Microfluidic Platform for High-throughput Screening of Leach Chemistry

Die Yang and Craig Priest*. Future Industries Institute, University of South Australia, Mawson Lakes, SA 5095, Australia. ABSTRACT: We demonstrate an ...
0 downloads 0 Views 935KB Size
Download PDF
Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Article

Microfluidic Platform for High-throughput Screening of Leach Chemistry Die Yang, and Craig Priest Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01413 • Publication Date (Web): 20 Jun 2018 Downloaded from http://pubs.acs.org on June 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Microfluidic Platform for High-throughput Screening of Leach Chemistry Die Yang and Craig Priest* Future Industries Institute, University of South Australia, Mawson Lakes, SA 5095, Australia. ABSTRACT: We demonstrate an optofluidic screening platform for studying thiosulfate leaching of Au in a transparent microchannel. The approach permits in situ (optical) monitoring of Au thickness, reduced reagent use, rapid optimization of reagent chemistry, screening of temperature, and determination of the activation energy. The results demonstrate the critical importance of the (1) preparation and storage of the leach solution, (2) deposition and annealing of the Au film, and (3) lixiviant chemistry. The density of sputter deposited Au films decreased with depth resulting in accelerating leach rates during experiments. Atomic leach rates were determined and were constant throughout each experiment. Annealing above 270 °C was found to prevent leaching, which can be attributed to diffusion of the chromium adhesion layer into the Au film. The optofluidic analysis revealed leach rates that are sensitive to the stoichiometric ratio of thiosulphate, ammonia and copper in the leach solution, and optimized for 10 mM CuSO4, 1 M Na2S2O3 and 1 M NH4OH. The temperature dependence of the leach rate gave an apparent activation energy of ~ 40 kJ.mol-1, based on Arrhenius’ relationship.

Screening of mineral leaching is presently time-consuming and requires large amounts of both lixiviant (leach solution) and sample. Increasingly challenging ore bodies and a pressing need for alternative (safe and selective) reagents demands rapid and inexpensive screening of leaching conditions (reagents, stoichiometry, temperature). Electrochemical methods (e.g. rotating disk electrodes) offer in situ monitoring of leach reactions; however, many methods require post-leach analysis of the spent lixiviant, which is slow and gives an averaged result. The delayed reaction-to-analysis time raises concerns for reliable scale-up to industrial processes1 when leaching solution chemistry changes with time. Further, mined ores are inherently porous or deliberately fractured to create new surfaces in micro/nano-scale environments that are chemically reactive and critical to the effectiveness of leach strategies.2,3 Microfluidic chips that mimic the chemistry and geometry of real leaching scenarios and require tiny volumes of reagent and sample offer potential advantages for screening of chemical and physical parameters. Previously, we reported an optofluidic method for in situ, continuous monitoring of Au leaching in a high aspect ratio microchannel, i.e. a model crack.4 Here, we apply this method for the investigation of the effects of (1) lixiviant preparation, (2) physical properties of the gold, and (3) optimization of the lixiviant stoichiometry. The results demonstrate that the optofluidic approach is capable of rapid screening and optimization of physical and chemical parameters using minimal reagent volumes. Gold is extremely valuable and found in its metallic form in various ores at concentrations ranging from < 5 g/t (lowgrade) to 30 g/t.5 As grade reduces, improved strategies for

recovery will be demanded, including the liberation and recovery from micro- and nano-scale environments, such as cracks and fractures created during blasting and comminution. Crystallographic orientation, lattice defects, grain size, phase composition, and surface morphology cannot be ignored6 and interfacial chemistry is important7. Chemical dissolution of deposited thin films depends on the preparation of the film and post-treatment conditions.8-10 The choice of lixiviant must also be considered carefully in terms of leach rate, reaction products (that may passivate the surface), environmental and human safety, and cost. Thiosulphate leaching of gold as a non-toxic alternative to cyanide has attracted considerable interest in recent years, which have been overviewed in recent reviews11-13. Most of these studies focused on the gold ores, showing that optimum leach conditions are strongly ore-dependent.12 Here, we choose the thiosulphate-gold system for this screening study and exploit the ability to create artificial gold ‘deposits’ in micro-cracks (7.5 µm x 4 mm x 50 mm), tune the complex lixiviant chemistry (CuSO4, Na2S2O3, and NH4OH), and adjust the leach temperature. The preparation of the lixiviant is shown to be important, resulting in leach rates that differ by up to 60%. The leach rate (nm/s) trends with the gold density depth profile, while the atomic leach rate (nAu/s) is constant. Finally, we report optimization of the relative concentration of the reagents, using small (< 1 mL) lixiviant volumes and small amount of (< 0.2 mg) gold per measurement, and short experiments time (2-12 mins).

EXPERIMENTAL

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Materials. Na2S2O3 (A. R.), CuSO4 (A. R.), and NH4OH solution (30%) were provided by Chem-Supply, Australia. Milli-Q water (18 MΩ.cm resistivity) was used to prepare all solutions. Borosilicate glass (Borofloat® 33) plates coated with 12.5 nm chromium, 50 nm gold, and S1800 photoresist (101.6 x 101.6 x 1.1 mm3) were supplied by TELIC (Valencia, CA). Solution Preparation. In this study, different concentrations of lixiviant solution were prepared by first mixing the desired amount of ammonium hydroxide solution (30%) with copper sulphate before adding sodium thiosulphate. The volumetric flask was then filled with water and mixed thoroughly. The lixiviant was freshly prepared before the leach experiments. The stability of the lixiviant was monitored using UV-Vis spectrophotometry (Ocean Optics QE65000), revealing a broad peak at 618 nm. The UV-vis results are discussed in detail later. Chip Fabrication. The chip design and the fabrication procedure has been described elsewhere4. The PDMS channel and the Cr/Au-coated glass halves of the chip were prepared in parallel, then contact bonded after treatment with oxygen plasma. The glass substrate differed slightly in this study: (i) the plate was diced into 30 x 70 mm2 substrates; (ii) cleaned with acetone to remove the protecting layer of photoresist from the surface; (iii) coated with AZ1518 photoresist; (iv) patterned by UV-photolithography in a mask aligner (EVG 620); (v) exposed Au and Cr layers were removed using KI/I2 and ammonium cerium nitrate etchant, respectively; (vi) the substrate was then cleaned with acetone to remove the AZ1518 photoresist. After plasma bonding, a post-bake for 1 h at 60 ̊C was used to strengthen the bond. The chip is then ready for the experiments. The final channels were 4 mm wide, 50 mm long, and 7.5±0.1 µm high. Experimental Set-up. The etching experimental set-up was shown in Figure 1A. The ammoniacal thiosulphate lixiviant was introduced into the channel by a syringe pump at 3 mL/h. The waste solution was collected. The thickness of the gold film was determined from the intensity of the transmitted light through the gold layer, as reported elsewhere. Briefly, there is a correlation between the thickness of the etched Au films, which can be tracked in real time. This correlation and its use will be further explained, see Characterization of Au Layer section. A white LED backlight was used to illuminate the device through a defined illumination window (3 mm x 48 mm). The images of the gold film were captured during the leach experiment at approximately 15 mm downstream of the inlet using a microscope camera (LEICA-DFC310-FX) mounted to a macroscope (LEICA-Z16-APO). Video analysis software (Tracker 4.11.0) was used for measurement of greyscale data from the image files. For each channel and liquid combination, a series of three independent measurements were conducted, showing good agreement between experiments. The bulk etching experiments were carried out by immersing Au plates (30 cm x 70 cm, surface was activated by oxygen plasma before etching) in the excess of stirring etchant (0.5 M Na2S2O3, 1 M NH4OH, and 10 mM CuSO4) at different time intervals. The thickness of the etched Au films was determined from the intensity of light transmitted through the gold layer.

Page 2 of 7

The mass change between each leach experiments was determined from the concentration of Au in the lixiviant after leaching, as measured by inductively-coupled plasma mass spectrometry (ICP-MS) (Agilent 8800). Density of gold was calculated, based on the change in mass of Au compared with the change in film thickness. Characterization of Au Layer. The Au layer thickness was calibrated by measuring the light transmitted through Aucoated glass plates (grey-scale values) with known Au thickness. The calibration plates were prepared by activating the Au surface by oxygen plasma and immersing them in bulk lixiviant (0.5 M Na2S2O3; 1 M NH4OH; 10 mM CuSO4) at different time intervals. The thickness of the calibration plates was measured by optical profilometry (Wyko NT9100) and correlated with the grey-scale measured from microscopy images using the optical method given in this study, Figure 1B. The calibration result differs to that reported previously because a different source of Au film was used. The dashed line is a linear fit between 50-15 nm, where saturation of the light is avoided. Experimental data for Au thickness less than 15 nm are therefore discarded in this study. The etched gold film thickness was then determined from the calibration curve in the defined thickness range. For selected experiments, the chemical composition of the Au layer was determined by X-ray photoelectron spectroscopy (XPS). The x-ray source was an aluminium monochromatic source with a photon energy of 1486.8 eV and a power of 225 W. A charge neutraliser was used. The survey step size was 0.5 eV and the dwell time used was 55 ms. The depth and area of analysis was 10-15 nm and 0.7 x 0.3 mm2, respectively.

Figure 1. (A) Schematic representation of leaching experimental set-up. Inset: schematic representation of a single high aspect ratio channel of the Au leaching chip. (B) Au thickness calibration curve (grey values versus Au thickness). The solid black dots represent the experimental data. The blue vertical dash lines designate the boundaries of the reliable calibration free of error re-

2 ACS Paragon Plus Environment

Page 3 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

gion. The red dash line refers to the linear fitting of the experimental data in the defined thickness calibration range (R2=0.98).

RESULTS AND DISCUSSION Lixiviant Preparation. Thiosulphate leaching is a sensitive process and the detailed chemistry of the oxidative ammoniacal leaching is complex11,12. Briefly, Cu(II) amine complex acts as a catalyst for the oxidation of gold in the presence of the thiosulphate, presumably by being reduced to the Cu(I) amine, see Eq. (1).14 Au + 2 S2O32- + [Cu(NH3)4]2+ →[Au(S2O3)2]3- + [Cu(NH3)2]+ + 2 NH3 (1) Meanwhile, Cu(II) can also react with thiosulfate to form a series of stable or metastable sulphur-oxygen species that could affect leaching kinetics.12 The reduction of Cu(II) by thiosulphate is rapid but is suppressed in the presence of excess ammonia.11 Thus, directly mixing CuSO4 solution with an excess of thiosulphate reduces Cu(II) to Cu(I). This can be observed; the blue solution becomes colorless. Subsequent addition of the ammonia solution forms the light blue Cu(I) amine complex, which does not favor Au leaching, see Eq. (1). Therefore, the preparation of the lixiviant was always carried out via pre-mixing Cu(II) and NH4OH solutions before the addition of Na2S2O3 to maximize the presence of Cu(II) in our leach experiments. Nonetheless, the Cu(II) in the prepared lixiviant did gradually reduce to Cu(I) over time. This can be tracked using the characteristic UV-Vis peak for Cu(NH3)42+ at 618 nm15, see Figure 2. After 2 and 5 h, approximately 96% and 86% of the Cu(II) complex remained, respectively. After 1 week, equilibrium was reached and approximately 26% of the Cu(II) complex remained. The reactivity of the aged solution was evaluated by leaching of gold in the microchannels, showing that a significant reduction (~ 60%) in the average leach rate was observed after aged one week but no obvious degradation in 2 h.

stored for 2 h, 5 h, 1 days, 3 days, and 6 days respectively at room temperature. Inset: figure showing the percentage of the Cu(II) complex remaining in the solution as a function of time (the solid line is an exponential fit of the experimental data).

According to the literature, the rate of thiosulphate oxidation is at least 40 times greater in the presence of oxygen12 and the mechanism is complex and not yet well understood. For long term use, it is recommended that the lixiviant be prepared using freshly boiled, double-distilled water and stored in an air tight bottle. Here, we freshly prepared and used the solutions within 2 h to minimize thiosulphate oxidation effects. Gold Density and Annealing. The density profile of the gold film was characterized by leaching a large area sample to different depths (measured by transmitted light) and comparing with the amount (moles) of Au leached. The latter was determined from the concentration of Au in the lixiviant after leaching, as measured by ICP-MS. The density of the Au film increased with film thickness, starting at ~ 4 g.cm-3 near the substrate (40 nm depth) and increasing to ~ 15 g.cm-3 at the exposed surface, Figure 3A. This density profile is consistent with previous studies.16 Low density underlayers are due to a high density of defects and porous grain boundaries at low temperature deposition conditions.6 Later in the deposition (thicker films), kinetic energy of plasma species has been converted to heat at the substrate surface, allowing greater mobility of arriving atoms and, therefore, less defects and grain boundaries. In our case, the density at the surface (15 g.cm-3) is less than bulk gold (19.3 g.cm-3) but this is not unusual for sputter-coated Au films.17

Figure 2. Absorption spectra of Cu(II) in copper (II)-thiosulphateammonia solution (10 mM CuSO4, 0.5 M Na2S2O3, and 1 M NH4OH) from top to bottom: freshly prepared solution (0 h),

3 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. (A) Density profile of the gold film. The Au surface before leaching is located at zero depth (Dashed line serves as a guide to the eye). (B) Comparison of Au leach rate versus film depth for experiments in bulk solution and microchannels (10 mM CuSO4, 0.5 M Na2S2O3, and 1 M NH4OH). Results are shown for both the thickness and amount of Au removed, with the latter determined from the density profile in figure part A.

Figure 3B compares leaching rates for the same Au films located in a microchannel and bulk-scale experiment. The leach rates were consistently faster in the microchannel. With all other parameters held constant, this effect appears to be due to the refreshment of the lixiviant (3 mL/h), which continuously removes reaction products (the lixiviant residence time is ~ 1.8 s). It is not clear what the flow rate in, for example, a heap leach scenario would be; however, this result shows that flow rate and geometry must play a significant role.

Page 4 of 7

zinc oxide film8 and tin-doped indium oxide films9, which was ascribed to the change of the micro-crystallinity in the film. The above results suggest that thermal annealing could affect the density and therefore impact the leach performance. Figure 4 shows results for leaching of annealed gold films using the same reagent. The atomic leach rates were determined at several depths (6, 8.5, 16.3, and 26.4 nm) and the average value is plotted against annealing temperature (error bars represent the standard deviation of the obtained atomic etch rates). No significant difference in the leach rate was observed up to 200 °C (473 K). At 250 °C (523 K), the leach rate reduced slightly. At 270 °C (543 K), no leaching was observed. This morphological evolution of the gold film upon thermal annealing has been reported in the literature19-21. As the annealing temperature increases, the lattice defects decrease, crystal size increases, and film becomes denser. This may be what we observe for annealing at 250 °C. Beyond 270 °C, the Au atoms in the film are more mobile and migration of the Cr adhesion layer into the Au is possible.22 In Table 1, the XPS analysis of thermal treated Au (at 270 °C) showed a considerable amount of Cr at the surface (scan depth 10-15 nm) compared to that of untreated one, confirming the migration of Cr into the Au layer. Note that the detected high ratio of C and O might be due to the surface contamination caused by direct exposure of the Au surface to the air and a change of C and O ratio after thermal treatment possibly because of the reduced surface contamination and increased amount of Cr oxides on the surface. Miller et al. demonstrated this for temperatures between 275-350 o 22 C. It is therefore likely that annealing above ~ 270 °C forms a Cr/Au alloy that is not leached by the thiosulphate chemistry.

Figure 4. Leach rate plotted against Au film depth for samples annealed at different temperatures. The lixiviant chemistry was 10 mM CuSO4, 0.5 M Na2S2O3, and 1 M NH4OH.

Figure 3B shows the leach rate plotted against the etch depth. The leach rate varied with depth for both the channel and bulk leach experiments, with initial rates slow and then accelerating as gold was leached from the surface. This is consistent with the density profile of the gold layer. Once the higher density ‘crust’ of the gold layer (< 5 nm) is removed, the leach rate (nm/s) increased with increasing porosity of the underlayer. Note that sputter-coated metal films are widely known to exhibit a higher density ‘crust’, with a porous tubular structure underneath.6,18 In short, to validate this reasoning, the measured density (Figure 3A) was used to replot the leach rate in terms of moles Au per unit area, rather than thickness. The result is a constant atomic leach rate (3.8 µmol.m-2s-1 with standard error of 0.2) over the full range of film thicknesses. Similar etch behavior has been observed on aluminum-doped

4 ACS Paragon Plus Environment

Page 5 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 5. (A) Au atomic leach rate for varied CuSO4 concentration (0.5 M Na2S2O3; 1 M NH4OH). (B) Au atomic leach rate for varied NH4OH (10 mM CuSO4; 1.0 M Na2S2O3) and Na2S2O3 (10 mM CuSO4; 1 M NH4OH) concentration.

Table 1. Near-surface (