4420
Langmuir 2005, 21, 4420-4424
Formation of {Cu6[S2P(OC2H5)2]6} on Cu2S Surfaces from Aqueous Solutions of the KS2P(OC2H5)2 Collector: Scanning Electron Microscopy and Solid-State 31P Cross-Polarization/Magic Angle Spinning and Static 65Cu NMR Studies Daniela Rusanova, Willis Forsling, and Oleg N. Antzutkin* Division of Chemistry, Luleå University of Technology, SE-971 87 Luleå, Sweden
Kevin J. Pike† and Ray Dupree Department of Physics, University of Warwick, Coventry CV4 7AL, United Kingdom Received December 3, 2004. In Final Form: February 10, 2005 The interactions of synthetic chalcocite surfaces with diethyldithiophosphate, potassium salt, K[S2P(OC2H5)2], were studied by means of 31P cross-polarization/magic angle spinning (CP/MAS) NMR spectroscopy and scanning electron microscopy (SEM). To identify the species formed on the Cu2S surfaces, a polycrystalline {CuI6[S2P(OC2H5)2]6} cluster was synthesized and analyzed by SEM, powder X-ray diffraction techniques and solid-state 31P CP/MAS NMR and static 65Cu NMR spectroscopy. 31P chemical shift anisotropy (CSA) parameters, ∆CS and ηCS, were estimated and used for assigning the bridging type of diethyldithiophosphate ligands in the {CuI6[S2P(OC2H5)2]6} cluster. The latter data were compared to 31P CSA parameters estimated from the spinning sideband patterns in 31P NMR spectra of the collector-treated mineral surfaces: formation of polycrystalline {CuI6[S2P(OC2H5)2]6} on the Cu2S surfaces is suggested. The secondorder quadrupolar line shape of 65Cu was simulated, and the NMR interaction parameters, CQ and ηQ, for the copper(I) diethyldithiophosphate cluster were obtained.
Introduction Alkyl-substituted derivatives of dithiophosphoric acid are frequently used as collectors in the froth flotation enrichment of metal sulfide ores. Potassium and sodium salts of dialkyldithiophosphates are known to provide higher selectivity than xanthates and dithiocarbamates, which is required in monoflotation and differential flotation of mixed sulfide ores (Cu2S; Cu-Zn; Cu-Mo; Cu-Ni,Co; Cu-Pb,Zn).1 Surfaces of chalcocite, treated with potassium diethyldithiophosphate in aqueous solutions, have been studied by means of diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy, X-ray photoelectron spectroscopy (XPS), and time-of-flight secondary-ion mass spectrometry (TOF-SIMS).2,3 Data for flotation recovery, contact angle measurements, and the electrochemical and wetting behavior of the system have also been reported.4,5 These studies have shown that a concomitant adsorption process occurs on the mineral surface when it is treated with thiol collectors. Buckley et al. have performed an XPS investigation of preoxidized chalcocite surfaces treated with copper-saturated 0.02 mM diethyldithiophosphate solution for periods ranging from * To whom correspondence should be addressed. E-mail:
[email protected]. † Current address: ANSTO NMR Facility, c/o Materials & Engineering Science, Lucas Heights Research Laboratories, Private Mail Bag 1, Menai NSW 2234, Australia. (1) Cheminova Agro A/S database, www.cheminova.com. (2) Valli, M.; Malmsten, B.; Persson, I. Colloids Surf., A 1994, 83, 227. (3) Buckley, A. N.; Goh, S. W.; Lamb, R. N.; Woods, R. Int. J. Miner. Process. 2003, 72, 163. (4) Woods, R.; Kim, D. S.; Yoon, R. H. Int. J. Miner. Process. 1993, 39, 101. (5) Hanson, J. S.; Fuerstenau, D. W. Int. J. Miner. Process. 1991, 33, 33.
40 s to 10 min.3 For samples treated for 40 s, XPS spectra indicated that all copper(II) oxidation products had been removed from the surface while the chemisorbed collector species had been formed.3 For treatment periods longer than 40 s the presence of molecular CuDTP species adsorbed at the chalcocite surface has been revealed.3 However, a more detailed description concerning the identification and composition of these copper(I) diethyldithiophosphate species has not been published, to our knowledge, and the processes leading to their formation are still not fully understood. This work is focused on the formation and assignment of copper(I) diethyl dithiophosphate species on synthetic chalcocite surfaces. For this purpose a bulk copper(I) diethyldithiophosphate compound was synthesized and studied by means of XRD (powder) and solid-state 31P CP/MAS and 65Cu static NMR. The main strategy was to obtain and compare both the isotropic chemical shift and the chemical shift anisotropy (CSA) 31P NMR parameters for the bulk copper(I) compound and for the species formed on the Cu2S mineral surfaces. 31P CSA data were used to aid the reliable assignment of the species formed on the Cu2S surfaces because the range of isotropic chemical shifts for the phosphorus sites in copper(I) dialkyldithiophosphate compounds was relatively small. The CSA parameters, ∆CS and ηCS, were estimated (using the Mathematica-based program of Levitt and co-workers)6 by analyzing the intensities of the spinning sidebands in the 31P spectral patterns for the samples of the polycrystalline copper(I) diethyldithiophosphate compound and of the collector-treated chalcocite powder. (6) Antzutkin, O. N.; Lee, Y. K.; Levitt, M. H. J. Magn. Reson. 1998, 135, 144.
10.1021/la047026e CCC: $30.25 © 2005 American Chemical Society Published on Web 04/07/2005
Formation of {Cu6[S2P(OC2H5)2]6} on Cu2S Surfaces
Langmuir, Vol. 21, No. 10, 2005 4421
The quadrupolar coupling constants (CQ), asymmetry parameters (ηQ), and isotropic chemical shifts of the 65Cu sites in the polycrystalline copper(I) diethyldithiophosphate compound were obtained from simulations of the central transition line shape of the 65Cu NMR spectrum. To our knowledge, there are very few reports on solidstate 65Cu NMR spectroscopy, and we have therefore explored the possibility of applying this method to copper(I) dialkyldithiophosphate systems. Morphological changes on the chalcocite surfaces treated with a potassium diethyldithiophosphate were observed using scanning electron microscopy (SEM). Experimental Section Materials and Methods. All solvents (p.a. grade) were obtained from Merck. The solvent CH2Cl2 was distilled prior to use. [Cu(CH3CN)4]PF6 was synthesized and used as a source of copper.7 The ligand O,O′-diethyldithiophosphate potassium salt, K[S2P(OC2H5)2], a commercial collector Danafloat 123K, was provided by Cheminova Agro A/S and was used as received. Synthetic Cu2S was obtained from Strem Chemicals, and its structural identity was confirmed by X-ray powder diffraction data. {Cu6[S2P(OC2H5)2]6} was isolated after a reaction of K[S2P(OC2H5)2], in a slight excess, with [Cu(CH3CN)4]PF6 in CH2Cl2/H2O (1:1) solution at pH 9.5 (NH4OH(c)). A pale-yellow polycrystalline solid was obtained (60% yield) (mp 190 °C). No 31P signal from PF - was observed in the 31P NMR, which would 6 be expected if the sample contained PF6- in the lattice. The product was examined by X-ray powder diffraction to confirm its identity and single-phase character. All operations were performed at ambient temperature. Details of the preparation of the copper(I) diethyldithiophosphate cluster have been reported elsewhere.8 As a final step, the samples were washed with methanol to remove concomitant products. Surface-Formatted Complexes. A 1 g sample of synthetic chalcocite (Cu2S) was immersed in an aqueous solution of potassium diethyldithiophosphate with pH 9.2 (borate buffer) and stirred for 30 min. Aqueous solutions of the ligand (prepared with deionized water) with concentrations of 1, 5, and 10 mM were used in this study. After the conditioning, the samples were centrifuged at 4000 rpm for 5 min using Hettich Universal, rinsed with deionized water, centrifuged again, and dried in a vacuum desiccator. The procedure was repeated, with KOH being used to adjust the pH (instead of the borate buffer), and no change in the results was observed. 31P CP/MAS NMR. Solid-state 31P MAS NMR spectra were recorded at 145.73 MHz using a Varian/Chemagnetics Infinity CMX-360 (B0 ) 8.46 T) spectrometer with CP from the protons and proton-decoupling.9 The CP mixing time was 3 ms, and the 90° pulse width was 5 µs. Totals of 32 (for the pure compound) and 10000 (for the surface compound) signal transients were accumulated with 3 s relaxation delays. The samples were packed in standard ZrO2 4 mm or 7.5 mm rotors. The NMR experiments were performed at ambient temperature (298 K). The 31P NMR spectra were obtained at three different spinning frequencies (3-5 kHz) for reliable simulations of the chemical shift anisotropy parameters. All spectra were externally referenced to 85.5% H3PO4.10 Static 65Cu NMR. Solid-state 65Cu NMR spectra were recorded at 170.40 MHz using a Varian/Chemagnetics Infinity CMX-600 spectrometer equipped with a wide-bore 14.1 T magnet. The static spectra were collected using a Bruker 5 mm static probe. In all cases, a spin-echo sequence was applied (1.1 µs-τ-1.1 µs, τ ) 15 µs, γB1/2π ) 110 kHz). The spectra were obtained with a pulse delay of 40 ms, and 30000-50000 transients were averaged. NMR data were processed in two different ways: (i) the whole (7) Kubas, G. J. Inorg. Synth. 1979, 19, 90. (8) Liu, C. W.; Stubbs, T.; Fackler, J. P., Jr. J. Am. Chem. Soc. 1995, 117, 9778. (9) Pines, A.; Gibby, M. G.; Waugh, J. S. J. Chem. Phys. 1972, 56, 1776. (10) Harris, R. K. Encyclopedia of Nuclear Magnetic Resonance; Wiley: New York, 1996; Vol. 6, p 3612.
Figure 1. 31P CP/MAS NMR spectra at 8.46 T of copper(I) diethyldithiophosphate compounds: polycrystalline {Cu6[S2P(OC2H5)2]6} compound (a) and surface-formatted compound on the synthetic chalcocite treated with 10 (b) and 1 (c) mM K[S2P(OC2H5)2]. Totals of 32 and 12000 signal transients were accumulated in (a) and (b), (c), respectively. The MAS frequency was 3 kHz. Centerbands are shown in the insets. echo was Fourier transformed (FT); (ii) FID data were left-shifted to the echo maximum prior to FT (half-echo). It was acquired by stepping the transmitter frequency by 100 kHz, retuning, acquiring a new set of data, and adding all the subspectra together. All spectra were externally referenced to a secondary standard of powdered CuCl.11 SEM Images. The morphology of the samples was studied with a Philips XL 30 SEM instrument equipped with a LaB6 emission source. XRD Pattern. The X-ray diffraction (XRD) powder pattern was collected using a Siemens D5000 diffractometer and Cu KR radiation. The calculated XRD powder pattern for {Cu6[S2P(OC2H5)2]6} was obtained with the CrystalDiffract XRD simulation program using the single-crystal X-ray data reported for the copper(I) diethyldithiophosphate cluster.8
Results and Discussion 31 P CP/MAS NMR spectra (for a spinning frequency of 3 kHz) of the synthesized polycrystalline copper(I) diethyldithiophosphate cluster and a sample of the synthetic Cu2S, treated with the K[S2P(OC2H5)2] collector (1 and 10 mM, aqueous solution), are shown in Figure 1. The isotropic chemical shifts of the phosphorus sites can be read from the positions of the centerbands (δiso ≈ 100 ppm for all samples; see the insets in Figure 1), while the 31P chemical shift anisotropies can be estimated from the intensities of the spinning sidebands, which flank the centerbands at δiso (k(νr/ν0) × 106 (ν0 is the Larmor resonance frequency for the 31P nuclei, νr is the sample spinning frequency, and k ) 1, 2, 3, ...).12 The absence of a central resonance line at 84 ppm indicates that there are no traces of the oxidized collector (disulfide of the diethyldithiophosphate, (EtO)2P(S)-S-S-(S)P(OEt)2) after washing of the samples.13 This compound is a concomitant product in the self-redox reaction of copper dithiophosphate complexes:14
(11) Mackenzie, K. J. D.; Smith, M. E. K.; Dunn, J. E. Multinuclear Solid-State Nuclear Magnetic Resonance of Inorganic Materials; Pergamon Press: New York, 2002. (12) Herzfeld, J.; Berger, R. G. J. Chem. Phys. 1980, 73, 6021. (13) Haiduc, I.; Goh, L. Y. Coord. Chem. Rev. 2002, 224, 151. (14) Yordanov, N. D.; Alexiev, V.; Macicek, J.; Glowiak, T.; Russel, D. Transition Met. Chem. (Dordrecht, Neth.) 1983, 8, 257.
4422
Langmuir, Vol. 21, No. 10, 2005
2CuII(R2dtp)2 a [CuII(R2dtp)2‚‚‚CuII(R2dtp)2] a 2CuI(R2dtp) + (R2dtp)2 (1) Its presence was also not detected on the collector-treated chalcocite surfaces. The spinning sideband patterns of the samples (shown in Figure 1) are very similar and provide a qualitative indication of the similarities of the chemical shift anisotropies of the phosphorus sites in the copper(I) cluster and the copper(I) dithiophosphate species on the chalcocite surface. The 31P CP/MAS spectrum of {Cu6[S2P(OC2H5)2]6} is shown in Figure 1a. A single resonance line at 101.0 ppm is observed. It suggests that the copper(I) cluster has a high symmetry with all six dithiophosphate ligands structurally and magnetically equivalent. All phosphorus sites are chemically equivalent; i.e., PO2S2 tetrahedra in all six ligands are not additionally distorted by a different orientation of the hydrocarbon chains. According to the reported single-crystal structure of {Cu6[S2P(OC2H5)2]6}, values of P-S and P-O bond lengths vary only within (0.001 Å for the six phosphorus atoms, and S-P-S and O-P-O bond angles are 118.41° and 101.42°, respectively, for all six ligands.8 Thus, the magnetic equivalence for all phosphorus sites can be expected, which in turn gives rise to a single resonance line in the solid-state 31P NMR spectrum of this compound. To examine the geometry of the chemical environments around the phosphorus sites, a chemical shift anisotropy analysis was performed and 31P CSA parameters, ∆CS ) δzz - δiso (anisotropy) and ηCS ) (δyy - δxx)/∆CS (asymmetry parameter), were estimated from the intensities of the spinning sidebands in the 31P NMR spectra of copper(I) diethyldithiophosphate compounds.6 From the values of these two parameters the principal values of the chemical shift tensor, δxx, δyy, and δzz, were calculated and tabulated in Table 1 (together with δiso, ∆CS, and ηCS). For the phosphorus sites in the {Cu6[S2P(OC2H5)2]6} cluster ∆CS(31P) has a negative value (-74 ppm) and ηCS is close to 0.3 (a nonaxial symmetry of the chemical shift tensor). Recent ab initio calculations of 31P chemical shift tensors in a model [PO2S2]- fragment, performed by de Dios and co-workers, have shown that negative values of ∆CS(31P) correspond to large S-P-S bond angles, and are characteristic of a bridging-type bonding, while positive values of ∆CS correspond to phosphorus sites in the terminal ligands with smaller S-P-S bond angles in fourmembered heterocycles, MeS2(P).15 These conclusions were also confirmed by 31P CSA experimental data for mononuclear Ni[S2P(OR)2]2, binuclear Zn2[S2P(OR)2]4, and tetranuclear Zn4[S2P(OR)2]6(O) complexes.15 Therefore, similar correlations (i.e., negative values of ∆CS) suggest that the {Cu6[S2P(OC2H5)2]6} cluster has only the bridging type of bonding of diethyldithiophosphate ligands to the copper(I) atoms. The 31P CP/MAS spectra of the synthetic chalcocite treated with 10 and 1 mM aqueous solutions of the collector K[S2P(OC2H5)2] are shown in parts b and c of Figure 1, respectively. A broad centerband at 100.6 ppm with a line width of ∼3 ppm is flanked by the spinning sidebands, with the whole spinning sideband pattern resembling that for {Cu6[S2P(OC2H5)2]6} (see Figure 1a). Therefore, formation of a polycrystalline copper(I) cluster on the surface of Cu2S is indicated at both low and high concentrations of the ligand. Results of a deconvolution procedure and the following 31P CSA simulations using the integrated intensities of the spinning sidebands reveal that 31P CSA parameters (together with δiso) are similar
Rusanova et al.
(within the confidence limits) to those for the polycrystalline {Cu6[S2P(OC2H5)2]6} cluster discussed above, and ∆CS has a negative value (-72 ppm), confirming the presence of only the bridging-type coordination of the dialkyldithiophosphate ligands in copper(I) species formed on the mineral surface. The asymmetry parameter of the surface compound, η, exhibits a somewhat higher value than that estimated for the corresponding bulk copper(I) cluster, but this difference is within that confidence limit for η. A similar arrangement of the six dialkyldithiophosphate ligands around copper atoms in both the bulk copper(I) cluster and in a copper(I) compound formed on the surface of the collector-treated chalcocite is likely. The polycrystalline bulk copper(I) diethyldithiophosphate and chalcocite/collector-treated samples were further investigated by the SEM and powder XRD techniques. Figure 2 shows SEM images of {Cu6[S2P(OC2H5)2]6} (Figure 2a) and the surface of the synthetic Cu2S used in this study before (Figure 2b) and after the treatment of the chalcocite with 5 mM (Figure 2c) and 10 mM (Figure 2d) aqueous solutions of K[S2P(OC2H5)2]. After the treatment, the initially smooth chalcocite surface became roughened with numerous polyhedron-shaped impregnations. When the concentration of the collector was doubled (10 mM), the observed polyhedra were larger with the more extensive coverage of the chalcocite surface (Figure 2d). Similar polyhedron-shaped crystals can be seen in an SEM image of the polycrystalline {Cu6[S2P(OC2H5)2]6} cluster (Figure 2a). Both SEM and 31P NMR data suggest the formation of a copper(I) polycrystalline compound on the surface of the synthetic chalcocite treated with the collector in the concentration range studied (5-10 mM). The sample treated with 1 mM potassium diethyldithiophosphate did not show significant morphological changes of the mineral surface (image not shown) even though the 31P NMR data confirmed the presence of the same copper(I) diethyldithiophosphate species. The disulfide of the collector that should be formed according to the equilibrium shown in eq 1 most possibly is involved in a reaction with other copper atoms at the chalcocite surface, leading also to the formation of copper(I) dialkyldithiophosphate species on the mineral surface.16 This is correlated to the fact that such a reaction product is not seen in the 31P NMR spectrum of the conditioned chalcocite (Figure 1b). The bulk copper(I) diethyldithiophosphate compound was examined by powder XRD. The powder pattern of the polycrystalline {Cu6[S2P(OC2H5)2]6} compound is shown in Figure 3a and compared to the calculated powder pattern using the known single-crystal X-ray diffraction structure of {Cu6[S2P(OC2H5)2]6} (Figure 3b).8 The peak positions in the calculated pattern were found to be identical to the experimentally obtained pattern for the sample. However, the experimental pattern of the {Cu6[S2P(OC2H5)2]6} cluster shows a more amorphous character, with a higher intensity background level, probably caused by the small size of the crystallites in the sample. No additional peaks corresponding to traces of the disulfide or other copper(I) dithiophosphate species were detected. To obtain more information about the local environment of the copper ions in the cluster, the {Cu6[S2P(OC2H5)2]6} system was investigated further with solid-state static (15) Larsson, A.-C.; Ivanov, A. V.; Forsling, W.; Antzutkin, O. N.; Abraham, A. E.; de Dios, A. J. Am. Chem. Soc. 2005, 127, 2218. (16) Buckley, A. N.; Goh, S. W.; Lamb, R. N.; Woods, R. Int. J. Miner. Process. 2003, 72, 163.
Formation of {Cu6[S2P(OC2H5)2]6} on Cu2S Surfaces Table 1.
31P
Langmuir, Vol. 21, No. 10, 2005 4423
Chemical Shift and Chemical Shift Anisotropy Data for the Copper(I) Diethyldithiophosphate Compounds (L ) S2P(OC2H5)2)
compound
δiso (ppm)
∆CSa (ppm)
ηCSa
δxxa
δyya
δzza
Cu6L6 surface-formatted compound
101.0 ( 0.1 100.6 ( 0.2
-74 ( 1 -72 ( 3
0.28 ( 0.06 0.3 ( 0.1
148 ( 2 148 ( 5
128 ( 5 125 ( 10
27 ( 1 28 ( 2
a
68.3% joint confidence limit. Table 2. NMR Parameters Used in Simulations of the 65Cu NMR Spectrum of the Polycrystalline Copper(I) Diethyldithiophosphate Cluster (L ) S2P(OC2H5)2)
compound
δiso (ppm) ( 50
CQ (MHz) ( 0.5
ηQ ( 0.02
∆CS (ppm) ( 50
ηCS ( 0.05
χa (deg) (1
Cu6L6
300
45.6
0.17
750
0.00
4
a
Figure 2. SEM images of the polycrystalline {Cu6[S2P(OC2H5)2]6} compound (a), a synthetic chalcocite sample (b), and a chalcocite sample treated with 5 mM (c) and 10 mM (d) K[S2P(OC2H5)2].
Figure 3. XRD powder patterns of copper(I) diethyldithiophosphate clusters: polycrystalline {Cu6[S2P(OC2H5)2]6} (a); calculated XRD powder pattern from the known single-crystal structure of {Cu6[S2P(OC2H5)2]6}8 (b).
http://crmht-europe.cnrs-orleans.fr/dmfit/default.asp.
{Cu6[S2P(OC2H5)2]6} obtained using this frequency-stepping mode. The line shape shown is characteristic of a second-order quadrupolar-perturbed central transition (1/2 T -1/2) and extends over 1.3 MHz. According to the singlecrystal X-ray diffraction structure of {Cu6[S2P(OC2H5)2]6}, values of Cu-S distances vary only within (0.001 Å for the six copper atoms, and S-Cu-S planar bond angles are 118.36°, 116.53°, and 124.07°, respectively, for all six atoms.8 Thus, very similar chemical environments for all CuS3 sites are expected that in turn may give rise to rather similar resonance line shapes in the solid-state 65Cu NMR spectrum. Indeed, the second-order quadrupolar line shape is well delineated (Figure 4a) despite the expected overlap of the six 65Cu resonance patterns. Therefore, the spectrum can be considered as a superposition of the central transitions of the six copper sites (CuS3). Simulations gave an estimate for the quadrupolar interaction parameter, CQ, of about 45 MHz, and the asymmetry parameter, ηQ, of 0.17 (see Figure 4b and Table 2). It is interesting that a single static quadrupolar line shape alone does not adequately simulate the experimental pattern of {Cu6[S2P(OC2H5)2]6} (δiso ) -300 ppm): there seems to be a considerable 65Cu CSA contribution (∆CS ) 750 ppm) to the line shape, with the fit suggesting that the principal axes of the electric field gradient and CSA tensors are nearly collinear (see Table 2). In summary, both solid-state CP/MAS 31P and static 65Cu NMR data combined with XRD powder pattern analysis are shown to be useful in the assignment of the copper dialkyldithiophosphate cluster. This approach can be applied to studies of various copper-containing mineral surfaces treated with dialkyldithiophosphate collectors and assist in clarifying flotation mechanisms that are not yet fully understood in these systems. Conclusions
Figure 4. Static solid-state 65Cu NMR spectra of polycrystalline {Cu6[S2P(OC2H5)2]6} obtained at 14.1 T (a) along with the simulation (b). 65 Cu NMR spectroscopy. It was found that the conventional “spin-echo” experiment could not irradiate the whole range of the 65Cu NMR spectra since strong pulses (of about 110 kHz amplitude in frequency units) only excite effectively a central region of ca. (100 kHz from the carrier frequency. “Full-range” spectra were successfully obtained by stepping the spectrometer’s carrier frequency from -1.5 to +1.0 MHz of the 65Cu Larmor frequency (see Figure 4). Figure 4a shows the solid-state static 65Cu spectrum of
Synthetic chalcocite surfaces treated with 1, 5, and 10 mM aqueous solutions of diethyldithiophosphate, potassium salt, K[S2P(OC2H5)2], were studied by means of 31P CP/MAS NMR spectroscopy and SEM. The 31P chemical shift tensor parameters of the copper(I)-collector species formed on the Cu2S surface were estimated. To identify the species formed on the surface, a polycrystalline sample of the {Cu6[S2P(OC2H5)2]6} cluster was synthesized and characterized by solid-state 31P CP/MAS NMR, powder XRD, and SEM techniques. Additional information about the local copper environments in the {Cu6[S2P(OC2H5)2]6} cluster was obtained from estimates of the 65Cu quadrupolar interaction parameters revealed by simulating its static 65Cu NMR spectrum. A comparative analysis of the 31P chemical shift data estimated from the experimental NMR spectrum of the
4424
Langmuir, Vol. 21, No. 10, 2005
polycrystalline copper(I) diethyldithiophosphate cluster was used for assigning resonance lines in 31P CP/MAS NMR spectra of species formed at the collector-treated Cu2S surfaces: the substance was assigned to a polycrystalline phase of the Cu6L6 cluster with L ) S2P(OC2H5)2. Acknowledgment. This work was financed by the Agricola Research Centre at the Luleå University of Technology and Marie Curie Industry Host Fellowship HPMT-CT-2001-00335-02. The Varian/Chemagnetics
Rusanova et al.
CMX-360 spectrometer was purchased with a grant from the Swedish Council for Planning and Coordination of Research (FRN) and further upgraded with a grant from the Foundation to the memory of J. C. and Seth M. Kempe. We also acknowledge CHEMINOVA AGRO A/S for the dialkyldithiophosphate collectors. We thank Dr. Andy Howes and Prof. Mark E. Smith, University of Warwick, U.K., and Dr. Bruce Johnson, La Trobe University, Australia, for discussions. LA047026E
