The size distribution of 'gold standard' nanoparticles

The spherical gold nanoparticle reference materials RM 8011, RM 8012, and RM 8013, with a nominal radius of 5, 15, and 30 nm, respectively, have been available since 2008 from NIST. These materials are recommended as standards for nanoparticle size measurements and for the study of the biological effects of nanoparticles, e.g., in pre-clinical biomedical research. We report on determination of the size distributions of these gold nanoparticles using different small-angle X-ray scattering (SAXS) instruments. Measurements with a classical Kratky type SAXS instrument are compared with a synchrotron SAXS technique. Samples were investigated in situ, positioned in capillaries and in levitated droplets. The number-weighted size distributions were determined applying model scattering functions based on (a) Gaussian, (b) log-normal, and (c) Schulz distributions. The mean radii are 4.36 ± 0.04 nm (RM 8011), 12.20 ± 0.03 nm (RM 8012), and 25.74 ± 0.27 nm (RM 8013). Low polydispersities, defined as relative width of the distributions, were detected with values of 0.067 ± 0.006 (RM 8011), 0.103 ± 0.003, (RM 8012), and 0.10 ± 0.01 (RM 8013). The results are in agreement with integral values determined from classical evaluation procedures, such as the radius of gyration (Guinier) and particle volume (Kratky). No indications of particle aggregation and particle interactions—repulsive or attractive—were found. We recommend SAXS as a standard method for a fast and precise determination of size distributions of nanoparticles.      

Evidence for hydrophobic attraction between latex particles in aqueous systems as investigated by turbidimetry

We report the evidence for attractive interaction of latex particles which are covered by poly(ethylene oxide) chains. These particles are suspended in aqueous solutions of ammonium sulfate. The interaction is probed by measurements of the turbidity of the suspensions up to 70 g/l. Turbidity is insensitive to multiple scattering and allows the static structure factor, S(q) [q=(4πn ₀/λ₀)sin(θ/2), where θ is the scattering angle, n₀ is the refractive index of the medium and λ₀ is the wavelength in vacuo], to be determined at small q values. The analysis of S(q) at small q values yields information about possible attraction of the particles. The analysis of the turbidity data furthermore shows that no aggregation took place in these systems. A weak but long-range attractive interaction was found at ammonium sulfate concentrations of 0.01 and 0.1 M. The relation of this attractive force to hydrophobic forces is discussed. Received: 9 March 2000/Accepted: 28 June 2000     

A Highly Sensitive Resonance Rayleigh Scattering Method for the Determination of Vitamin B1 with Gold Nanoparticles Probe

In weakly acidic buffer medium, vitamin B₁ (VB₁) interacts with gold nanoparticles to form a binding product, which resulted in a significant enhancement of resonance Rayleigh scattering (RRS) intensity and the appearance of a new RRS spectrum. The maximum RRS peak was at 368 nm, and there are three smaller scattering peaks that were at 284 nm, 440 nm and 495 nm, respectively. The enhanced RRS intensity (ΔI) was directly proportional to the concentration of VB₁ in the range of 0–2.8 × 10⁻⁷ mol L⁻¹. The method had high sensitivity and its detection limit (3σ) was 0.9 ng mL⁻¹. The optimum conditions and the influencing factors have been investigated. The method had good selectivity, which could be observed from the influence of coexisting substances. A sensitive, simple and fast RRS method for the determination of VB₁ with gold nanoparticle probe has been developed. In addition, the reasons for RRS enhancement were discussed.     

Electrophoresis of colloidal α-alumina

Measurements of the electrophoretic mobility (u _E) of particles of colloidal α-alumina were made as a function of pH, electrolyte concentration and electrolyte type (NaCl, NaNO₃ and KCl) using two similar instrumental techniques. Significant differences (50% or less) in the values of u _E of particles in NaCl were obtained from the two instruments; however, the isoelectric points (IEPs) (the pH at which u _E=0), estimated from the two sets of measurements, occurred at 7.5 ± 0.3 and 7.8 ± 0.05 and were not significantly different. The latter estimate corresponds with those for particles in KCl and NaNO₃ of 8.05 ± 0.11 and 7.95 ± 0.18, respectively, made using the same instrument and indicate that the IEP was a weak function of electrolyte type. When cations acted as counterions (pH > IEP), the absolute magnitudes and the ranges of u _E with electrolyte concentration were found to be significantly less than when anions acted as counterions (IEP > pH). Estimates of the zeta potential (ζ), made using various procedures, showed variations of up to 25% at low ratios of electrical-double-layer thickness (κ ⁻¹) to particle radius (a) (κa∼10) and were of a similar scale to differences in u _E, but no significant variations (95% confidence) in ζ were obtained at high values (κa∼200). Received: 12 July 2000 Accepted: 17 October 2000     

The size and shape of silica particles

Negatively charged silica particles were investigated at pH 10.0. They were found to be rod-shaped (cylinder) with a diameter of 5–5.5 nm and a full length of 44–67 nm depending on the rod model used. Moreover, the particles were found to be stable against aggregation in the region 0.4–50 mM NaCl. Received: 2 December 1998 Accepted in revised form: 2 February 1999     

Analysis of photocurrent spectra at polybithienyl film coated on Pt

Photoelectrochemical measurements have been performed at a polybithienyl (PBT) film (doping level of 1 × 10¹⁸/cm³) deposited on a platinum electrode. The cathodic photocurrents and negative slope of the Mott-Schottky plot indicate that the PBT film has the features of a p-type semiconductor. The cathodic photocurrents are interpreted in terms of the Gaertner-Butler model on the basis of the theory of the semiconductor|solution interface. The (i _ph hν)^2/n vs. hν plots taken from the photocurrent spectra show two linearities for n=1 in the wavelength range from 460 nm to 490 nm and for n=4 in the wavelength range λ > 490 nm. The band gaps of the PBT film were determined to be 2.05 ± 0.05 eV for n=1 and 1.55 ± 0.05 eV for n=4. The flat-band potential is 0.33 V (vs SCE). From the slope of the Mott-Schottky plot at the modulation frequency of 3 kHz, the dielectric constant ɛ of the film and the thickness of the depletion layer W ₀ of the PBT film were determined to be 7.4 and 0.29 μm, respectively. Received: 6 January 1999 / Accepted: 6 June 1999     

Determination of Yttrium by Resonance Light Scattering Technique

 Yttrium was determined by means of the resonance light scattering (RLS) technique. The characteristics of resonance light scattering spectra of yttrium-1, 6-bi(1′-phenyl-3′-methyl-5′-pyrazolone-4′-) hexanedione (BPMPHD), the effective factors and optimum conditions were studied. In the pH range of 5.0–6.1, yttrium-BPMPHD complex produces three characteristic peaks of RLS at 365, 402 and 467 nm. The enhanced intensity of RLS is proportional to the concentration of yttrium in the range of 1.0×10⁻⁸ to 1.0×10⁻⁵ mol · L⁻¹. The limit of detection is 5.9×10⁻⁹ mol · L⁻¹. The method has been used for the determination of Y³⁺ in mixed rare earths. Correspondence: Key Lab for Colloid and Interface Chemistry of Education Ministry, School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, People’s Republic of China. e-mail: wux@sdu.edu.cn Received July 3, 2002; accepted October 20, 2002     

Quantitative determination of large structures by small-angle light scattering

A new small-angle light scattering camera has been developed. In contrast to conventional detection the present system is based on a recently developed two-dimensional charge-coupled-device chip made by Thomson (France). The advantage of this chip is its excellent linear response and very low dark signal even at room temperature. The best linearity was obtained by leading each pixel signal through the same amplifying system. The optical system covered a diffraction angular region from about 1° to 15° (q = 0.2–2.6 μm⁻¹). The camera was calibrated with grids and pinholes and was tested with polymer latex particles in solution and with spherulites from polymer films. Received: 06 December 1999 Accepted: 04 February 2000     

Characterization of currently marketed heparin products: analysis of molecular weight and heparinase-I digest patterns

We evaluated polyacrylamide gel electrophoresis (PAGE) and size exclusion chromatography coupled with multi-angle laser light scattering (SEC-MALLS) approaches to determine weight-average molecular weight (M _w) and polydispersity (PD) of heparins. A set of unfractionated heparin sodium (UFH) and low-molecular-weight heparin (LMWH) samples obtained from nine manufacturers which supply the US market were assessed. For SEC-MALLS, we measured values for water content, refractive index increment (dn/dc), and the second virial coefficient (A ₂) for each sample prior to molecular weight assessment. For UFH, a mean ± standard deviation value for M _w of 16,773 ± 797 was observed with a range of 15,620 to 18,363 (n = 20, run in triplicate). For LMWHs by SEC-MALLS, we measured mean M _w values for dalteparin, tinzaparin, and enoxaparin of 6,717 ± 71 (n = 4), 6,670 ± 417 (n = 3), and 3,959 ± 145 (n = 3), respectively. PAGE analysis of the same UFH, dalteparin, tinzaparin, and enoxaparin samples showed values of 16,135 ± 643 (n = 20), 5,845 ± 45 (n = 4), 6,049 ± 95 (n = 3), and 4,772 ± 69 (n = 3), respectively. These orthogonal measurements are the first M _w results obtained with a large heparin sample set on product being marketed after the heparin crisis of 2008 changed the level of scrutiny of this drug class. In this study, we compare our new data set to samples analyzed over 10 years earlier. In addition, we found that the PAGE analysis of heparinase digested UFH and neat LMWH samples yield characteristic patterns that provide a facile approach for identification and assessment of drug quality and uniformity.     

Investigations on the melting behaviour of triglyceride nanoparticles

Suspensions of triglyceride nanoparticles have been proposed as carrier systems for intravenous administration of poorly water soluble drugs. Such nanosuspensions can easily be produced by homogenization of the melted triglyceride in an aqueous phase. Using special emulsifier blends it is possible to obtain suspensions with an average size of the recrystallized particles below 100 nm (photon correlation spectroscopy z-average). As can be observed by transmission electron microscopy the particles are very thin platelets with thicknesses in the range of only a few molecular layers. Nanoparticles of saturated monoacid triglycerides (smaller than 200 nm) exhibit uncommon melting behaviour, which is expressed in their differential scanning calorimetry curve by multiple endothermal peaks over a temperature range of about 10 °C. This effect was attributed earlier to the particle thickness distribution in the suspension rather than to polymorphic transitions since all the material exists in the stable β modification. Here we present experimental investigations on the correlation between the melting behaviour of trilaurin nanosuspensions and the particle thickness distribution determined by analysis of difference X-ray diffraction patterns recorded at progressively higher temperatures in the melting range of the particles. Because of the weak X-ray scattering of the systems investigated synchrotron radiation was used besides conventional sources. The Fourier analysis of the difference diffraction patterns is described in detail and the advantages and difficulties in using this method are discussed. It was observed that the melting temperatures of the nanoparticles increase with increasing particle thicknesses. Simultaneously a decrease in the interplanar (001) spacing with increasing particle thickness was found. Received: 27 July 1999 Accepted: 5 October 2000     

Desorption of BTEX from activated charcoal using accelerated solvent extraction: evaluation of occupational exposures

A procedure for the determination of benzene, toluene, ethylbenzene and o-xylene, m-xylene and p-xylene (BTEX) in occupational environments is proposed. These compounds are extracted from activated charcoal using accelerated solvent extraction. Operational parameters are optimized and quantitative recovery is obtained using acetonitrile as the extraction solvent and 1-mL extraction cells, a preheat time of 2 min, a temperature of 160 °C, a pressure of 1,500 psi, a static period of 5 min, a flush volume of 110%, two cycles and a purge time of 90 s. Determination of BTEX compounds is carried out by gas chromatography using a flame ionization detector. The recoveries, obtained for a confidence level of 95%, are 91 ± 4, 100 ± 3, 104 ± 2, 93 ± 4, 99 ± 2 and 99 ± 2% for benzene, toluene, ethylbenzene, o-xylene, m-xylene and p-xylene, respectively. The detection limits are 0.5 μg for benzene, 0.7 μg for toluene and 1.0 μg for the other compounds. The proposed procedure has been applied to real samples collected in several workplaces, like a microbiology laboratory, an analytical chemistry laboratory, a printer’s, a car repair shop and a petrol station. From the results obtained, it can be concluded that the occupational exposures determined are always acceptable because they are lower than the tenth part of the recommended exposure limits (VLA-ED and VLA-EC).     

Porous structure of the catalyst layers of electrodes in a proton-exchange membrane fuel cell: A stage-by-stage study

Porous structure is studied by standard contact porosimetry after each stage in the preparation of a catalyst layer, which contains a carbon substrate (CS), an ionomer in the form of Nafion resin, and a platinum catalyst. The influence of the ionomer on the porous structure of ten different CS is investigated. The structure of these samples is studied over the maximum range of their pore radii r ∼ 0.3–10⁵ nm. Pores of main volume within particles of the CS under investigation are mainly distributed over the maximum range of their radii from r ≤ 1 to ∼ 50 nm. Ionomer introduction into all the CS under investigation leads to an increase in the integral porosity due to the porosity of the intergranular structure. The change in porosity of the intragranular structure is caused by ionomer blocking small pores in the CS. In most CS, ionomer blocks pores of different sizes, from micropores with radii r ≤ 1 nm and up to r ∼ 1000 nm. It is concluded that the extent of blockage of CS pores is largely determined by the surface properties of the CS and Nafion resin and, more precisely, by the difference in resin adhesion to the CS surface because of the presence of different surface groups on the CS surface. When platinum is applied to CS, this leads to an increase in the specific volume of the micropores. The smallest dimensions of platinum particles are estimated to be on the order of 1 nm.     

Enthalpy relaxation in an epoxy-cycloaliphatic amine resin

Enthalpy relaxation in a system containing the diglycidyl ether of bisphenol A (DGEBA) resin and a diamine, 1,3-bisaminomethylcyclohexane (1,3-BAC) as curing agent, has been investigated by differential scanning calorimetry (DSC). Samples fully cured were annealed at temperature T_g–15 °C for periods of time from 1 h to a maximum of 168 h. The enthalpy relaxation is analyzed by the peak shift method, in which the sample is heated at 10 °C/min following cooling at various rates through the glass transition region. The key parameters of structural relaxation determined were the non-linearity parameter x=0.47 ± 0.02, the apparent activation energy Δh*=1264 ± 48 kJ/mol or Δh*/R=152 ± 6 kK and the non-exponentiality parameter β ≈ 0.3. The results, obtained by the same method, were compared with those for other systems based on fully cured DGEBA. The correlations among these parameters with the peak shift model should be considered with caution. However, the results show that a correlation between crosslink lengths and the value of Δh* can be considered. The relaxation process for DGEBA/1,3-BAC proves to be highly cooperative. Received: 28 June 2000 Accepted: 6 September 2000     

Characterising latex particles and fractal aggregates using image analysis

The fractal nature of latex particles and their aggregates was characterised by image analysis in terms of fractal dimensions. The one- and two-dimensional fractal dimensions, D ₁ and D ₂, were estimated for polystyrene latex aggregates formed by flocculation in citric acid/phosphate buffer solutions. The dimensional analysis method was used, which is based on power law correlations between aggregate perimeter, projected area and maximum length. These aggregate characteristics were measured by image analysis. A two-slopes method using cumulative size distributions of aggregate length and solid volume has been developed to determine the three-dimensional fractal dimension (D ₃) for the latex aggregates. The fractal dimensions D ₁, D ₂ and D ₃ measured for single latex particles in distilled water agreed well with D ₁ = 1, D ₂ = 2 and D ₃ = 3 expected for Euclidean spherical objects. For the aggregates, the fractal dimension D ₂ of about 1.67 ± 0.04 (±standard deviation) was comparable to the fractal dimension D ₃ of approximately 1.72 ± 0.13 (±standard deviation), taking the standard deviations into account. The measured three-dimensional fractal dimension for latex aggregates is within the fractal dimension range 1.6–2.2 expected for aggregates formed through a cluster-cluster mechanism, and is close to the D ₃ value of about 1.8 indicated for cluster formation via diffusion-limited colloidal aggregation. Received: 28 September 1998 Accepted: 29 October 1998     

Separation of molecular tracers sorbed onto atmospheric particulate matter by flash chromatography

In to order increase sensitivity and to reduce the background induced by matrix effects, a method was developed that uses flash chromatography to separate various compounds present in atmospheric aerosol samples prior to their analysis with different analytical techniques (GC–MS, GC–FID, HPLC). For this purpose, flash chromatography using a 4 g silica gel column crossed by eluent at a flow rate of 20 mL min⁻¹ was used. An eluent with enhanced polarity is needed to separate nonpolar (linear and branched alkanes), semipolar (PAH, nitro-PAH and cholesterol) and polar (methoxyphenols, alkanoic acids, and levoglucosan) compounds. Three combinations of solvents were used: hexane for the nonpolar fraction (F1), toluene/hexane for the semipolar fraction (F2) and dimethylformamide for the polar fraction (F3). The use of different eluents for each fraction allows separation of the sample to be accomplished with good repeatability and satisfying yields [85 ± 5% for F1, 81 ± 8% (PAHs), 89 ± 6% (nitro-PAHs) and 74 ± 7% (cholesterol) for F2 and 79 ± 7% (n-alkanoic acids), 40 ± 11% (methoxyphenols) and 77 ± 6% (levoglucosan) for F3]. The methoxyphenol yields were low due to losses during the concentration/evaporation step. This method was then applied to analyse the organic composition of particles collected at an urban site in Strasbourg (France).     

Simultaneous determination of harpagoside and cinnamic acid in rat plasma by high-performance liquid chromatography: application to a pharmacokinetic study

Radix Scrophulariae (Xuanshen) is one of the famous Chinese herbal medicines widely used to treat rheumatism, tussis, pharyngalgia, arthritis, constipation, and conjunctival congestion. Harpagoside and cinnamic acid are the main bioactive components of Xuanshen. The purpose of this study was to develop an HPLC–UV method for simultaneous determination of harpagoside and cinnamic acid in rat plasma and investigate pharmacokinetic parameters of harpagoside and cinnamic acid after oral administration of Xuanshen extract (760 mg kg⁻¹). After addition of syringin as internal standard, the analytes were isolated from plasma by liquid–liquid extraction. Separation was achieved on a Kromasil C₁₈ column, and detection was by UV absorption at 272 nm. The described assay was validated in terms of linearity, accuracy, precision, recovery, and limit of quantification according to the FDA validation guidelines. Calibration curves for both analytes were linear with the coefficient of variation (r) for both was greater than 0.999. Accuracy for harpagoside and cinnamic acid ranged from 100.7–103.5% and 96.9–102.9%, respectively, and precision for both analytes were less than 8.5%. The main pharmacokinetic parameters found for harpagoside and cinnamic acid after oral infusion of Xuanshen extract were as follows: C _max 1488.7 ± 205.9 and 556.8 ± 94.2 ng mL⁻¹, T _max 2.09 ± 0.31 and (1.48 ± 0.14 h, AUC_0–24 10336.4 ± 1426.8 and 3653.1 ± 456.4 ng h mL⁻¹, 11276.8 ± 1321.4 and 3704.5 ± 398.8 ng h mL⁻¹, and t _1/2 4.9 ± 1.3 and 2.5 ± 0.9 h, respectively. These results indicated that the proposed method is simple, selective, and feasible for pharmacokinetic study of Radix Scrophulariae extract in rats. Figure Radix Scrophulariae     

Small-angle neutron-scattering study of bis(quaternaryammonium bromide) surfactant micelles in water. Effect of the long spacer chain on micellar structure

The microstructure of the normal micelles formed by dimeric surfactants with long spacers, [Br⁻(CH₃)₂N⁺(C _m H_{2 m +1})-(CH₂) _S  -(C _m H_{2 m +1})N⁺(CH₃)₂Br⁻, m = 10 and s = 8, 10 and 12], has been investigated by small-angle neutron scattering and compared with previously reported results for micelles of the same dimeric surfactants with shorter spacers (m = 10 and s = 2, 3, 4 and 6). It was found that for dimeric surfactants with long spacers (s = 8 and 10), both micellar growth and variation in shape occur to only a small extent, if at all, compared with dimeric surfactants with short spacers. However, for the dimeric surfactant with the longest spacer, s = 12, the extent of micellar growth and shape variation is also large. These results are due to the differences in conformation of dimeric surfactants with short spacers (s = 2–6) compared with that of the surfactants with long spacers (s = 8–12). Received: 15 June 1998 Accepted: 22 July 1998     

Determination of molar absorptivity coefficients for major type-B trichothecenes and certification of calibrators for deoxynivalenol and nivalenol

This paper presents results from the European Commission-funded project Doncalibrant, the objective of which was to produce calibrators with certified mass fractions of the Fusarium toxins deoxynivalenol (DON), 3-acetyldeoxynivalenol (3-Ac-DON), 15-acetyldeoxynivalenol (15-Ac-DON), and nivalenol (NIV), in acetonitrile. The calibrators, available in ampoules, were sufficiently homogeneous, with between-bottle variations (s _bb) of less than 2%. Long-term stability studies performed at four different temperatures between −18 and 40 °C revealed no significant negative trends (at a confidence level of 95%). Molar absorptivity coefficients (in L mol⁻¹ cm⁻¹) were determined for all four toxins (DON: 6805 ± 126, NIV: 6955 ± 205, 3-Ac-DON: 6983 ± 141, 15-Ac-DON: 6935 ± 142) on the basis of a mini-interlaboratory exercise. The overall uncertainty of the calibrators’ target values for DON and NIV were evaluated on the basis of gravimetric preparation data and include uncertainty contributions from possible heterogeneity, storage, and transport. The Doncalibrant project resulted in the production of calibrators for DON (IRMM-315) and NIV (IRMM-316) in acetonitrile with certified mass fractions of 25.1 ± 1.2 μg g⁻¹ and 24.0 ± 1.1 μg g⁻¹, respectively. Both CRMs became commercially available from the Institute for Reference Materials and Measurements (IRMM, Geel, Belgium) at the beginning of 2007.     

Neutron scattering and electrophoresis of dodecyldimethylamine oxide micelles

Small-angle neutron scattering (SANS) measurements were carried out at 25 °C on dodecyldimethylamine oxide solutions under different conditions. In media with no added salt, the micelle aggregation number remained nearly constant (70–78) over the range of the degree of ionization, α_M, between 0 and 0.73 in contrast with the sharp critical micelle concentration increase in a narrow range of α_M from 0.35 to 0.40. This characteristic α_M dependence that deviates markedly from the prediction of the regular solution approach is thus shown to take place without a considerable change with respect to micelle size and shape. The surface electric potential employed to calculate the intermicellar repulsion under the Debye–Hückel approximation was found to be much lower than the actual surface potential determined from hydrogen ion titration. This inconsistency was solved by introducing a rescaled particle size for the hard-sphere interaction part. The surface potential from SANS was rather similar to the zeta potential determined by electrophoretic light scattering. In the range of NaCl concentrations higher than about 0.2 M, micelle growth was observed for both the hemihydrochloride (1:1 complex) and the cationic species and the growth into a cylindrical shape was confirmed. Received: 3 March 1999 Accepted in revised form: 22 April 1999     

Polydisperse fractal aggregate formation in clay mineral and iron oxide suspensions, pH and ionic strength dependence

pH- and ionic-strength-dependent aggregation of permanently and conditionally charged clay mineral (montmorillonite) and iron oxide (magnetite) particles was investigated by means of dynamic light scattering and rheology. An indifferent electrolyte (NaCl) was used. The surface charging of solids was determined by acid–base titration. The point of zero charge (PZC) of magnetite seemed to be at pH 8.0 ± 0.1. The permanent negative charges on the basal plane of montmorillonite influence the interfacial distribution of H⁺ and Na⁺ ions. The pH dependence of the electrophoretic mobility showed directly the dominance of negative charges on montmorillonite lamellae independently of pH, while for magnetite the sign of the mobility reversed at pH ˜ 8.0. Montmorillonite particles formed stable suspensions; coagulation did not take place below 35 mM 1:1 electrolyte independently of pH. The aggregation of magnetite sol becomes significant near the pH of the PZC even at low ionic strength. Colloidal stability in composite systems was investigated at pH ˜ 4, where oxide and clay mineral particles are oppositely charged. At the lowest NaCl concentrations (1, 5 mM) the mixed systems remained stable and aggregation of oppositely charged particles could not be observed at all. Heterocoagulation of dissimilar particles needed a definite amount of dissolved electrolytes (about 8 mM). Mixed clay mineral and oxide systems are more sensitive to electrolyte under acidic conditions than those separately. Rheological investigation of the mixed clay mineral–oxide suspensions at pH ˜ 4 provided proof for the absence of attractive particle interaction at low ionic strength (2 mM). A physical network of oppositely charged particles could form only at higher salt concentration, for example, in the presence of 10 mM NaCl. The yield value of plastic systems showed a significant maximum at 1:15 magnetite/montmorillonite mass ratio. Received: 21 November 2000 Accepted: 20 December 2000