Synthesis, X-ray and solid state NMR studies on silver(I) silanethiolates. Crystal structure of bis[(triphenylphosphine)-(tri-o-tolylsilanethiolato)silver(I)]-toluene(1/2), [{Ag[SSi(C6H4-Me-o)3](PPh3)} 2]·2C6H5Me

Article abstract of DOI:10.1039/a903694f

A series of silver(I) trialkoxysilanethiolates [Ag(SSiR₃)(PPh₃)₂], R = Bu^tO, Bu^sO or PrⁱO, and triarylsilanethiolates [Ag(SSiR′₃)(PPh₃)], R′ = o-MeC₆H₄ or Ph, were obtained by reaction of [Ag(acac)(PPh₃)₂] with the respective silanethiols. Products were studied by ¹H NMR, CP MAS ³¹P and ²⁹Si NMR in solution and/or in the solid state. A single-crystal X-ray structural analysis was carried on for [{Ag[SSi(C₆H₄Me-o)₃](PPh₃)} ₂]·2C₆H₅Me.

Full text of DOI:10.1039/a903694f

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis, X-ray and solid state NMR studies on silver(I)
silanethiolates. Crystal structure of bis[(triphenylphosphine)-
(tri-o-tolylsilanethiolato)silver(I)]–toluene(1/2), [{Ag[SSi(C6H4-
Me-o)3](PPh3)}2]ؒ2C6H5Me†
Jaroslaw Chojnacki,a Barbara Becker,*a Atoni Konitz,a Marek J. Potrzebowskib and
Wieslaw Wojnowskia
a Technical University of Gdansk, Department of Chemistry, G. Narutowicz Str. 11/12,
80-952 Gdansk, Poland
b Center of Molecular & Macromolecular Studies, Polish Academy of Sciences,
Sienkiewicz Str. 112, 90-363 Lódz, Poland
Received 10th May 1999, Accepted 13th July 1999
A series of silver() trialkoxysilanethiolates [Ag(SSiR3)(PPh3)2], R = ButO, BusO or PriO, and triarylsilanethiolates
[Ag(SSiRЈ3)(PPh3)], RЈ = o-MeC6H4 or Ph, were obtained by reaction of [Ag(acac)(PPh3)2] with the respective
silanethiols. Products were studied by 1H NMR, CP MAS 31P and 29Si NMR in solution and/or in the solid state.
A single-crystal X-ray structural analysis was carried on for [{Ag[SSi(C6H4Me-o)3](PPh3)}2]ؒ2C6H5Me.
SiSH, and triarylsilanethiols, Ph3SiSH and (o-MeC6H4)3SiSH,
have been used. Triphenylphosphine was bound as the second
ligand.
Introduction
The thiolato group RSϪ is widely recognized as a basic type of
ligand, and a huge number of copper group metal thiolates has
been obtained up to now. However, owing to the strong ten-
dency of sulfur to bridge metal centers and consequently form
insoluble polymers, their detailed characterization has proved
very difficult in many cases. The most common approaches to
overcome the irregular polymerization problem have been
increasing the steric demand of the thiol substituent R and/or
incorporation of heteroligands on the metal atom. To increase
the bulk of silicon containing thiols, for example triorganosilyl
substituted thiophenols2 or (triorganosilyl)methanethiols
Results and discussion
For the preparation of phosphine ligated silver silanethiolates
three methods may be of interest. (1) Direct reaction between
silanethiol, silver salt and phosphine. The reaction between
silver nitrate, phosphine and tri-tert-butoxysilanethiol under
heterogeneous conditions (water–CHCl3) in the presence of
sodium acetate gave the pure compound [Ag{SSi(OBut)3}-
(PPh3)2] 1 in good yield. This simple procedure could not how-
ever be applied for the other four silanethiols. Even the more
stable of them, (o-MeC6H4)3SiSH,8 was easily hydrolysed and
Ag2S precipitated from the reaction mixture. (2) Reaction of
homoleptic silver silanethiolate with phosphine. The procedure
is of limited importance because of its requirement for homo-
leptic silver silanethiolates, which besides [{AgSSi(OBut)3}4] are
still unknown. (3) Reaction of a silanethiolate ligand source
with an appropriate silver phosphine complex. The reaction of
tri-tert-butoxysilanethiol (in the presence of Et3N) with differ-
ent triphenylphosphine complexes of silver, [Ag(PPh3)4]NO3,
[Ag2Br2(PPh3)4] or [{AgI(PPh3)}4], gave invariably 1 which is
obviously the preferred product.
The only method which could be used for all silanethiols was
that utilizing silver phosphine complexes as substrates. The
complexes mentioned above required the presence of a base for
neutralization of the strong acid produced. However, too basic
conditions promote hydrolysis of silanethiols.9 Since silane-
thiols are more acidic than organic thiols10,11 a method with
substitution of a stronger acid for a weaker one was envisaged.
The preparation of phosphine ligated silver() silanethiolates
was best accomplished by the use of bis(triphenylphosphine)-
silver() acetylacetonate, eqn. (1). Besides 1 the following new
(R3Si)nCH3-nSH,3 were used as ligand sources. These however
are still based on a common ᎐C–SH theme.
᎐
᎐
Our continuing interest in silicon–sulfur compounds
prompted us to introduce a novel type of ligand where the
thiolate function rather than on carbon is bonded directly to
silicon. The first metal silanethiolate we obtained and character-
ized structurally was the silver() complex [{AgSSi(OBut)3}4].4
This was followed by several other neutral tri-tert-butoxysilane-
thiolates, which were invariably molecular. Among them were
also copper() and gold() derivatives, [{MSSi(OBut)3}4] (where
M = Cu or Au).5,6 Later we found that tetrameric copper() tri-
tert-butoxysilanethiolate reacts with triphenylphosphine giving
a neutral, heteroleptic copper() complex, [Cu{SSi(OBut)3}-
(PPh3)2].7 So far our investigations have been limited to tri-tert-
butoxysilanethiol (ButO)3SiSH, as the source of silanethiolate
ligand. It has one great advantage over other silanethiols.
Owing to the great steric hindrance caused by three bulky tert-
butoxy substituents bonded to silicon, under normal conditions
it does not undergo hydrolysis in which silanethiols more or less
easily decompose to the corresponding silanol and hydrogen
sulfide.8 Now we present some results of our investigations on
the synthesis and properties of heteroleptic silver complexes
where also other trialkoxysilanethiols, (BusO)3SiSH, (PriO)3-
[Ag(acac)(PPh3)2] ϩ R3SiSH →
R3SiSAg(PPh3)n ϩ Hacac ϩ (2 Ϫ n)PPh3 (1)
† Contributions to the chemistry of silicon–sulfur compounds. Part
71.1
Supplementary data available: preparations. For direct electronic
access see http://www.rsc.org/suppdata/dt/1999/3063/, otherwise avail-
able from BLDSC (No. SUP 57613, 2 pp.) or the RSC Library. See
Instructions for Authors, 1999, Issue 1 (http://www.rsc.org/dalton).
complexes were obtained: [Ag{SSi(OPri)3}(PPh3)2] 2, [Ag-
{SSi(OBus)3}(PPh3)2] 3, [Ag{SSi(C6H4Me-o)3}(PPh3)] 4 and
[Ag(SSiPh3)(PPh3)] 5. Contrary to trialkoxysilanethiols, triaryl-
silanethiols gave complexes with a silver:phosphine ratio of
J. Chem. Soc., Dalton Trans., 1999, 3063–3068
3063

View Article Online
Table 1 Selected bond lengths (Å) and angles (Њ) for [{Ag[SSi(C6H4Me-o)3](PPh3)}2]ؒ2C6H5Me 4
Ag(1)–S(2)
Ag(1)–S(21)
Ag(1)–P(3)
S(2)–Si(4)
2.4680(9)
2.6376(8)
2.3845(12)
2.1060(12)
1.815(2)
P(3)–C(32)
P(3)–C(38)
Si(4)–C(5)
Si(4)–C(12)
Si(4)–C(19)
1.822(2)
1.824(2)
1.899(2)
1.896(2)
1.883(2)
P(3)–C(26)
S(21)–Ag(1)–S(2)
Ag(11)–S(2)–Ag(1)
P(3)–Ag(1)–S(2)
P(3)–Ag(1)–S(21)
Si(4)–S(2)–Ag(11)
Si(4)–S(2)–Ag(1)
91.98(4)
88.02(4)
145.90(3)
122.12(3)
121.98(4)
105.34(5)
C(26)–P(3)–Ag(1)
C(32)–P(3)–Ag(1)
C(38)–P(3)–Ag(1)
C(5)–Si(4)–S(2)
C(12)–Si(4)–S(2)
C(19)–Si(4)–S(2)
113.06(9)
114.65(8)
115.11(8)
108.63(8)
110.11(8)
111.84(8)
Torsion angles
S(21)–Ag(1)–S(2)–Ag(11)
P(3)–Ag(1)–S(2)–Ag(11)
P(3)–Ag(1)–S(2)–Si(4)
S(2)–Ag(1)–P(3)–C(26)
S(2)–Ag(1)–P(3)–C(32)
S(2)–Ag(1)–P(3)–C(38)
0.0a
Ag(1)–P(3)–C(26)–C(31)
Ag(1)–P(3)–C(32)–C(33)
Ag(1)–P(3)–C(38)–C(39)
S(2)–Si(4)–C(5)–C(10)
S(2)–Si(4)–C(12)–C(13)
S(2)–Si(4)–C(19)–C(20)
Ϫ42.8(3)
Ϫ51.3(2)
Ϫ37.7(2)
Ϫ4.3(2)
Ϫ61.5(2)
Ϫ52.2(2)
Ϫ179.01(3)
Ϫ56.32(5)
149.79(9)
Ϫ89.85(9)
30.22(9)
Atoms related by the center of symmetry are indicated by superscript 1. a Flat Ag2S2 ring.
only 1:1, i.e. one phosphine molecule was eliminated from
silver during the reaction. Compounds 1–5 form colorless
crystals characterized by sharp melting points.
Compound 1 is stable and can be stored without special pre-
cautions. It is readily soluble in CHCl3, tetrahydrofuran and
benzene, sparingly soluble in heptane, acetone, acetonitrile and
practically insoluble in methanol or butan-1-ol. Solutions
in more polar solvents (THF or CHCl3) are not stable and
multiple crystallization leads to separation of PPh3 and the
known [{Ag[SSi(ButO)3]}4].4 The mass spectrum of 1 (FAB;
m-nitrobenzyl alcohol matrix, NBA): positive ions, m/z 1270
ϩ
(Ag2(PPh3)4ϩ, 65), 1019 (Ag2SR(PPh3)2ϩ, 10), 633 (Ag(PPh3)2
,
100) and 370 (Ag(PPh3)2ϩ, 92%); negative ions, m/z 279 (RSϪ,
45), 539 (Ag(NBA)SRϪ, 30), 667 (Ag(SR)2Ϫ, 30) also 1053
(Ag2(SR)3Ϫ, <2) and 1301 (Ag2(PPh3)2(SR)2Ϫ, <2%); where
R = SSi(OBut)3. The cited intensities are given for the most
intense peaks because of the wide isotropic profile of each
signal. All the profiles are consistent with the proposed
formulas.
The two other trialkoxysilanethiolates 2 and 3 are much less
stable and should be stored under a dry neutral atmosphere.
Under normal laboratory conditions the crystals underwent
visible decomposition within a few days, even in a refrigerator,
with formation of Ag2S. Their solubilities are similar to that of
1 but in solutions, especially in more polar solvents, the com-
pounds decompose quite easily. The tri-o-tolylsilanethiolate 4
can be dissolved in aromatic hydrocarbons and, as shown by a
subsequent structure determination, the crystals obtained from
toluene contained the solvent. The triphenylsilanethiolate 5
is completely insoluble in all common organic solvents and
as such may be a polymer. Its elemental analysis suggests a
benzene solvate. This was the reason why we turned to the more
bulky tri-o-tolylsilanethiol as substrate.
Fig. 1 Molecular structure of [{Ag[SSi(C6H4Me-o)3](PPh3)}2]ؒ2C6-
H5Me 4. General shape of the dimer part and atom numbering scheme
(toluene atom numbers and all hydrogen atoms omitted, thermal
ellipsoids 20%).
[Ag{SSi(OBut)3}(PPh3)2] 1. Details of the measurement,
complete set of positional parameters and view of the molecule
have been reported.12 There are no silver complexes that
could be directly compared with 1. From the geometric point
of view the closest related structure is represented by [Ag-
(PPh3)2(Hpymt)]NO3 (Hpymt = pyrimidine-2-thione) reported
recently.13 However, this complex is ionic with Ag co-ordinated
by thione sulfur and thus having a much longer Ag–S bond
2.573(1) as compared to 2.4495(11) Å found in 1. It is worthy to
note that the Si–S bond shortens on going from homoleptic
[{AgSSi(OBut)3}4], 2.119(8) Å where sulfur is three-bonded
to heteroleptic silanethiolate, to 1 with sulfur two-bonded,
2.0738(14) Å. The same type of shortening was observed
previously for the pair [{CuSSi(OBut)3}4]5 and [Cu{SSi-
(OBut)3}(PPh3)2].7
Molecular structures
The different types of stoichiometric formulas found for the
trialkoxy- and triaryl-silanethiolate complexes of silver called
for elucidation and attempts have been made to determine their
structures. Good quality single crystals were obtained for 1, 2
and 4, but 2 survived only two days under X-ray radiation at
room temperature. Despite numerous efforts the crystals of 5,
although well shaped, grew too small to be considered for X-ray
measurements. The compound is hard to characterize because
of its very low solubility in all popular solvents. No solution
spectral data are available. Elemental analysis and solid state
NMR data suggest its structure is similar to that of its o-tolyl
homologue 4.
[{Ag[SSi(C6H4Me-o)3](PPh3)}2] 4. Bis[(triphenylphosphine)-
(tri-o-tolylsilanethiolato)silver()–toluene (1/2) crystallizes in
the triclinic system as colorless prisms. Selected bond lengths
and angles are given in Table 1 and the molecular structure and
atom numbering scheme are illustrated in Fig. 1. Alternating
3064
J. Chem. Soc., Dalton Trans., 1999, 3063–3068

View Article Online
Table 2 Coupling constants for compounds 1, 2, 4 and 5 and
phosphorus–silver bond lengths
Compound
1JP-Ag/Hz
2JP-P/Hz
dP-Ag/Å
4
1
549.3
268.6
390.6
354.0
538.0
—
146.5
—
109.9
—
2.3844
2.4939
2.4475
—
2
5
—
(µ-O3SCF3)2(PPh3)4]17 also contains an Ag2S2 core with two
more (exocyclic) silver atoms held together not only by sulfur
bridges but also by triflate (O3SCF3) groups. In fact 4 seems to
be the only known neutral bimetallic silver thiolate complex
with only three-co-ordinating metal centers.
Solid state NMR studies
High-resolution 31P solid state NMR spectra of silver()
silanethiolates 1, 2, 4 and 5 are shown in Fig 2. The 31P CP
MAS spectra present different patterns owing to the changes in
indirect spin–spin coupling J. The sample 4 (Fig. 2a), in the
isotropic part of spectrum, exhibits a simple AX doublet due to
strong phosphorus–silver splitting (1JP-Ag = 549 Hz) (Table 2).
At natural abundance silver has two I = 1/2 isotopes, 107Ag
(51.82%) and 109Ag (48.18%), and their magnetogyric ratios are
similar in magnitude [γ(107Ag)/γ(109Ag) = 0.87]. Thus, the result-
ing spectrum should show two doublets in which the separate
couplings of 31P to the 107Ag and 109Ag nuclei are resolved. Such
coupling pattern is often observed for silver() complexes of
phosphorus donor ligands in solution. The difference between
1J
and 1J
is of the order of 40 Hz. For solid
31P-107Ag
31P-109Ag
Fig. 2 The 121.49 MHz 31P CP MAS NMR spectra of (a) compound
4, νrot = 2.0 kHz, (b) 1, νrot = 2.33 kHz, (c) 2, νrot = 2.0 kHz and (d) 5,
νrot = 2.0 kHz. The spectra have 4K data points, a contact time of 1 ms
and 100 scans.
state samples the spectral resolution is usually below 40 Hz and
the separate splitting due to the 107Ag and 109Ag nuclei is not
observed. Only in a few cases such a distinction was reported in
the solid phase.18
The 31P CP MAS spectrum of sample 1 (Fig. 2b) shows a
more complex pattern compared to that of 4. Eight well
resolved lines in the isotropic part correspond to an AMX
spin system in which two crystallographically and magnetically
non-equivalent phosphorus ligands are geminally split and
further split by silver. The origin of the splitting pattern is indi-
cated by the stick diagram above the spectrum. The 2J
silver and sulfur atoms in the central Ag2S2 core of the molecule
form a structure close to a rectangle. The two Ag–S bond
lengths are different and the two inside angles are close to the
right angle: 92.0 and 88.0Њ. The molecule, consisting of two
monomer units, possesses a center of symmetry in the middle
of the Ag2S2 ring. Both Ag–P bonds are in the ring plane,
whereas the silanethiolate groups are above and below the
plane. Well ordered toluene molecules are trapped in the struc-
ture at niches in the vicinity of the silanethiolate groups. One
dimeric silanethiolate molecule and two solvent molecules are
present in the unit cell.
The Ag–S bond lengths in compound 4 (2.468 and 2.638 Å)
are comparable with that (2.449 Å) found in 1. They are longer
than values usually observed for two-co-ordinated silver (2.38–
2.40 Å), e.g. [{AgSSi(OBut)3}4],5 [(AgSBut)14(PPh3)4],14 [{AgSC-
(SiMe3)3}4] and [{AgSCH(SiMe3)2}8],3 [(AgSC6H11)12]3
or [{AgSC6H4SiMe3-o}4]2 and are close to the values of
ca. 2.50 Å observed in complexes where Ag is three-co-
ordinated, e.g. [Ag4(SPh)4(PPh3)4]15 or [(AgSCMeEt2)8(PPh3)2]
and [(AgSBut)14(PPh3)4].14
The Ag–P bond in compound 4 (2.3845(12) Å) is significantly
shorter than both such bonds in 1 (2.4950(12) and 2.4470(9) Å),
however the compounds differ in the type of co-ordination.
There are some remarkable features in the structure of 4. Its
bimetallic central Ag2S2 ring is rare among neutral thiolates. We
can recall here two similar copper-group complexes. One is
neutral, [Cu2(SPh)2(PPh3)4], with a central Cu2S2 core.16 The
two copper atoms have tetrahedral environments with two
phosphines attached to each and both bridging thiolate groups
are coplanar with the central Cu2S2 ring which precludes more
detailed comparisons. A recently reported silver thiolate,
of some resemblance to 4, is a derivative of o-carboranethiol,
1-(SH)-1,2-C2B10H11. The molecule of [Ag4(µ3-SC2B10H11)-
31P-31
P
coupling is found to be 146 Hz and is in the range for other
silver() complexes of phosphorus donor ligands. Of interest is
1
the significant difference in JP-Ag couplings for 1 (269 and 391
Hz). Moreover these values are much smaller than that of
4. The question of the relationship between the magnitude
1
of JP-Ag and geometrical parameters of phosphorus–silver()
complexes was discussed by several research groups.19 For
phosphole–silver() compounds Attar et al.20 have shown the
simple relationship between the inverse cube of the Ag–P bond
1
length (d) and JP-Ag; 1/(d /Å)3 = (AJ /Hz) ϩ B where A and B
are constants dependent on the class of complexes. It is appar-
ent that a similar correlation will be valid for phosphinesilver()
silanethiolates. The values of J and corresponding Ag–P
distances are collected in Table 2.
1
Fig. 3 shows the trend in JP-Ag versus the phosphorus–silver
bond length. It is obvious that for a quantitative description of
this relationship a larger experimental database is necessary.
Nevertheless, it is worthy of note that even with such a limited
data set the A and B parameters (3.32 × 10Ϫ5 and 5.54 × 10Ϫ2
,
respectively) are very similar to those reported elsewhere.20
Owing to the lack of X-ray data for compounds 2 and 5 their
structures were deduced from analysis of NMR parameters.
The 31P CP MAS spectrum of 2 (Fig. 2c) displays triplet like
patterns which are due to partial overlap of AMX spin systems.
Such coupling confirms that two phosphine ligands are bonded
2
31P-31
P
to silver. The geminal J
splitting is 110 Hz. Moreover,
only one value of the 1JP-Ag coupling constant equal to 354 Hz is
J. Chem. Soc., Dalton Trans., 1999, 3063–3068 3065

View Article Online
Table 3 The 31P principal elements of effective dipolar/chemical shift tensors for compounds 1, 2, 4 and 5
Compound
Spin
δiso
T11/ppm
T22/ppm
T33/ppm
Ω/ppm
κ
4
1
Xϩ
XϪ
Zϩ
8.3
3.8
8.2
7.1
4.9
3.8
0.6
Ϫ0.5
Ϫ1.6
Ϫ2.7
2.6
Ϫ0.3
6.4
33
29
46
39
41
34
29
27
26
24
39
36
33
28
12
8
8
7
5
Ϫ20
53
54
75
63
72
60
49
48
47
48
72
71
55
55
0.21
0.23
0.00
0.00
0.00
Ϫ25
Ϫ29
Ϫ24
Ϫ31
Ϫ26
Ϫ20
Ϫ21
Ϫ21
Ϫ24
Ϫ33
Ϫ35
Ϫ22
Ϫ27
ZϪ
Yϩ
YϪ
Xϩ
XϪ
Wϩ
WϪ
Xϩ
XϪ
Xϩ
XϪ
4
0.00
Ϫ7
Ϫ8
Ϫ9
Ϫ9
2
Ϫ2
9
Ϫ0.46
Ϫ0.47
Ϫ0.47
Ϫ0.40
Ϫ0.03
Ϫ0.07
0.14
2
5
2.0
5
0.16
Estimated errors in T11, T22 and T33 are 2 ppm; errors in δiso are 0.2 ppm. The principal components of the chemical shift tensor are
T11 > T22 > T33. The isotropic chemical shift is given by δiso = (T11 ϩ T22 ϩ T33)/3.
J anisotropy is in the range of the isotropic J coupling value
and in our case this effect should not influence the dipolar
phosphorus–silver coupling. As the line shape and spinning
sideband pattern are affected by chemical shift and dipolar
interactions, according to the notation used by Zilm and
Grant21 the effective dipolar/chemical shift tensor components
are labeled as Tii. Subsequent analysis of the sideband inten-
sities gives principal values of the tensors Tiiϩ and TiiϪ. The
values of δiso and of the isotropic coupling constants Jiso can be
obtained directly. Unfortunately, it is not possible to obtain the
details of the Tii orientation with respect to the principal axis
system. The calculated values of the 31P principal components
Tii, span Ω and skew κ are given in Table 3.
The accuracy of calculation was confirmed by comparison of
experimental and theoretical spectra. Deconvolution of indi-
vidual lines for sample 2 was impossible due to partial overlap
Fig. 3 Plot of the reciprocal of cube of the distance d(Ag–P) vs.
coupling constant 1JP-Ag
.
found. Hence, these results suggest that d(Ag–P) for both lig-
ands are similar and according to the proposed equation should
be ca. 2.46 Å. Sample 5 (Fig. 2d) exhibits an AX doublet. The
line shape and magnitude of JP-Ag are very similar to those
observed for complex 4. Thus the structural resemblance of
of resonances in AMX multiplets. However, with line broaden-
2
31P-31
P
ing of 110 Hz the J
splitting was not observed. Only the
AX subspectrum was used for further calculation.
1
The span parameter reflects the distortion of the geometry
from ideal tetrahedral or cubic symmetry. The values found for
compounds 1, 2, 4, 5 (48–75 ppm) are in the range typical for
silver() complexes with phosphine ligands. The skew parameter
reflects the distribution of electron density around the nuclei
under investigation. It ranges from Ϫ1 for δ22(T22) = δ33(T33) to
ϩ1 for δ11(T11) = δ33(T33). The κ magnitudes, established for
silver() silanethiolates, indicate that electron density is not
localized in any particular bond but is spread over the entire
tetrahedron. On the other hand, of interest is the linear correl-
these two compounds is apparent.
Interesting information about the molecular structure of
complexes 1, 2, 4, 5 can be obtained from an analysis of
chemical shift tensor elements (δii) and chemical shift tensor
parameters, span (Ω) and skew (κ). The static 31P NMR solid
state spectra of phosphinesilver() silanethiolates 1, 2, 4, 5 are
relatively broad (ca. 9 kHz). With sample spinning higher than
9 kHz the broad lines can be narrowed to isotropic signals.
Under slow MAS condition (ca. 2 kHz) the isotropic lines are
symmetrically flanked by spinning sidebands, which can be
further used to calculate chemical shift anisotropy and other
NMR parameters. However, for the systems under investigation
several precautions have to be taken. It is apparent that for
spectra of dipolar-coupled I = 1/2 spin pairs the relationship
between sample spin rotation, dipolar coupling and chemical
shift anisotropy is crucial for the analysis of MAS spectra.
Dipolar interaction constants RDD expressed by [(γPγAgh)/
2π] 〈rP-AgϪ3〉(µ0/4π) are dependent only on the inverse cube of
the separation of phosphorus and silver. The RDD values calcu-
lated using the P–Ag distances taken from X-ray crystallo-
graphic data were found to be for 109Ag from Ϫ146.8 Hz to
Ϫ168.1 Hz and for 107Ag from Ϫ127.6 to Ϫ146.0 Hz. The
homonuclear 31P–31P dipolar coupling effect for sample 1
was also considered. With a phosphorus–phosphorus separ-
ation equal to 4.3074 Å the RDD is 247 Hz. For static spectra
when the J and D tensors are collinear the dipolar coupling
constant is diminished in effect by one third of the anisotropy
of the J coupling, Reff = RDD Ϫ (∆J/3), or increased if the
anisotropy has a negative value. It is known, however, that
1
ation between κ and JP-Ag. As shown in Fig. 4 for a large
coupling constant the skew is positive, becoming zero for
1
intermediate values and is negative for small values of JP-Ag
.
There is not such tendency for the span parameter. Compared
to other silver() complexes, silver() silanethiolates provide an
additional, attractive probe for structural studies, which allow
extended application of the NMR technique. Silicon-29, for the
samples under investigation, is considered as an isolated
nucleus. Splitting by the 107Ag, 109Ag and 31P nuclei is below the
detectable limit. Experimental 29Si CP MAS NMR spectra for
4, 1, 5 samples are displayed in Fig. 5. The calculated values
of the 29Si principal components of the chemical shift tensor
δii, span Ω and skew κ are collected in Table 4.
Inspection of 29Si data provides interesting information
about significant distinction of the local environment of the
silicon centers in 4, 5 and 1. The isotropic chemical shift δiso
is extremely sensitive to the character of the substituent. For
sample 1 the δiso is upfield by about 55 ppm compared to 4 and
5. Moreover, the span parameter for 1 is over 20 ppm larger
compared to 4 and 5. The differences in Ω are very likely related
with distortion of the tetrahedral geometry around silicon. For
3066
J. Chem. Soc., Dalton Trans., 1999, 3063–3068

View Article Online
Table 4 The 29Si principal elements of chemical shift tensors for com-
plexes 1, 4 and 5
from 31P CP MAS data about the similarity of their structures.
The effect of 29Si chemical shift anisotropy is not observed for
sample 2. This complex exhibited single resonance at spinning
rate 1 kHz. The δiso is found to be Ϫ43.5 ppm.
Compound
δiso
δ11
δ22
δ33
Ω
κ
The preliminary, 29Si CP MAS NMR results presented in this
work unambiguously show that this technique is potentially
useful in structural studies of silver() silanethiolate complexes.
4
1
5
Ϫ2.6
Ϫ57.6
Ϫ2.1
23
Ϫ20
23
10
Ϫ42
10
Ϫ41
Ϫ111
Ϫ40
64
91
63
0.59
0.51
0.58
Estimated errors in δ11, δ22 and δ33 are 2 ppm; errors in δiso are 0.2
ppm. The principal components of the chemical shift tensor are
Experimental
Miscellaneous
δ11 > δ22 > δ33. The isotropic chemical shift is given by δiso = (δ11 ϩ δ22
δ33)/3.
ϩ
Trialkoxysilanethiols (ButO)3SiSH, (BusO)3SiSH and (PriO)3-
SiSH were obtained by alcoholysis of SiS2 as described previ-
ously.22,23 Triarylsilanethiols were obtained by insertion of
elemental sulfur into silanes.24 Acetylacetonatobis(triphenyl-
phosphine)silver(), [Ag(acac)(PPh3)2], was prepared by reac-
tion of Ag2O with 2,5-pentanedione and triphenylphosphine
according to the literature procedure25 (developed for fluorin-
ated analogues). The crude product was purified by passing its
toluene solution through a short (4 cm) silica gel layer. Thin
layer chromatography was performed on silica gel plates, Silufol
(Kavalier): mobile phase, hexane–benzene (4:1, v/v); develop-
ers, iodine vapors or hydrogen sulfide.
Preparation of compounds
[Ag{SSi(OBut)3}(PPh3)2] 1. The compound [Ag(acac)-
(PPh3)2] (366 mg) was dissolved in 15 ml of toluene and a solu-
tion of 140 mg of tri-tert-butoxysilanethiol in 15 ml of toluene
was added. The reagents were allowed to react for 15 min, then
most of the solvent was removed. Upon layering of hexane 470
mg of crude product deposited. Recrystallization from hexane–
ethanol (7:5) yielded 310 mg (68%) of 1. Found: C, 63.26;
H, 6.27; S, 3.39. C48H57AgO3P2SSi requires C, 63.22; H, 6.30;
S, 3.52%). NMR: δH (C6D6) 1.58 (27 H, s), 7.0–7.1 (18 H, m,
m- and p-H of Ph) and 7.3–7.5 (12 H, m, o-H of Ph), (CDCl3)
1.40 (27 H, s) and 7.37 (30 H, m, Ph); δC (C6D6) 32.126 (s, Me),
62 (s, CMe3), 128.73, 128.86, 128.94, 133.98, 134.37 and 137
[d, J(C–P) 15 Hz, Ph]; δP (CDCl3) 3.083 (s, PPh3), ν(1/2) = 20
Hz; at Ϫ60 ЊC 4.849 (s, PPh3), ν(1/2) = 400 Hz. Single crystals
of 1 suitable for diffraction measurements were obtained by
recrystallization from octane solution to which a few drops of
chloroform were added. Other synthetic procedures leading to 1
are given in SUP 57613.
Fig. 4 Plot of skew vs. coupling constant 1JP-Ag
.
[Ag{SSi(OPri)3}(PPh3)2] 2. To a solution of [Ag(acac)-
(PPh3)2] (1 mmol) in 15 ml of benzene 0.260 ml of 80% tri-iso-
propoxysilanethiol was added. The contents turned light brown
and were allowed to react for 30 min. The solvent was evapor-
ated to a volume of about 3 ml and then 4 ml of hexane were
slowly added. The deposit was filtered off, washed with hexane,
dried and weighed. The procedure yielded 700 mg (80%) of
compound 2; mp 161–162 ЊC. Found: Ag, 12.5. C45H51AgO3-
P2SSi requires Ag, 12.4%. The product is not stable and grad-
ually darkens in the air at ambient temperatures as well as in a
refrigerator. NMR: δH (C6D6) 1.26 (18 H, d, J = 6, CH3), 4.63 (3
H, septet, J = 6 Hz, CH), 7.0–7.1 (18 H, m, m- and p-H of Ph)
and 7.45–7.6 (12 H, m, o-H of Ph).
[Ag{SSi(OBus)3}(PPh3)2] 3. Preparation was analogous to
that of compound 2 except the use of half of the substrates
(mp 126 ЊC). Compound 3 is not stable in the air and gradually
darkens during storage.
Fig. 5 The 59.63 MHz 29Si CP MAS NMR spectra of (a) compound
4, νrot = 1.24 kHz, (b) 1, νrot = 1.20 kHz and (c) 5, νrot = 1.24 kHz. The
spectra have 4K data points, a contact time of 1 ms and 400 scans.
1 the S–Si–O valence angles are 107.41, 112.64 and 115.78Њ
while for 4 the corresponding S–Si–C valence angles are 108.62,
110.09 and 111.80Њ. The spread of valence angles for 1 is much
larger. This is probably a result of the steric hindrance of the
tert-butyl groups. The resemblance of 29Si δii parameters for 4
and 5 provides further evidence in support of the hypothesis
[{Ag[SSi(C6H4Me-o)3](PPh3)}2]ؒ2C6H5Me 4. To a solution of
340 mg of [Ag(acac)(PPh3)2] in 15 ml of toluene a solution of
167 mg of (o-MeC6H4)3SiSH in 5 ml of toluene was added.
After 15 min of reaction 2 ml of hexane were added, then
another 4 ml after half an hour. On the next day the tiny
crystals were filtered off and recrystallized from 10 ml of hot
J. Chem. Soc., Dalton Trans., 1999, 3063–3068
3067

View Article Online
Table 5 Summary of crystal data, intensity collection and structure
refinement for compound 4
width of 50 kHz were used and 8 K data points represented the
FID. Spectra were accumulated 100 times. The Hartman–Hahn
match was set up using Q8M8 standard (Bruker). The 29Si
chemical shifts were calibrated indirectly through TMS set at
0.0 ppm. The principal elements of the 31P and 29Si chemical
shift tensor and shielding parameters were calculated employ-
ing WIN-MAS program. The details describing the method
and accuracy of calculations are exhaustively discussed
elsewhere.28,29 The principal components δii were used for calcu-
lation of the chemical shift parameters: Ω = δ11 Ϫ δ33 and
κ = 3(δ22 Ϫ δiso)/Ω.
Chemical formula
Formula weight
C46H44AgPSSi
795.80
Crystal system
Space group
Triclinic
P1
¯
a/Å
b/Å
c/Å
α/Њ
12.824(3)
13.667(3)
14.524(3)
115.15(3)
99.76(3)
107.71(3)
2059.0(8)
293(2)
β/Њ
γ/Њ
V/Å3
T/K
Acknowledgements
Financial support from the Polish State Committee of Scientific
Research (project No. 3 TO9A 072 15) is gratefully acknowl-
edged.
Z
2
µ/mmϪ1
0.638
Number of reflections, measured
unique
R1, wR2 for I > 2σ(I)
all reflections
8402
8111
0.0267, 0.0716
0.0927, 0.0899
References
1 Part 70, B. Becker, K. Radacki and W. Wojnowski, J. Organomet.
Chem., 1996, 521, 39.
2 E. Block, M. Gernon, H. Kang, G. Ofori-Okai and J. Zubieta, Inorg.
Chem., 1989, 28, 1263.
toluene giving crystals suitable for the X-ray study. Yield 305
mg (75%) of compound 4, mp 183 ЊC. NMR: δH (C6D6) 2.1
(s, C6H4CH3) and 6.85–7.2 (m, aryl).
3 K. Tang, M. Aslam, E. Block, T. Nicholson and J. Zubieta, Inorg.
Chem., 1987, 26, 1488.
4 W. Wojnowski, M. Wojnowski, K. Peters, E. M. Peters and
H. G. von Schnering, Z. Anorg. Allg. Chem., 1985, 530, 79.
5 B. Becker, W. Wojnowski, E. M. Peters, K. Peters and H. G. von
Schnering, Polyhedron, 1990, 9, 1659.
6 W. Wojnowski, B. Becker, J. Saβmannshausen, E. M. Peters,
K. Peters and H. G. von Schnering, Z. Anorg. Allg. Chem., 1994, 620,
1417.
7 B. Becker, W. Wojnowski, K. Peters, E. M. Peters and H. G. von
Schnering, Inorg. Chim. Acta, 1993, 214, 9.
8 A. Herman, B.Becker and W. Wojnowski, Z. Anorg. Allg. Chem.,
1979, 450, 178.
9 J. Pikies and W. Wojnowski, J. Organomet. Chem., 1990, 386, 305.
10 W. Wojnowski and A. Herman, Z. Anorg. Allg. Chem., 1976, 425, 91.
11 B. Becker, E. W. Felcyn, A. Herman, J. Pikies and W. Wojnowski,
Z. Anorg. Allg. Chem., 1982, 488, 229.
[{Ag(SSiPh3)(PPh3)}2]ؒ2C6H6 5. To a solution of 370 mg of
[Ag(acac)(PPh3)2] in benzene a solution of 150 mg of Ph3SiSH
in 5 ml of benzene was added. Immediately a microcrystalline,
colorless product deposited (345 mg, 75%). In order to obtain
bigger crystals the reaction was carried out in a special test-tube
with a capillary narrowing at its middle. One substrate solution
was placed in the lower part of the tube and the second in the
upper part. This considerably slowed the mixing rate and
should slow the reaction rate and eventually the crystallization
process. Nevertheless, the crystals (mp 164–166 ЊC), although
bigger, were unsuitable for an X-ray study. Found: C, 67.7;
H, 4.8; Ag, 14.4; S, 4.2. C42H36AgPSSi requires C, 68.2; H, 4.9;
Ag, 14.58; S, 4.3%.
12 B. Becker, J. Chojnacki, A. Konitz and W. Wojnowski, Z.
Kristallogr. NCS, 1998, 213, 697.
13 P. Aslanidis, P. Karagiannidis, P. D. Akrivos, B. Krebs and M. Läge,
Inorg. Chim. Acta, 1997, 254, 277.
14 I. G. Dance, L. J. Fitzpatrick, D. C. Craig and M. L. Scudder, Inorg.
Chem., 1989, 28, 1853.
15 L. S. Ahmed, J. R. Dilworth, J. R. Miller and N. Wheatley, Inorg.
Chim. Acta, 1998, 278, 229.
16 I. G. Dance, P. J. Guerney, A. D. Rae and M. L. Scudder, Inorg.
Chem., 1983, 22, 2883.
17 M. M. Artigas, O. Crespo, M. C. Gimeno, P. G. Jones, A. Laguna
and M. D. Villacampa, Inorg. Chem., 1997, 36, 6454.
18 G. Wu, R. E. Wasilishen, H. Pan, C. W. Liu, J. P. Fackler, Jr and
M. Shang, Magn. Reson. Chem., 1995, 33, 734.
19 G. A. Bowmaker, Effendy, P. J. Harvey, P. C. Healy, B. W. Skelton
and A. H. White, J. Chem. Soc., Dalton Trans., 1996, 2449, 2459.
20 S. Attar, N. W. Alcock, G. A. Bowmaker, J. S. Frye, W. H. Bearden
and J. H. Nelson, Inorg. Chem., 1991, 30, 4166.
21 K. W. Zilm and D. M. Grant, J. Am. Chem. Soc., 1981, 103, 2913.
22 W. Wojnowski and R. Piekos, Z. Anorg. Allg. Chem., 1962, 318, 189.
23 R. Piekos and W. Wojnowski, Z. Anorg. Allg. Chem., 1962, 318, 212.
24 B. Becker and W. Wojnowski, Synth. React. Inorg. Metal Org.
Chem., 1982, 12, 565.
Crystal structure determination
Diffraction measurements were carried on using a KUMA
KM4 diffractometer with Mo-Kα radiation (λ 0.71073 Å) at
ambient temperature. The crystal structure was solved and
refined using the SHELX-97 program package26,27 with full-
matrix least-squares refinement based on F 2. Crystal data are
given in Table 5.
CCDC reference number 186/1572.
See http://www.rsc.org/suppdata/dt/1999/3063/ for crystallo-
graphic files in .cif format.
NMR Spectroscopy
Spectra of compounds in solutions (1H, 31P and 13C) were
recorded on Gemini 200 (200 MHz) or Unity Plus (500 MHz)
spectrometers; C6D6 or CDCl3 was used as solvent. Cross-
polarization magic-angle spinning solid state 31P and 29Si NMR
spectra were recorded on a Bruker 300 MSL instrument with
high-power proton decoupling at 121.496 MHz for 31P and
59.627 MHz for 29Si. Powder samples were placed in a cylindri-
cal rotor and spun at 2.0–4.5 kHz. For the 31P experiments
the field strength for 1H decoupling was 1.05 mT, a contact time
of 5 ms, a repetition of 6 s and spectral width of 50 kHz were
used and 8 K data points represented the free induction decay
(FID). Spectra were accumulated 100 times which gave a
reasonable signal-to-noise ratio. The 31P chemical shifts were
calibrated indirectly through bis(dineopentoxythiophos-
phoryl) disulfide at δ 84.0 as secondary reference; 85% H3PO4
was used as primary reference for 31P and set at 0 ppm. For the
25 Z. Yuang, N. H. Dryden, J. J. Vittal and R. J. Puddephatt, Can. J.
Chem., 1993, 72, 1605.
26 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467.
27 G. M. Sheldrick, SHELX 97. Programs for the solution and the
refinement of crystal structures from diffraction data, University of
Göttingen, 1997.
28 G. Jeschke and G. Grossmann, J. Magn. Reson., Ser. A, 1993, 103,
323.
29 J. Herzfeld and A. Berger, J. Chem. Phys., 1980, 73, 6021.
29Si experiments the field strength for H decoupling was 1.05
1
mT, a contact time of 3 ms, a repetition of 10 s and spectral
Paper 9/03694F
3068
J. Chem. Soc., Dalton Trans., 1999, 3063–3068