View Article Online / Journal Homepage / Table of Contents for this issue

THE
ANALYST

FULL PAPER
Separation of polythionates and the gold thiosulfate complex in
gold thiosulfate leach solutions by ion-interaction
chromatography

John W. OReilly,a Matthew J. Shaw,a Greg W. Dicinoski,a Andrew C. Grosse,a

Yasuyuki Miurab and Paul R. Haddad*a www.rsc.org/analyst

a Australian Centre for Research on Separation Science, School of Chemistry, University of

Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

Tasmania, GPO Box 252-75, Hobart 7001 Tasmania, Australia.

E-mail: [email protected]; Fax: +61 3 6226 2858; Tel: +61 3 6226 2179
b Department of Chemistry, Faculty of Science, Tokai University, Hiratsuka, Kanagawa

259-1292, Japan

Received 2nd May 2002, Accepted 17th May 2002

First published as an Advance Article on the web 11th June 2002

A method for the separation of the polythionates (SxO622, x = 35) in gold thiosulfate leach solutions using
ion-interaction chromatography with conductivity and ultraviolet (UV) detection is described. Polythionates were
eluted within 18 min using an eluent comprising an acetonitrile step gradient at 0.0 min from 15% v/v to 28% v/v,
3 mM TBAOH, and 2.5 mM sodium carbonate, operated using a Dionex NS1-5 m column with guard. The
developed method was capable of separating the gold thiosulfate complex ion in standard solutions, but
quantification of this species in realistic leach solutions proved impractical due to a self-elution effect that caused
the gold peak to be eluted as a broad band. Detection limits for polythionates using a 10 L injection volume
ranged between 16 mg L21 (523 M) for conductivity and 0.813 mg L21 (468 M) for UV detection, based
on a signal-to-noise ratio of 2. Calibration was linear over the ranges 52000, 10-2000 and 252500 mg L21 for
trithionate, tetrathionate and pentathionate, respectively. The technique was applied successfully to leach liquors
containing 0.5 M ammonium thiosulfate, 2 M ammonia, 0.05 M copper sulfate and 20 % m/v gold ore.

1. Introduction The process still has several problems to overcome before it can
be considered to be an economic alternative to cyanide. The
There have been extensive efforts in recent years to find a less chemistry is more complicated and less understood than the
toxic alternative to cyanide for the leaching of gold from cyanide system, with reagent consumption often being prohibi-
ores.13 These efforts have been motivated by concern over the tively high. Extraction of the gold from the leach is also more
use of cyanide and the potential of this highly toxic substance to difficult. A recent review of thiosulfate leaching for gold by
damage the environment, highlighted by the tailings dam breach Aylmore and Muir1 discusses these points in more detail.
at the AURUL gold mine in Baia Mare, Romania (January Polythionates (SxO622), primarily tri- (S3O622) and tetra-
2000). A second motivating factor for this research is the desire thionate (S4O622), comprise one important group of compounds
to leach certain ores types that can often not be handled present in the leach solution and these species are formed
economically by cyanide, such as carbonaceous and sulfidic predominantly as degradation products of the thiosulfate ion. A
ores. One of the most promising alternatives under considera- major cause of this decomposition is the copper(II) catalyst
tion is thiosulfate, with which leaching is conducted in required to enable appreciable leaching, which in the absence of
ammoniacal solution using copper(II) as catalyst. The proposed oxygen oxidises thiosulfate to tetrathionate by the simplified
mechanism for the leach is given in eqns. (1) and (2).1,4 overall reaction shown in eqn. (5).1,6,7
Anodic reaction: 2Cu(NH3)42+ + 8S2O322 ? 2Cu(S2O3)352 + S4O622 + 8NH3
(1) (5)
Au + 2S2O3 22 ? Au(S2O3)2 32 + e2
This process is more complex in the presence of oxygen.8
Cathodic reaction: Tetrathionate can, in turn, decompose to trithionate and
thiosulfate through the reactions given in eqns. (6)(8).1,6
Cu(NH3)42+ + 3S2O322 + e2 ? Cu(S2O3)352 + 4NH3 (2)
S4O622 + S2O322 ? S5O622 + SO322 (6)
Some have also argued that the anodic reaction occurs with the
gold entering solution as an ammine intermediate (Au(NH3)2+) S4O622 + SO322 ? S3O622 + S2O322 (7)
prior to formation of the more stable thiosulfate complex.5 To S5O622 + 3OH2 ? 5/2 S2O322 + 3/2 H2O (8)
complete the process, copper(I) is reoxidised to copper(II) with
This results in the overall reaction given by eqn. (9), with the
oxygen as shown in eqn. (3).
process being highly catalysed by the presence of thiosulfate
4Cu(S2O3)352 + O2 + 16NH3 + 2H2O ? ions.
4Cu(NH3)42+ + 12S2O322 + 4OH2 (3) 2S4O622 + 3OH2 ? S3O622 + 5/2 S2O322 + 3/2 H2O (9)
This gives the overall leaching reaction (eqn. (4)) as: Monitoring polythionate concentrations in leach solutions is
essential, both as an indicator for the level of thiosulfate
4Au + 8S2O322 + O2 + 2H2O ? decomposition and to assist in efforts to understand and
4Au(S2O3)232 + 4OH2 (4) optimise the leaching process. For systems utilising commercial

906 Analyst, 2002, 127, 906911 DOI: 10.1039/b204118a

This journal is The Royal Society of Chemistry 2002
 View Article Online

ion-exchange resins for gold recovery a knowledge of tri- and College, PA, USA) was used to deliver the PCR reagent at a
tetrathionate concentrations is additionally critical since these flow rate of 1.0 mL min21. A reaction coil (150 cm 3 0.3 mm
have been identified as ions that strongly compete with the gold i.d.) was fitted between the Teflon mixing tee and the visible
thiosulfate complex for sites on the resin,9 thereby influencing detector.
the gold recovery. The columns used throughout this work were a Dionex NG1
Analytical methods for polythionates are widespread in the and NS1-5 m in series, except for the PDA study in which a
literature, with ion-chromatography (IC) being the most standard NS1 column (10 m packing) was used.
common for mixtures of these ions.10,11 However, application
of these methods to thiosulfate leach liquors has been minimal
2.2 Reagents
to date with the only report being that of Zou et al.12 who
applied their IC method to heavily diluted liquors. Few details
Potassium trithionate (K2S3O6) was synthesised according to
of the analysis or leach were provided however, and trithionate
the method of Stamm et al.14 and recrystallised from water
was the only polythionate quantified. There has been a similar
initially at 35 C. Sodium tetrathionate (Na2S4O62H2O, 98%,
Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

lack of research conducted on the determination of the gold

Aldrich Chemicals, Milwaukee, WI, USA) was further purified
thiosulfate complex by chromatographic techniques. The only
by dissolving in a minimal amount of water, filtering the
reference to this ion has been a qualitative study using paper
solution (0.45 m, Gelman Scientific, Lane Cove, NSW,
chromatography, thin layer chromatography and paper electro-
Australia) to remove the insoluble elemental sulfur present,
phoresis.13 There is therefore a requirement for a robust,
followed by precipitation through the addition of ethanol to the
selective and sensitive chromatographic technique, capable of
solution. Potassium pentathionate (K2S5O61.5H2O) was pre-
providing essential information to better understand and
pared according to the method of Goehring and Feldmann.15
develop the thiosulfate leaching process for gold ores.
The pentathionate salt was recrystallised twice from 2 M
In this study we will demonstrate the first separation of gold
hydrochloric acid (BDH chemicals, Kilsyth, VIC, Australia)
thiosulfate together with the polythionates (x = 35) using ion-
below 60 C. All polythionates used were stored below 25 C
interaction chromatography. The problems encountered with
after drying at room temperature.
the analysis of gold thiosulfate by this technique in leach
All other chemicals used in this work were of analytical
solutions will be discussed, and the developed method applied
reagent grade unless otherwise specified. Acetonitrile (Hi-
to the determination of polythionates in these samples. The
PerSolv, far UV grade), copper sulfate pentahydrate, sodium
chromatography of copper in the optimised system is also
carbonate, sodium thiosulfate pentahydrate and nitric acid were
examined.
obtained from BDH chemicals. Sulfuric acid and ammonia
(28% w/w) were obtained from APS chemicals (Auburn, NSW,
Australia). Tetrabutylammonium hydroxide (TBAOH, 40% w/
2. Experimental w) and 4-(2-pyridylazo)resorcinol monosodium salt hydrate
(PAR) were obtained from Aldrich, sodium gold thiosulfate
2.1 Instrumentation dihydrate (99.9% on metal basis) from Alfa Aesar (Ward Hill,
MA, USA), and ammonium thiosulfate from Reidel-de Han
The ion-chromatograph used in this work was a Dionex DX-500
(Seelze, Germany).
(Sunnyvale, CA, USA) system consisting of a GP50 gradient
The ore concentrates used were supplied by Osleach Pty. Ltd.
pump, an AS50 autosampler with thermal compartment, a
(Currumbin, QLD, Australia). All solutions and eluents were
CD20 conductivity detector and an AD20 ultra-violet/visible
prepared in water purified using a Millipore Milli-Q (Bedford,
(UV/VIS) absorbance detector. If a second UV detector was
MA, USA) water purification system. Sodium carbonate was
required, a Waters (Milford, MA, USA) 486 detector was used,
dried at 110 C prior to use. Eluents were filtered through a
connected to the data system via a Dionex UI20 universal
nylon 0.45 m filter (Alltech Associates Pty. Ltd., Baulkham
interface. A pump flow rate of 1.0 mL min21 and column oven
Hills, NSW, Australia) prior to use.
temperature of 35 C were used throughout this work. For UV
The regenerant used for the suppressed conductivity detec-
detection a wavelength of 215 nm was used except where
tion was 5 mM sulfuric acid. The post-column reagent used for
otherwise specified. Conductivity detection was used with
copper analysis was based on that used by Shaw et al.16 and
suppression provided by a Dionex ASRS-Ultra in the chemical
contained 0.5 mM PAR, 2.6 M ammonia and 0.85 M
suppression mode. The regenerant was delivered by a Waters
ammonium nitrate.
510 high performance liquid chromatography (HPLC) pump at
a rate of 4.0 mL min21. All chromatographic data were
collected using Dionex PeakNet software version 5.1. A 100 L 2.3 Leaching experiments
injection loop was used throughout this work. For experiments
using a 10 L injection the autosampler was programmed to Leaching experiments were conducted in a 1000 mL, 3-neck,
perform a partial loop injection. flat-bottomed flask. A sample (100 g) of the ore under
When photodiode array (PDA) detection of the gold investigation was slurried in water and added to the flask,
thiosulfate complex was required, a Waters Alliance 2690 followed by 6.24 g of copper sulfate and 69 mL of concentrated
HPLC instrument was used, fitted with a Waters 996 PDA aqueous ammonia. The solution was made to a volume of ~ 470
detector. The chromatograms obtained were scanned between mL with distilled water and heated to 50 3 C with overhead
200 and 600 nm at 1.2 nm intervals, with the data collected mechanical stirring at ~ 140 rpm. The stirring arm used was
using Waters Millennium software (version 3.05.01). glass with Teflon paddles. A condenser was fitted to the flask to
The atomic absorption spectrometer (AAS) used for the minimise evaporation. When the solution reached the required
fraction collection study was a Varian SpectrAA-800 (Varian, leach temperature, 37.05 g of ammonium thiosulfate was added
Mulgrave, VIC, Australia), using an air (13.68 L min21) with the aid of distilled water to give a final volume of 500 mL.
acetylene (1.93 L min21, BOC Gases, Chatswood, NSW, The time at which the thiosulfate was added to the leach was
Australia) flame. The gold concentration was measured at 242.8 denoted time zero (t = 0).
nm using a Photron (Photron, Narre Warren, VIC, Australia) For sampling, the condenser was removed and approximately
gold hollow cathode lamp, using standards prepared in the 6 mL samples were removed from the leach solution with
chromatographic eluent. stirring still in progress. The solid present in the sample was
For the post column reaction (PCR) study into copper elution, removed by filtration through a 0.45 m nylon syringe filter,
a Model 350 HPLC pump (Scientific Systems Inc., State and the samples were analysed immediately.

Analyst, 2002, 127, 906911 907

 View Article Online

3. Results and discussion shown in Fig. 2 for both conductivity and UV detectors. These
conditions were chosen to elute the gold thiosulfate between
3.1 Optimisation of separation conditions tetra- and pentathionate, since the gold thiosulfate peak became
extremely tailed with increasing residence time on the col-
A recent review of the literature11 has shown that one of the umn.
most common IC systems employed for the separation of
polythionates is the use of ion-interaction chromatography
employing a reversed phase Dionex NS1 column (with NG1 3.2 Analysis of synthetic leach solutions
guard), TBAOH as the ion-interaction reagent, sodium carbon-
ate as the primary eluting ion, and acetonitrile as the organic A wide range of thiosulfate leach conditions has been reported
modifier. It may seem surprising that such an alkaline eluent has in the literature1 varying between the extremes of 0.12 M for
been used since it is well documented10,17,18 that tetrathionate thiosulfate, 0.16 M for ammonia and 0.0010.1 M for copper.
and pentathionate are unstable at alkaline pH through reaction Based on a recent review1 and our own experience it was
Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

with the hydroxide ion. However, the rate of decomposition concluded that approximately 70% of leaching regimes use
appears to be slow enough to prevent it from hindering the 50.5 M thiosulfate, 52 M ammonia and 550 mM copper. In
analysis. It can be expected that the use of an alkaline eluent order to ensure the chromatographic method was able to
would cause decomposition of a small portion of the injected separate the species present in realistic leach conditions a
polythionate as a continuous process on the column, thereby synthetic leach solution containing the above concentrations of
slightly increasing the detection limit. However, this eluent does these species was used. While it is unlikely that these extreme
have the advantage in that it is compatible with both suppressed conditions would be used widely in any real leach solution, it
conductivity and UV detection modes, which is useful since provides a very challenging matrix in which to evaluate the
trithionate has a weak UV chromophore11,19 and is more suited method.
to detection by conductivity, while UV detection is more Initial studies involving the elution of standard solutions of
sensitive for tetra- and pentathionate.19 the gold thiosulfate complex from the NS1 column demon-
Preliminary work, combined with evaluation of the previous strated that the observed peak area was significantly affected by
literature, suggested that the dominant factors in the separation the concentration of thiosulfate in the current and previously
process were the acetonitrile and carbonate concentrations of injected sample. This effect could be minimised by adding a
the eluent. It was therefore decided to optimise the eluent using small amount of thiosulfate to the eluent. For the ensuing
these parameters, keeping the TBAOH concentration of the studies, 40 M sodium thiosulfate was added to the optimised
eluent constant at 3 mM, which was found suitable to maintain eluent to account for this problem. As expected, this caused an
a stable dynamic loading on the column. In addition, it was increase in baseline noise for both detectors, but did not
found that when using isocratic eluents, the trithionate appeared significantly affect the separation. It is proposed that the
as a split peak under conditions for which the other ions of
interest were eluted quickly. To resolve this problem it was
necessary to insert an acetonitrile step gradient from 15% to the
higher separation concentration during the analysis, which
will be discussed later.
The UV spectra of the polythionates are known, with lmax
values of < 200 nm for trithionate and between 210 and 220 nm
for tetra- and pentathionate.19 The absorption spectrum (over
the region 200600 nm) of the gold thiosulfate complex showed
an absorbance maximum at ~ 205 nm , although in view of the
additional baseline noise observed at this and lower wave-
lengths it was decided to use a wavelength of 215 nm for this
work to achieve an improved signal-to-noise ratio.
Fig. 1 demonstrates the effect of acetonitrile and carbonate
concentrations on the separation. Acetonitrile provides no
means of changing the separation selectivity, but affects the
analyte retention by influencing the amount of adsorbed
TBAOH on the stationary phase. On the other hand the
carbonate concentration strongly influences the selectivity for
the gold thiosulfate complex with retention orders of
Au(S2O3)232 > S5O622 > S4O622 > S3O622 being observed
when no carbonate is present and S5O622 > S4O622 >
Au(S2O3)232 > S3O622 at 10 mM carbonate. This effect can be
explained by the higher charge (23) on the gold thiosulfate
complex compared to the polythionates (22), indicating that the
gold thiosulfate complex will be more strongly influenced by
eluent concentration.20 This ability to move the gold thiosulfate
peak relative to the peaks for the polythionates is advantageous
as it can be used to decrease interferences when required.
Based on the results from this study, and with the gradient
step time set to occur at injection, the optimum eluent was
determined to consist of a acetonitrile step gradient from 15% to
28% v/v occurring at injection, with 3 mM TBAOH and 2.5 mM
sodium carbonate maintained in the eluent at all times. After 14 Fig. 1 Effect of (a) acetonitrile and (b) carbonate concentrations in the
eluent on retention and separation of polythionates and the gold thiosulfate
min, the acetonitrile concentration was reduced to 15% and held complex. Eluent compositions: (a) 3 mM TBAOH, 2.2 mM sodium
for a period of 4 min, yielding a total analysis time of 18 min, carbonate, acetonitrile step gradient at 2.5 min from 15% v/v to the indicated
including the time required to re-equilibrate the column with the final composition; (b) 3 mM TBAOH, acetonitrile step gradient at 2.5 min
initial conditions. The separation attained using this eluent is from 15% v/v to 30% v/v. For remaining conditions, see text.

908 Analyst, 2002, 127, 906911

 View Article Online

beneficial effects on reproducibility of adding thiosulfate to the is probably within the uncertainty of the experiment. For the
eluent arose from minimising dissociation of the gold thio- leach sample the gold was spread across the first three fractions
sulfate complex as it travels through the column. collected, indicating that some kind of self-elution effect
Fig. 3(a) shows a chromatogram of 100 L of the artificial influences the separation in the sample matrix. These results
leach solution containing 100 mg L21 gold(I) as the thiosulfate demonstrate that gold thiosulfate cannot be determined in the
complex (A), overlaid with a thiosulfate standard containing the leach solutions by this method and the species was therefore not
same concentration of gold (B). Fig. 3(b) shows an injection of considered further. As a result, thiosulfate was not added to the
the leach solution but using only a 10 L injection. Fig. 3(a) eluent in the remainder of the work discussed in this paper.
highlights a major problem for the determination of the gold The only difficulty that the leach matrix provides for the
thiosulfate peak in these mixtures since it is apparent that the polythionate determination is the unknown shoulder peak on the
gold peak is greatly reduced in size when injected together with front of the tetrathionate peak (Fig. 3(b)). To test whether this
the leach matrix. peak was caused by an anionic copper species such as
To establish the fate of the gold, column fractions were Cu(S2O3)232 or Cu(S2O3)352 the elution of copper in the
Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

collected during analysis of samples containing 100 mg L21 system was also examined. A post-column reaction system
gold (as the thiosulfate complex) and analysed using AAS. For using PAR followed by visible detection at 510 nm was used to
each analysis four fractions were collected, corresponding to monitor copper elution in further injections of the artificial
04, 48, 812 and 1216 min, where t = 0 was the injection leach solution. The results are shown in Fig. 4. The retention of
time. The results of this study are provided in Table 1, which copper varied according to the injection volume, with different
shows the average concentration of gold in each fraction and the results being obtained for injections of 10 and 100 L. In both
total gold recovery. Theoretically, all the gold should be cases there was a significant peak in the region of the void
contained in the 812 min fraction (see Fig. 3(a)) with a volume, presumably corresponding to cationic complexes of
concentration of 2.5 mg L21. Gold recovery was high for both copper such as the tetra-ammine copper(II) complex,
the standard (115%) and leach (122%) solutions, although this Cu(NH3)42+. However, there was also a diffuse peak observed

Fig. 3 (a) Overlay of UV chromatograms for: (A) an artificial leach

Fig. 2 (a) Conductivity and (b) UV chromatograms of optimised solution containing 0.5 M ammonium thiosulfate, 2 M ammonia, 0.05 M
separation for the polythionates and gold thiosulfate. Optimum eluent copper sulfate and 100 mg L21 gold (as thiosulfate complex); and (B) a 100
composition, acetonitrile step gradient at 0.0 min from 15% v/v to 28% v/v, mg L21 gold standard (as thiosulfate complex), using the optimised
3 mM TBAOH, 2.5 mM sodium carbonate. Sample composition, 20 mg L21 separation conditions. Injection volume 100 L. (b) UV chromatogram of
thiosulfate, 40 mg L21 trithionate, 21 mg L21 tetrathionate, 20 mg L21 gold the artificial leach solution under the same conditions except using a 10 L
(as thiosulfate complex), 21 mg L21 pentathionate. injection volume. The conductivity chromatograms were similar.

Table 1 Average gold concentrations observed in IC column fractions, and total gold recovery for 100 L injections of 100 mg L21 gold thiosulfate standard
and artificial leach solutions containing 100 mg L21 gold as thiosulfate, measured by AAS [n = number of replicates]

[Au]/mg L21
Average total
Sample 04 min 48 min 812 min 1216 min gold recovery (%)

100 mg L21 gold standard (as thiosulfate complex) [n = 2] 0.03 0.03 2.78 0.05 115
Artificial leach spiked with 100 mg L21 gold (as thiosulfate complex) [n = 3] 0.58 1.03 1.39 0.04 122

Analyst, 2002, 127, 906911 909

 View Article Online

for the 10 L injection (Fig. 4(a)) between 3.45.0 min, and a significantly from the line for the low concentration poly-
small peak at 3.3 min, followed by a large poorly shaped peak thionate solutions. It was therefore more accurate in most cases
between 4.0-5.8 min for the 100 L injection (Fig. 4(b)). These to define two lines of best fit, one for low and another for high
may correspond to the copper thiosulfate species mentioned concentrations. Detection limit and linearity data (without the
earlier. Recoveries were not quantitative, since after many presence of thiosulfate in the eluent) are summarised in Table 2.
injections of leach solutions, flushing the column with 0.5 M Before the linearity and detection data were determined, it was
ammonia produced a major response from the PCR system, deemed necessary to ensure that the polythionates did not
which took some hours to dissipate. For both injection volumes interact with one another in standard solutions, as this would
no copper was observed in the region of the shoulder peak and have affected quantification. This was investigated by preparing
this peak remains unidentified. a 203 mg L21 trithionate, 201 mg L21 tetrathionate and 298 mg
L21 pentathionate solution and comparing the peak area
obtained against standards of the same concentration injected
3.3 Linearity and detection limits individually. No significant differences were observed.
Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

Reproducibility data was calculated from triplicate 10 L

The detection limits for the polythionates using the optimised injections of the linearity standards. Using conductivity detec-
method were determined using a 10 L injection volume, with tion, polythionate concentrations of 420 (S3O622), 4100
the limit taken as the concentration of analyte registering a peak (S4O622) and 4125 (S5O622) mg L21 all yielded peak area
two times the peak-to-peak baseline noise. Linearity was tested reproducibility values of < 2% RSD. The equivalent concentra-
from the detection limit to 2000 mg L21 for trithionate, 2000 mg tions for UV detection were 800, 50 and 100 mg L21. The high
L21 for tetrathionate and 2500 mg L21 for pentathionate. Least concentration for trithionate using UV detection reflects the
squares lines of best fit for the data gave R2 values of > 0.999, comparatively high detection limit for this ion.
for both conductivity and UV detection. Closer investigation of If required, the detection limits could be reduced by using a
the calibration plots showed that the data points deviated larger injection volume. This may be necessary for the
determination of pentathionate, which is present in much lower
concentrations than the tri- and tetrathionate in the leach
solutions. Equally, if polythionate concentrations increased
above the tested range, a smaller injection volume or dilution
may be required.
In view of the complex nature of the leach solution, it would
be normal to consider matrix matching between the standards
and samples. This is inappropriate here since the matrix
contains the species under examination and for reasons
discussed in the Introduction, the concentrations present change
with time, prohibiting any use of a standard additions method.
Matrix matching the thiosulfate content of the leach was also
not possible due to eqns. (6)(8) whereby the thiosulfate will
catalyse decomposition of the polythionates. For these reasons,
all standards were prepared in Milli-Q water only.

3.4 Analysis of an actual leach solution

Fig. 5 shows UV chromatograms from thiosulfate leach liquors

of: (a) oxide and (b) sulfide gold ore concentrates, with the
sample in each case taken 5 h after leaching was commenced.
Polythionate concentrations were determined from these chro-
matograms by comparison to three-point calibration curves. The
calculated concentrations for the oxide ore were 792 mg L21 for
tetrathionate and 19 mg L21 pentathionate with the trithionate
concentration above the calibration range. The corresponding
values for the sulfide ore were 1716 mg L21 trithionate, 822 mg
Fig. 4 (a) Chromatogram obtained at 510 nm for the artificial leach L21 tetrathionate and 11 mg L21 pentathionate. These chroma-
solution (0.5 M ammonium thiosulfate, 2 M ammonia, 0.05 M copper
sulfate, 100 mg L21 gold as thiosulfate) using the optimised separation
tograms show that the method was able to handle the leach
conditions, with PAR as post-column reagent. (b) Identical chromatogram conditions and both ore types quite well, although the shoulder
of the same solution using a 100 L injection volume. (c) UV chromatogram peak previously noticed in the artificial leach solutions was still
of the same leach solution for comparison (10 L injection volume). present, in addition to another unknown peak at 5.25 min. This

Table 2 Polythionate detection limit and linear range data for the optimised method by both conductivity and UV detection

Detection limit/mg L21 Linear range/mg L21 (R2 values)

Analyte Conductivity UV Conductivity UV

Trithionate 1 13 5100 (0.9998) 502000 (0.9999)

502000 (0.9999)
Tetrathionate 3 0.8 10200 (0.9998) 51200 (0.9999)
1002000 (0.9994) 502000 (0.9999)
Pentathionate 6 1 25500 (0.9999) 6500 (0.9999)
5002500 (0.9995) 5002500 (0.9992)

910 Analyst, 2002, 127, 906911

 View Article Online

second peak does not interfere with the analysis. Using the PCR develop a method that can determine this ion in the sample
system described previously, no peak corresponding to the matrix in addition to the polythionates.
retention time of this unknown peak was observed, indicating
that it was not a copper complex or other metal that reacts with
PAR. 5. Acknowledgements
The authors wish to thank the Australian Research Council,
Newcrest Mining Ltd. and Dionex Corporation for financial
4. Conclusions support, and Osleach Pty. Ltd. for donation of the ore
samples.
A method for the successful determination of polythionates in
complex leach liquors with good sensitivity and selectivity has
been developed. While the method has also been used to References
Published on 11 June 2002. Downloaded by University of Nevada - Reno on 08/02/2017 20:29:26.

demonstrate the first reported chromatographic separation of

gold thiosulfate in standard solutions, the results for leach 1 M. G. Aylmore and D. M. Muir, Miner. Eng., 2001, 14, 135174.
solutions were disappointing and a self-elution effect prohibited 2 I. M. Ritchie, M. J. Nicol and W. P. Staunton, in Cyanide: Social,
successful analysis. Currently work is being undertaken to Industrial and Economic Aspects, Proceedings of a Symposium held
at Annual Meeting of TMS, New Orleans, Louisiana, USA, Feb.
1215, 2001, ed. C. A. Young, L. G. Twidwell and C. G. Anderson,
Minerals, Metals and Minerals Society, Warrendale, PA, 2001, pp.
427440.
3 G. J. Sparrow and J. T. Woodcock, Miner. Process. Extr. Metall.
Rev., 1995, 14, 193247.
4 P. L. Breuer, M. I. Jeffrey and W. L. Choo, in Cyanide: Social,
Industrial and Economic Aspects, Proceedings of a Symposium held
at Annual Meeting of TMS, New Orleans, Louisiana, USA, Feb.
1215, 2001, pp. 455468.
5 T. Jiang, J. Chen and S. Xu, in Proceedings of XVIII International
Mineral Processing Congress, Sydney, Australia, May 2328, 1993,
pp. 11411146.
6 J. A. Byerley, S. A. Fouda and G. L. Rempel, J. Chem. Soc., Dalton
Trans., 1973, 889893.
7 P. L. Breuer and M. I. Jeffrey, Miner. Eng., 2000, 13, 10711081.
8 J. A. Byerley, S. A. Fouda and G. L. Rempel, J. Chem. Soc., Dalton
Trans., 1975, 13291338.
9 G. P. OMalley and M. J. Nicol, in Cyanide: Social, Industrial and
Economic Aspects, Proceedings of a Symposium held at Annual
Meeting of TMS, New Orleans, Louisiana, USA, Feb. 1215, 2001,
pp. 469483.
10 T. Koh, Anal. Sci., 1990, 6, 314.
11 J. W. OReilly, G. W. Dicinoski, M. J. Shaw and P. R. Haddad, Anal.
Chim. Acta, 2001, 432, 165192.
12 H. F. Zou, Z. J. Jia, Y. K. Zhang and P. C. Lu, Anal. Chim. Acta, 1993,
284, 5965.
13 M. Lederer, J. Chromatogr., 1978, 153, 302304.
14 H. Stamm, M. Goehring and U. Feldmann, Z. Anorg. Allg. Chem.,
1942, 250, 226228.
15 M. Goehring and U. Feldmann, Z. Anorg. Chem., 1948, 257,
223226.
16 M. J. Shaw, S. J. Hill and P. Jones, Anal. Chim. Acta, 1999, 401,
6571.
17 A. B. Roy and P. A. Trudinger, The Biochemistry of Inorganic
Compounds of Sulfur, Cambridge University Press, Cambridge,
1970.
18 E. Rolia and C. L. Chakrabarti, Environ. Sci. Technol., 1982, 16,
Fig. 5 UV chromatograms from gold thiosulfate leach solutions of: (a) an 852857.
oxide and (b) a sulfide ore concentrate, obtained from samples extracted 5 19 Y. Miura and A. Kawaoi, J. Chromatogr., A, 2000, 884, 8187.
h after leaching was commenced. For leach conditions and results of the 20 P. R. Haddad and P. E. Jackson, Ion Chromatography: Principles and
analysis, see text. Applications, Elsevier, Amsterdam, 1990, ch. 56.

Analyst, 2002, 127, 906911 911