DaltonTransactions

PAPER

Cite this: Dalton Trans., 2013, 42, 11212

Received 27th March 2013,Accepted 23rd May 2013

DOI: 10.1039/c3dt50829c

www.rsc.org/dalton

Synthesis and antiproliferative activity of theheterobimetallic complexes [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N’-MCl2] (M = Co, Ni, Zn; dmoPTA= 3,7-dimethyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]-nonane)†

Manuel Serrano-Ruiz,a Luis Manuel Aguilera-Sáez,a Pablo Lorenzo-Luis,b

José M. Padrónc and Antonio Romerosa*a

The synthesis and characterization of three novel heterobimetallic derivatives of general formula

[RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N’-MCl2] (M = Co (1), Ni (2) and Zn (3), dmoPTA = 3,7-dimethyl-1,3,7-

triaza-5-phosphabicyclo[3.3.1]nonane) are described. The sizes of their diffusing units are in agreement

with the solid-state structures of 2 and 3, supporting a similar solid-state structure for complex 1. The

diamagnetic ruthenium–zinc derivative 3 was fully characterized in solution at 193 K by NMR. Electro-

chemical properties of the complexes were investigated by cyclic voltammetry, supporting the influence

of the {MCl2} moiety on the electronic properties of the Ru-unit. For the three complexes the GI50 values

were in the range 0.8–6.5 μM and comparable to those of the standard anticancer drug cis-platin in the

same cell lines.

1. Introduction

The wide coordination possibilities and interesting propertiesof the water soluble phosphines1 have induced us to focusour research activity during the last few years in synthesizingnew water soluble complexes containing this kind of ligand.Among the known aqua-soluble phosphines we have beeninterested in studying the coordinative properties of the airstable and economically obtained ligand 1,3,5-triaza-7-phos-phaadamantane (PTA) and its derivatives,2 some of them(Scheme 1) featuring high water solubility and interestingchemical and physical properties.3

The interesting ligand dmPTA was synthesized by us only afew years ago along with the new water-soluble phosphineligands dmoPTA and HdmoPTA (dmPTA = N,N′-dimethyl-1,3,5-triaza-7-phosphaadamantane; dmoPTA = 3,7-dimethyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]nonane; HdmoPTA = 3,7-H-3,7-

dimethyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]nonane) (Scheme 2).These ligands showed interesting coordinative properties pro-viding a way to obtain heterometallic complexes in which it is

Scheme 1 PTA molecule and some of its derivatives.

Scheme 2 Ligands dmPTA, HdmoPTA and dmoPTA.

†CCDC 929129 (2) and 929128 (3). For crystallographic data in CIF or other elec-tronic format see DOI: 10.1039/c3dt50829c

aÁrea de Química Inorgánica, Facultad de Ciencias Experimentales, Universidad

de Almería, 04071 Almería, Spain. E-mail: romerosa@ual.esbDepartamento de Química Inorgánica, Facultad de Química, Universidad de La

Laguna, La Laguna, Tenerife, Spain. E-mail: plorenzo@ull.escBioLab, Instituto Universitario de Bio-Orgánica “Antonio González” (IUBO-AG)

Universidad de La Laguna, C/Astrofísico Francisco Sánchez 2, 38206 La Laguna,

Spain. E-mail: jmpadron@ull.es

11212 | Dalton Trans., 2013, 42, 11212–11219 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article OnlineView Journal | View Issue

possible to combine the properties of different metal centressuch as in [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,-O′)2] (M = Co, Ni, Zn).4

Remarkably, the complexes containing dmoPTA haveshown a better anticancer activity than cis-platin as well as incell lines sensitive to this complex.5 To have more informationon the influence of the second metal bonded to the dmoPTAin the chemical properties of the {RuCpCl(dmoPTA)(PPh3)}-moiety, the new bimetallic complexes [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-MCl2] (M = Co (1), Ni (2) and Zn (3))were synthesized, containing the simple and more reactive{MCl2}-unit. These complexes were fully characterized andtheir anticancer activity assessed on a representative panel ofhuman solid tumor cell lines.

2. Results and discussion2.1. Synthesis and characterization of 1–3

The bimetallic complexes 1–3 were synthesized by a similarprocedure (Scheme 3).

The proton between the two HdmoPTA-NMe atoms in[RuClCp(HdmoPTA)(PPh3)](OSO2CF3) was removed with pot-assium tert-butoxide and the resulting deprotonated complex[RuClCp(dmoPTA-κP)(PPh3)] reacted with MCl2 (M = Co, Niand Zn) to give rise 1, 2 and 3, respectively. The single crystalX-ray structures of 2 and 3 (vide infra) show that the {RuClCp-(dmoPTA-κP)}-moiety is bonded through its dmoPTA-NMe

atoms to the metal of {NiCl2} and {ZnCl2}-units, respectively.The properties of the complexes 1–3 are in agreement withthose of the previously studied bimetallic complexes [RuClCp-(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2] (M = Co (a), Ni(b), Zn (c)).4a,b In particular, for complexes 1–3 it is also clearthat the ruthenium atom of the {RuCpCl(dmoPTA)(PPh3)}-moiety is sensitive to the metal bonded to the dmoPTA-NMe.Additionally, the comparison of the properties of 1–3 withthose of bimetallic complexes containing acac-ligand alsosuggests that the ancillary ligands on the metal unit havesome influence on the ruthenium atom.

NMR spectroscopic studies of 1–3 show that solid-statestructures are preserved in solution. Indeed the 31P{1H} NMRspectra of the complex 3 shows two doublets with 2JPP in agree-ment with the pseudo-cis arrangement of the dmoPTA andPPh3 around the Ru-center. The 31P{1H} NMR for PPh3 reson-ance arises at about 50 ppm while that for dmoPTA is a broadand unresolved signal in both 1 and 2, the complexes contain-ing the paramagnetic metals. The chemical shift fordmoPTA-P atom of 1 and 2 were low field shifted of more than

270 and 150 ppm, respectively, by reference to the samesignals in the starting complex [RuClCp(HdmoPTA)(PPh3)]-(OSO2CF3) (δP = −2.93 ppm),4c which are much more largerchemical shifts than those observed for the bimetallic com-pounds containing acac-ligand (ca. 50 ppm).4a,b For the samereason the 1H NMR of dmoPTA were observed as broad un-resolved peaks in an extremely wide spectral region (1: 140 to−15 ppm; 2: 70 to −5 ppm). It was not possible to obtainenough quality 13C{1H} NMR spectra for complexes 1 and 2 asthey are not soluble enough in the solvents used and alsoprobably by the effect on the signals of the carbons closer tothe paramagnetic metal. In contrast, well-shaped and resolvedsignals were observed in the 1H NMR spectrum of the diamag-netic complex 3.

We monitored the spin echo intensity of selected PPh3

signals that are briefly influenced by the paramagnetism ofthe metal bonded to dmoPTA-NMe atoms. The experimentswere carried out in CDCl3 as this is the solvent in which thethree compounds are soluble and stable. The structurallyrelated [RuClCp(PTA)(PPh3)]

6 derivative was used as a bench-mark as well as the parent bimetallic complexes [RuClCp-(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2] (M = Co (a), Ni(b), Zn (c)).4a Calculated diffusion coefficients (D) are given inTable 1.7 The obtained diffusion coefficients for a–c aresimilar while a bigger diffusion coefficient for the benchmarkcomplex [RuClCp(PTA)(PPh3)] was observed. As far as thehydrodynamic radii (rH) are concerned, they were calculated byusing the Stokes–Einstein equation (arbitrarily assuming aspherical shape for the diffusing species) and are also listed inTable 1.7 Several studies on the hydrodynamic dimensions ofthe molecules have been carried out, including metal com-plexes with the PTA-molecule,8 indicating the convenience ofthis kind of study to determine the behavior and dimensionsof the complexes in solution. Interestingly, binuclear com-plexes 1–3 exhibit very similar rH in agreement with the strongsimilarity solid-state structures of 2 and 3, and regardless ofthe nature of the metal bonded to the dmoPTA-NMe atoms.The coefficient diffusion of the starting complex [RuClCp-(HdmoPTA)(PPh3)](OSO2CF3) is bigger than that for complex 1but smaller than those of complexes 2 and 3. As a

Scheme 3 Synthetic procedure for 1–3.

Table 1 Diffusion coefficient (D) and hydrodynamic radii (rH) for 1–3 com-plexes, [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N’-M(acac-κ2O,O’)2] [M = Co (a), Ni(b), Zn (c)], [RuClCp(HdmoPTA)(PPh3)](OSO2CF3) and [RuCpCl(PPh3)(PTA)]

a

Complex D (× 106 cm2 s−1) rH (Å)

1 6.15 6.52 6.68 6.03 7.15 5.6a 6.81 5.9b 6.92 5.8c 6.82 5.9[RuClCp(HdmoPTA)(PPh3)]

+ 6.28 6.4[RuCpCl(PTA)(PPh3)] 7.26 5.5

aMeasurements were obtained in CDCl3 0.02 M at 298 K (viscosityη = 0.543 Kg m−1 s−1 at 298 K, from http://www.bruker-biospin.com).The experimental error was estimated to be ±2%.7

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 11212–11219 | 11213

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

consequence, the hydrodynamic radii of the starting complexis somewhat smaller than that for 1 but significantly largerthan those for complexes 2 and 3.

Complex 3 was fully characterized by NMR in order tocompare its structural behavior in solution with the parentcomplex [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-Zn(acac-κ2O,-O′)2].

4a The 1H and 31P{1H} NMR spectra at room temperatureand at −60 °C showed that there was no exchange between theligands around the metal center {ZnCl2}, in contrast to thatfound for the parent complex containing the {Zn(acac)2}-moiety. The 1H,1H COSY and 1H,13C HSQC spectra of 3 werenecessary for reaching the correct assignment of the signalsin 1H NMR and 13C{1H} NMR spectra. In solution the NMR ofcomplex 3 is more complicated than expected as one of thedmoPTA ligands has lost the C2 symmetry.

2.2. Electrochemical properties

As observed for parent complexes [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2] (M = Co (a), Ni (b), Zn (c)),4a thedichloride-metal moiety coordinated to the dmoPTA-NMe

atoms has a clear influence on the electrochemical propertiesof {RuClCp(dmoPTA)} (Fig. 1). Complex 1 shows one reversibleruthenium-centred monoelectronic wave at 840(5) mV (ΔEp =16 mV, icat/ian = 1; ian

2/ν vs. ν constant). Similarly, complexes 2and 3 display redox waves at 770(5) mV (ΔEp = 25 mV, icat/ian =1; ian

2/ν vs. ν constant) and 798(5) mV (ΔEp = 22 mV, icat/ian = 1;ian

2/ν vs. ν constant) indicating that these processes are alsoreversible and can reasonably be attributed to the Ru(II)/Ru(III)couple. Under similar experimental conditions,4a reversibleone-electron processes were observed for [RuClCp(HdmoPTA)-(PPh3)](OSO2CF3) at 710 mV and [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2] at 915 mV (M = Co), 812 mV (M =Ni) and 900 mV (M = Zn).

Therefore the different metals bonded to the dmoPTA-NMe

atoms but also the ligands bonded to this metal influence theelectrochemical properties of the Ru atom.

2.3. Crystal structures of 2 and 3

A perspective view of isostructural complexes 2 and 3 is shownin Fig. 2. Selected bond distances and angles are reported in

Table 2. The Cp-rings are essentially planar, the largest devi-ation from the overall-plan-Cp being 0.0071 Å (C52) for 2 and0.0150 Å (C53) for 3. The angle between the Cpcent plane andthe P1–Ru1–P2 plane was found to be 53.8(8)° for 2, which isvirtually identical to that observed in 3 (53.7(2)°), but consider-ably shorter than in the bimetallic complexes [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2] (M = Co, Ni or Zn)4a,b

(range 56.62(4)–56.85(0) Å, mean value 56.73°). The Ru1–Cpcentdistance is 1.853(2) Å for 2 and 1.845(5) Å for 3, near thatexpected for similar complexes containing {M(acac)2} moieties(range 1.836(1)–1.855(1) Å, mean value 1.843 Å).4a,b The rest ofthe bond distances found are in agreement with other pianostool complexes described by our research group.4,9 Besidesthe pseudo-octahedral geometry around the ruthenium atom,two six-membered rings [(M–N–C–P–C–N) + (M–N–C–N–C–N)]are formed upon chelation of the {RuClCp(PPh3)(dmoPTA-1κP)}-entity to the Ni or Zn centers (Fig. 2). The six-memberedrings around the metal, with a pseudo-chair conformation, arecompleted with two coordinated chloride atoms as additionalligands. In regard to the coordination of the metal, a little dis-tortion from the tetrahedral geometry is represented by thespread of the Cl–M–NMe-dmoPTA angles between 105.3(6)° for2 and 112.3(1)° for 3 (Table 2). Besides the bond lengthsbetween dmoPTA-NMe and Ni or Zn atoms are significantlyshorter (ca. 0.2 Å) than that observed in the bimetallic com-plexes (2.182(6)–2.278(7) Å, mean value 2.226 Å)4a,b probably

Fig. 1 Cyclic voltammograms in CH2Cl2 of 1 (black line), 2 (red line) and 3(green line).

Fig. 2 Perspective view of the isostructural complexes 2 and 3, showing thethree six-membered rings around the metal with a pseudo-chair conformation.For the sake of clarity, only methyl hydrogen atoms are included. Dashed linesrepresent the significant intermolecular interactions.

Table 2 Selected bond lengths [Å] and angles [°] for 2 and 3

2 3

Ru1–P1 2.297(7) Ru1–P1 2.295(1)Ru1–P2 2.279(7) Ru1–P2 2.276(2)Ru1–Cl1 2.445(7) Ru1–Cl1 2.447(2)Ru1–Cpcent 1.853(2) Ru1–Cpcent 1.845(5)Cp–ML2 53.8(8) Cp–ML2 53.7(2)Ni–N41 2.008(2) Zn–N1 2.081(4)Ni–N42 2.028(2) Zn–N2 2.067(5)Cl2–Ni–N41 108.9(6) Cl2–Zn–N1 110.4(1)Cl2–Ni–N42 105.3(6) Cl3–Zn–N1 108.9(1)Cl3–Ni–N42 109.6(6) Cl3–Zn–N2 112.3(1)

Paper Dalton Transactions

11214 | Dalton Trans., 2013, 42, 11212–11219 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

due to the trans influence of the acac-ligand. The dmoPTAtorsion angles for both complexes lie in the range 53.8(6)–55.9(3)°. The mean value (54.9°) is somewhat greater thanthose observed for complexes containing acac-ligand (52.2(1)–54.9(2)°, mean value 53.7°),4a,b and significantly less thandmPTA (58.0°) and HdmoPTA (av. 56.8°).4c In 2 and 3 the inter-molecular interactions play a precise role in supporting thenetwork as responsible for the crystal packing, due to closecontacts between the hydrogen on the PPh3 and dmoPTA-groups and the chloride atoms on the coordination of themetals (H⋯Cl, 2.800–2.896 Å, Fig. 2). These close contacts areless than the sum of the van der Waals radii H (1.20 Å) andCl (1.75 Å).4,9

2.4. Antiproliferative activity

The antiproliferative activity of compounds 1–3 was studied inthe panel of human solid tumour cells HBL-100, HeLa,SW1573, and WiDr. The results expressed as GI50 (Table 3)were obtained using the SRB assay10 after 48 hours of exposureto the drugs. Overall, the data show that all complexes inducegrowth inhibition in all cell lines with GI50 values in the range0.8–6.5 μM. When compared to the reference drug cis-platin,compounds 1–3 showed a superior activity profile. In fact, theRu(II) complexes were more active against the breast T-47D andthe colon WiDr cells, which were less affected by the exposureto cis-platin. This observation is consistent with previousstudies in which these cells proved to be more drug resistant.5

The results point out that the three compounds are active inboth drug-sensitive (HBL-100) and drug-resistant (T-47D andWiDr) cell lines. Although the differences in activity are smallwe can observe that compounds 1 and 2 show more potentactivity than the ruthenium–zinc complex 3. Complexes 1 and2 are transformed quickly in DMSO–water to provide thecomplex5 [RuClCp(HdmoPTA)(PPh3)]

+ while 3 is stable enoughover the antiproliferative experiment time. Therefore it can beassumed that the observed anticancer activity of 1 and 2 isreally that for the [RuClCp(HdmoPTA)(PPh3)]

+ complex. Theparent bimetallic complexes containing the acac-ligand, whichare stable in a mixture DMSO–water, and the complex [RuClCp-(HdmoPTA)(PPh3)](OSO2CF3) showed a somewhat smalleractivity.5

This fact supports that the main antiproliferative active sitein the bimetallic complexes is the ruthenium atom but theinfluence of the other metal unit must not be excluded,

modulating the antiproliferative activity of the complexes.Complex 3 displayed the smallest cell activity but also thesmallest diffusion radius as well as being the only one stablein DMSO–H2O. At present, the mechanism of action of thesecompounds remains unclear. Ongoing studies will clarify this,and will be reported elsewhere, as for example studies focusedon providing information on the interaction of the complexeswith purines and the synthesis of parent bimetallic complexescontaining antiproliferative active metals such as Pt.

3. Conclusions

Three novel bimetallic [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-MCl2] (M = Co (1), Ni (2), Zn (3); dmoPTA = 3,7-dimethyl-1,3,7-triaza-5-phosphabicyclo[3.3.1]nonane) complexes showed asimilar size by diffusing experiments in CDCl3 in agreementwith their solid-state structures. The ruthenium–zinc derivative3 is not involved in an equilibrium with other species in solu-tion, preserving its solid state structure. The electrochemicalproperties of the {RuClCp(dmoPTA)(PPh3)}-moiety is clearlydependent on the {MCl2}-unit bonded to the dmoPTA-NMe

atoms. Comparison of the obtained results with the electrochem-ical properties found for previously published parent com-plexes [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-M(acac-κ2O,O′)2](M = Co (a), Ni (b), Zn (c)) confirms the suspicion that thecoordination of a metal to the dmoPTA-NMe atoms leads to anew PTA-resembling ligand whose properties can be tuned bythe adequate selection of the metal, its oxidation state and theancillary ligands. The GI50 values obtained for the three com-plexes were in the range 0.8–6.5 μM. Although there are smalldifferences in activity those for 1 and 2 are larger than that for3. The obtained results support that the Ru(II)-moiety is themost antiproliferative active part of the molecule. At present,the mechanism of action of these compounds remains unclearbut efforts are targeted to clarify it and to synthesize newbimetallic complexes with possibly improved antiproliferativeproperties.

4. Experimental section4.1. Materials and methods

All chemicals were reagent grade and, unless otherwise stated,were used as received by commercial suppliers. All reactionswere carried out in a pure argon atmosphere by using standard

Table 3 GI50 values of complexes 1–3, [RuClCp(HdmoPTA)(PPh3)]+ and cis-platin against human solid tumor cells

Compound

Cell line (type)a

HBL-100 (breast) T-47D (breast) SW1573 (lung) HeLa (cervix) WiDr (colon)

1 1.0 (±0.2) 1.4 (±0.2) 1.1 (±0.1) 1.3 (±0.1) 1.5 (±0.3)2 0.8 (±0.2) 1.2 (±0.2) 1.0 (±0.1) 1.4 (±1.1) 1.1 (±0.2)3 3.5 (±0.7) 6.5 (±3.0) 4.6 (±0.1) 4.0 (±0.1) 5.7 (±1.4)[RuClCp(HdmoPTA)(PPh3)]

+ — 1.9 (±0.5) 1.5 (±0.1) 2.6 (±0.2) 1.7 (±0.4)cis-platin 1.9 (±0.2) 1.5 (±2.3) 3.0 (±0.4) 2.0 (±0.3) 2.6 (±5.3)

a Values are given in μM and are means of three experiments (±standard deviation).

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 11212–11219 | 11215

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

Schlenk-tube techniques with freshly distilled and oxygen-freesolvents.

The hydrosoluble phosphine-PTA and the complexes[RuClCp(PTA)(PPh3)] and [RuClCp(HdmoPTA)(PPh3)] (OSO2CF3)were prepared as described in the literature.4,6,9 1H and13C{1H} NMR spectra were recorded on Bruker AV300 andAV500 spectrometers operating at 300.13 and 500.13 MHz (1H),respectively. The 1H,1H 2D COSY and 1H,13C{1H} HSQC NMRexperiments were routinely conducted on the Bruker DRX300instruments in the absolute magnitude mode using a 45 or90° pulse after the increment delay. Peak positions are relativeto tetramethylsilane and were calibrated against the residualsolvent resonance (1H) or the deuterated solvent multiplet(13C{1H}). Chemical shifts for 31P{1H} NMR spectra weremeasured relative to external 85% H3PO4. Infrared spectrawere recorded on KBr disks using a Bruker Vertex 70 FT-IRspectrometer, and the intensity of the bands is indicated as:s (strong), m (medium) and weak (w). UV-vis spectra wereobtained with a Jasco V-650 spectrometer from solutions of10−3–10−5 M in quartz cuvettes (1 cm optical path). Specificabsorption coefficients were obtained by linear regression over5–7 points (R2 ≥ 0.999). Elemental analyses (C, H, N, S) wereperformed on a Fisons Instrument EA 1108 elementalanalyzer.

4.1.1. Synthesis of [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-MCl2] (M = Co (1), Ni (2), Zn (3)). Complexes 1–3 wereobtained by using a similar procedure. Potassium tert-butoxide(0.051 g, 0.456 mmol for 1, 3; 0.059 g, 0.525 mmol for 2) wasadded into a solution of [RuClCp(HdmoPTA)(PPh3)](OSO2CF3)(0.359 g, 0.456 mmol for 1; 0.411 g, 0.522 mmol for 2; 0.356 g,0.453 mmol for 3) in 150 mL of EtOH. The resulting mixturewas stirred at room temperature. When the reagents were com-pletely dissolved, the respective metal chloride (CoCl2: 0.058 g,0.444 mmol for 1; NiCl2·6H2O: 0.123 g, 0.517 mmol for 2;ZnCl2: 0.060 g, 0.440 mmol for 3) was added. The resultingmixture was refluxed for 3 h for 1 and 3 while 6 h were neededfor synthesizing 2 in good yield. Then, the solvent was reducedto 5 mL at room temperature under vacuum. The obtainedgreen (1), brown (2) and yellow (3) powders were filtered,washed with cold diethyl ether (2 × 5 mL) and vacuum dried.The obtained solids were dissolved into 10 mL of CHCl3 andfiltered through Celite. The solutions were concentrated to2 mL and then 10 mL of EtO2 were added. The precipitateswere filtered and dried under vacuum.

4.1.2. [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-CoCl2] (1).Yield: 0.255 g, 75%. S25 °C,CHCl3 = 12.5 mg mL−1, S25 °C,H2O =1.0 mg mL−1. Elemental analysis C30H36N3P2Cl3RuCo (766.94 gmol−1): Found C 46.72; H 4.71; N 5.39%; calcd C 46.98; H 4.73;N 5.48%. IR (KBr, cm−1): ν(ArH) 3071, 3057; ν(CH) 2961, 2915,2861; δas(CH3), δ (CH2) 1434 (m); ν(C–N) 1029 (m), 1071 (m);δfdp(CarH) 757 (m), 745 (m); δfdp(CvCar) 696 (s). 1H NMR(300.13 MHz, 25 °C, CDCl3): δ (ppm) 7.0–7.6 (bs, PPh3, 15 H),10 (bs, Cp, 5 H), 94–138 (bm, dmoPTA, 16 H). 13C{1H} NMR(75.47 MHz, 22 °C, CDCl3): δ (ppm) 92.8 (s, Cp), 128–139 (m,aromatic). 31P{1H} NMR (121.49 MHz, 22 °C, CDCl3): δ (ppm)50.28 (d, 2JPP = 33.3 Hz, PPh3), 277 (bs, dmoPTA).

31P{1H} NMR (121.49 MHz, 22 °C, DMSO-d6): δ(ppm) 54.12(bs, PPh3), 220 (vbs, dmoPTA). UV-vis (solid state, 22 °C)λmax/nm: 370 (s), 553 (m), 574 (m), 627 (m), 901 (w), 1275 (m).UV-vis (CHCl3, 22 °C) λmax/nm (ε/dm3 mol−1 cm−1): 241(16 420), 329 (2512), 358 (2000), 418 (864), 549 (147), 568 (229),632 (411). Cyclic voltammetry (CH2Cl2, 22 °C): Eox = 0.92 mV;Ered = 0.76 mV; E1/2 = 0.84 mV.

4.1.3. [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-NiCl2] (2).Yield: 0.250 g, 64%. S25 °C, CHCl3 = 16.7 mg mL−1, S25 °C, H2O =1.0 mg mL−1. Elemental analysis for powder sampleC30H36N3P2Cl3RuNi (766.7 g mol−1): Found C 46.98; H 4.70; N5.62%; calcd C 46.99; H 4.73; N 5.48%. IR (KBr, cm−1): ν(CarH)3071, 3057; ν(CH) 2961, 2915, 2861; δas(CH) 1434 (m); ν(C–N)1029 (m), 1071 (m); δoop(CarH) 757 (m), 745 (m); δoop(CvCar)696 (s). 1H NMR (300.13 MHz, 22 °C, CDCl3): δ (ppm)7.35–7.47 (bs, PPh3, 15H), 7.57 (bs, Cp, 5H), 60–74 (bm,dmoPTA, 16H). 13C{1H} NMR (75.47 MHz, 22 °C, CDCl3):δ(ppm) 80.3 (s, Cp), 128–139 (m, PPh3).

31P{1H} NMR(121.49 MHz, 22 °C, CDCl3): δ(ppm) 47.3 (d, 2JPP = 40.5 Hz,PPh3), 152 (bm, dmoPTA). 31P{1H} NMR (121.49 MHz, 22 °C,DMSO-d6): δ(ppm) 49.38 (bs, PPh3), 81 (vbs, dmoPTA). UV-vis(solid state, 22 °C) λmax/nm: 373 (s), 840 (m), 851 (m), 882 (m),948 (m). UV-vis (CHCl3, 22 °C, nm) λmax/nm (ε/dm3 mol−1

cm−1): 242 (21 289), 336 (2345), 359 (2418), 417 (987). Cyclicvoltammetry (CH2Cl2, 22 °C): Eox = 0.908 mV; Ered = 0.688 mV;E1/2 = 0.798 mV.

4.1.4. [RuClCp(PPh3)-μ-dmoPTA-1κP:2κ2N,N′-ZnCl2] (3).Yield: 0.235 g, 69%. S25 °C, CHCl3 = 16.7 mg mL−1, S25 °C, H2O =1.0 mg mL−1. Elemental analysis C30H36N3P2Cl3RuZn (773.39 gmol−1): Found C 46.33; H 4.86; N 5.31%; calcd C 46.59; H 4.70;N 5.43%. IR (KBr, cm−1): ν(CarH) 3071, 3057; ν(CH) 2961 (m),2915 (m), 2861 (m); δas(CH) 1434 (m); ν(C–N) 1029 (m), 1071(m); δfdp(CarH) 757 (m), 745 (m); δfdp(CvCar) 696 (s). 1H NMR(500.13 MHz, 22 °C, CDCl3): δ (ppm) 7.38–7.56 (m, aromaticprotons, 15 H), 4.53 (s, Cp, 5H), 4.38 (d, 2JHiHc = 12.2 Hz, Hi),4.33 (d, 2JHgHb = 12.1 Hz, 2JHjHe = 12.1 Hz, Hg y Hj), 4.1(ABEFX system, 2JHhHd = 13.7 Hz, 4JHhHa = 1.85 Hz, 4JHhHj =1.1 Hz, 2JPH = 3.4 Hz, Hh), 3.58 (d, 2JHcHi = 12.1 Hz, Hc), 3.56(ABX system, 2JHbHg = 13.9 Hz, 2JPH = 2.2 Hz, Hb), 3.33 (ABEXsystem, 2JHaHf = 15.3 Hz, 4JHaHh = 1.8 Hz, 2JPH = 2.55 Hz, Ha),3.08 (d, 2JHeHj = 11.9 Hz, He), 2.75 (ABEX system, 2JHdHh =13.8 Hz, 4JHdHb = 1.3 Hz, 2JPH = 2.3 Hz, Hd), 2.351–2.461 (s + s,CH3N, 6H), 1.79 (ABEX system, 2JHfHa = 15.3 Hz, 4JHfHg =2.1 Hz, 2JPH = 3.4 Hz, Hf).

13C{1H} NMR (75.47 MHz, 22 °C,CDCl3): δ (ppm) 44.9 (d, 3JCP = 4.8 Hz, NMe′), 44.6 (d,3JCP = 4.9 Hz, NMe′′), 45.1 (d, 1JCP = 21.0 Hz, C1), 58.2(dd, 1JCPdmoPTA = 14.5 Hz, 3JCPPPh3 = 4.8 Hz, C2), 58.5 (d,1JCPdmoPTA = 13.4 Hz, 3JCPPPh3 = 1.2 Hz, C3), 74.9 (d, 3JCP =2.9 Hz, C5), 75.3 (d, 3JCP = 2.8 Hz, C4), 79.8 (t, 2JCP = 2.2 Hz,Cp), 128–138.1 (m, PPh3).

31P{1H} NMR (121.49 MHz, 22 °C,CDCl3): δ (ppm) 2.56 (d, 2JPP = 41.7 Hz, dmoPTA), 46.4 (d,2JPP = 41.7 Hz, PPh3).

31P{1H} NMR (121.49 MHz, 22 °C, DMSO-d6): δ(ppm) 47.01 (d, 2JPP = 44 Hz, PPh3), −1.49 (d, 2JPP =44 Hz, dmoPTA). The labels for NMR assignation are showedin Scheme 4. UV-vis (solid state, 22 °C) λmax/nm: 374. UV-vis(CHCl3, 22 °C) λmax/nm (ε/dm3 mol−1 cm−1): 241 (15 061), 336

Paper Dalton Transactions

11216 | Dalton Trans., 2013, 42, 11212–11219 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

(1866), 360 (1927), 421 (741). Cyclic voltammetry (CH2Cl2,22 °C): Eox = 0.899 mV; Ered = 0.65 mV; E1/2 = 0.77 mV.

4.2. Stability of 1–3 in CDCl3

Into a 5 mm NMR tube were dissolved 10 mg of 1–3 under N2

atmosphere in 0.5 mL of CDCl3. The solution was recorded by1H and 31P{1H} NMR at room temperature and 40 °C. After2 days no significant decomposition of the complexes wasobserved at both temperatures and then air was bubbled intothe solutions. The compounds were fully decomposed after3 days at room temperature and 1 day at 40 °C.

4.3. Stability of 1–3 with DMSO and water

The complexes (10 mg) were dissolved in a mixture of 0.5 mLof DMSO-d6 and the resulting solution studied by NMRat room temperature and 40 °C. The three complexes werestable at room temperature for more than 1 day whilesome decomposition (<5%) is observed at 40 °C after 12 h.Nevertheless, when 0.1 mL of H2O was introduced in the solu-tion, 1 was decomposed quickly at room temperature (5 min)while 2 needed more than 0.2 mL of H2O to be decomposedwithin 10 min and 3 decomposed partially (<20%) with 0.5 mLafter 15 h. The complete decomposition of 3 was achievedafter 3 days at room temperature and within 2 days at 40 °C.The resulting final compound obtained from the three com-plexes was characterized by NMR as the ruthenium complex[RuClCp(HdmoPTA)(PPh3)], which is stable under the reactionconditions for more than 3 days. The UV-vis absorptionanalysis of the resulting solutions showed the absorptionbands of CoCl2 (1) and NiCl2 (2) in water.11

4.4. Diffusion measurements

The diffusion experiments were performed in CDCl3 by thestimulated echo pulse sequence without spinning. The shapeof the gradient pulse was rectangular and its strength variedautomatically in the course of the experiments. The diffusioncoefficients (D, Table 1) were determined from the slope ofthe regression line ln(I/I0) versus G2 according to ln(I/I0) =−(γδ)2(Δ − δ/3)DG2, in which: I and I0 are the observed spinecho intensity with and without gradients, respectively; G isthe gradient strength; Δ is the delay between the mid-points ofthe gradients; D is the diffusion coefficient; δ is the gradientlength. The calibration of the gradients was carried out by adiffusion measurement of HDO in D2O. The experimentalerror in the D values was estimated to be ±2%.7 All the dataleading to the reported D values afforded lines with correlation

coefficients of >0.999 and 6–8 points were used for regressionanalysis. The value for δ parameter was set to 1.75–2 ms.Additionally three different measurements with differentdiffusion parameters (Δ = 25–100 ms) were always carried outto check the reproducibility of the data obtained. The gradientstrength was increased in 10–15% steps from 10%. A measure-ment of 1H NMR spectrum and T1 was carried out beforeeach diffusion experiment and the recovery delay (ca. 5–8 s) setto five times T1.

The number of scans varied between 32 and 64 perincrement. Typical experiment duration is 1.5–3 h.

4.5. Cyclic voltammetry experiments

Electrochemical experiments were recorded with a VersaSTAT3apparatus. A standard disposition for the measurement cellwas used including three-electrode glass cell consisting ofa platinum-disk working electrode, a platinum-wire auxiliaryelectrode and an Ag reference electrode. The supportingelectrolyte solution (NBu4PF6, 0.1 M) was scanned over thesolvent window to verify the absence of electro-activeimpurities. A similar concentration of the analyte (2.5 10–4 M)was employed in all the measurements and HPLC gradeCH2Cl2 as solvent.

4.6. X-ray analysis

Single crystals with suitable dimensions were mounted ona glass fiber with cyanoacrylate at room temperature. Datacollection was performed on a Bruker APEX diffractometer(XDIFRACT service of the University of Almeria) in the range4.668 ≤ 2θ ≤ 43.593 and 4.682 ≤ 2θ ≤ 50.03 for 2 and 3. Datawere collected at 100(2) K using graphite-monochromatizedMo-Kα (λ = 0.71073) in the range −17 ≤ h ≤ 17, −11 ≤ k ≤ 11,−26 ≤ l ≤ 20 for 2 and −14 ≤ h ≤ 17, −11 ≤ k ≤ 11, −26 ≤ l ≤26 for 3. Lorentz and polarization corrections were appliedto 17 306 for 2 and 5506 for 3 reflections collected of which4762 for 2 and 4085 for 3 were unique with I0 > 2σ(I0). Thestructures were determined by direct methods (SIR97 orSHELXS-XTL)12,13 and refined by least-squares procedureson F2 (SHELX-XTL).13 The function minimized during therefinement was w = 1/[σ2(F0

2) + (0.0368P)2 + 0.0000P] for 2 andw = 1/[σ2(F0

2) + (0.0436P)2 + 0.0000P] for 3. The final geometri-cal calculations, the graphical manipulations and the analysisof H-bond network and other crystallographic calculationswere carried out with SHELXS-XTL package.13 In the finalleast squares cycles all non-H atoms were allowed to vibrateanisotropically. The hydrogen atoms were located at the calcu-lated positions and assigned a fixed displacement and con-strained to ideal geometry. Crystal data and data collectiondetails are given in Table 4.

4.7. Growth inhibition assays

The human solid tumor cell lines HBL-100, HeLa, SW1573,T-47D and WiDr were used in this study. These cell lineswere a kind gift from Prof. G. J. Peters (VU Medical Center,Amsterdam, Netherlands). Cells were maintained in 25 cm2

culture flasks in RPMI 1640 supplemented with 5% heat

Scheme 4

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 11212–11219 | 11217

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

inactivated fetal calf serum and 2 mM L-glutamine in a 37 °C,5% CO2, 95% humidified air incubator. Exponentially growingcells were trypsinized and re-suspended in antibiotic contain-ing medium (100 units of penicillin G and 0.1 mg of strepto-mycin per mL). Single cell suspensions displaying >97%viability by trypan blue dye exclusion were subsequentlycounted. After counting, dilutions were made to give theappropriate cell densities for inoculation onto 96-well microti-ter plates. Cells were inoculated in a volume of 100 μL per wellat densities 10 000 (HBL-100, HeLa and SW1573) of 15 000(T-47D), and 20 000 (WiDr) cells per well, based on their dou-bling times. Compounds were initially dissolved in DMSO at400 times the desired final maximum test concentration.Control cells were exposed to an equivalent concentration ofDMSO (0.25% v/v, negative control). Each agent was testedin triplicate at different dilutions in the range of 1–100 μM.The drug treatment was started on day 1 after plating. Drugincubation times were 48 h, after which time cells were precipi-tated with 25 μL ice-cold TCA (50% w/v) and fixed for 60 minat 4 °C. Then the SRB assay was performed.10 The opticaldensity (OD) of each well was measured at 492 nm, usingBioTek’s PowerWave XS Absorbance Microplate Reader. Valueswere corrected for background OD from wells only containingmedium.

Acknowledgements

Financial support co-financed by the EU FEDER: the SpanishMINECO (CTQ2010-20952), the Spanish Instituto de SaludCarlos III (PI11/00840) and Junta de Andalucía through PAI(research teams FQM-317) and Excellence ProjectsP07-FQM-03092 and P09-FQM-5402. Thanks are also given toCOST Action CM0802 (WG2, WG3, WG4). M. Serrano-Ruiz and

L. M. Aguilera are grateful to Excellence project P09-FQM-5402for a postdoctoral and a technic-predoctoral contract,respectively.

Notes and references

1 (a) Aqueous-Phase Organometallic Catalysis. Concepts andApplications, ed. B. Cornils and W. A. Herrmann, Wiley-VCH, Weinheim, Germany, 2nd edn, 2004, pp. 52–59;(b) F. Joó, Aqueous Organometallic Catalysis, Kluwer,Dordrecht, The Netherlands, 2001, pp. 32–39; (c) AqueousOrganometallic Chemistry and Catalysis, ed. I. T. Horváthand F. Joó, NATO ASI 3/5, Kluwer, Dordrecht, 1995, vol. 5,pp. 61–80; (d) S. Bolaño, M. M. Rodríguez-Rocha, J. Bravo,J. Castro, E. Oñate and M. Peruzzini, Organometallics, 2009,28, 6020; (e) F. Mohr, E. Cerrada and M. Laguna, Organo-metallics, 2006, 25, 644; (f ) F. Mohr, S. Sanz,E. R. T. Tiekink and M. Laguna, Organometallics, 2006, 25,3084; (g) B. J. Frost and C. A. Mebi, Organometallics, 2004,23, 5317; (h) D. J. Darensbourg, C. G. Ortiz andJ. W. Kamplain, Organometallics, 2004, 23, 1747;(i) D. J. Darensbourg, F. Joó, M. Kannisto, A. Kathó,J. H. Reibenspies and D. J. Daigle, Inorg. Chem., 1994, 33,200.

2 (a) C. Lidrissi, A. Romerosa, M. Saoud, M. Serrano-Ruiz,L. Gonsalvi and M. Peruzzini, Angew. Chem., Int. Ed., 2005,44, 2568; (b) D. N. Akbayeva, L. Gonsalvi, W. Oberhauser,M. Peruzzini, F. Vizza, P. Brüggeller, A. Romerosa, G. Savaand A. Bergamo, Chem. Commun., 2003, 264.

3 (a) J. Bravo, S. Bolaño, L. Gonsalvi and M. Peruzzini, Coord.Chem. Rev., 2010, 254, 555; (b) A. García-Fernández, J. Díez,M. P. Gamasa and E. Lastra, Inorg. Chem., 2009, 48, 2471;(c) A. T. Ekubo, M. R. J. Elsegood, A. J. Lake andM. B. Smith, Inorg. Chem., 2009, 48, 2633; (d) S. Miranda,E. Vergara, F. Mohr, D. de Vos, E. Cerrada, A. Mendía andM. Laguna, Inorg. Chem., 2008, 47, 5641; (e) B. J. Frost,C. M. Bautista, R. Huang and J. Shearer, Inorg. Chem.,2006, 45, 3481; (f ) B. J. Frost, C. A. Mebi andP. W. Gingrich, Eur. J. Inorg. Chem., 2006, 1182; (g) F. Mohr,L. R. Falvello and M. Laguna, Eur. J. Inorg. Chem., 2006,3152; (h) D. J. Darensbourg, J. C. Yarbrough and S. J. Lewis,Organometallics, 2003, 22, 2050.

4 (a) A. Mena-Cruz, P. Lorenzo-Luis, V. Passarelli,A. Romerosa and M. Serrano-Ruiz, Dalton Trans., 2011, 40,3237; (b) A. Mena-Cruz, P. Lorenzo-Luis, A. Romerosa andM. Serrano-Ruiz, Inorg. Chem., 2008, 47, 2246; (c) A. Mena-Cruz, P. Lorenzo-Luis, A. Romerosa, M. Saoud andM. Serrano-Ruiz, Inorg. Chem., 2007, 46, 6120.

5 C. Ríos-Luci, L. G. León, A. Mena-Cruz, E. Pérez-Roth,P. Lorenzo-Luis, A. Romerosa and J. M. Padrón, Bioorg.Med. Chem. Lett., 2011, 21, 4568.

6 A. Romerosa, T. Campos-Malpartida, C. Lidrissi, M. Saoud,M. Serrano-Ruiz, M. Peruzzini, J. A. Garrido-Cárdenas andF. García-Maroto, Inorg. Chem., 2006, 45, 1289.

Table 4 Selected crystallographic data for 2 and 3

2 3

Empirical formula C30H36Cl3N3P2RuNi C30H36Cl3N3P2RuZnFormula weight 766.69 773.35Cryst dimensions (mm) 0.100 × 0.085 ×

0.0700.099 × 0.083 ×0.065

Cryst system Monoclinic MonoclinicSpace group P21/n P21/na [Å] 14.8944(9) 14.847(4)b [Å] 9.4708(6) 9.499(2)c [Å] 22.5306(14) 22.587(6)β [°] 102.3250(10) 102.151(6)Z 4 4Volume [Å3] 3105.0(3) 3114.1(13)ρcalcd. [mg m−3] 1.640 1.650μ [mm−1] 1.480 1.642F(000) 1560 1568Data/restrains/parameters

5470/0/363 5506/0/363

R1 [I > 2σ(I)] (all) 0.0284 (0.0335) 0.0558, 0.1055wR2 [I > 2σ(I)] (all) 0.0685 (0.0704) 0.0833, 0.1150GOF 0.978 1.02Δρ [e Å−3] 0.960; −0.430 1.053; −0.698

Paper Dalton Transactions

11218 | Dalton Trans., 2013, 42, 11212–11219 This journal is © The Royal Society of Chemistry 2013

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online

7 J. M. Casas-Solvas, E. Ortiz-Salmerón, I. Fernández,L. García-Fuentes, F. Santoyo-González and A. Vargas-Berenguel, Chem.–Eur. J., 2009, 15, 8146: D = (kBT)/(6πηrH),in which D is the diffusion coefficient, kB is the Boltzmanconstant, T is the temperature, and η is the viscosity of thesolvent.

8 S. Bolaño, G. Ciancaleoni, J. Bravo, L. Gonsalvi,A. Macchioni and M. Peruzzini, Organometallics, 2008, 27,1649.

9 (a) M. Serrano-Ruiz, P. Lorenzo-Luis, A. Romerosa andA. Mena-Cruz, Dalton Trans., 2013, 42, 7622;(b) B. González, P. Lorenzo-Luis, P. Gili, A. Romerosa andM. Serrano-Ruiz, J. Organomet. Chem., 2009, 694, 2029.

10 P. O. Miranda, J. M. Padrón, J. I. Padrón, J. Villar andV. S. Martín, ChemMedChem, 2006, 1, 323.

11 See for example: (a) S. F. A. Kettle, Physical InorganicChemistry: A Coordination Chemistry Approach, Spektrum,1996; (b) M. Gerloch and E. C. Constable, Transition MetalChemistry, VCH, Weinheim, 1994; (c) D. Sutton, ElectronicSpectra of Transition Metal Complexes, McGraw-Hill,New York, 1968.

12 A. Altamore, G. Cascarano, C. Giacovazzo, A. Guagliardi,G. Moliterni, M. C. Burla, G. Polidori, M. Camalli andR. Spagna, Sir 97 Package Program, 1997.

13 G. M. Sheldrick, SHELXTL Version 6.14, Bruker-AXSMadison, WI, 2003.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2013 Dalton Trans., 2013, 42, 11212–11219 | 11219

Publ

ishe

d on

28

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f M

isso

uri -

St L

ouis

on

17/0

9/20

13 2

2:15

:23.

View Article Online