReV. 1993, 191-197.
(18) Cambridge Structural Database; Cambridge Crystallographic Data
Centre, University Chemical Laboratory: Cambridge, U.K., 1996. See,
for instance: Holloway, R. G.; Penfold, B. R.; Colton, R.; McCormick,
M. J. J. Chem. Soc., Chem. Commun. 1976, 485-486.
(19) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5525-
5534.
A single-crystal X-ray diffraction study of 5b was attempted.
Although the reflections were collected at 200 K and the crystal
seemed to be of reasonable quality, a good model could not be
(20) Goggin, P. L.; Mink, J. J. Chem. Soc., Dalton Trans. 1974, 534-
540.

dppm-Bridged Dipalladium(I) Complexes
Inorganic Chemistry, Vol. 36, No. 9, 1997 1917
Table 5. Atomic Coordinates (×104) and Equivalent Isotropic
Displacement Parameters (Å2 × 103) for 5ca
x
y
z
U(eq)a
Pd(1A)
Pd(2A)
P(1A)
P(2A)
P(3A)
P(4A)
S(1A)
C(31A)
9540(1)
10309(1)
8816(1)
9630(1)
10407(1)
10879(1)
9876(1)
9861(4)
8699(1)
9417(1)
8468(1)
7778(1)
8731(1)
10379(1)
9713(1)
8178(5)
10000(1)
9710(1)
10387(1)
9446(1)
9028(1)
9667(1)
10389(1)
8892(4)
23(1)
23(1)
26(1)
25(1)
24(1)
29(1)
27(1)
27(2)
Pd(1B)
Pd(2B)
P(1B)
P(2B)
P(3B)
P(4B)
S(1B)
C(31B)
11605(1)
10831(1)
12327(1)
11503(1)
10725(1)
10258(1)
11271(1)
11247(4)
6438(1)
5717(1)
6677(2)
7356(1)
6382(2)
4753(2)
5445(1)
6960(6)
11389(1)
11655(1)
11005(1)
11944(1)
12342(1)
11685(1)
10973(1)
12497(4)
26(1)
29(1)
29(1)
23(1)
29(1)
41(1)
35(1)
29(2)
a U(eq) is defined as one-third of the trace of the orthogonalized Uij
tensor.
Figure 3. Cation A of compound 5c in the crystal (ellipsoids at 50%
electron probability level; other atoms as circles of arbitrary radii; H
atoms omitted).
Table 6. Selected Bond Lengths (Å) and Angles (deg) for 5c
Pd(1A)-P(2A)
Pd(1A)-S(1A)
Pd(1A)-P(1A)
Pd(1A)-Pd(2A)
Pd(2A)-P(3A)
2.255(3)
2.290(3)
2.318(3)
2.6136(11)
2.265(3)
Pd(2A)-S(1A)
Pd(2A)-P(4A)
P(2A)-C(31A)
P(3A)-C(31A)
2.298(3)
2.328(3)
1.811(10)
1.845(11)
Pd(1B)-P(2B)
Pd(1B)-S(1B)
Pd(1B)-P(1B)
Pd(1B)-Pd(2B)
Pd(2B)-P(3B)
2.255(3)
2.296(3)
2.312(3)
2.6090(11)
2.261(3)
Pd(2B)-S(1B)
Pd(2B)-P(4B)
P(2B)-C(31B)
P(3B)-C(31B)
2.309(3)
2.332(3)
1.834(11)
1.826(11)
P(2A)-Pd(1A)-S(1A) 147.19(11) P(3A)-Pd(2A)-S(1A) 150.65(10)
P(2A)-Pd(1A)-P(1A) 106.64(10) P(3A)-Pd(2A)-P(4A) 105.42(11)
S(1A)-Pd(1A)-P(1A) 105.91(10) S(1A)-Pd(2A)-P(4A) 103.42(10)
P(2A)-Pd(1A)-Pd(2A) 92.71(8) Pd(1A)-S(1A)-Pd(2A) 69.45(8)
S(1A)-Pd(1A)-Pd(2A) 55.42(7) P(2A)-C(31A)-P(3A) 108.9(5)
P(1A)-Pd(1A)-Pd(2A) 159.91(7)
P(2B)-Pd(1B)-S(1B) 147.40(11) P(3B)-Pd(2B)-P(4B) 105.07(12)
P(2B)-Pd(1B)-P(1B) 107.09(10) S(1B)-Pd(2B)-P(4B) 103.93(12)
S(1B)-Pd(1B)-P(1B) 104.91(11) Pd(1B)-S(1B)-Pd(2B) 69.02(9)
P(3B)-Pd(2B)-S(1B) 150.53(11) P(3B)-C(31B)-P(2B) 108.9(6)
obtained (best R1 > 9.5%) partly due to pseudosymmetry
problems (a noncrystallographic inversion center relates the
heavy atoms and most of the carbon atoms of the two
independent molecules), but the connectivity was established.
Since the crystal system was found to be orthorhombic (as
confirmed by oscillation photographs) with a ) c, twinning
(racemic as well as pseudotetragonal) was tried with no better
results (the coefficients of the additional components vanished
in the least-squares fitting). The perchlorate counterions are
disordered over multiple sites, which are assumed to be partially
occupied by dichloromethane of solvation as well.
The crystal structure of the homologous compound 5c, with
CF3SO3- as the counterion, was determined at 150 K. Atomic
positional parameters for the significant atoms are listed in Table
5, while selected bond distances and angles are given in Table
6. The compound is isostructural with 5b, crystallizing in the
orthorhombic space group Pca21. The crystal was a racemic
twin, with a distribution of 0.5 for each domain. A possible
tetragonal twinning was included in the model and discarded,
since the population of the additional twinned domains con-
verged to values of 0, within experimental uncertainty.
Figure 4. Cation B of compound 5c in the crystal (ellipsoids at 50%
electron probability level; other atoms as circles of arbitrary radii; H
atoms omitted).
to be more complex. Attempts to refine more accurate models
did not converge. Given the high number of parameters already
used in the model, we did not attempt to further account for
the C(38B)-C(43B) phenyl ring, which is clearly disordered.
Similarity restraints on bond distances, angles, and displacement
parameters were applied among both triflate anions and all
solvent molecules.
There are two chemically equivalent but symmetry-indepen-
dent cations per asymmetric unit (see Figures 3 and 4 for
structure and numbering scheme), related by a noncrystallo-
graphic inversion center, which lies displaced 2.5 Å from the
21 axis, where it should be located in the corresponding
centrosymmetric group Pbcm.
The cations comprise the Pd(µ-SC6F5)(µ-dppm)Pd unit
present in the octanuclear precursor 1 and the neutral compound
2, with the two triphenylphosphine ligands trans to the metal-
metal bond (Figures 3 and 4); the palladium-palladium bond
lengths (2.614(1), 2.609(1) Å) are equivalent to the one
determined in 2 and slightly longer than that of the starting
octanuclear complex 1. The terminal ligands are not collinear18
with the palladium-palladium bond, the P-Pd-Pd angles
The asymmetric unit contains four molecules of dichlorome-
tane, three of them disordered. These three were modeled in
two alternative sites, with 50% occupancy for each site, but
the large displacement parameters along with the high residual
electron density in these areas show the situation in the crystal

1918 Inorganic Chemistry, Vol. 36, No. 9, 1997
Uso´n et al.
Figure 6. Optimized structures of the [Pd2] and [Pd2S] models.
trifluoromethanethiolate anion. Preliminary calculations on this
structure showed that the computational effort was still too large
and that further simplifications would be in order. Single-point
calculations, carried out at idealized experimental geometries,
-
showed that replacement of the SCF3 bridge by HS- did not
significantly affect the bond properties of the PdI2S cycle;
therefore, the [(H3P)Pd(µ-H2PCH2PH2)(µ-SH)Pd(PH3)]+ cluster
was finally adopted as the model. The system without the sulfur
bridge was then simply represented by [(H3P)Pd(µ-H2PCH2-
PH2)Pd(PH3)]2+. These models are depicted in Figure 6 and
will be designated hereafter as [Pd2S] and [Pd2], respectively.
(b) Computational Details. Ab initio Hartree-Fock calcula-
tions were undertaken, using the effective core potentials (ECP)
reported by Stevens et al.21 to describe inner electrons of heavy
atoms. For the PdI ion, the 4s2, 4p6, and 4d9 electrons were
considered, while for P, C, and S, only the valence electrons
were explicitly taken into account (70 and 78 electrons for [Pd2]
and [Pd2S], respectively). The basis set was of sp type with
the following contractions: (8s8p5d)/[4s4p3d] for Pd, (4s4p)/
[2s2p] for P, C, and S, and (4s)/[2s] for H. For P and S, a set
of d polarization functions was added. Molecular geometries
were optimized using standard analytical gradient techniques.
All the calculations were performed using the HONDO-8
program22 running on an HP-735 work station.
(c) Molecular Structure. The structural data for compound
5c (see above) show that the Pd(µ-SC6F5)(µ-dppm)Pd moiety
is puckered, with a larger twist than envelope component.
However, the puckering parameters are quite different for cations
A and B (compare Figure 5) showing this to be a packing effect.
Besides, the chemical equivalence observed in the 31P NMR
spectrum of compound 5c for the phosphorus nuclei of the
triphenylphosphine ligands, as well as for those of the diphos-
phine, favors a (mean) envelope conformation. Orpen16,17 has
found this to be the most preferred geometry in the solid state
for five-membered M2P2C rings. According to these data, the
geometrical structure of the [Pd2S] and [Pd2] models was
optimized under Cs constraint; selected geometrical parameters
are reported in Table 7, which shows a quite satisfactory
agreement with experiment. The interatomic bond distances
are systematically found somewhat shorter than the experimental
values, although this result seems reasonable considering that
the models are much less hindered that the actual complexes.
Concerning bond angles, the agreement between the computed
values for [Pd2S] and the mean experimental values is excellent.
When the geometrical parameters of the two models are
compared, the most outstanding feature is that the structures of
Figure 5. Fit of the two crystallographically independent cations in
compound 5c (mean deviation for SPd2P2P2C ) 0.027 Å).
ranging from 158.0 to 159.9°; the P(1)-Pd(1)-Pd(2)-P(4)
torsion angles are 9.5 and 4.6°, respectively, for molecules A
and B).
The five-membered Pd2P2C rings are puckered:15 83% twist
with the axis through Pd(2A) and P(3A) (66% with the axis
through Pd(2B) and P(3B)) and 17% envelope with the flap at
P(2A) (34% with the flap at P(2B)); these quite different
conformations demonstrate that no authentic inversion center
relates both cations in the crystal (as illustrated in Figure 5).
These results are again in contrast to the statistically preferred
conformation found by Orpen16,17 in similar systems; accord-
ingly, the P(2)-Pd(1)-Pd(2)-P(3) torsion angles (11.1 for
species A and -12.1° for species B) are far from zero. For the
SPd2P2C core, C(31A) lies 0.48 Å under (0.39 Å for C(31B))
and P(2A) 0.39 Å above (0.44 Å for P(2B)) the best plane,
described by the sulfur, the remaining phosphorus, and both
palladium atoms (mean deviation 0.02 Å, in both cases).
The aromatic systems in the thiolato ligands are roughly
parallel with one phenyl ring of each PPh3 ligand, describing
angles between 6.2 and 13.0°, and the distances between planes
are around 3.6 Å. The perfluorophenyl rings are tilted about
the S-Cipso bonds, forming dihedral angles of 87.4 (molecule
A) and 88.5° (molecule B) with the respective mean SPd2P2C
planes; the nonbonding distances to the nearest ortho-fluorine
atoms are Pd(1A)-F(63A) ) 3.078 and Pd(1B)-F(67B) )
3.109 Å.
Given the stability of the Pd(µ-SC6F5)(µ-dppm)Pd unit, which
is maintained in the above described reactions, theoretical
calculations were undertaken on this atom assembly. In order
to compare the predicted structure with experimentally found
results, compound 5c was chosen; modeling was, however,
necessary to diminish the number of variables and make the
computations affordable.
Theoretical Calculations
(a) The Models. Because of the high number of atoms and
electrons in the [(Ph3P)Pd(µ-SC6F5)(µ-dppm)Pd(PPh3)] unit,
some simplification of the model was necessary. In a first stage,
this unit was modeled as a cluster of formula [(H3P)Pd(µ-H2-
PCH2PH2)(µ-SCF3)Pd(PH3)]+, in which the dppm bridge was
replaced by diphosphinomethane, the tryphenylphosphine by
phosphine, and the pentafluorobenzenethiolato ligand by a
(21) Stevens, W. J.; Basch, H.; Krauss, M. J. Chem. Phys. 1984, 81, 6026-
6033. Stevens, W. J.; Krauss, M.; Basch, H.; Jasien, P. G. Can. J.
Chem. 1992, 70, 612-630.
(22) Dupuis, M.; Chin, S.; Ma´rquez, A. CHEM-Station and HONDO:
Modern Tools for Electronic Structure Studies Including Electron
Correlation. In RelatiVistic and Electron Correlation Effects in
Molecules and Clusters; Malli, G. L., Ed.; NATO ASI Series; Plenum
Press: New York, 1994.

dppm-Bridged Dipalladium(I) Complexes
Inorganic Chemistry, Vol. 36, No. 9, 1997 1919
Table 7. Selected Geometrical Parameters for [Pd2] and [Pd2S]a
Pd-Pd
Pd-P
Pd-Pexo
P-C
Pd-S
Pd-Pd-P
Pd-Pd-Pexo
C-P-Pd
P-C-P
Pd-S-Pd
τC
τS
[Pd2S]
[Pd2]
expb
2.504
2.500
2.61
2.157
2.158
2.26
2.177
2.235
2.32
1.850
1.842
1.83
2.222
96.3
96.5
94
161.4
167.1
159
112.7
111.8
111
107.0
108.5
109
68.6
23
25
25
174
2.30
69
173
a Bond distances are in Å, and bond angles in deg. a Experimental mean values for units A and B.
Table 8. Mulliken Population Analysis over Shells
s
p
d
total
[Pd2S]
1.606
2.897
2.960
3.224
3.853
Pda
P
Pexo
C
1.145
1.277
1.298
1.084
1.617
8.134
0.314
0.288
10.886
4.488
4.547
4.309
5.642
S
0.171
[Pd2]
1.165
2.851
2.988
3.228
Pda
P
Pexo
C
1.184
1.292
1.313
1.098
8.127
0.316
0.257
10.477
4.459
4.557
4.326
a For palladium s and p shells, the values reported are those obtained
after subtraction of the 4s2 and 4p6 electrons.
the five-membered rings are practically the same. This agrees
with the well-known stability of the (µ-dppm)PdI2 ring. In fact,
significant variations are observed only for bond distances and
bond angles involving the exo PH3 ligand.
(d) The PdI-PdI and PdI-S Bonds. The electronic
properties of the PdI-PdI bond have been examined in a small
number of theoretical papers, mostly performed using extended
Hu¨ckel molecular orbital theory (EHMO).4-7 The main conclu-
sion that arises from these calculations is that the major
contribution to the PdI-PdI σ-bond is from the interaction
between palladium d atomic orbitals, with small participation
of s and p orbitals (less than 5%). These results suggest that
the palladium(I) ion should preserve its d9 configuration without
mixing with (sp)1d8 type configurations; however, as is well-
known, 5s and 5p atomic orbital energies are close to the 4d
level and some incorporation of s and p character into the bond
would clearly improve the overlap. A second aspect to be
considered is that every palladium(I) center is bound to two
phosphines (with high σ-donor capabilities23) and, therefore, the
5s and 5p PdI orbitals have to be partially occupied, while back-
donation from PdI to phosphorus d orbitals would decrease the
d population on the palladium(I) centers.
These expectations are fully confirmed by our ab initio
Hartree-Fock calculations. As can be observed in Table 8,
Mulliken population analysis on the [Pd2] model gives 8.13
electrons on palladium 4d orbitals and 1.18 and 1.16 on the 5s
and 5p orbitals, respectively. Although the use of Mulliken
absolute values has been questioned,24 they clearly agree better
with an (sp)1d8 configuration than with a d9 configuration.
The analysis of the molecular orbitals (MO’s) obtained for
[Pd2] gives a description of the metal-metal bond clearly
different from that previously proposed. The MO’s involved
in the PdI-PdI bond are the highest nine occupied orbitals.
Among these, eight correspond to bonding-antibonding ar-
rangements of d orbitals, while one of them is of bonding sp
nature and is easily identified as a σsp MO. Taking the
molecular plane as xy and defining the symmetry plane as xz,
Figure 7. Contour plots for the Pd-Pd σsp molecular orbitals of [Pd2]
(top) and [Pd2S] (bottom) structures.
these nine highest MO’s are πxy, πyz, δxz, δxz*, πxy*, σsp, πzz,
πzz*, and πyz* (ordered by increasing energy). Contour plots
for the σsp MO are reported in Figure 7. As can be seen, this
MO lies entirely along the P-Pd-Pd-P line and the main
contributions are the py orbitals of palladium, while the
2
2
participation of the Pd dx -y orbitals is minor, since they are
more involved in the bond with phosphines.
It is worthy of note, on the other hand, that EHMO
calculations carried out on both [Pd2] and [Pd2S] give a
(23) Puddephatt, R. J.; Dignard-Bailey, L.; Bancroft, G. M. Inorg. Chim.
Acta 1985, 96, L91-L92.
(24) As is well-known, Mulliken population analyses are basis set depend-
ent, and therefore, care should be exercised in their use. See, for
instance: Ammeter, J. H.; Bu¨rgi, H. B.; Thibeault, J. C.; Hoffmann,
R. J. Am. Chem. Soc. 1978, 100, 3686-3692. Whangbo, M.-H.;
Hoffmann, R. J. Chem. Phys. 1978, 68, 5498-5500.
2
2
description of the palladium-palladium bond in which the dx -y
metal orbitals are preponderant, as had been found by others.4-6
However, as shown in the present work, higher level calculations
lead to a qualitatively different interpretation.
The analysis of the Pd-S bond appears to be quite involved

1920 Inorganic Chemistry, Vol. 36, No. 9, 1997
Uso´n et al.
Figure 8. Contour plots for the molecular orbitals arising from the interaction between [Pd2] and HS- fragments: (top) πxy - spx; (bottom) dx -y
- spx.
2
2
since the interaction between the palladium and sulfur orbitals
occurs through several MO’s and also because the palladium d
orbitals are considerably mixed among them. In order to
highlight what we think are the main interactions, let us first
consider the MO’s of an isolated HS- ligand. The electronic
configuration of HS- is (C∞V) 1σ21σ21π4. Under Cs constraint,
the σ MO’s transform as a′ and the π MO’s split into a′ and a′′:
and the spx lone pair of HS-; when the mix of these orbitals is
in phase (πxy-spx combination), the typical three-center MO is
obtained:
However, for the out-of-phase combination (πxy + spx), there
is still some bonding overlap between Pd and S, although in
this case such an overlap is of π type and involves the other
lobes of palladium dxy orbitals. These interactions can be clearly
observed in the contour plots of Figure 8, where bonding overlap
is observed in both MO’s. A similar feature is found for the
-
2
2
In terms of localized orbitals, the σSH bond and the spz lone
pair are not involved in bonding with palladium centers, and
only the spx (a′) and the py (a′′) lone pairs have the suitable
directionality. Starting with the a′ symmetry orbitals, the main
interaction arises from the overlap between the πxy MO of [Pd2]
interaction of HS with the palladium dx -y orbitals. Depending
on the relative sign of the mix, the three-center interaction can
be viewed as of σ or π type (see Figure 8).
Combinations of the sulfur py orbital with a′′ orbitals of [Pd2],
giving rise to a bonding interaction, are depicted in Figure 9.

dppm-Bridged Dipalladium(I) Complexes
Inorganic Chemistry, Vol. 36, No. 9, 1997 1921
HS- ligand. This MO is just slightly destabilized upon
coordination with sulfur and appears almost “pure” as shown
in Figure 7, where the Pd-Pd σsp MO’s of [Pd2] and [Pd2S]
models are plotted together for the sake of comparison. On
the other hand, a careful inspection of the contour plots of the
MO’s here reported shows that their maxima do not lie on the
Pd-S axes. In other words, the charge centroids of these MO’s
fall out of the ideal three membered ring. This result shows
that a nonlinear bond between the palladium and the sulfur
centers is established, just like the classical “banana” bonds
proposed for the cyclopropane-based rings.
Finally, the charge transfer from the HS- ligand to the [Pd2]
complex will be addressed. As shown in Table 8, after
complexation the total Mulliken populations on Pd centers rise
by about 0.4 electron. The largest effect is observed for
palladium p orbitals, which increase their population by 0.43
electron each. This trend confirms the preeminent role played
by the p metal orbitals in the coordination with the HS- anion.
An additional point concerns the absolute values of the Mulliken
populations obtained for the palladium atoms. As can be seen,
the total populations are 10.48 and 10.87; i.e., they are two units
larger than those expected for a PdI cation. Though these values
are likely to be overestimated, since in the actual compounds
the ligands are phenylphosphines instead of plain PH3 groups,
they reflect the well-known high donor capabilities of phos-
phines.23
(e) Pd-Pd and Pd-S Vibrational Frequencies. Although
a full vibrational analysis for both the [Pd2] and [Pd2S] models
has been carried out, only the frequencies associated with Pd-
Pd and Pd-S stretchings will be considered. The vibrational
frequencies ωPd-Pd are computed to be 244 and 253 cm-1 for
[Pd2] and [Pd2S] (harmonic), respectively. These values are in
reasonable agreement with an experimental peak observed at
213 cm-1 in the Raman spectrum of compound 5c; they are at
higher energy than those found25 for Pd2(dppm)2X2 compounds,
whose palladium-palladium bonds are however longer. Al-
though the difference between the theoretical values found for
[Pd2] and [Pd2S] is almost negligible, the increment found after
coordination with sulfur seems to disagree with the fact that
the calculated bond order diminishes from 1.14 for [Pd2] to 0.95
for [Pd2S] and also with the slight lengthening in the Pd-Pd
interatomic bond distance. This apparent discrepancy may be
understood by considering that the normal coordinate associated
with the Pd-Pd stretching incorporates some Pd2-S stretching
internal coordinate; i.e., the normal coordinate involves dis-
placements of the entire three-membered ring. This mix gives
rise to two a′ normal modes: one in which the atoms move in
phase (mainly rPd-S) and the other with displacements out of
phase (mainly rPd-Pd).
Figure 9. Contour plots for the molecular orbitals arising from the
interaction between [Pd2] and the py orbital of HS-: (top) interaction
2
2
with palladium spx orbitals; (bottom) interaction with dx -y orbitals.
The main contribution arises from the overlap between the HS-
py orbital and an antisymmetric combination of spx palladium
orbitals (top part of Figure 9). The analysis of this MO may
be confusing, since a nodal plane between the palladium and
the sulfur centers is observed. However, this node corresponds
to the different signs of the inner and outer basis functions
representing the 5px atomic orbital of palladium. The highest
coefficient is that of the outer function, which strongly mixes
with the sulfur py orbital. The second largest interaction of a′′
symmetry type arises from the overlap with an antisymmetric
As a consequence, the ωPd-Pd frequency rises while ωPd-S
decreases. The value computed for the latter is 530 cm-1, in
excellent agreement with a peak at 528 cm-1 observed in the
Raman spectrum.
2
2
combination of the dx -y metal orbitals (bottom part of Figure
9).
Conclusions
The reaction of {[Pd(µ-SC6F5)(µ-dppm)Pd](µ-SC6F5)}4 with
neutral group 15 ligands L affords neutral [LPd(µ-SC6F5)(µ-
Furthermore, as we have seen in the preceding discussion,
the Pd-Pd σsp MO is not involved in the interaction with the
(25) Alves, O. L.; Vitorge, M.-C.; Sourisseau, C. New J. Chem. 1983, 7,
231-238.

1922 Inorganic Chemistry, Vol. 36, No. 9, 1997
Uso´n et al.
dppm)Pd(SC6F5)] or cationic [LPd(µ-SC6F5)(µ-dppm)PdL]+
complexes. These binuclear complexes retain the palladium-
palladium bond and the two dissimilar bridging ligands, as
demonstrated by the X-ray structural determinations of [(Ph3P)-
Pd(µ-SC6F5)(µ-dppm)Pd(SC6F5)]‚1.4CH2Cl2 and [(Ph3P)Pd(µ-
SC6F5)(µ-dppm)Pd(PPh3)]SO3CF3‚2CH2Cl2. The surprising
stability of the mixed bridge permits the incorporation of two
different ligands, trans to the palladium-palladium bond, which
is unprecedented in palladium(I) chemistry.
arises from palladium 5s and 5p orbitals and sulfur lone pairs,
giving rise to a typical three-atom “banana” bond.
Acknowledgment. This work was supported by the Direc-
cio´n General de Investigacio´n Cient´ıfica y Te´cnica (Proyectos
PB95-1247 and PB94-0597). I.U. is grateful to the EU for an
HCM postdoctoral fellowship (ERB CHBG 93 0338), and S.H.
thanks the Ministerio de Educacio´n y Ciencia for the award of
an FPI scholarship.
Contrary to the currently accepted d-d formulation of the
palladium-palladium bond, based on EHMO calculations, ab
initio calculations on the model systems [(H3P)Pd(µ-H2PCH2-
PH2)(µ-SH)Pd(PH3)]+ and [(H3P)Pd(µ-H2PCH2PH2)Pd(PH3)]2+
show that this bond arises mainly from the interaction between
the 5s and 5p orbitals. Coordination of the [Pd2] unit with sulfur
involves three-center MO’s in which the largest participation
Supporting Information Available: Complete listings of crystal-
lographic details, atomic coordinates, bond lengths and bond angles,
thermal displacement parameters, and calculated H positional parameters
for 2 and 5c (32 pages). Ordering information is given on any current
masthead page.
IC960818E