A novel hexanuclear silver(i) cluster containing a regular Ag6 ring with short Ag-Ag distances and an argentophilic interaction

Article abstract of DOI:10.1039/c3dt33003f

The hexanuclear complex [HQ][Ag(p-mpspa)] (H₂-p-mpspa = 3-(4-methoxyphenyl)-2-sulfanylpropenoic acid) was prepared by reacting the precursor [Ag(H-p-mpspa)] with diisopropylamine (Q). The complex was characterized by spectroscopic techniques and the structure was solved by a single crystal X-ray diffraction study. The crystal contains hydrogen-bonded diisopropylammonium cations and [Ag₆(p-mpspa)₆] ^6- anions that are based on a regular Ag₆ ring with each S-donor atom of the sulfanylcarboxylate ligand bridging two Ag atoms. The Ag-Ag bond distances, 2.8036(6) A, are very short and suggest a closed shell d¹⁰...d¹⁰ argentophilic interaction. To analyze the relative role of this interaction and that of the S-bridging atom the anionic [Ag₆(p-mpspa)₆]^6- moiety has been studied theoretically at the Hartree-Fock (HF) and 2^nd order Moller-Plesset perturbation theory (MP2) levels on a very simple [Ag ₆(SH)₆] A model system. A large model system [Ag ₆(p-mpspa)₆]^6-B has also been studied by applying the ONIOM (QM/MM) approach using HF/UFF and MP2/UFF combinations as levels of theory. The six experimentally observed Ag(i)...Ag(i) supported interactions are reproduced when dispersion-type interactions are considered in the theory levels MP2 and ONIOM MP2/UFF for models A and B, respectively. The use of HF and ONIOM HF/UFF levels led to a similar hexanuclear structure but displayed a large hexagonal disposition without argentophilic contacts for both models A and B. The steric hindrance exerted by the ligands did not preclude the formation of argentophilic interactions, as observed experimentally.

Full text of DOI:10.1039/c3dt33003f

Dalton
Transactions
View Article Online
View Journal
PAPER
A novel hexanuclear silver(I) cluster containing a
regular Ag6 ring with short Ag–Ag distances and
an argentophilic interaction†
Cite this: DOI: 10.1039/c3dt33003f
Elena Barreiro,a José S. Casas,a María D. Couce,b Antonio Laguna,*c
José M. López-de-Luzuriaga,d Miguel Monge,d Agustin Sánchez,a José Sordo*a
and Ezequiel M. Vázquez Lópezb
The hexanuclear complex [HQ][Ag(p-mpspa)] (H2-p-mpspa = 3-(4-methoxyphenyl)-2-sulfanylpropenoic
acid) was prepared by reacting the precursor [Ag(H-p-mpspa)] with diisopropylamine (Q). The complex
was characterized by spectroscopic techniques and the structure was solved by a single crystal X-ray diffr-
action study. The crystal contains hydrogen-bonded diisopropylammonium cations and [Ag6(p-mpspa)6]6−
anions that are based on a regular Ag6 ring with each S-donor atom of the sulfanylcarboxylate ligand
bridging two Ag atoms. The Ag–Ag bond distances, 2.8036(6) Å, are very short and suggest a closed
shell d10⋯d10 argentophilic interaction. To analyze the relative role of this interaction and that of the
S-bridging atom the anionic [Ag6(p-mpspa)6]6− moiety has been studied theoretically at the Hartree–
Fock (HF) and 2nd order Møller–Plesset perturbation theory (MP2) levels on a very simple [Ag6(SH)6] A
model system. A large model system [Ag6(p-mpspa)6]6− B has also been studied by applying the ONIOM
(QM/MM) approach using HF/UFF and MP2/UFF combinations as levels of theory. The six experimentally
observed Ag(I)⋯Ag(I) supported interactions are reproduced when dispersion-type interactions are
considered in the theory levels MP2 and ONIOM MP2/UFF for models A and B, respectively. The use of
HF and ONIOM HF/UFF levels led to a similar hexanuclear structure but displayed a large hexagonal dis-
position without argentophilic contacts for both models A and B. The steric hindrance exerted by the
ligands did not preclude the formation of argentophilic interactions, as observed experimentally.
Received 14th December 2012,
Accepted 8th February 2013
DOI: 10.1039/c3dt33003f
www.rsc.org/dalton
with the bonding interaction between the metal atoms. The
Introduction
study of this weak attractive closed-shell d10–d10 interaction,
Interest in the biological activity and medicinal properties,1 i.e. metallophilicity, in the group 11 elements has mainly been
together with a wide range of other technological appli- focused on gold(I) compounds, where the relativistic effects are
cations,2 has been a driving force in the study of silver(I) more significant. The influence of this interaction, aurophili-
coordination compounds,3a among which cluster compounds city, on the molecular structures and on the chemical
represent a relevant and current field of research.3b–f
and physical properties of gold compounds was recently
A relevant aspect to be studied in these polynuclear com- reviewed.4a This interaction in silver(I) compounds, argento-
pounds is the metal–metal bond distance and its connection philicity, has been less widely studied. Furthermore, in com-
pounds that contain an additional bridging atom between the
silver atoms, i.e. ligand-supported or -assisted complexes, the
aDepartamento de Química Inorgánica, Facultade de Farmacia, Universidade de
connection between a short Ag–Ag distance and argentophili-
city is normally only suggested.4b,c,5 However, recent theoreti-
Santiago de Compostela, 15782 Santiago de Compostela, Spain.
E-mail: jose.sordo@usc.es; Fax: +34 981547102; Tel: +34 981528074
bDepartamento de Química Inorgánica, Facultade de Química, Universidade de Vigo,
cal and structural evidence4d–f underlines the relevance of this
interaction in unsupported complexes when compared with
36310 Vigo, Spain
cDepartamento de Química Inorgánica, Instituto de Ciencia de Materiales de Aragón,
Universidad de Zaragoza-CSIC, 50009-Zaragoza, Spain. E-mail: alaguna@unizar.es;
Fax: +34 976761187; Tel: +34 976761185
dDepartamento de Química, Universidad de La Rioja, Centro de Investigación en
Síntesis Química (CISQ), Complejo Científico-Tecnológico, 26004-Logroño, Spain
†CCDC 914133. For crystallographic data in CIF or other electronic format see
DOI: 10.1039/c3dt33003f
aurophilicity.
In the search for new antibacterial agents, we have pre-
viously described5 new sulfanylcarboxylates of silver(I) that
contain tetranuclear units with Ag–S and Ag–O bonds along
with short Ag–Ag bond distances. Besides these tetranuclear
species, the ESI/MS spectra of some of these compounds show
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 1 3-(4-Methoxyphenyl)-2-sulfanylpropenoic acid: H2-p-mpspsa.
Scheme 2 Synthesis of the complexes.
the existence in solution, albeit with low abundance, of
other species, of higher nuclearity. In the work described here,
reaction of the complex [Ag(H-p-mpspa)] (1) (H2-p-mpspa =
3-(4-methoxyphenyl)-2-sulfanylpropenoic acid, Scheme 1)
with diisopropylamine (Q) in ethanol led to the isolation of
[HQ][Ag(p-mpspa)] (2), the X-ray diffraction study of which
showed it to contain the [Ag6(p-mpspa)6]6− hexanuclear anion.
This anion, along with the counterpart diisopropylammonium
cation (HQ), configures the crystal structure of 2.
the NMR data obtained in dmso-d6. In the second step, reac-
tion of this solid (its solubility in commonly used solvents is
low, and we employed a suspension in ethanol) with diisopro-
pylamine led to deprotonation of the aforementioned COOH
group (the typical bands in the IR spectrum were replaced by
the typical bands of the carboxylate group) and gave rise to
[HQ][Ag(p-mpspa)], which is soluble in ethanol, methanol,
acetone and dmso.
Structural analysis of the anion revealed an internal core
with a regular hexagonal Ag6 ring in which each pair of
Ag atoms is bridged by an S-atom of the sulfanylcarboxylate
ligand, which is also O-coordinated. A significant parameter in
the ring is the short Ag–Ag bond distance, which could be due
to an argentophilic interaction, the effect of the S-bridge or a
combination of both factors. In an effort to analyse the relative
influence of these potential supporters, we carried out a theor-
etical study at different levels of complexity. Thus, the use of
non-correlated (Hartree–Fock level, HF) and correlated (2nd
order Møller–Plesset level, MP2) levels of theory allows a com-
parison among the theoretically predicted geometries in the
absence of argentophilic interactions (HF) and in the presence
of argentophilic interactions (MP2 level) with the experimental
one. It is also possible to account for the energetic stability
induced by the presence of these metallophilic dispersive
forces. The study revealed the strong influence of the argento-
philic interaction on the structural arrangement of the hexago-
nal anion.
Crystal structure of [HQ][Ag(p-mpspa)] (2)
The compound crystallizes from a methanol/dmso solution in
ˉ
the rhombohedric space group R3, with the crystal containing
diisopropylammonium cations and hexanuclear [Ag6(p-
mpspa)6]6− anions. The structure of the anion is shown in
Fig. 1 together with the numbering scheme and selected bond
distances and angles are listed in Table 1. As can be seen in
the figure, the anion is centered on a three-fold rotoinversion
axis (a position in the Wyckoff notation) and it is built on the
basis of a central regular Ag6 ring with each Ag atom 0.103(6) Å
from the ideal plane (rms = 0.1025). Six p-mpspa2− ligands are
bonded to this ring through the S atom and one of the O
atoms of the carboxylate group [O(1)–Ag(1)–S(1): 76.80(11)°];
the S and O atoms are alternately located above and below the
ring. Each ligand is significantly planar, with the O(1)C(1)C(2)-
S(1) and the C(4)–C(9) planes making an angle of 15.0(5)°
and the S atom bridging two Ag atoms and the O atom
Results and discussion
Synthesis and characterization
The synthesis of [HQ][Ag(p-mpspa)] (2) was carried out as
described in detail in the Experimental section, according to
the route shown in Scheme 2. The overall deprotonation of the
H2-p-mpspa ligand was carried out in two steps. In the first
step the ligand was reacted with AgNO3 and triethylamine and
the solid [Ag(H-p-mpspa)] (1) was liberated. This solid was iso-
lated in high yield and was characterized analytically. The
IR spectrum showed the disappearance of the ν(S–H) band
present in the spectrum of the free ligand6 at 2570 cm−1
,
whereas the COOH bands at 1662, ν(CvO), 1427, δ(OH) and
1259, ν(C–O), cm−1 in the free ligand remained at similar wave-
numbers, suggesting that this group is not deprotonated in
Fig. 1 The molecular structure and numbering scheme of the
the complex. This lack of deprotonation was also suggested by [Ag6(p-mpspa)6]6− anion.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Taking into account the S- and O-coordination only, each
silver atom adopts an AgS2O kernel that can be described as
distorted trigonal. The distortion is mainly due to the bite of
the ligand but the sum of the angles around the silver atom is
essentially 360°. However, in the ring, close to each Ag atom,
there are two other Ag atoms located a short distance away.
The Ag(1)–Ag(1)# bond distance, 2.8036(6) Å, is lower than that
found in metallic silver (2.889(6) Å), in the range expected for
the sum of the covalent radii of the two silver atoms (2.90 Å)7
and significantly lower than the sum of the van der Waals
radii for this metal (3.44 Å)9. This distance suggests a signifi-
cant argentophilic interaction and this was investigated
further (vide infra). If this interaction is considered, the Ag
atom is then located in a very distorted pentacoordinated
Ag2S2O environment.
Table 1 Selected bond distances (Å) and angles (°) for [HQ][Ag(p-mpspa)] (2)
Ag(1)–S(1)#1
Ag(1)–S(1)
2.3973(15)
2.4752(15)
2.415(4)
2.3973(15)
2.8036(6)
2.8036(6)
S(1)#1–Ag(1)–S(1)
O(1)–Ag(1)–S(1)
146.56(6)
76.80(11)
111.57(4)
85.36(12)
53.58(4)
Ag(1)–O(1)
S(1)#1–Ag(1)–Ag(1)#2
O(1)–Ag(1)–Ag(1)#2
S(1)–Ag(1)–Ag(1)#2
S(1)#1–Ag(1)–Ag(1)#1
O(1)–Ag(1)–Ag(1)#1
S(1)–Ag(1)–Ag(1)#1
Ag(1)#2–Ag(1)–Ag(1)#1
Ag(1)–S(1)–Ag(1)
Ag(1)#2–S(1)
Ag(1)–Ag(1)#1
Ag(1)–Ag(1)#2
56.19(4)
148.83(12)
101.98(4)
119.475(5)
70.23(4)
#1x − y, x, −z, #2y, −x + y, −z.
Information from the CSD10 shows that, although less
common structures have been described,11a,b the usual struc-
tural arrangement of the silver atoms for discrete hexanuclear
units in silver(I) complexes containing S-donor ligands is the
Ag6 octahedron, which shows a variable degree of distortion.
Ligands such as simple thiolates {[Ag(pfbt)(PPh3)] (Hpfbt =
pentafluorobenzenethiol),12a}, S,S′-donors like dithiocarba-
mates {[Ag(S2CNR2)] (R = propyl,12b diisopropyl,12c or butyl12d},
dithiolate {[Ag{S2C = C(CN)2}]12e} or dithiophosphate {[Ag{S2P-
(OC3H7)2}]12f}, S,N-donors containing a thione/thiol C–S bond
for which the N-donor atom is included in a ring {in
complexes of the type [Ag(L)]: HL1 = 4-benzyl-thiazolidine-2-
Fig. 2 A schematic view of the Ag6 ring showing the positions of the S- and
O-coordinated atoms.
coordinating to one of these two atoms. The core of the struc-
ture can thus be schematically represented as shown in Fig. 2.
The S atoms are located 1.992(2) Å from the Ag6 plane and
the angle between the Ag2S and the Ag6 planes is 89.75(5)°.
The Ag–S bond distances [Ag(1)–S(1)#1: 2.3973(15) Å; Ag(1)–
S(1): 2.4752(15) Å] are slightly different, and thus the S atom is
slightly out of the plane normal to the silver ring that divides
the two opposite sides of the hexagon. The values are close to,
or lower than, the sum of the covalent radii of Ag and S atoms
(2.50 Å)7 and are close to those previously found in silver(I)
complexes with similar sulfanylcarboxylate ligands of different
nuclearity5 and in [(AgPPh3)2(p-mpspa)],6 where this same
ligand coexists with triphenylphosphine in a tetranuclear
silver(I) unit. Similar values were also found in silver(I) com-
plexes with different S-donor ligands in which more or less
symmetric Ag–S–Ag bridges are also present.8 The average
value of the two Ag–S distances, 2.44 Å, is close to the value
obtained from a search of the Cambridge Structural Database
(CSD)8e for three-coordinate silver present in compounds con-
taining at least one Ag–S bond and where no metal–metal
bonds were considered.
The O atoms are also alternately arranged above and below
the silver plane and are located 0.445(6) Å away from this
plane. In this case the O atoms are significantly out (1.314(6)
Å) of the normal plane to the silver ring that passes through
two opposite corners of the hexagon. The Ag–O bond distances
[Ag(1)–O(1): 2.415(4) Å] are larger than the sum of the covalent
radii for Ag and O atoms (2.11 Å)7 but significantly shorter
than the sum of the van der Waals radii for both atoms
(3.20 Å)9 and also in the range previously found in other
silver(I) sulfanylcarboxylates.5,6
thione;12g HL2 = 4-isopropyl-thiazolidine-2-thione;12h HL3
4-ethyl-5-pyridin-2-yl-2,4-dihydro-[1,2,4]triazol-3-thione;12i HL4
=
=
6-(tert-butyldimethylsilyl)pyridine-2-thione;12j HL5 = 2-benzimi-
dazole thiol8d and H2L6
= 2-mercaptonicotinic acid for
which complexes [Ag(Hmna)],12k [Ag(mna)]−,8b [Ag6(Hmna)4-
(mna)2]2− 12l are known} or S,N-donor thiosemicarbazones
{[Ag(L)]: HL1 = salicylaldehydethiosemicarbazone;8c HL2
=
2-pyridinecarboxaldehyde 4-N-ethylthiosemicarbazone4b} are
present in these units.
As was mentioned, a general feature for these compounds
is the presence of an Ag6 distorted octahedral skeleton. The
usual distortion is trigonal and gives rise to two staggered Ag3
triangles placed at variable distances.12f In all the cases, when
the ligand has two donor atoms (S, D), the S atom of each
ligand bridges two Ag atoms occupying the three edges of
these staggered Ag3 triangles. The second donor atom, D, is
terminally coordinated to only an Ag atom. In this way the
ligands act as tridentate and the silver atoms, excluding poten-
tial Ag–Ag bonds, are trigonally coordinate. The structure of
the [Ag6(p-mpspa)6]6− anion shows an alternative way to place
six Ag atoms and six ligands having S- and D- as donor atoms.
The ligands are again tridentate, the S-atoms also bridge
two Ag atoms and the D (O in this case) atom also terminally
coordinates to each of the six Ag atoms, which now, in the Ag6
hexagonal ring, are equivalent.
To the best of our knowledge, a core similar to that reported
in this paper has not previously been described, although two
previously reported Ag6 rings do show some similarities. In
the complex {(Ph3PS)4·[Ag(OC(O)CF3]6}13a there is a central Ag6
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Paper
Dalton Transactions
ring with a twisted boat conformation and with Ag–Ag dis-
tances ranging from 2.9133(6) to 3.1001(6) Å. Each CF3COO2
group bridges two Ag atoms whereas the S-atom of each Ph3PS
group is coordinated to three contiguous Ag atoms in the ring.
In addition to these Ag–O and Ag–S bonds there are argento-
philic interactions contributing to the stability of the structure.
An Ag6 ring with a chair conformation and Ag–Ag distances
of 2.9099(8) Å was stabilized by the deprotonated 4-methyl-2-
((trimethylsilyl)amino)pyridine (HAP) ligand in the [Ag(AP)]
complex.13b Each N,N′-donor AP ligand bridges two contiguous
atoms of the ring, which occupy alternate positions above and
below the ring. Despite some similarities, mainly related with
the structural parameters of the ring, both annular structures
are significantly different from that described in this paper. In
the first case the ring is supported by an O,O′-bidentate bridge
−
CF3CO2 ligand and Ph3PS molecules, which offer an S-donor
atom to coordinate three neighboring Ag atoms. In the second
case an N,N′-bidentate ligand binds two contiguous Ag atoms Fig. 3 A view from the ‘c’ axis of the sheet containing the hydrogen-bonded
[Ag6(p-mpspa)6]6− anions and the HQ+ cations.
of the ring. These two contiguous atoms in the Ag6 ring
described in this paper are only bonded to the same S-donor
atom that acts as a bridge.
In the crystal the [Ag6(p-mpspa)6]6− anion and the HQ+
diisopropylammonium cations are hydrogen bonded, with the
O atoms of the carboxylate groups and the N–H groups of the
diisopropylammonium group being involved in these inter-
actions (Table 2). These hydrogen bonds give rise to sheets, as
shown in Fig. 3.
Theoretical studies
Firstly, in order to check whether the hexanuclear [Ag6S6]
structural disposition depends on the argentophilic inter-
actions or whether the steric hindrance of the p-mpspa ligands
exerts any influence on the structure, we fully optimized the
model system [Ag6(SH)6] A at the HF and MP2 levels of theory
(Fig. 4). Comparison of the results at both levels of theory
allows the evaluation of the role of the dispersive Ag(I)⋯Ag(I)
interactions in the absence of ligands since at the HF level the
dispersion forces are not included and at the MP2 level, where
correlation effects are included, the dispersive nature of the
Fig. 4 Optimized model systems A at the HF and MP2 levels and model B at
metallophilic interactions is taken into account. In order to
the ONIOM HF/UFF and MP2/UFF levels of theory.
analyze the role of the ligands in the structural arrangement
we next carried out ONIOM HF/UFF and MP2/UFF calculations
on the model system [Ag6(p-mpspa)6]6− B. This level of theory
permits the calculation of very large metallic systems since the
ligands can be evaluated at a lower level of theory, which sig-
nificantly reduces the computational cost, but the steric effects
are taken into account. We proceeded to optimize at the
ONIOM MP2/UFF level of theory the hexanuclear structure
including both the dispersive forces that are responsible for
the intermetallic interactions and the steric effects of the
S-donor ligands. At the ONIOM HF/UFF level only the ligand
steric effects are evaluated.
Representative structural data for model A at the HF and
MP2 levels and model B at the ONIOM HF/UFF and MP2/UFF
levels of theory are listed in Table 3. In the case of model A,
full optimization at the HF level led to very large Ag(I)⋯Ag(I)
intermetallic distances of 4.05 Å, i.e. greater than the sum of
the van der Waals radii for Ag(I), and this rules out the for-
mation of intermetallic interactions. The Ag–S distances
are slightly larger than the experimental ones (2.50 Å vs.
Table 2 Hydrogen bonds between the [Ag6(p-mpspa)6]6− anions and the HQ+
cations in (2)
D–H⋯A
d(D–H)
d(H⋯A)
d(D⋯A)
<(DHA)
N(1)–H(1B)⋯O(1)#3
N(1)–H(1A)⋯O(2)
0.90
0.90
1.83
1.79
2.712(6)
2.679(6)
165.1
171.0
#3x + 1, −y + 1, −z.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 3 Selected experimental structural parameters for anion complex [Ag6(p-mpspa)6]6− and optimized parameters for model system [Ag6(SH)6] A at HF and
MP2 levels of theory and [Ag6(p-mpspa)6]6− model system B at ONIOM HF/UFF and MP2/UFF levels of theory (distances in Å and angles in °)
Ag⋯Ag
Large Ag–Ag
Ag–S
S–Ag–S
Ag–S–Ag
Experimental
Model A (HF)
Model A (MP2)
Model B (HF/UFF)
Model B (MP2/UFF)
2.80
4.05
2.78
3.68
2.77
5.60
8.10
5.57
7.37
5.53
2.40–2.47
2.50
2.36
2.51
2.35
146.5
179.8
179.1
179.3–179.7
175.5–179.3
70.2
107.8
72.3
94.5–94.7
71.7–71.8
Table 4 Computed HF and MP2 absolute energies for model systems A and B at each level of theory
HF optimised
model A
MP2 optimised
model A
ONIOM HF/UFF optimised
model B
ONIOM MP2/UFF optimised
model B
HF energy (a.u.)
MP2 energy (a.u.)
−940.4393164
−943.8016974
−940.3422141
−943.9187277
−940.4345105
−943.8203779
−940.3381401
−943.9184040
2.40–2.47 Å) and the S–Ag–S angles are close to 180°. The latter argentophilic interactions is responsible for the structural dis-
are larger than the experimental angles (ca. 146.5°) since the position. Once again the theoretically predicted S–Ag–S angles
theoretical model did not take into account the coordination of 179.1° are far from the experimental ones due to the
site at each Ag(I) center being occupied by an oxygen atom of absence of the sulfanylcarboxylate ligands. In the case of
each sulfanylcarboxylate ligand. The absence of argentophilic model B the most important structural parameters also match
interactions does not preclude the formation of a hexagonal perfectly with the experimental ones. For example, the
Ag6 arrangement with the sulfur atoms of the sulfanylcarboxy- theoretical argentophilic interactions of 2.77 Å, the largest
late functional groups acting as bridging ligands, but of a Ag(I)–Ag(I) distance in the hexanuclear core of 5.53 Å and the
larger size. A similar situation was observed when the com- Ag–S distances of 2.35 Å are very similar to the experimental
plete model system B was optimized at the ONIOM HF/UFF ones. The S–Ag–S angles of 175.5–179.3° also deviate from the
level of theory. In this case the intermetallic Ag–Ag distance is experimental value of 146.5° since theoretical model B was
still larger than the experimental one (3.68 Å vs. 2.80 Å), since built up using an S6 symmetry in order to save the compu-
dispersion effects are not taken into account at this level, but tational cost by placing the O atoms equidistant from the Ag
is slightly shorter than for model A (4.05 Å). This shorter inter- centers. In the structural disposition the O atom of each sulfa-
action distance is probably due to the inclusion of the ligands nylcarboxylate ligand is placed closer to one of the Ag(I)
in the theoretical model. The presence of these bulky sulfanyl- centers. The theoretical Ag–S–Ag angle of 71.7°, similar to
carboxylate ligands in an alternate disposition above and both the experimental one and that obtained in model A at the
below the Ag6 plane would make the Ag–S–Ag angles shorter MP2 level in the absence of the sulfanylcarboxylate ligands,
(107.8° in model A at the HF level and 94.5° in model B at the suggests that the ligands do not have a significant influence
ONIOM HF/UFF level) in order to avoid, in part, the ligand on the structural arrangement of the hexanuclear metallic
steric effects and this change would lead to slightly shorter core and that their alternate disposition should avoid the
Ag–Ag distances. In any case, none of the optimizations steric repulsion between them.
carried out for models A and B at non-correlated levels of
theory gave rise to a local minimum with a structural disposi- gies for model systems A and B at each level of theory, includ-
tion close to the experimental one. ing the hypothetical MP2 energy for the HF-optimised model
Table 4 depicts the computed HF and MP2 absolute ener-
In contrast, when correlation effects are included in the system A and the corresponding MP2/UFF energy for the HF-
level of theory (MP2 for model A and ONIOM MP2/UFF for optimised model B, both displaying an open disposition with
model B), the resulting Ag6 hexagonal arrangement is very very large Ag–Ag distances. With these results we can establish
close to the experimental one. The calculated argentophilic several interesting conclusions. If we have a look at the HF
interaction distances for model A at the MP2 level are very energies for model A optimised at HF and MP2 levels we
similar to the experimental ones (2.78 Å vs. 2.80 Å), as is the observe that when dispersion forces are not taken into account
largest Ag(I)–Ag(I) distance in the planar Ag6 arrangement the HF-optimised structure is more stable than the MP2 one in
(5.57 Å vs. 5.60 Å). The theoretically predicted Ag–S distances 0.0971 a.u., which corresponds to ca. 255 kJ mol−1, which
are 2.36 Å (exp. 2.40–2.47 Å) and the Ag–S–Ag angles are means that a high repulsion between the Ag(I) centers is
around 72.3° (exp. 70.2°). The high structural similarity described at the HF level when Ag centres are close to each
between the simple theoretical model A and the experimental other. Nevertheless, when correlation effects are included the
[Ag6S6] core suggests that the presence of the six Ag(I)⋯Ag(I) MP2 energies show that the MP2-optimised structure, that is
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Paper
Dalton Transactions
the Ag6 cluster displaying short argentophilic interactions, is
now more stable than the HF-optimised one in 0.117 a.u.,
which corresponds to ca. 307 kJ mol−1. If we divide this energy
Experimental section
General
difference by the six silver interactions observed in the cluster 4-Methoxybenzaldehyde, rhodanine and triethylamine (all
we observe that for each new Ag⋯Ag interaction formed from from Aldrich), silver nitrate (Prolabo) and diisopropylamine
the open Ag6 cluster optimized at the HF level to the (Merck) and NaOH (Probus) were all used as supplied. 3-(4-
more compact Ag6 cluster an additional stabilization energy of Methoxyphenyl)-2-sulfanylpropenoic acid was prepared by
ca. 51 kJ mol−1 is reached. This energy not only corresponds to condensation of 4-methoxybenzaldehyde with rhodanine,14
each argentophilic interaction but also to the new overall followed by hydrolysis in an alkaline medium and acidification
cluster conformation. In any case, this additional stabilization with aqueous HCl.6
of 307 kJ mol−1 obtained when dispersion forces are taken
Elemental analyses were performed with
a
Fisons
into account in the level of calculation can be considered 1108 microanalyzer. Melting points were determined with a
argentophilically-driven. A similar result is obtained when Büchi apparatus and are uncorrected. IR spectra (KBr pellets
model B is analysed. In this case the computed energies or Nujol mulls) were recorded on a Bruker IFS66V FT-IR spec-
correspond to the QM (quantum mechanics) part of the trophotometer and are reported in the synthesis section using
ONIOM calculation, which is the same model as model A. The the following abbreviations: vs = very strong, s = strong, m =
1
energy differences arise from the influence exerted by the medium, br = broad. H and 13C NMR spectra were recorded
ligands computed at the UFF level of theory in the MM (mole- in dmso-d6 at room temperature on a Bruker AMX 300
cular mechanics) part of the ONIOM calculation that slightly spectrometer operating at 300.14 and 75.40 MHz, respectively,
change the overall structure. Thus, the computed HF energies using 5 mm o.d. tubes; chemical shifts are reported relative to
show that the HF-optimised structure is 253 kJ mol−1 TMS using the solvent signals (δ 1H = 2.50 ppm; δ 13C =
more stable than the MP2-optimised one. In contrast the MP2 39.5 ppm) as references. Signals are described as follows:
energies show that the MP2-optimised structure displaying chemical shift; multiplicity (s = singlet, d = doublet, m = multi-
argentophilic interactions is 257 kJ mol−1 more stable than the plet, brs = broad singlet); relative integral (for 1H).
HF-optimised one (42.9 kJ mol−1 per Ag⋯Ag interaction
formed).
Synthesis of the compounds
[Ag(H-p-mpspa)] (1). A solution of H2-p-mpspa (0.15 g,
7.1 × 10−4 mol) and triethylamine (0.10 cm3, 7.1 × 10−4 mol) in
Concluding remarks
ethanol (15 cm3) was added to a solution of AgNO3 (0.12 g, 7.1
The complex [HQ][Ag(p-mpspa)] (H2-p-mpspa = 3-(4-methoxy- × 10−4 mol) in acetonitrile (8 cm3). The mixture was stirred for
phenyl)-2-sulfanylpropenoic acid) was prepared by reacting the 1 h and the resulting yellow solid was filtered off, washed with
precursor [Ag(H-p-mpspa)] with diisopropylamine (Q). This ethanol, distilled water and ether. The solid was dried under
complex is hexanuclear and represents one of the very few vacuum. Yield: 87%. Mp: 148 °C (dec.). Elem. Anal. (Found: C
examples of silver(I) complexes containing S-donor ligands for 37.1, H 2.6, S 10.6%. Calc. for C10H9O3SAg: C 37.9, H 2.9, S
which the hexanuclear Ag6 unit adopts a ring structure. The 10.1%). IR (cm−1): 1667 s, 1638 s, ν(CvO); 1384 m, δ(OH);
1
ring is regular, and the silver atoms are placed at short dis- 1254 vs, br, ν(C–O). NMR (dmso-d6): H, δ 13.33 (brs, 1H, C(1)
tances, which suggests an argentophilic interaction. The role OH), 7.75 (s, 1H, C(3)H), 8.09 (d, 2H, C(5)H, C(9)H), 6.87 (d,
of this interaction to define the experimental structural dis- 2H, C(6)H, C(8)H), 3.67 (s, 3H, OCH3); 13C, δ 171.1 C(1), 125.8
position was confirmed using theoretical methods. The calcu- C(2), 139.2 C(3), 128.8 C(4), 133.0 C(5) and C(9), 113.8 C(6) and
lations show that at the HF level of theory, where dispersion C(8), 159.7 C(7), 55.4 C(OCH3).
forces are not included, the Ag6 hexagonal arrangement is kept
[HQ][Ag(p-mpspa)] (2). Diisopropylamine (0.03 cm3, 2.0 ×
but leaving the Ag atoms at very large distances. When dis- 10−4 mol) was added to a suspension of [Ag(H-p-mpspa)]
persion interactions are considered at the MP2 correlated level (0.07 g, 2.0 × 10−4 mol) in ethanol (7 cm3). The mixture was
of theory the six Ag centres are kept closer at short interacting stirred for 24 h and the resulting pale orange solid was filtered
distances. The role of the ligands is evaluated using the off and dried under vacuum. Yield: 73%. Mp: 135 °C (dec.).
ONIOM approach and permits to conclude that the experimen- Elem. Anal. (Found: C 45.6, H 6.0, S 7.4, N 3.3%. Calc. for
tal Ag6 ring structure arises from the argentophilic interactions C16H24O3SNAg: C 45.9, H 5.8, S 7.7, N 3.3%). IR (cm−1): 1617 s,
+
rather than from the possible steric hindrance. The stabiliz- ν(NH2 ); 1545 vs, νasym(COO); 1346 vs, νsym(COO). NMR (dmso-
1
ation energies observed at both levels of theory also support d6): H, δ 7.47 (s, 1H, C(3)H), 8.01 (d, 2H, C(5)H, C(9)H), 6.89
the structural optimizations for models A and B. The MP2-opti- (d, 2H, C(6)H, C(8)H), 3.73 (s, 3H, OCH3), 1.20 (d, 12H,
mized structure for model A bearing six argentophilic inter- [Q]CH3), 3.25 (m, 2H, [Q]CH), 8.91 (s, 2H, [Q]NH2+); 13C,
actions is 307 kJ mol−1 more stable than the HF-optimized one δ 171.2 C(1), C(2), 135.7 C(3), 130.6 C(4), 131.7 C(5) and C(9),
(without Ag⋯Ag interactions) when the energy is evaluated at 113.0 C(6) and C(8), 157.7 C(7), 55.9 C(OCH3), 45.5 CH[HQ],
the correlated MP2 level, where the dispersive forces are taken 19.0 CH3[HQ]. Crystals appropriate for X-ray analysis were
into account.
obtained from a solution in MeOH/dmso.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 5 Crystal data for [HQ][Ag(p-mpspa) (2)
Acknowledgements
Compound
[HQ][Ag(p-mpspa)] (2)
C96H144Ag6N6O18S6
2509.75
Rhombohedric
ˉ
R3
We thank the Dirección Xeral de I+D, Xunta de Galicia, Spain
(IN845B-2010/121), and Ministerio de Economía y Competitivi-
dad, Spain (MEC/FEDER CTQ2010-20500), for financial
support and the Centro de Supercomputación de Galicia
(GESGA) for computational resources.
Empirical formula
M
Crystal system
Space group
a (Å)
15.313(2)
b (Å)
c (Å)
15.313(2)
43.690(10)
A (°)
Β (°)
90
90
References
γ (°)
120
8873(3)
3
1.409
1.138
0.37 × 0.28 × 0.21
1.40–28.07
−13 ≤ h ≤ 20, −20 ≤ k ≤ 17,
−52 ≤ l ≤ 57
19 636
4765 [R(int) = 0.1158]
0.0519, 0.0988
0.1754, 0.1202
V (Å3)
Z
1 (a) N. C. Kasuga, A. Sugie and K. Nomiya, Dalton Trans.,
2004, 3732–3740; (b) I. J. Chopra, Antimicrob. Chemother.,
2007, 59, 587–590; (c) V. Sambhy, M. M. MacBride,
B. R. Peterson and A. Son, J. Am. Chem. Soc., 2006, 128,
9798–9808; (d) O. A. Baranova, N. I. Kuźmin,
T. I. Samsonova, I. S. Rebetskaya, O. P. Petrova,
P. M. Pakhomov, S. D. Khizhnyak, P. V. Komarov and
M. M. Ovchinnikov, Fibre Chem., 2011, 43, 90–103.
2 (a) R. S. Eachus, A. P. Marchetti and A. A. Muenter, Annu.
Rev. Phys. Chem., 1999, 50, 117–144; (b) R. P. Evstigneeva
and V. P. Pchelkin, Pharm. Chem. J., 2006, 40, 211–221;
(c) M. Péter, T. Schüller, F. Furthner, P. A. Rensing,
G. T. van Heck, H. T. M. Schoo, R. Möller, W. Fritzsche,
A. J. J. M. van Breemen and E. R. Meinders, Langmuir,
2009, 25, 5384–5390; (d) K. Kim, Y. M. Lee, H. B. Lee and
K. S. Shin, Biosens. Bioelectron., 2009, 24, 3615–3621.
3 (a) M. C. Gimeno and A. Laguna, Comprehensive Coordi-
nation Chemistry II, Elsevier, Oxford, 2004, vol. 6,
pp. 911–991; (b) D. Fenske, N. Zhu and T. Langetepe,
Angew. Chem., Int. Ed., 1998, 37, 2640–2644;
(c) T. C. W. Mak, X.-L. Zhao, Q.-M. Wang and G.-C. Guo,
Coord. Chem. Rev., 2007, 251, 2311–2333; (d) S. D. Bian,
J.-H. Jia and Q.-M. Wang, J. Am. Chem. Soc., 2009, 131,
3422–3423; (e) J.-H. Liao, H.-W. Chang, H.-Ch. You,
Ch.-Sh. Fang and C. W. Liu, Inorg. Chem., 2011, 50, 2070–
2072; (f) S. C. K. Hau, P.-Sh. Cheng and T. C. W. Mak,
J. Am. Chem. Soc., 2012, 134, 2922–2925.
4 (a) H. Schmidbaur and A. Schier, Chem. Soc. Rev., 2012, 41,
370–412 and references therein (b) R. Pedrido,
M. J. Romero, M. R. Bermejo, M. Martínez-Calvo,
A. M. González-Noya and G. Zaragoza, Dalton Trans., 2009,
8329–8340; (c) M. O. Awaleh, A. Badia and F. Brisse, Inorg.
Chem., 2007, 46, 3185–3191; (d) M. A. Omary, T. R. Webb,
Z. Assefa, G. E. Shankle and H. H. Patterson, Inorg. Chem.,
1998, 37, 1380–1386; (e) E. O’Grady and N. Kaltosoyannis,
Phys. Chem. Chem. Phys., 2004, 6, 680–687; (f) L. Ray,
M. M. Shaikh and P. Ghosh, Inorg. Chem., 2008, 47, 230–
240.
Dc (Mg m−3
)
m (mm−1
)
Crystal size (mm3)
θ range for data collection (°)
Index ranges
Reflections collected
Unique reflections, R
Final R1, wR2 [I > 2σ(I)]
(All data)
Crystallography. Data collection and structure determination
Single crystals of [HQ][Ag(p-mpspa)] (2) were mounted on
glass fibres for data collection in a Bruker Smart CCD auto-
matic diffractometer at 293 K using Mo-Kα radiation (λ =
0.71073 Å). Corrections for Lorentz effects, polarization15 and
absorption16 were made. The structures was solved by direct
methods and refined by full-matrix least squares on F2 using
SHELX-97.17a The contribution from a disordered solvent
molecule (water) on the data was removed by the SQUEEZE
routine of PLATON.17b The crystal data, experimental details
and refinement results are summarized in Table 5.
Theoretical methods
All calculations were performed using the Gaussian 09
package.18 The molecular geometry of model system [Ag6(SH)6]
A was optimized at the HF and MP2 levels of theory, whereas
model system [Ag6(p-mpspa)6]6− B was fully optimized taking
advantage of the ONIOM QM/MM approach,19 which allows
the calculation of the hexanuclear silver-thiolate core at the HF
or MP2 levels while keeping the ligands at a lower compu-
tational cost using the UFF level of calculation. The electronic
correlation effects, keeping the core orbitals frozen, were
included by using 2nd order Møller–Plesset perturbation theory
on both model systems A and B.
The following basis set combination was employed: for
silver, the 19-VE pseudopotential from Stuttgart and the corre-
sponding basis sets20 augmented with two f polarization func-
tions were used.21 The S atoms were treated by Stuttgart
pseudopotentials,22 including only the valence electrons for
each atom. For these atoms double-zeta basis sets from ref. 22
were used, augmented by d-type polarization functions.23 For
the H atom, a double-zeta, plus a p-type polarization function
was used.24
5 (a) E. Barreiro, J. S. Casas, M. D. Couce, A. Sánchez,
R. Seoane, J. Sordo, J. M. Varela and E. M. Vázquez-López,
Dalton Trans., 2007, 3074–3085; (b) E. Barreiro, J. S. Casas,
M. D. Couce, A. Sánchez, R. Seoane, J. Sordo, J. M. Varela
and E. M. Vázquez-López, Eur. J. Med. Chem., 2008, 43,
2489–2497; (c) E. Barreiro, J. C. Casas, M. D. Couce,
A. Laguna, J. M. López de Luzuriaga, M. Monge,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Paper
Dalton Transactions
A. Sánchez, J. Sordo, J. M. Varela and E. M. Vázquez-López, 13 (a) B. Djordjevic, D. Schuster and H. Schmidbaur, Inorg.
Eur. J. Inorg. Chem., 2011, 1322–1332.
6 E. Barreiro, J. S. Casas, M. D. Couce, A. Sánchez, J. Sordo,
Chem., 2005, 44, 673–676; (b) G. Glatz, G. Motz and
R. Kempe, Z. Anorg. Allg. Chem., 2008, 634, 2897–2902.
J. M. Varela and E. M. Vázquez-López, Dalton Trans., 2005, 14 E. Barreiro, J. S. Casas, M. D. Couce, A. Sánchez, J. Sordo,
1707–1715.
J. M. Varela and E. M. Vázquez-López, Cryst. Growth Des.,
2007, 7, 1964–1973.
7 B. Cordero, V. Gómez, A. E. Platero-Plats, M. Revés,
J. Echeverría, E. Cremades, F. Barragán and S. Alvarez, 15 Bruker Analytical Instrumentation, SAINT: Sax Area Detector
Dalton Trans., 2008, 2832–2838. Integration, 1996.
8 (a) P. Pérez-Lourido, J. A. García-Vázquez, J. Romero, 16 G. M. Sheldrick, SADABS, version 2.03, University of Göttin-
A. Sousa, E. Block, K. P. Maresca and J. Zubieta, Inorg. gen, Germany, 2002.
Chem., 1999, 38, 538–544; (b) I. Tsyba, M. B. Bun-kit, 17 (a) G. M. Sheldrick, SHELX-97. Program for the Solution and
R. Bau, R. Noguchi and K. Nomiya, Inorg. Chem., 2003, 42,
8028–8032; (c) L. J. Ashfield, A. R. Cowley, J. R. Dilworth
and P. S. Donnelly, Inorg. Chem., 2004, 43, 4121–4123;
(d) Ch. Yue, Ch. Yan, R. Feng, M. Wu, L. Chen, F. Jiang and
Refinement of Crystal Structures, University of Göttingen,
Germany, 1997; (b) A. L. Spek, PLATON, A Multipurpose
Crystallographic Tool, Utrecht University, Utrecht, The
Netherlands, 2001.
M. Hong, Inorg. Chem., 2009, 48, 2873–2879; (e) R. Silva, 18 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. D. Smith and J. R. Gardinier, Inorg. Chem., 2006, 45,
2132–2142.
9 A. Bondi, J. Phys. Chem., 1964, 68, 441–451.
10 F. H. Allen, Acta Crystallogr., Sect. B: Struct. Sci., 2002, 58,
380–388.
11 (a) I. G. Dance, L. J. Fitzpatrick and M. L. Scudder, Inorg.
Chem., 1984, 23, 2276–2281; (b) W. Petz, B. Neumüller and
I. Z. Krossing, Z. Anorg. Allg. Chem., 2006, 632, 859–865.
12 (a) R. Noguchi, A. Hara, A. Sugie and K. Nomiya, Inorg.
Chem. Commun., 2006, 9, 60–63; (b) R. Hesse and L. Nilson,
Acta Chem. Scand., 1969, 23, 825–845; (c) X. Yin, M.-B. Xie,
W.-G. Zhang, J. Fan and M. Zeller, Acta Crystallogr.,Sect. E:
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
GAUSSIAN 09 (Revision A.1), Gaussian, Inc., Wallingford,
CT, 2009.
Struct.
Rep.
Online,
2007,
63,
m2063–m2064;
(d) W.-G. Zhang, Y. Zhong, M.-Y. Tan, W.-S. Liu and
Ch.-Y. Su, Chin. J. Chem., 2002, 20, 420–423; (e) H. Dietrich,
W. Storck and G. Manecke, J. Chem. Soc., Chem. Commun.,
1982, 1036–1037; (f) C. W. Liu, J. T. Pitts and J. P. Fackler
Jr., Polyhedron, 1997, 22, 3899–3909; (g) M. Shi and
J.-K. Jiang, Chirality, 2003, 15, 605–608; (h) M. Shi, 19 T. Vreven and K. Morokuma, J. Comput. Chem., 2000, 21,
J.-K. Jiang and G.-L. Zhao, Eur. J. Inorg. Chem., 2002, 3264– 1419–1432.
3267; (i) A. Castiñeiras, I. García-Santos, S. Dehnen and 20 D. Andrae, U. Häusserman, M. Dolg, H. Stolland and
P. Sevillano, Polyhedron, 2006, 25, 3653–3660; H. Preuss, Theor. Chim. Acta, 1990, 77, 123–141.
( j) P. A. Pérez-Lourido, J. A. García-Vázquez, J. Romero, 21 P. Pyykkö, N. Runeberg and F. Mendizabal, Chem.–Eur. J.,
M. S. Louro, A. Sousa, Q. Chen, Y. Chang and J. Zubieta, 1997, 3, 1451–1457.
Chem. Soc., Dalton Trans., 1996, 2047–2054; (k) K. Nomiya, 22 A. Bergner, M. Dolg, W. Küchle, H. Stoll and H. Preuss,
S. Takahashi and R. Noguchi, J. Chem. Soc., Dalton Trans., Mol. Phys., 1993, 80, 1431–1441.
2000, 2091–2097; (l) P. C. Zachariadis, S. K. Hadjikakou, 23 S. Huzinaga, Gaussian Basis Sets for Molecular Calculations,
N. Hadjiliadis, A. Michaelides, S. Skoulika, Y. Ming and
Elsevier, Amsterdam, 1984, p. 16.
Y. Xiaolin, Inorg. Chim. Acta, 2003, 343, 361–365.
24 S. Huzinaga, J. Chem. Phys., 1965, 42, 1293–1302.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013