Novel Cu(I)-selective chelators based on a bis(phosphorothioyl)amide scaffold

Article abstract of DOI:10.1021/ic500465z

Bis(dialkyl/aryl-phosphorothioyl)amide (BPA) derivatives are versatile ligands known by their high metal-ion affinity and selectivity. Here, we synthesized related chelators based on bis(1,3,2-dithia/dioxaphospholane-2- sulfide)amide (BTPA/BOPA) scaffolds targeting the chelation of soft metal ions. Crystal structures of BTPA compounds 6 (N^-R₃NH ⁺) and 8 (NEt) revealed a gauche geometry, while BOPA compound 7 (N^-R₃NH⁺) exhibited an anti-geometry. Solid-state ³¹P magic-angle spinning NMR spectra of BTPA 6-Hg(II) and 6-Zn(II) complexes imply a square planar or tetrahedral geometry of the former and a distorted tetrahedral geometry of the latter, while both BTPA 6-Ni(II) and BOPA 7-Ni(II) complexes possibly form a polymeric structure. In Cu(I)-H ₂O₂ system (Fenton reaction conditions) BTPA compounds 6, 8, and 10 (NCH₂Ph) were identified as most potent antioxidants (IC₅₀ 32, 56, and 29 μM, respectively), whereas BOPA analogues 7, 9 (NEt), and 11 (NCH₂Ph) were found to be poor antioxidants. In Fe(II)-H₂O₂ system, IC₅₀ values for both BTPA and BOPA compounds exceeded 500 μM indicating high selectivity to Cu(I) versus the borderline Fe(II)-ion. Neither BTPA nor BOPA derivatives showed radical scavenging properties in H₂O₂ photolysis, implying that inhibition of the Cu(I)-induced Fenton reaction by both BTPA and BOPA analogues occurred predominantly through Cu(I)-chelation. In addition, NMR-monitored Cu(I)- and Zn(II)-titration of BTPA compounds 8 and 10 showed their high selectivity to a soft metal ion, Cu(I), as compared to a borderline metal ion, Zn(II). In summary, lipophilic BTPA analogues are promising highly selective Cu(I) ion chelators.

Full text of DOI:10.1021/ic500465z

Article
pubs.acs.org/IC
Novel Cu(I)-Selective Chelators Based on a
Bis(phosphorothioyl)amide Scaffold
Aviran Amir,†,§ Alon Ezra,†,§ Linda J. W. Shimon,‡ and Bilha Fischer*,†
†Department of Chemistry, Bar Ilan University, Ramat-Gan 52900, Israel
‡Department of Chemical Research Support, The Weizmann Institute of Science, Rehovot 76100, Israel
S
* Supporting Information
ABSTRACT: Bis(dialkyl/aryl-phosphorothioyl)amide (BPA) derivatives are
versatile ligands known by their high metal-ion affinity and selectivity. Here, we
synthesized related chelators based on bis(1,3,2-dithia/dioxaphospholane-2-
sulfide)amide (BTPA/BOPA) scaffolds targeting the chelation of soft metal
ions. Crystal structures of BTPA compounds 6 (N−R3NH+) and 8 (NEt)
revealed a gauche geometry, while BOPA compound 7 (N−R3NH+) exhibited
an anti-geometry. Solid-state 31P magic-angle spinning NMR spectra of BTPA
6-Hg(II) and 6-Zn(II) complexes imply a square planar or tetrahedral
geometry of the former and a distorted tetrahedral geometry of the latter, while
both BTPA 6-Ni(II) and BOPA 7-Ni(II) complexes possibly form a polymeric
structure. In Cu(I)-H2O2 system (Fenton reaction conditions) BTPA
compounds 6, 8, and 10 (NCH2Ph) were identified as most potent
antioxidants (IC50 32, 56, and 29 μM, respectively), whereas BOPA analogues
7, 9 (NEt), and 11 (NCH2Ph) were found to be poor antioxidants. In Fe(II)-
H2O2 system, IC50 values for both BTPA and BOPA compounds exceeded 500 μM indicating high selectivity to Cu(I) versus the
borderline Fe(II)-ion. Neither BTPA nor BOPA derivatives showed radical scavenging properties in H2O2 photolysis, implying
that inhibition of the Cu(I)-induced Fenton reaction by both BTPA and BOPA analogues occurred predominantly through
Cu(I)-chelation. In addition, NMR-monitored Cu(I)- and Zn(II)-titration of BTPA compounds 8 and 10 showed their high
selectivity to a soft metal ion, Cu(I), as compared to a borderline metal ion, Zn(II). In summary, lipophilic BTPA analogues are
promising highly selective Cu(I) ion chelators.
INTRODUCTION
■
Metal-ion chelators containing sulfur atoms have been
suggested as a treatment for Alzheimer’s and Wilson’s diseases,
as well as for heavy-metal poisoning (by soft metal ions such as
Hg(II), Pb(II), and Cd(II)).1 Dimercaptosuccinic acid and
dimercaprol are used to treat heavy-metal poisoning by metal
chelation and secretion through the urine.2 Penicillamine3 and
5′-nucleoside-phosphorthioate derivatives4 have been proposed
for the treatment for Wilson’s and Alzheimer’s disease,
respectively. Related metal chelators containing sulfur atoms
are also important for industrial uses as metal extractors5 and
lubricant additives.6
In the quest for new tools for chelation therapy and industrial
applications that exhibit both high metal-ion affinity and
selectivity, we recently prepared several bisphosphonate
analogues bearing four sulfur atoms. Specifically, we synthesized
O,O′-diester-methylene-diphonsphono-tetrathioate derivative
17 and methylene diphosphonotetrathioate, MDPT, 2, (Figure
1).8 Compounds 1 and 2 were highly effective and selective
soft/borderline metal-ion chelators. For instance, MDPT-
Zn(II) complex was highly stable, log K 10.84, and was 10
000 000 times more stable than MDPT-Ca(II) complex.
MDPT, 2, was found to inhibit, via chelation, the Cu(I)-
catalyzed Fenton reaction, Cu(I)/H2O2, (IC50 26 μM) 2.5
Figure 1. Structures of O,O′-diester-methylenediphosphonotetra-
thioate derivative 1 and methylene diphosphonotetrathioate
(MDPT), 2.
times more potently than the Fe(II)-catalyzed Fenton reaction.
Furthermore, 2 was 8 and 2.5 times more potent than
glutathione (GSH) and EDTA in the Cu(I)/H2O2 system,
respectively.8 Analogue 1, R = benzyl, proved to be a potent
antioxidant (IC50 53 μM), inhibiting, via chelation, the Cu(I)-
catalyzed Fenton reaction at a 4-fold lower concentration than
GSH. Analogue 1, R = CH3(CH2)7, inhibited the Fe(II)-
catalyzed Fenton reaction at about the same concentration as
ascorbic acid (IC50 83 vs 93 μM).7
Received: March 4, 2014
Published: July 17, 2014
© 2014 American Chemical Society
7901
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
The related metal-ion chelators, bis(dialkyl/aryl-
phosphorothioyl)amide (BPA) analogues, 3a (Figure 2),
by BTPA/BOPA derivatives; investigation of their antioxidant
properties, and evaluation of their metal-ion selectivity.
RESULTS AND DISCUSSION
■
Design of Metal Ion Chelators Based on Bis-
(phosphorothioyl)amide Scaffold. We synthesized bis-
(phosphorothioyl)amide analogues bearing dithiaphospholane
and dioxaphospholane rings 6−11 with a view to develop novel
metal ion-selective chelators (Figure 4). Our design of the new
chelators involved modification of three BPA moieties: (1)
Addition of two sulfur atoms or two oxygen atoms on the
phosphorus atom, aiming to tune the “soft” nature of the
phosphorothioate moiety, and hence its metal-ion selectivity.
(2) Inclusion of the phosphorothioate or phosphorotrithioate
moiety into a five-membered ring, aiming to reduce degrees of
freedom of the phosphorothioate moiety and help the correct
orientation of the metal-coordinating atoms.24,25 (3) Substitut-
ing ethyl or benzyl groups on the bridging nitrogen atom (or
having a bridging nitrogen anion), aiming to tune solubility and
to add M2+ coordination sites (in the cases of benzyl and
nitrogen anion as in 6/7).
Synthesis of BTPA/BOPA Derivatives 6−11. The
synthesis of derivatives 6−11 was achieved by reaction of 2
equiv of 2-chloro-1,3,2-dithia(dioxa)phospholane 12a,b with
various amines in tetrahydrofuran (THF) at −78 °C to room
temperature (RT), to form the dithia(dioxa)phospholane
derivatives 13a,b (Scheme 1). The formation of intermediates
13a,b was verified by 31P/1H NMR. Intermediates 13a,b were
then treated with elemental sulfur, S8, in pyridine at −78 °C to
obtain the oxidized bis(phosphorothioyl)amide analogues 6−
11.
Compounds 6 and 7 were obtained as pure white solids upon
filtration of the reaction mixtures and crystallization of the
yellowish oily residue in hot EtOH. Compound 6 was obtained
at a 69% yield (31P NMR: 75.1 ppm), and compound 7 was
obtained at a 91% yield (31P NMR: 69.6 ppm). Compounds 8
and 9 were crystallized from CHCl3/n-hexane to give yellowish
solids that, after successive washings with CS2, provided pure
white solids. Compounds 8 and 9 were obtained at a 77% yield
(31P NMR: 102.8 ppm) and a 45% yield (31P NMR: 87.1 ppm),
respectively. Compound 10 was crystallized from CHCl3 to
give a white solid at a 60% yield (31P NMR: 108 ppm).
Compound 11 was obtained upon separation over silica gel
column with hexane/EtOAc to give a white solid at a 30% yield
(31P NMR: 86.4 ppm).
Figure 2. General structure of bis(dialkyl/aryl-phosphorothioyl)amide
(BPA) derivatives, 3a, and general structure of BPA-M2+ complex, 3b.
allow delocalization of electrons and therefore are expected
to be more polarizable than the corresponding bisphosphonate
analogues.9 BPA derivatives are soft/borderline metal ligands
that bind a variety of metal ions (e.g., Zn(II),10 Ni(II),11,12
Cu(I)/(II),13,14 and Mn(II)15,16). These compounds are
versatile ligands, displaying a broad variety of coordination
patterns leading to a great diversity of molecular and
supramolecular structures.17 Anions of BPA derivatives were
usually found to be coordinated to a metal center by both sulfur
atoms in a 2:1 ligand-to-metal ratio, thus leading to two six-
membered MS4 inorganic chelate rings, 3b (Figure 2).16
The geometry and stoichiometry of various neutral BPA
ligand complexes with metal ions were studied.16 For instance,
the Zn(II) complex of BPA adopts a distorted tetrahedral
geometry with a 2:1 ligand/metal ion ratio, 4,10 and BPA-Cu(I)
2−
complex, formed due to a redox reaction within (BPA)2
-
Cu(II) complex,13,18 adopted multinuclear structures such as
[Cu4L3]+ and [Cu3L3], 5a and b,9,19 respectively (Figure 3).
The affinity of bis(dialkyl/aryl-phosphorothioyl)amide to
metal ions is used for metal extraction;9,20 design of new
materials, for example, long-lived, highly luminiscent lantha-
nide−imidodiphosphinate complexes with potential applica-
tions in photonic devices and sensors;21 and treatment of
antifungal22 and antimicrobial infections.23
Here, we targeted to improve the soft metal-ion selectivity of
the metal-ion chelator, 1, by replacing its methylene bis-
(phosphonothioyl) scaffold by a bis(phosphorothioyl)amide
moiety. Specifically, we synthesized a novel family of metal-ion
chelators based on bis(1,3,2-dithia/dioxaphospholane-2-
sulfide)amide (BTPA/BOPA) scaffolds, analogues 6−11. We
report on a novel synthetic approach for the preparation of
these BTPA/BOPA chelators; X-ray crystal structures of the
novel compounds; determination of geometries of BTPA/
BOPA-M(II) complexes; evaluation of metal-ion binding mode
X-ray Crystal Structures of BTPA/BOPA Analogues 6,
7, and 8. Single crystals of analogues 6 and 7 were obtained by
slow perfusion of ether into dichloromethane (DCM) solutions
Figure 3. Examples of BPA derivatives forming metal complexes with various geometries: Zn(bis(di-iPr-phosphorothioyl)amide)2, 4,
Cu4((Ph2(S)P)2N)3, 5a, and Cu3((Ph2(S)P)2N)3, 5b.
7902
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
Figure 4. Structures of BPTA and BOPA Analogues 6−11.
between the two dithiaphospholane rings. The average PS
and P−C bond lengths were 2.044 and 1.827 Å, respectively,
longer than PS and P−N bonds in 6−8 due to the lack of
electron delocalization in 14.
Scheme 1. Synthesis of BPTA and BOPA Analogues 6−11
In both crystal structures of PNP-based analogues 6 and 8
and the corresponding PCP-based analogue 14, the dithio-
phospholane rings dictated gauche configuration, while in the
PNP-based compound 7, dioxaphospholane rings forced an anti
conformation. The difference of conformation between 6 and 7
is possibly due to the larger dipole in P−O versus P−S bond
(Δ electronegativity 1.3 vs 0.3, respectively), resulting in dipole
repulsion and an anti conformation in compound 7.
To study metal-ion chelation by BTPA and BOPA analogues,
we first formed metal complexes of 6 and 7 with Zn(II),
Hg(II), and Ni(II) in ethanol at a 2:1 ligand-to-metal ion ratio.
Complexes of compound 6 with Zn(II), Hg(II), or Ni(II)
precipitated from ethanolic solution. Hg(II) and Ni(II)
complexes were insoluble in any organic and aqueous media.
Addition of Hg(II) or Ni(II) to compound 7 resulted in a black
paste or a brownish solid, respectively. A Zn(II) complex with 7
was soluble in ethanol. Solid complexes of chelators 6 or 7 with
Hg(II), Ni(II), or Zn(II), were investigated by solid-state NMR
spectroscopy applying the magic-angle spinning (MAS)
method. First, a 31P-MAS NMR spectrum of the free chelator
6 was measured (Figure 7a). The chelator isotropic (iso)
signals are found at 69.4 ppm, J = 255 Hz, in addition to six
aniso signals. In the 31P-MAS NMR spectrum of 6-Hg(II)
complex (Figure 7b) the iso signals move downfield to 83.4
ppm, and the iso signals are in 1:1 ratio that may be attributed
to a square planar26 or tetrahedral geometry.9 The two aniso
signals are barely seen due to the micro crystalline solid and the
of compounds 6 and 7. Compound 8 was crystallized from
chloroform by slow evaporation of the solvent. The structures
of compounds 6, 7, and 8 were solved by X-ray diffraction.
Compound 6 exhibited gauche conformation between P1S1/
P2S4 moieties (Figure 5A), consistent with that reported for
iPr2P(S)NHP(S)iPr2.10 The average PS and P−N bond
lengths of 6 were 1.960 and 1.593 Å, respectively. Likewise,
compound 8 exhibited gauche conformation between P1S3/
P2S6 (Figure 5B), and the average PS and P−N bond
lengths were 1.938 and 1.698 Å, respectively. Surprisingly, the
crystal structure of compound 7 indicated an anti-conformation
with P1S1/P2S2 moieties inverted to each other (Figure
5C). The average P−N bond length for compounds 6 and 7
was 1.587 Å, while for compound 8 it was 1.698 Å. These
differences are attributed to electron delocalization in
compounds 6 and 7 over SPNPS atoms. Selected bond lengths
and angles are listed in Table 1.
Previously we reported the crystal structure of the related
PCP-based compound methylene-bis(1,3,2-dithiaphospholane-
2-sulfide), 14 (Figure 6).7 A gauche conformation was observed
Figure 5. Structures of BTPA and BOPA analogues (A) 6, (B) 8, and (C) 7, shown in a Newman projection (50% ellipsoid probability). The
counter cations of compounds 6 and 7 were omitted for clarity.
7903
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
Table 1. Selected Bond Lengths and Angles of Compounds 6, 7, and 8
compound
atoms
length [Å]
atoms
angle [deg]
6
6
6
6
7
7
7
7
8
8
8
8
S(1)−P(1)
P(1)−S(3)
P(1)−S(2)
P(1)−N(1)
S(1)−P(1)
P(2)−O(3)
P(1)−O(2)
P(1)−N(1)
P(1)−S(3)
S(1)−P(1)
P(1)−S(2)
P(1)−N(1)
1.9635(13)
2.1025(13)
2.1158(13)
1.597(3)
S(1)−P(1)−S(3)
S(1)−P(1)−N(1)
S(2)−P(1)−S(3)
S(5)−P(2)−S(6)
S(1)−P(1)−O(1)
S(2)−P(2)−O(3)
O(1)−P(1)−O(2)
O(3)−P(2)−O(4)
S(2)−P(1)−N(1)
S(1)−P(1)−S(3)
S(2)−P(1)−S(3)
S(5)−P(2)−S(6)
113.83(5)
111.0(1)
98.52(5)
97.96(6)
112.54(8)
110.8(8)
96.6(1)
1.9631(10)
1.621(2)
1.6113(19)
1.572(2)
96.3(1)
1.9410(5)
2.0768(5)
2.1085(5)
1.7124(13)
112.74(5)
110.16(2)
119.73(2)
112.16(2)
Figure 6. (A) Methylene-bis(1,3,2-dithiaphospholane-2-sulfide), 14. (B) X-ray crystal structure of 14.7
Figure 8. 31P-MAS NMR spectra at 202 MHz. (a) Compound 6. (b)
6-Ni(II) complex.
Figure 7. 31P-MAS NMR spectra at 202 MHz. (a) Compound 6. (b)
6-Hg(II) complex. (c) 6-Zn(II) complex.
BTPA Analogues Prefer Cu(I) Versus Fe(II) Chelation.
We indirectly studied Cu(I) and Fe(II) chelation by BTPA and
BOPA analogues, by assessing their antioxidant capacity in
Fenton reactions. Specifically, we monitored, using electron
paramagnetic resonance (EPR), their ability to modulate OH•
radical formation from H2O2 in Fe(II)/Cu(I)-induced Fenton
reaction. Because of insolubility of these compounds in water,
all analogues were dissolved in dimethyl sulfoxide (DMSO). α-
Phenyl N-tertiary-butyl nitrone (PBN) was used as a radical
spin trap.27 The hydroxyl radical generated by Fenton reaction
is converted into a methyl radical via its reaction with DMSO,
which undergoes oxidation under aerobic conditions to
complex geometry dictated by the metal ion. Zn(II)-6 complex
exhibited two iso signals at 83.1 ppm in a 3:1 ratio, which is
characteristic of a distorted tetrahedral geometry (Figure 7c).16
Changes in the dithiaphosphalane rings due to M(II)
coordination were observed by 13C-MAS NMR as well
(Supporting Information, Figure S1).
The 31P-MAS NMR spectrum of 6-Ni(II) exhibits broad
signals (Figure 8b), unlike the narrow signals of spectra of 6-
Zn(II) and 6-Hg(II) ion complexes. The broad aniso signals
may possibly be attributed to a polymeric structure.
Similarly to the case of 6, the 31P-MAS NMR spectrum of
compound 7 (Supporting Information, Figure S2) exhibits iso
peaks at 74 ppm, J = 1250 Hz, and an aniso pattern, and the
spectrum of 7-Ni(II) complex exhibits very large anisotropy
signals, possibly due to a polymeric structure.
27
•
produce methoxy radical OCH3 . The long-lived PBN/
•
OCH3 adduct was then detected by EPR. The addition of
chelators to Fe(II)/Cu(I)-H2O2 mixture lowered the PBN/
•
OCH3 signal due to metal-ion chelation and/or radical
7904
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
scavenging (Figure 9).28 The inhibition of radical production
by compounds 6−11, expressed in IC50 values, was compared
“softer” character than dioxaphospholane compounds 7, 9, and
11, due to the presence of six sulfur atoms versus two sulfur
atoms in the latter compounds. An appealing explanation for
the reduced activity of 7, 9, and 11 is that the electronegative
oxygen atoms, unlike sulfur atoms, withdraw electrons from the
SPNPS chelating system to the dioxaphosphospholane rings.
The poor electron availability of the SPNPS moiety reduces
chelation of metal ions. Yet, the presence of oxygen atoms
significantly affects also the conformation of the BOPA
analogue (Figure 5). In dioxaphospholane analogue 7, the
sulfur atoms are arranged in an anti-conformation, thus making
the Cu(I) chelation less effective as compared to gauche
conformation in dithiaphospholane 6. Indeed, compound 6 was
5 times more potent than compound 7 (32 vs 160 μM,
respectively) although both share the same charge delocaliza-
tion on the SPNPS moiety.
Substitution of a benzyl group on the central BTPA nitrogen
atom resulted in the best antioxidant studied here, compound
10, being 3-fold more active than EDTA (29 vs 86 μM).
Compound 10 was a 2-fold more potent antioxidant than
compound 8 possibly due to additional π-cation interactions
between the aromatic ring and Cu(I).
Figure 9. Inhibition of OH radical formation in Cu(I)/H2O2 system
by derivatives, 6, 7, 8, and 10 vs EDTA, as reflected by the relative
amount of OCH3 .
•
to the inhibitory effect of a metal-ion chelator, ethyl-
Interestingly, in the Fe(II)-H2O2 system, we could not
determine IC50 values for BTPA/BOPA compounds because
the minimal amount of radical production exceeds 50% even at
500 μM BTPA/BOPA analogue. On the other hand, in Cu(I)-
H2O2 system all BTPA analogues inhibited OH radical
production. This observation indicates the clear selectivity of
BTPA compounds to Cu(I) versus the borderline Fe(II) ion.
BTPA Analogues Are Cu(I) Chelators and Not Radical
Scavengers. Since Fenton reaction can be inhibited by both
metal-ion chelation and/or radical scavenging, we also
evaluated analogues 6−11 as potential radical scavengers. We
produced OH radicals by photolysis of H2O2 with UV
radiation29 and monitored radical scavenging by compounds
6−11 using EPR. The radical scavenging activity was
determined as a function of the antioxidant concentration
and was calculated relative to a control measurement, which is
the maximal production of OH radical via photolysis of H2O2,
in the absence of a scavenger. The experiment conditions were
the same as for the above-mentioned Fenton reaction inhibition
experiment except for the absence of the metal ion. The
mixture was irradiated (at 365 nm) for 2 min with a UV lamp
followed by EPR measurement. The inhibition of radical
enediaminetetraacetic acid (EDTA) (Table 2).
Table 2. Inhibition of Fe(II)/Cu(I)-Induced Fenton
Reaction by Analogues 6−11 Versus EDTA
a
IC50 [μM]
compound
Fe(II)
Cu(I)
6
>500 μM
>500 μM
>500 μM
>500 μM
>500 μM
>500 μM
32
1
7
160
56
6
8
4
9
>500 μM
29 0.8
>500 μM
10
11
EDTA
77
1
86
4
a
IC50 values represent the compound concentration that inhibits 50%
•
of the OCH3 radical amount produced in the control reaction. The
latter includes Cu(I)/Fe(II)-H2O2 in DMSO in the presence of PBN.
The best antioxidant activity was measured for the
dithiaphospholane compounds 6, 8, and 10, with IC50 values
of 32, 56, and 29 μM, respectively (Table 2). These results
meet the expected, since those compounds have a much
Figure 10. 31P NMR spectra in DMSO-d6 of (A) 8-Zn(II) complex and (B) 10-Cu(I) complex at 241 MHz.
7905
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
production by compounds 6−11 was compared to the
inhibitory effect of the known radical scavenger Trolox.30
All compounds showed no radical scavenging activity, even at
high ligand concentration (>500 μM). This finding indicates
that the mechanism of action of BTPA/BOPA analogues as
antioxidants involves chelation, namely, capturing the metal ion
and preventing electron transfer from Cu(I), thus inhibiting the
production of hydroxyl/methoxyl radicals.
Unlike BOPA analogues, BTPA analogues 6, 8, and 10 were
found to be potent antioxidants inhibiting OH radical
production in Cu(I)-H2O2 Fenton system (IC50 32, 56, and
29 μM, respectively). Both BTPA and BOPA analogues 6−11
were extremely poor antioxidants in Fe(II)-H2O2 system. None
of the derivatives 6−11 showed radical scavenging properties,
which led us to conclude that inhibition of OH radical
production occurred predominantly through metal-ion chela-
tion. The selectivity of BTPA to Cu(I) versus Fe(II) exceeded
20-fold. The reduced antioxidant activity of BOPA analogues 7,
9, and 11 is possibly due not only to oxygen atoms, which
withdraw electrons from the SPNPS chelating system, but also
to an anti conformation of the PS moieties, thus making
metal-ion chelation less effective versus dithiaphospholane-
sulfide derivatives 6, 8, and 10 arranged in a gauche
conformation.
Solution NMR-monitored Cu(I) and Zn(II) titration of
compounds 8 and 10 revealed high selectivity for a soft metal
ion, Cu(I), versus a borderline metal ion, Zn(II), with the
former being directly coordinated via a (R2S)PS moiety, and
the latter possibly via bridging water molecules. Although
sharing similar potency (IC50 values) with MDPT, 2,
compound 10 proved to be by far more selective to Cu(I)
versus Fe(II) (≫ 20-fold vs 2.5-fold). Likewise, BTPA
analogues were found to be more effective Cu(I)- chelators
than the related lipophilic MDPT−diester chelators, 1.7
Furthermore, MDPT is prone to air-oxidation,8 whereas
BTPA is highly stable under these conditions.
BTPA Analogues 8 and 10 Chelate Preferentially Cu(I)
Versus Zn(II). The above EPR data showed that BTPA
analogues 6, 8, and 10 were most potent and Cu(I)-selective
antioxidants acting by Cu(I)-chelation rather than radical
scavenging. To establish the mode of borderline versus soft
metal-ion chelation by these analogues, we titrated compound 8
with either Zn(II) or Cu(I) ions and monitored the titration by
31P/1H NMR in DMSO-d6.
Upon the addition of ZnCl2 (dissolved in D2O) the 31P
NMR signal of 8 at 104 ppm neither shifted nor broadened
significantly. The maximal shift measured upon the addition of
2 equiv of ZnCl2 was 0.3 ppm (Figure 10A). Hence, it appears
that there is an indirect chelation of Zn(II) by the ligand,
1
causing an insignificant shift in 31P NMR spectrum. H NMR
showed no change even upon the addition of up to 2 equiv of
Zn(II).
We repeated the above NMR-monitored metal-ion titration,
this time with (Cu(I)(CH3CN)4PF6) solution in CD3CN.
Addition of up to 0.1−0.5 equiv of Cu(I) neither changed the
chemical shift of phosphor and proton signals of analogue 8 nor
broadened those signals (Supporting Information, Figure S3).
However, upon addition of 1 equiv of Cu(I), the phosphor
signal shifted from 104.2 to 105.5 ppm, and its width
broadened dramatically.
In summary, in this study, we successfully developed a novel
class of stable, lipophilic BTPA chelators that are highly Cu(I)-
selective versus borderline ions such as Zn(II) and Fe(II).
Next, we titrated compound 10 with Cu(I) (Figure 10B).
Again, no change was observed in the 1H NMR spectra;
however, in the 31P NMR spectra the phosphor signal shifted
and broadened, even upon addition of 0.1 equiv of Cu(I). After
an addition of 0.5−2 equiv of Cu(I), the signal broadened and
shifted from 109.5 to 111.5 ppm, a more significant change
than that observed for compound 8 (Supporting Information,
Figure S3). Yet, the shift of chemical shift, Δδ, of compound 10
is still small as compared to that observed for a related BPA
EXPERIMENTAL SECTION
■
General. All air- and moisture-sensitive reactions were carried out
in flame-dried, N2-flushed, two-neck flasks sealed with a rubber
septum, and the reagents were introduced with a syringe. All
commercial reagents were used without further purification, unless
otherwise noted. All reactants in moisture-sensitive reactions were
dried overnight in a vacuum oven. Diisopropyl ethyl amine (DIPEA)
was distilled from CaH2. Pyridine was distilled from KOH. THF was
distilled from sodium in the presence of benzophenone, and benzyl
amine was distilled from NaOH. Cu(CH3CN)4PF6 was purified before
use by dissolving the salt in acetonitrile (HPLC grade) and filtering the
unsoluble Cu(II) salt. The filtrate was flushed with argon stream. The
concentration of the Cu(I) salt was determined by UV spectroscopy
by the addition of a specific Cu(I) indicator, bicinchoninic acid (BCA)
(ε562 = 7700 M−1).31 Progress of reactions was monitored by thin-
layer chromatography on precoated Merck silica gel plates (60F-254).
10
analogue, Pr2(S)PNHP(S)iPr2, where a Δδ value of ca. 30
ppm was observed upon coordination of Zn(II). This
phenomenon may be because this BPA analogue is negatively
charged, whereas analogues 8 and 10 are neutral ligands.
Thus, solution NMR data indicated the preference of 8 for
Cu(I) versus Zn(II) and that chelation of Cu(I) ion occurred
through the (R2S)PS moiety, whereas Zn(II) coordination
occurred possibly via bridging water molecules. In addition, in
compound 10, Cu(I) is possibly coordinated also by the phenyl
group via π-cation interactions.
i
Visualization was accomplished by UV light. H NMR and 31P NMR
1
spectra were measured with Bruker DPX-300 (300 MHz for 1H), AC-
200 (80.3 MHz for 31P), or DMX-600 (600 and 240.9 MHz for H
1
and 31P, respectively) spectrometers. High-resolution mass spectra
(HR-MS) were recorded on an AutoSpec-E FISION VG mass
spectrometer by chemical ionization and a high-resolution matrix-
assisted laser desorption ionization time-of-flight mass spectrometer
(MS-MALDI-TOF) with autoflex TOF/TOF instrument (Bruker,
Germany). Analysis of the amount of OH radicals produced in Cu(I)
and Fe(II)-H2O2/tested compound systems was performed by
solution EPR spectroscopy using a Bruker ER 100d X-band
spectrophotometer. The spectra were processed by Bruker WIN-
EPR software version 2.11. The signal intensities were evaluated by
integration of the second signal of the PBN−OCHH3 spin adducts and
reported as percentage of the control (the signal intensity obtained
without the addition of tested compound).
CONCLUSIONS
■
Bis(1,3,2-dithia/dioxaphospholane-2-sulfide)amide (BTPA/
BOPA) analogues 6−11 were synthesized via a new and
efficient synthetic methodology. Crystal structures of BTPA
compounds 6 and 8 revealed a gauche geometry, while BOPA
compound 7 exhibited an anti geometry. Solid-state 31P-MAS
NMR spectra of 6-Hg(II) and 6- Zn(II) complexes imply a
square planar or tetrahedral geometry of the former and a
distorted tetrahedral geometry of the latter. The 31P-MAS
NMR spectra of both 6-Ni(II) and 7-Ni(II) complexes suggest
a polymeric structure.
Typical Procedure for the Synthesis of N-ethyl-N-isopropyl-
propan-2-aminium-bis(2-sulfido-1,3,2-dithiaphospholan-2-yl)-
7906
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
amide, 6, and N-ethyl-N-isopropylpropan-2-aminium-bis(2-
sulfido-1,3,2-dioxaphospholan-2-yl)amide, 7. 2-Chloro-1,3,2-di-
thiaphospholane, 12a, or 2-chloro-1,3,2-dioxaphospholane, 12b,
(11.23 mmol) in dry THF (4 mL) was cooled to −78 °C. Dry
DIPEA (1.45 g, 11.23 mmol) was added dropwise so that reaction
temperature did not exceed −45 °C, followed by a slow addition of 0.5
M NH3 solution in THF (45 mL, 22.46 mmol). After 0.5 h at −78 °C,
the temperature was raised to RT for another 0.5 h. 31P NMR analysis
of the reaction mixture indicated that the reaction was completed. The
mixture was cooled to −78 °C, and dry DIPEA (1.45 g, 11.23 mmol)
was added followed by a slow addition of 12a or 12b (11.23 mmol).
The reaction was stirred at RT overnight. 31P NMR showed that the
reaction was completed. The mixture was cooled to −78 °C, dry
pyridine (8 mL) and S8 (0.77 g, 24.06 mmol) were added, and the
temperature was raised to RT. After an additional 2 h, 31P NMR
indicated the formation of compound 6 or 7. The mixture was vacuum
filtered, and the filtrate was evaporated to give yellowish oil. The oil
was crystallized by dissolving it in hot EtOH. The clear mixture was
cooled to RT and then cooled to −78 °C to obtain a white solid. The
solid was filtered, washed with cold EtOH, and dried under vacuum to
give a white solid. Compound 6 was obtained at a 69% yield (3.52 g).
mmol), in dry THF (8 mL), was cooled to −78 °C. Dry DIPEA
(2.41 g, 18.66 mmol) was slowly added followed by the addition of dry
benzylamine (1 g, 9.33 mmol). The mixture was stirred at RT
overnight. After 24 h pyridine (8 mL) and S8 (0.657 g, 20.52 mmol)
were added to the mixture at −78 °C to obtain a light yellow turbid
suspension. The mixture was allowed to reach RT, and after 2 h the
mixture was vacuum filtered. The filtrate was evaporated to give
yellowish oil that was dissolved in DCM (20 mL) and washed with
water and a saturated solution of NaHCO3 and NaCl. Finally, the
solution was dried over MgSO4 and evaporated to give yellowish oil.
The organic phase was crystallized from CHCl3 to give a white solid in
1
60% yield (2.3 g). mp 150 °C. H NMR (400 MHz, CDCl3): 7.46−
7.36 (m, 5H), 5.01 (t, J = 20 Hz, 2H), 3.62−3.43 (m, 8H) ppm. 13C
NMR (100 MHz, CDCl3): 137.0, 128.4, 127.6, 127.4, 53.6, 41.9 ppm.
31P NMR (162 MHz, CDCl3): 108.03 ppm. HR-MS (Cl) m/z calcd.
for C11H16NP2S6 [M + H]+: 415.908, found: 415.911. FT-IR (ATR):
2963w, 1494w, 1448s, 1412w, 1353w, 1281w, 1244w, 1149m, 1023s,
1001m, 938m, 737s cm−1.
2,2′-(Benzylazanediyl)bis(1,3,2-dioxaphospholane-2-sul-
fide), 11. 2-Chloro-1,3,2-dioxaphospholane, 12b (2.13 g, 16.84
mmol), in dry THF (8 mL), was cooled to −78 °C. Dry DIPEA
(1.98 g, 15.30 mmol) was slowly added followed by the addition of dry
benzylamine (0.82 g, 7.65 mmol). The mixture was stirred at RT
overnight. After 24 h pyridine (8 mL) and S8 (0.539 g, 16.83 mmol)
were added to the mixture at −78 °C to obtain a light yellow turbid
suspension. The mixture was allowed to reach −15 °C, and after 2 h
the mixture was vacuum filtered. The filtrate was evaporated to give
yellowish oil that was dissolved in DCM (20 mL) and washed with
water and a saturated solution of NaHCO3 and NaCl. Finally, the
solution was dried over Na2SO4 and evaporated to give yellowish oil.
The crude product was separated over silica gel column chromatog-
raphy, hexane/EtOAc (60%:40%), to give a white solid in 30% yield
1
mp 146 °C. H NMR (600 MHz, CDCl3): δ 3.95 (sep, J = 6.7 Hz,
2H), 3.74−3.64 (m, 4H), 3.64−3.57 (m, 4H), 3.29 (q, J = 7.61 Hz,
2H), 1.53−1.43 (m, 15H) ppm. 13C NMR (150 MHz, CDCl3): δ 54.9,
42.9, 41.6, 19.3, 18, 12.2 ppm. 31P NMR (243 MHz, CDCl3): δ 75.1
ppm. HR-MS (MALDI) m/z calcd for C4H8NP2S6 [M − C8H20N]−:
323.846, found: 323.848. Fourier transform infrared (FT-IR)
attenuated total reflection (ATR): 2984w, 2912w, 2691w, 1457w,
1414w, 1177s, 1143m, 938w, 614vs cm−1. Compound 7 was obtained
at a 91% yield (4 g). mp 105 °C. 1H NMR (600 MHz, CDCl3): 4.34−
4.24 (m, 8H), 3.94 (sep, J = 6.7 Hz, 2H), 3.32 (q, J = 7.44 Hz, 2H),
1.52−1.46 (m, 15H) ppm. 13C NMR (150 MHz, CDCl3): 63.5, 55.1,
43.0, 18.7, 12. 31P NMR (243 MHz, CDCl3): 69.6 ppm. HRMS
(MALDI) m/z calcd for C4H10NO4P2S2 [M + H − C8H19N]+:
261.952, found: 261.951. FT-IR (ATR): 2974w, 2699w, 1466w,
1235m, 1024s, 917m, 807m, 761m, 664m cm−1.
2,2′-(Ethylazanediyl)bis(1,3,2-dithiaphospholane-2-sulfide),
8. 2-Chloro-1,3,2-dithiaphospholane, 12a (3.06 g, 19.29 mmol), in dry
THF (8 mL), was cooled to −78 °C. Dry DIPEA (2.49 g, 19.29
mmol) was slowly added followed by the addition of 2 M ethylamine
in THF (0.44 g, 9.65 mmol). The mixture was stirred at RT overnight.
31P NMR of the reaction mixture indicated the formation of 13a (31P
1
(0.8 g). mp 145 °C. H NMR (600 MHz, CDCl3): 7.48−7.26 (m,
5H), 4.97 (t, J = 16.8 Hz, 2H), 4.42−4.22 (m, 8H) ppm. 13C NMR
(150 MHz, CDCl3): 137.8, 128.5, 127.9, 127.8, 66.5, 53.4 ppm. 31P
NMR (243 MHz, CDCl3): 86.43 ppm. HR-MS (Cl) m/z calcd. for
C11H16NO4P2S2 [M + H]+: 351.999, found: 351.996. FT-IR (ATR):
2970w, 2905w, 1494s, 1457w, 1317w, 1230w, 1202w, 1031m, 1011m,
1001s, 924m, 822s, 791s, 773s, 734s, 695s, 651m, 625m cm−1.
Evaluation of Inhibition of OH Radical Production in Fenton
Reactions by BPA Analogues 6−11 Using EPR. EPR settings for
OH radical detection were as follows: microwave frequency, 9.76
GHz; modulation frequency, 100 kHz; microwave power, 6.35 mW;
modulation amplitude, 1.2 G; time constant, 655.36 ms; sweep time,
83.89 s; and receiver gain, 2 × 105 in experiments with Cu(I) and
Fe(II).
NMR: 106 ppm). Pyridine (8 mL) and S8 (1.32 g, 41.25 mmol) were
added to a mixture of 13a at −78 °C, the mixture was allowed to reach
RT, and after 2 h the mixture was vacuum filtered. The filtrate was
evaporated to give yellowish oil that was dissolved in DCM (20 mL)
and washed with water and a saturated solution of NaHCO3 and NaCl.
Finally, the solution was dried over Na2SO4 and evaporated to give
yellowish oil. The oil was crystallized from CHCl3/n-hexane to give a
yellowish solid that, after successive washings with CS2, provided a
pure white solid, which was dried in air. Compound 8 was obtained in
Fenton Reaction Monitored by EPR Using PBN as a Radical
Spin Trap. Cu(CH3CN)4PF6 (2 mM) in acetonitrile (10 μL) or 0.5
mM FeCl2·4H2O in DMSO (10 μL) were added to 5−500 mM tested
compound in DMSO (10 μL). Afterward, DMSO (60−70 μL) was
added to the mixture. After mixing for 2 s, 50 mM PBN (10 μL) in
DMSO (1 mL) was quickly added followed by the addition of 100
mM H2O2 (10 μL). Each EPR measurement was performed 150 s after
the addition of H2O2. All experiments were performed at room
temperature, in a final volume of 100 μL.
Photolysis of H2O2. EPR settings for OH radical detection were as
described above. The sample, composed of 5−500 μM tested
compound (1−10 μL), was followed by an addition of DMSO (70−
80 μL). 50 mM PBN (10 μL) was quickly added followed by the
addition of 100 mM H2O2 (10 μL) to the mixture. A short mixing of
the sample was performed after the addition of each component. The
sample was then irradiated for 2 min with a UV lamp (model VL-315
BL, wavelength: 365 nm, power: 30 W, Vilber Lourmat, Marne-la-
1
77% yield (2.63 g). mp 154 °C. H NMR (600 MHz, acetone-d6):
3.9−3.7 (m, 10H), 1.04 (t, J = 7.1 Hz, 3H) ppm. 13C NMR (150 MHz,
acetone-d6): 47.7, 42.6, 16.8. 31P NMR (243 MHz, acetone-d6): 102.8
ppm. HR-MS (MALDI) m/z calcd. for C6H14NP2S6 [M + H]+:
353.892, found: 353.889. FT-IR (ATR): 2989w, 2952w, 2922w,
1438w, 1409w, 1290w, 1147w, 1031m, 925m, 861m, 748m, 663s
cm−1.
2,2′-(Ethylazanediyl)bis(1,3,2-dioxaphospholane-2-sulfide),
9. The reaction was performed following the above-mentioned
procedure for the synthesis of compound 8. The product was
obtained from 12b (1.42 g, 11.23 mmol) as a white powder in 45%
1
yield (0.73 g). mp 105 °C. H NMR (600 MHz, CDCl3): 3.02−2.96
(m, 10H), 1.19 (t, J = 7.1 Hz, 3H) ppm. 13C NMR (150 MHz,
CDCl3): 65.93, 37.02, 16.8 ppm. 31P NMR (243 MHz, CDCl3): 87.16
ppm. HR-MS (MALDI) m/z calcd. for C6H14NO4P2S2 [M + H]+:
289.984 found 290.951. FT-IR (ATR): 2989w, 2952w, 2922w, 1438w,
1409w, 1290w, 1147w, 1031m, 925m, 861m, 748m, 663s cm−1.
2,2′-(Benzylazanediyl)bis(1,3,2-dithiaphospholane-2-sul-
fide), 10. 2-Chloro-1,3,2-dithiaphospholane, 12a (2.95 g, 18.66
́
Vallee, France) and was immediately transferred to a narrow Teflon
tube for EPR measurement. The control measurement composed from
DMSO (80 μL), PBN (10 μL), and H2O2 (10 μL) excluding the
ligand. All experiments were performed in triplicate at RT, in a final
volume of 100 μL.
Identification of Metal Ion Binding Sites by 31P/1H NMR
Measurements. Titration of Compound 8 by Zn(II). ZnCl2 (0.1−2
7907
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908

Inorganic Chemistry
Article
equiv) in D2O solution was added to compound 8 (1.27 mg, 0.0036
mmol) in DMSO-d6 (0.6 mL), and 1H/31P NMR spectra were
measured at 200 and 81 MHz, respectively.
Titration of Compounds 8 and 10 by Cu(I). Cu(CH3CN)4PF6 was
purified before use by dissolving the salt in acetonitrile-d3 and filtering
the unsoluble Cu(II) salt. The filtrate was flushed with argon stream to
dryness. The concentration of the Cu(I) salt was determined by UV
spectroscopy by the addition of the specific Cu(I) indicator,
bicinchoninic acid (BCA) (ε562 = 7700 M−1).31
Lippolis, V.; Ogilvie, H. R.; Verani, G. Eur. J. Inorg. Chem. 2006, 2006,
200−206.
(19) Huber, C. P.; Post, M. L.; Siiman, O. Acta Crystallogr., Sect. B
1978, 34, 2629−2632.
(20) Hong, P.; Li, C.; Banerji, S. K.; Wang, Y. J. Hazard. Mater. 2002,
94, 253−272.
(21) Magennis, S. W.; Parsons, S.; Corval, A.; Woollins, J. D.;
Pikramenou, Z. Chem. Commun. 1999, 61−62.
(22) Ishikawa, H.; Kido, T.; Umeda, T.; Ohyama, H. J. Pestic. Sci.
1992, 56, 1882−1883.
Cu(CH3CN)4PF6 (0.1−2 equiv) in CD3CN (47.5 mM) was added
to compound 8 (1.27 mg, 0.0036 mmol) or 10 (1.5 mg, 0.0036 mmol)
(23) Kumar, V.; Ahamad, T.; Nishat, N. Eur. J. Med. Chem. 2009, 44,
785−793.
1
in DMSO-d6 (0.6 mL), and H/31P NMR spectra were measured at
200 and 81 MHz, respectively.
(24) Morley, J. O.; Charlton, M. H. J. Phys. Chem. A 1998, 102,
6871−6876.
ASSOCIATED CONTENT
* Supporting Information
(25) Birdsall, D. J.; Slawin, A. M. Z.; Woollins, J. D. Polyhedron 2001,
20, 125−134.
■
S
(26) Bharara, M. S.; Parkin, S.; Atwood, D. A. Inorg. Chem. 2006, 45,
2112−2118.
Tables of data including crystallographic and structural
information; NMR spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
(27) Britigan, B. E.; Coffman, T.; Buettner, G. J. Biol. Chem. 1990,
265, 2650−2656.
(28) Richter, Y.; Fischer, B. J. Biol. Inorg. Chem. 2006, 11, 1063−
1074.
AUTHOR INFORMATION
Corresponding Author
*E-mail: bilha.fischer@biu.ac.il. Fax: 972-3-6354907. Phone:
972-3-5318303.
■
(29) Anzai, K.; Aikawa, T.; Furukawa, Y.; Matsushima, Y.; Urano, S.;
Ozawa, T. Arch. Biochem. Biophys. 2003, 415, 251−256.
(30) Bohm, V.; Puspitasari-Nienaber, N. L.; Ferruzzi, M. G.;
Schwartz, S. J. J. Agric. Food Chem. 2002, 50, 221−226.
(31) Brenner, A. J.; Harris, E. D. Anal. Biochem. 1995, 226, 80−84.
Author Contributions
§These authors equally contributed to this work
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors wish to thank Dr. K. Adamsky from the Chemistry
Department at BIU for her help with solid-state NMR
measurements.
REFERENCES
■
(1) Jomova, K.; Valko, M. Toxicology 2011, 283, 65−87.
(2) Rooney, J. P. K. Toxicology 2007, 234, 145−156.
(3) Delangle, P.; Mintz, E. Dalton Trans. 2012, 41, 6359−6370.
(4) Amir, A.; Shmuel, E.; Zagalsky, R.; Sayer, A. H.; Nadel, Y.;
Fischer, B. Dalton Trans. 2012, 41, 8539−8549.
(5) Crisponi, G.; Nurchi, V. M.; Fanni, D.; Gerosa, C.; Nemolato, S.;
Faa, G. Coord. Chem. Rev. 2010, 254, 876−889.
(6) Wang, X.; Li, W.; Meng, S.; Li, D. J. Chem. Technol. Biotechnol.
2006, 81, 761−766.
(7) Amir, A.; Sayer, A. H.; Zagalsky, R.; Shimon, L. J. W.; Fischer, B.
J. Org. Chem. 2012, 78, 270−277.
(8) Amir, A.; Sayer, A. H.; Ezra, A.; Fischer, B. Inorg. Chem. 2013, 52,
3133−3140.
(9) Ly, T. Q.; Woollins, J. D. Coord. Chem. Rev. 1998, 176, 451−481.
(10) Cupertino, D.; Keyte, R.; Slawin, A. M. Z.; Williams, D. J.;
Woollins, J. D. Inorg. Chem. 1996, 35, 2695−2697.
(11) Maganas, D.; Grigoropoulos, A.; Staniland, S. S.;
Chatziefthimiou, S. D.; Harrison, A.; Robertson, N.; Kyritsis, P.;
Neese, F. Inorg. Chem. 2010, 49, 5079−5093.
(12) Bolundut, L.; Haiduc, I.; Ilyes, E.; Kociok-Kohn, G.; Molloy, K.
C.; Gomez-Ruiz, S. Inorg. Chim. Acta 2010, 363, 4319−4323.
(13) Moore, P.; Errington, W.; Sangha, S. P. Helv. Chim. Acta 2005,
88, 782−795.
(14) Bereman, R. D.; Wang, F. T.; Najdzionek, J.; Braitsch, D. M. J.
Am. Chem. Soc. 1976, 98, 7266−7268.
(15) Preda, A. M.; Silvestru, A.; Farcas, S.; Bienko, A.; Mrozinski, J.;
Andruh, M. Polyhedron 2008, 27, 2905−2910.
(16) Silvestru, C.; Drake, J. E. Coord. Chem. Rev. 2001, 223, 117−216.
(17) Haiduc, I. J. Organomet. Chem. 2001, 623, 29−42.
(18) Aragoni, M. C.; Arca, M.; Carrea, M. B.; Demartin, F.;
Devillanova, F. A.; Garau, A.; Hursthouse, M. B.; Huth, S. L.; Isaia, F.;
7908
dx.doi.org/10.1021/ic500465z | Inorg. Chem. 2014, 53, 7901−7908