Relative contributions of individual phoretic effect in the below-cloud scavenging process

We systematically estimate the thermophoretic and diffusiophoretic effect to the below-cloud scavenging coefficient based on the below-cloud scavenging coefficient expression derived in Bae, Jung, and Kim [(2006). Development and evaluation of an expression for polydisperse particle scavenging coefficient for the below-cloud scavenging as a function of rain intensity using the moment method. J. Aerosol Sci. 37, 1507]. The temperature difference between the ambient air temperature (T) and raindrop surface temperature (T_s), denoted (T?T_s), is the most important factor to the thermophoretic effect. The scavenging coefficient by diffusiophoresis is mainly affected by T, (T?T_s), and relative humidity (RH). The diffusiophoresis increases with the decrease of (T?T_s). The driving force of diffusiophoresis increases with the increase of T. As RH becomes lower, diffusiophoresis is more effective. The relative contribution of thermophoresis and diffusiophoresis shows opposite trend to the changes of both T and (T?T_s) at all T except (T?T_s) being 1 °C. Thermophoresis and diffusiophoresis are important to the scavenging process when particle diameter is between 0.1 and 10 μm, especially around 1 μm.     

Spray drying formulation of hollow spherical aggregates of silica nanoparticles by experimental design

The present work employs an experimental design methodology to optimize the spray-drying production of micron-size hollow aggregates of biocompatible silica nanoparticles that are aimed to serve as drug delivery vehicles in inhaled photodynamic therapy. To effectively deliver the nanoparticles to the lung, the aerodynamic size (d_A) of the nano-aggregates, which is a function of the geometric size (d_G) and the degree of hollowness, must fall within a narrow range between 2 and 4 μm. The results indicate that (1) the feed concentration, (2) the feed pH, and (3) the ratio of the gas atomizing flow rate to the feed rate are the three most significant parameters governing the nano-aggregate morphology. Spray drying at a low pH (<7) and at a low feed concentration (<1%, w/w) generally results in nano-aggregates having small geometric and aerodynamic sizes (d_A = d_G ≈ 3 μm) with a relatively monodisperse size distribution. Spray drying at a higher feed concentration produces nano-aggregates having a larger d_G but with a multimodal particle size distribution. A trade-off therefore exists between having large d_G to improve the aerosolization efficiency and obtaining a uniform particle size distribution to improve the dose uniformity.     

Ball milling assisted hydrothermal synthesis of ZrO2 nanopowders

The formation of ZrO₂ nanopowders under various hydrothermal conditions such as temperature, time, autoclave rotation speed, heating rate and particularly assistance of ball milling during reaction was investigated. Full ZrO₂ formation (with monoclinic phase) from zirconium solution was completed at shorter times with increasing temperature such as after 4 h at 150 °C, 2 h at 175 °C and less than 2 h at 200 °C. Crystallite size increased from 2.9 to 4 nm with increasing reaction temperature from 125 °C to 200 °C, respectively. Ball milling assisted hydrothermal runs were performed to understand the effect of mechanical force on phase formation, crystallinity and particle size distribution. Monoclinic ZrO₂ was formed in both milled and non-milled runs when zirconium solution was used. Mean particle size for the 2 M solution was measured to be 94 nm for the milled and 117 nm for the non-milled powders. However, when amorphous aqueous zirconia gels (precipitated at pH 5.8) were used, tetragonal phase was also formed in addition to monoclinic phase. Mean particle size was measured to be 0.7 μm (d₉₀≅1.3 μm) for the milled and 7.9 μm (d₉₀≅13 μm) for the non-milled powders. Ball milling during hydrothermal reactions of both zirconium solution and aqueous zirconium gel resulted in smaller crystallite size and mean particle size and, at the same time, effectively controlled particle size distribution (or agglomeration) of nanopowders.     

Grain size effect on electrical properties of Mn-modified 0.67BiFeO3–0.33BaTiO3 lead-free piezoelectric ceramics

To investigate how grain size affects the dielectric, ferroelectric, and piezoelectric properties of Mn-modified 0.67BiFeO₃–0.33BaTiO₃ ceramics, we prepared samples with a wide variety of grain sizes from 4.1 μm to 0.59 μm via a conventional solid-state process that use the normal and the two-step sintering methods. Small-signal dielectric measurements show that all the samples exhibit a relaxor-like behavior and that grain size has little influence on the room-temperature dielectric permittivity. For grain sizes below 2 μm, the remanent polarization P_r and piezoelectric coefficient d₃₃ decrease with the grain size, whereas they remain almost constant near P_r = 27 μC/cm² and d₃₃ = 70 pC/N in samples with grain sizes exceeding 2 μm. The mechanism underlying the observed grain size effect is discussed in terms of the electric-field-induced formation of macroscopic ferroelectric domains.     

Application of Box–Behnken design and response surface methodology for modeling of some Turkish coals

The aim of our research was to apply Box–Behnken experimental design and response surface methodology for modeling of some Turkish coals. As a base for this study, standard Bond grindability tests were initially done and Bond work indexes (W_i) values were calculated for three Turkish coals. The Box–Behnken experimental design was used to provide data for modeling and the variables of model were Bond work index, grinding time and ball diameter of mill. Coal grinding tests were performed changing these three variables for three size fractions of coals (−3350 + 1700 μm, −1700 + 710 μm and −710 μm).Using these sets of experimental data obtained by mathematical software package (MATLAB 7.1), mathematical models were then developed to show the effect of each parameter and their interactions on product 80% passing size (d₈₀). Predicted values of d₈₀ obtained using model equations were in good agreement with the experimental values of d₈₀ (R² value of 0.96 for −3350 + 1700 μm, R² value of 0.98 for −1700 + 710 μm and R² value of 0.94 for −710 μm). This study proved that Box–Behnken design and response surface methodology could efficiently be applied for modeling of grinding of some Turkish coals.     

A kinetic study of the quartz–cristobalite phase transition

Cristobalite is a common silica polymorph in ceramics, as it can crystallize in SiO₂-rich systems during high temperature processes. Its occurrence in final traditional ceramic bodies remarkably affects their thermal expansion, thus playing an important role in the shrinkage upon cooling. The quartz–cristobalite transformation kinetics is investigated by in-situ isothermal X-ray powder diffraction experiments and then correlated to the average particle size (〈d〉) of the starting quartz using a model here developed. An Avrami-like rate equation, i.e. α(t) = 1 ? exp(? k × t)ⁿ, in which the n-term is assumed to account for the dependence on the average particle size, has provided the best fitting of theoretical to experimental data, yielding activation energy values that range from 181 to 234 kJ mol^?1, and exponential n-coefficients from 0.9 to 1.5. Ex-situ observations have demonstrated that the formation of cristobalite from quartz after 50 min, 2, 4 and 6 h at 1200 and 1300 °C, exhibits a remarkable dependence on 〈d〉 of quartz, showing comparable behaviours in the case of 〈d〉 equal to 15.8 and 28.4 μm, but significant differences for 〈d〉 of 4.1 μm. The formation of cristobalite is boosted remarkably at temperature higher than 1200 °C, with an increase by weight even of 500%, with respect to its content at lower temperature. The method of sample preparation (dry powder, wet powder and tablet of compressed dry powder) seems to influence the results only at temperature > 1200 °C and in the case of fine powder.     

Exfoliation/delamination of kaolinite by low-temperature washing of kaolinite–urea intercalates

The effects of short milling (2 × 10 min) of two fractions of kaolinite (< 5 μm (S5) and 5–45 μm (S)), previously expanded by intercalation with 20 wt.% and 40 wt.% urea (U) and exfoliated/delaminated using low-temperature procedures, were studied. Untreated and treated kaolinite samples were examined by X-ray powder diffraction (XRPD), and their specific surface area (SSA), particle size distribution (PSD) and Fourier-transform infrared (FTIR) spectra were measured and interpreted. Samples intercalated with different concentrations of urea and subsequently exfoliated during thermal or low-temperature washing exhibit a variable intensity of the 001 diffractions of kaolinite (K) and interstratified intermediate kaolinite (H) with randomly interstratified layers of grafted urea and water. Low-temperature washing procedures yield H structures interstratified with layers grafted with urea and water with a mean d₀₀₁ = 8.168(50) Å. After thermal washing procedures, H structures form with a mean d₀₀₁ = 8.322(60) Å, apparently as a result of prevailing interstratifications of water over grafted layers of urea. The H/K peak intensity ratio for exfoliated/delaminated samples S and S5 previously intercalated with 20% U is significantly higher than for those intercalated with 40% U. After all washing procedures, fine fraction S5 increases its SSA from 20.2 m²/g to 25–35 m²/g, whereas the grain-size diameter remains approximately constant, with a mean M_D = 2.80(20) μm. Exfoliated/delaminated coarse kaolinite of the S fraction not only increases its SSA from 9.4 m²/g to SSA between 12.5 m²/g and 20.5 m²/g, but also decreases its M_D from 12 μm to between 8.73 μm and 5.60 μm due to the breaking of particles. This procedure is quicker than milling and thermal washing operations reported and used so far, and, moreover, it yields SSA values comparable to those obtained for exfoliated/delaminated kaolinite.     

Fabrication and characterization of reaction-bonded silicon carbide with poly(methyl methacrylate) as pore-forming agent

Reaction-bonded silicon carbide (RBSC) ceramics were prepared by liquid silicon infiltration at 1600 °C using a carbon biscuit. The green body was made by slip casting a stabilized carbon powder slurry, followed by pyrolysis in a vacuum furnace at 1000 °C for 2 h; the density of the biscuit (ρ_b) was controlled using poly(methyl methacrylate) (PMMA) powder as pore former in mass proportion from 30% to 50% in 5 percentage point intervals. The particle size of the PMMA had significant effects on the microstructure, distribution of residual silicon, and the mechanical properties of the ceramic. For 40 mass% PMMA with d₅₀=1.17 μm and d₅₀=0.51 μm, ρ_b was 0.81 and 0.82 g/mL, with corresponding biscuit porosities of 51% and 50%, which gave peak values of both RBSC ceramic density of 3.07 and 3.10 g/mL, and flexural strength of 741 and 794 MPa, respectively. XRD analysis showed that the main phase was β-SiC, with a small quantity of α-SiC. Using PMMA with d₅₀=0.51 μm, a small quantity of residual Si was well dispersed with grain size <1 μm. “Black core” residual carbon in the RBSC was successfully avoided when ρ_b≤0.82 g/mL (mass proportion PMMA≥35%). PMMA as pore former favored the elimination of the detrimental black core and the preparation of dense RBSC with good mechanical properties.     

Effect of high energy milling process on microstructure and piezoelectric/dielectric properties of K0.475Na0.475Li0.05NbO3 solid solution

K_0.475Na_0.475Li_0.05NbO₃ (abbreviated as NKLN) ceramic of near the morphotropic phase boundary (MPB) composition was synthesized by two different processes. The first one is the high energy milling [sometimes abbreviated as HEM hereafter] process, which involves mixing the starting materials and milling the calcined powder using a high energy nano-mill, in order to obtain nano-sized particles. The second one is a conventional mixed oxide method. The HEM process of the starting materials lowered the calcination temperature to the extent of 200 °C as compared with conventionally fabricated NKLN. The particle size of the powder, exposed to the HEM process, reduced to 40 nm, whereas the conventionally ball-milled powder had a larger size of 420 nm after the mixing process. Furthermore, the HEM process improved the reaction activity and homogeneity of the materials used throughout the process, accompanying the enhancement of the sintering density, grain uniformity, and the decrease of grain size. In order to investigate the effects of the HEM process on the electric properties of NKLN ceramics, the dielectric and piezoelectric properties of sintered specimens fabricated by two different processes were evaluated. It was found that the properties of the nano-sized NKLN ceramic near the MPB composition were increased by the modified method, showing the maximum values of d₃₃=179 pC/N, k_p=34% and K₃₃^T=440 compared with 132 pC/N, 29%, and 400, respectively in the conventional process. Further evidence for the grain size effect was investigated by the polarization–electric field curve at room temperature. The remnant polarization for the nano-sized NKLN specimen had a higher value of 24.3 μC/cm² compared with that of 13.7 μC/cm² for conventional NKLN, whereas the coercive field had a similar value. The modified mixing and milling method was considered to be a new and promising process for lead-free piezoelectric ceramics owing to their excellent piezoelectric/dielectric properties.     

High temperature creep behaviour of 4 mol% yttria tetragonal zirconia polycrystals (4-YTZP) with grain sizes between 0.38 and 1.15 μm

The high temperature mechanical behaviour of 4% mol yttria tetragonal zirconia polycrystals (4-YTZP) with different grain sizes (0.38 < d < 1.15 μm) has been analyzed by means of compression creep tests. The working temperature was 1350 °C and the strain rates ranged between 5 × 10⁻⁷ and 2 × 10⁻⁴ s⁻¹. Experimental results have been fitted to the conventionally accepted creep law for superplastic ceramics. Thus, stress exponents and activation energies have been measured as a function of the grain size. The dependence of strain rate on grain size has also been determined. The experimental data are discussed with respect to the existing theoretical models for these materials.     

Impactor designed to increase mass output rate of nanoparticles from a pneumatic nebulizer

A modular impactor was designed to remove large droplets from aerosols generated by a pneumatic nebulizer, the Six-Jet Atomizer from TSI Inc. (Shoreview, MN), with the aim of generating dry nanoparticles. Three interchangeable nozzle heads were designed to provide droplet cutoff diameters of 0.5, 1, and 2 μm at an air flow rate of 8.3×10^?4 m³ s^?1 (50 L min^?1), which corresponds to all six jets of the nebulizer operated at 25 °C and an air pressure of 241 kPa (35 psi). The collection and output characteristics of the 0.5 μm impactor were evaluated from dry particle size distributions produced by nebulizing an aqueous solution with a NaCl mass fraction of 1% both with and without the impactor present. The impactor characteristic cutoff curve was sharp (impactor geometric standard deviation, GSD_imp=1.15–1.19) with a 50% cutoff diameter d₅₀ that ranged from 0.48 μm at 3.0×10^?4 m³ s^?1 to 0.74 μm at 11.7×10^?4 m³ s^?1. The rate of dry NaCl particle generation ranged from 0.5 to 5 g s^?1 (0.04 to 0.4 g day^?1) with mass median diameters MMD_p=80–123 nm and geometric standard deviations GSD_p=1.6–1.8 (depending on flow rate). Anomalous negative impactor efficiencies were observed at flow rates >8.3×10^?4 m³ s^?1 for 100 to 400 nm droplets and at all flow rates for droplets smaller than 100 nm. This phenomenon will be investigated further as a way to increase the generation rate of nanoparticles. A step-by-step procedure is presented for the selection of an appropriate impactor design and operating flow rate for a desired maximum aerosol particle size.     

How to increase the hydration degree of CA — The influence of CA particle fineness

It is shown that the hydration degree of CA is directly dependent on the fineness of CA-particles. Finer particles lead to an increased degree of hydration and also an increased hydration rate.The reaction of a sample with mainly coarse particles of CA (d₅₀ = 50 μm) is characterized by a low hydration rate and only 34 rel.-% of CA dissolved after 22 h. Whereas in a very fine CA-sample (d₅₀ = 4 μm) hydration starts delayed but then shows the highest hydration rate and a dissolution of 62 rel.-% CA. The behaviour is explained by the coverage of CA-particles with a dense hydrate layer of C₂AH_x and AH_x. This reacted CA-rim is supposed to have the same thickness for different sized CA-particles. Optimization of Gauss distribution curves, which were applied to simulate a more realistic particle size distribution, leads to a reacted rim thickness of 1.3 μm until reaction is stopped.     

Structure and density of deposits formed on filter fibers by inertial particle deposition and bounce

The morphology and packing density of particle deposits formed by accumulation on thin steel fibers suspended in an aerosol stream were studied by confocal microscopy. Measurements were made with electrically neutral polystyrene spheres (d_P=1.3, 2.0, 2.6 and 5.2 μm) as a function of flow velocity (v=0.7–5 m/s) and fiber diameters (d_F=8 and 30 μm). Deposition under these conditions was dominated by inertia (Stokes number St=0.3–9), interception (interception parameter R=0.04–0.35) and particle bounce, with a negligible contribution from diffusion.The experiments show a systematic transition of deposit morphologies with a newly introduced particle bounce parameter β～St/R, where St and R are based on the diameter d_F of the bare fiber. Compact, forward facing deposit structures dominate in case of significant particle bounce (i.e. for β>β?, where β* represents the critical conditions for the onset of bounce on the bare fiber). Dendritic structures with pronounced sideways branching are formed at β<β*. R is of relatively little influence as an independent parameter, probably because interception occurs mostly on preexisting deposit structures with dimensions in the order of d_P.The mean porosity ε of the deposit structures was determined on the basis of contour measurements by confocal microscopy, in combination with data on the accumulated particle volume per unit fiber length (known accurately from a previous paper by Kasper, Schollmeier, Meyer, and Hoferer (2009). Once noticeable deposits had formed, ε was found to attain stable values between 0.80 at d_P=1.3 μm and 0.55 at d_p=5.2 μm.     

Comparison of the wear particle size distribution of different a-C coatings deposited by vacuum arc

For biomedical application in the field of artificial hip joints diamond-like carbon (DLC) coatings have been widely studied due to their tribological properties. The wear particles as the main factor limiting the life expectancy of hip joints have attracted more and more interest, not only the number of them, but also the distribution of their size. In this study we have deposited DLC coatings on stainless steel (P2000) by a vacuum arc adjustable from anodic to cathodic operation mode, with the anode–cathode diameter ratio of d_a/d_c = 3/1 at a DC bias of − 250 V to − 1000 V. To improve the adhesion of the DLC coating on P2000, titanium as a metallic interlayer was deposited by cathodic vacuum arc evaporation. The internal structure of the coating was investigated by the visible Raman spectroscopy with the four-Gaussian curve fitting method. Comparing the results with the previous work (coatings deposited with d_a/d_c = 1/1), it was found that the anode–cathode diameter ratio has an effect on the structure (e.g. I_D/I_G) as well as the wear particle size distribution. It was shown that the maximum of the frequency distribution e.g. at − 1000 V bias can be shifted to below 1 μm with increasing d_a/d_c.     

High power density in a piezoelectric energy harvesting ceramic by optimizing the sintering temperature of nanocrystalline powders

Piezoelectric energy harvesting is the research hotspot in the field of new energy, and its core is to prepare piezoelectric ceramics with high transduction coefficient (d₃₃ × g₃₃) and large mechanical quality factor (Q_m) as well. In addition, the miniaturization of the piezoelectric energy harvester also requires the material to have a submicron fine grain structure. In this work, submicron-structured ternary system, MnO₂-doped Pb(Zn_1/3Nb_2/3)O₃-Pb(Zr_0.5Ti_0.5)O₃ was constructed by pressureless sintering of nanocrystalline powders, which has been synthesized for the first time by high-energy ball milling route thereby evading the calcination stage. The microstructure and the energy harvesting characteristics were tailored through changing the sintering temperature. It was found that 1000 °C sintered fine-grained specimen (mean grain size ∼0.95 μm) showed the maximum d₃₃ × g₃₃ value of 9627 × 10⁻¹⁵ m²/N, meanwhile Q_m was as large as 774, which was almost seven times larger than pure counterpart. In the mode of the cantilever-type energy harvester, a high power density of 1.5 μW/mm³ were obtained for 1000 °C sintered specimen at a low resonance frequency of 90 Hz and acceleration of 10 m/s², which were further increased to 29.2 μW/mm³ when the acceleration increased to 50 m/s², showing the potential applications as a next generation high power multilayer energy harvester.     

Structural and electrical characterization of (K0.48Na0.52)0.96Li0.04Nb0.85Ta0.15O3 synthesized by spray drying

The synthesis of lead-free ferroelectric materials with composition (K_0.48Na_0.52)_0.96Li_0.04Nb_0.85Ta_0.15O₃ has been achieved by spray drying technique. Pure perovskite phase was obtained after calcination of as-prepared powders at 800 °C for 1 h. Crystallized powders have a particle size average of 100 nm. Well sintered samples were obtained at 1120 and 1130 °C for 2 h in air with densities of 4.58 and 4.49 g/cm³, respectively. We attributed this improvement as the result of sintering process and the very small particle size in powders. Sintered samples have promising piezoelectric parameters, k_p 0.40–0.41, d₃₁ 50–55 pC/N and dielectric losses around 1%.     

Study on orientation in EVA/Fe3O4 composite hot-melt adhesives

EVA hot melt adhesives have tackiness to both external anticorrosion coating made by cross-linked polyethylene and iron-base petroleum pipeline. But, traditional EVA hot melt adhesives cannot meet the requirement of external anticorrosion coating for weaker tackiness. Two types of Fe₃O₄ particles with different particle sizes and magnetic strengths were added in adhesives. One is of 3.665 μm^{Laser particle size analyzer (LPSA)} and 8.403×10¹ emu/g^{vibrating sample magnetometer (VSM)} and the other one is of 0.426 μm^LPSA and 3.997×10¹ emu/g^VSM. The result of peel test indicated that peel strength of composite adhesives increased as Fe₃O₄ content increased when particles size was 3.665 μm^LPSA but the tackiness of composite adhesive decreased as Fe₃O₄ content increased when particles size was 0.426 μm^LPSA. Also, microphotos of SEM revealed that the composite adhesive with 3.665 μm^LPSA Fe₃O₄ was more likely to distribute in a region near the tackiness surface between the adhesive and iron layers, but the one with 0.426 μm^LPSA Fe₃O₄ was more likely to aggregate in the middle region of adhesive. The movement of 3.665 μm^LPSA Fe₃O₄ particles could induce EVA molar chain orientation and this orientation was confirmed by infrared dichroism and XRD. Results of infrared dichroism and XRD showed that the orientation degree of EVA increased as 3.665 μm^LPSA Fe₃O₄ content increased. Furthermore, crystallinity tests by XRD and DSC indicated that crystallinity of PE segment of EVA also increased as 3.665 μm^LPSA Fe₃O₄ content increased, which could support increase of orientation tested by infrared dichroism and XRD.     

Micromoulding of PZT film structures using electrohydrodynamic atomization mould filling

In this work, electrohydrodynamic atomization combined with a photolithography polymeric micromoulding technique was used to form PZT ceramic structures. PZT thick film structures consisting of squares and rectangles of various sizes and separations were produced and used to evaluate the process. An expansion effect of approximately 10 μm on the ceramic structure width relative to the 200 μm wide mould design was observed. The minimum continuous gap between features achieved using this process was 13.5 μm, and the smallest regular PZT square structure obtainable was 106 μm in width. A sloping side wall of the PZT structures caused by the shielding of the photoresist mould was also observed in the process. The resulting PZT structures had a homogenous microstructure and exhibited a relative permittivity of 250, d_33, f of 67 pCN^?1 and remnant polarisation of 8.8 μC/cm².     

Low shear compounding process for thermoplastic fabrication of ferroelectric lead-free fibres

The thermoplastic ceramic extrusion process involves the shaping of a polymer highly filled with inorganic powder, the so-called ceramic–thermoplastic feedstock. The limitation faced with the process is the amount of raw material required to produce the feedstock. Depending on the density and desired volume of the materials used, the typical amount of ceramic powder required is a minimum of ∼100 g. The validation of a low shear feedstock preparation method against a standard high shear mixing method occurred. Microstructure investigation and single electromechanical fibre characterization of low shear produced KNN (d₃₃ – 49 pC/N; P_r – 3.7 μC/cm³) and PZT (d₃₃ – 392 pC/N; P_r – 32.4 μC/cm³) fibres, in terms of PE, SE loops and d₃₃ measurements, demonstrating the reproducibility of the results when compared to a standard ceramic–thermoplastic high shear mixing process. The repeatability of the measurements showed the proposed procedure to be robust, validating the new compounding method for wide-scale use.     

Preparation,characterisation and properties of sulphoxide modified polystyrene resins for solid-phase extraction of PtIV,RuIII and RuIV from hydrochloric acid

New sulphoxide modified resins were synthesized using poly(styrene-co-divinylbenzene) (PS-DVB) as matrix. Infrared spectroscopy and elemental analysis were used for characterisation. Solid-phase extraction of Pt^IV, Ru^III and Ru^IV from acidic chloride solutions was performed via batch experiments. Influence of spacer length between sulphoxide and matrix (ethylene, hexamethylene), substitution of sulphoxide (R¹: ethyl, hexyl, phenyl) and bead size of PS-DVB (spherical beads: d₅₀ < 155 μm, d₅₀ < 80 μm; powder: d₅₀ < 30 μm) on adsorption was investigated subjected to acidity. Experimental results showed that ethyl substituted sulphoxide immobilised onto ground PS-DVB and hexamethylene spacer exhibited best adsorption properties. Different kinetic models and isotherms were fitted to the experimental data to identify extraction mechanism. Pt^IV was quantitative sorbed at [HCl] ⩽ 0.1 mol/L whereas Ru^III and Ru^IV sorption ranged between 90% and 95% at [HCl] ⩾ 5 mol/L. Desorption was reached using a solution of 0.5 M thiourea (Tu) in 0.1 M HCl at 90 °C. Separation of Pt^IV and Ru^III occurred at [HCl] ⩽ 0.1 mol/L whereas Pt^IV was extracted and Ru^III remained in solution. A further separation was achieved by extracting Pt^IV and Ru^IV at 5 M HCl followed by sequential elution of Pt^IV with concentrated HCl and Ru^IV with 0.5 M Tu in 0.1 M HCl at 90 °C.