Journal of Colloid and Interface Science 360 (2011) 508–518

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

The effect of copper on the precipitation of scorodite (FeAsO4
2H2O) under

hydrothermal conditions: Evidence for a hydrated copper containing ferric
arsenate sulfate-short lived intermediate
M.A. Gomez ⇑, L. Becze, M. Celikin, G.P. Demopoulos ⇑
McGill University, Department of Mining and Materials Engineering, 3610 University Street, Montreal, QC, Canada H3A2B2

a r t i c l e i n f o a b s t r a c t

Article history: The effect of copper sulfate on scorodite precipitation and its mechanism of formation at 150 °C was
Received 25 February 2011 investigated. Scorodite was determined to be the dominant phase formed under all conditions explored
Accepted 4 May 2011 (0.61 < Fe(III)/As(V) < 1.87, 0.27–0.30 M Fe(SO4)1.5, 0–0.3 M CuSO4, 0–0.3 M MgSO4, at 2.5 h and 150 °C).
Available online 11 May 2011
The produced scorodite was found to incorporate up to 5% SO4 and 61% Cu or Mg in its structure. The
precipitation of scorodite was stoichiometric, i.e. the Fe/As molar ratio in the solids was equal to one
Keywords: independent of the starting Fe/As ratio in the solution. The presence of excess ferric sulfate in the initial
Scorodite
solution (Fe/As > 1) was found to slow down the ordering of the H-bond structure in scorodite. Precipi-
Arsenic
Pressure leaching
tation under equimolar concentrations (As = Fe = Cu = 0.3 M), short times and lower temperatures (30–
Copper 70 min and 90–130 °C) revealed the formation of a Cu–Fe–AsO4–SO4–H2O short lived gelatinous interme-
Gelatinous intermediate diate that closely resembled the basic ferric arsenate sulfate (BFAS) type of phase, before ultimately con-
Basic ferric arsenate sulfate verting fully to the most stable scorodite phase (96 min and 138 °C). This phase transition has been traced
XPS throughout the reaction via elemental (ICP-AES, XPS), structural (PXRD, TEM) and molecular (ATR-IR,
ATR-IR Raman) analysis. ATR-IR investigation of an arsenic containing industrial residue produced during pres-
Raman sure leaching of a copper concentrate (1 h and 150 °C) found evidence of the formation of an arsenate
mineral form resembling the intermediate basic ferric arsenate sulfate phase.
Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction cated for the immobilization of arsenic (V) from arsenic-rich

Arsenic is a common toxic element encountered in the non-fer-
rous metallurgical industry and its removal-fixation remains an
important topic [1,2]. Of interest to the present work is the precip-
itation behavior of arsenic (V) during the processing of ores or con-
centrates in autoclaves under hydrothermal conditions (>100 °C).
Of particular relevance is the pressure leaching (PL) of copper sul-
fide concentrates as exemplified by the Cominco Engineering Ser-
vices Ltd. (CESL) process developed by Teck Metals [3,4]. The
CESL process typically employs relatively mild pressure leaching
conditions (150 °C, pH < 3, 1400 kPa oxygen pressure and 60 min
retention time).
During autoclave processing of sulfide feedstocks, arsenic (as
arsenate) may combine with ferric iron to form crystalline ferric
arsenate phases such as scorodite (FeAsO4
2H2O), ferric arsenate
sub-hydrate (FeAsO4
0.75H2O, FAsH) or basic ferric arsenate sulfate
(Fe[(AsO4)1 x(SO4)x(OH)x]
wH2O), BFAS) [5–7]. By far, the ferric
arsenate phase that has been most widely considered and advo-
⇑ Corresponding authors.
E-mail addresses: mario.gomez@mail.mcgill.ca (M.A. Gomez), george.demopou-
los@mcgill.ca (G.P. Demopoulos).
industrial feedstocks is scorodite (FeAsO4
2H2O) [2,8]. Detailed
information on scorodite’s (natural and synthetic) crystal structure
[7,9–11] and stability–solubility evaluation has been reported [12–
14]. In a previous paper [6] we reported scorodite to be the domi-
nant form in the temperature range 150–175 °C with Fe (III)/As (V)
ratios 0.75–1.5 and retention time of 2–24 h (without presence of
extra cations). Of interest, in addition to the type of ferric arsenate
phase that forms, is the possible loss of valuable metals (in this
case Cu) via co-precipitation; as well as the effect copper can have
on arsenic precipitation and type of phase formed. The effect of
cations (such as Cu, Zn, Ni, Co, Mn) and excess sulfate anion on
the precipitation of scorodite have been mainly examined at
atmospheric pressure conditions (95 °C) [15,16]. Under the latter
conditions, scorodite was found to contain up to 1% Cu and 4% SO4
with no measurable effect on its precipitation kinetics/yield.
Some limited data (1 study) on the effect of copper during
hydrothermal precipitation of scorodite was reported by Monhem-
ius and Swash [17]. Their tests in the presence of copper were done
at 190 °C (and not 150 °C which is the CESL operating temperature)
using a Fe/As = 1 and the presence of 0–15 g/L Cu and 20 g/L H2SO4.
Under those conditions, copper was found to have a negative effect
on the kinetics and yield of scorodite precipitation. For example,

0021-9797/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2011.05.010
 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518 509

arsenic precipitation efficiency decreased from 95% to 80% in
the presence of copper, this being further decreased when acid
was added. Moreover, they reported that the scorodite product
was in mixture with the ferric arsenate sub-hydrate (FAsH or ‘‘Type
1’’ according to the authors) [6,7] phase and that the solids con-
tained up to 3.4% copper attributing this behavior to copper’s influ-
ence on the nucleation and growth mechanism of scorodite.
It is the aim of this study to: (a) investigate how the formation
of scorodite (FeAsO4
2H2O) is influenced by the presence of foreign
cations (Cu2+ or Mg2+) under hydrothermal conditions (150 °C); (b)
verify if copper co-precipitates with scorodite (either as a substitu-
ent element or as a separate copper arsenate–sulfate phase); (c)
investigate the mechanism of scorodite (FeAsO4
2H2O) formation;
and (d) relate the findings to industrial practice via the analysis
of a Teck Metals industrial pressure leaching residue produced un-
der the CESL process conditions.
The presented work contributes to the new understanding of
ion association and crystallization events that precede the forma-
tion of crystalline structured scorodite (ferric arsenate dihydrate)
out of multi-ion complex composition solutions close to the ones
encountered in industry. Since the stability of arsenic in industrial
residues depends on the type and crystallinity of ferric arsenate
phase [6,14,18], understanding the nucleation and phase transition
phenomena [19,20] are equally important from fundamental and
practical points of view. The discovery, isolation/characterization
and monitoring of a short-lived gelatinous Fe(III)–Cu(II)–AsO4–
SO4 intermediate and its conversion towards an ordered semi-crys-
talline to crystalline phase that is reported herein provides rich in-
sight of greater implications for the broader study of colloid
nucleation and phase transition systems.
2. Experimental
2.1. Materials and procedure
The precipitation tests were conducted using 1 L aqueous solu-
tions made of reagent grade chemicals in a 1.6 L mechanically agi-
tated (300 rpm) Buchi glass autoclave. All solutions contained
0.3 M Fe(III) and variable amounts of As(V) and Cu(II) as sulfate
salts. The Fe(III)/As(V) molar ratio was varied from 0.7 to 1.9. The
heat up time from room temperature to the target temperature
of 150 °C took 2.5 h. The progress of the precipitation reactions
was monitored via sample collection and analysis at different
times. The initial pH of all solutions was 61 (free acidity 0.16 M
(15.2–17.2 g/L) H2SO4). The experimental conditions for the vari-
ous precipitation tests-cases investigated are summarized in Ta-
ble 1. Note that duplicate tests were run for all cases. The
industrial residue investigated in this study was provided by Teck
Metals and was produced in a bench scale pressure leaching auto-
clave under the CESL conditions (200 g/L solids, 15 g/L Cu, 15 g/L
free acid, 60 min, 3000 kPa total pressure and 150 °C) using a chal-
copyrite–bornite concentrate from their Antamina mine in Peru.
The major elemental components of the residue were: 0.54% As,
29.3% Fe, 1.92% Cu, and 36.3% S(tot) = 32.1% S0 + 4.2% SO4.
2.2. Characterization
All solutions were analyzed for As, Fe, Cu, and Mg concentra-
tions by ICP-AES using a Thermo Jarrel Ash Trace Scan machine.
The chemical composition of the final solids was determined by
HCl digestion followed by ICP-AES analysis. The powder (PXRD)
and selected area (SXRD) X-ray diffractograms were recorded with
a Bruker D8-DISCOVER diffractometer (Co Ka1 radiation,
k = 1.78897 Å) operating at 40 kV and 35 mA. The scans were mea-
sured from 5° to 113° 2h with 0.02° step and acquisition time of
70 s per frame (three frames were taken). The morphological char-
acterization of the produced solids was done with a Variable Pres-
sure Scanning Electron Microscope (VP-SEM) Hitachi S-3000N and
a Philips CM-200 Transmission Electron Microscope operating at
200 kV. X-ray photoelectron spectroscopy (XPS) measurements
were conducted in a Thermo Scientific K-Alpha, using a Al Ka
X-ray source at 1486.6 eV. A pass energy of 20 eV with a 0.1 eV
step size was used for all samples analyzed. The scale of the bind-
ing energy was calibrated against the Au (4f7/2) peak at 84.0 eV and
no effects due to charging or X-ray damage were observed. Colloi-
dal gels were analyzed using polarized light microscopy with a Ni-
kon DS Camera control unit DS-U2 on a Nikon Polarizing
Microscope (PM) Eclipse LV100POL. Infrared spectra were obtained
using a Perkin Elmer FTIR (Spectrum BX model) spectrometer with
a Miracle single bounce diamond ATR cell from PIKE Technologies.
Spectra over the 4000–550 cm 1 range were obtained by the co-
addition of 100 scans with a resolution of 4 cm 1. Raman spectra
were collected with an InVia Raman microscope from Renishaw
using a polarized 514 nm laser with a spot size of 65 lm with
the 50 objective. Averages of 10 scans were obtained from 1700
to 150 cm 1 with an energy resolution of 4 cm 1 using 10% of
the laser power to avoid radiation damage. More detailed informa-
tion may be found in Supplementary information section.
3. Results and discussion
3.1. Precipitation of scorodite
3.1.1. Equilibrium results
A summary of the precipitation results based on the analysis of
the final solutions and solids can be found in Tables 1 and 2. It
should be noted that these results correspond to the solutions
and solids after 10 h of reaction at 150 °C (or 12.5 h including the
heat up period). Given the elevated temperature at which the tests
were run, these results for all practical purposes are considered to

Table 1
Initial solution compositions and precipitation efficiencies at the end of the reaction period (12.5 h).

Test Initial solution Fe/As molar ratio As (M) Fe (M) Cu (M) Mg (M) Precipitation (%) efficiency Fe/As molar ratio (solid)
Fe As Cu Mg
Case 1a 1.87 0.16 0.30 – – 59.8 97.5 – 1.12
Case 1b 1.86 0.15 0.28 59.5 98.2 1.09
Case 2a 0.96 0.29 0.28 – – 98.7 91.4 – 1.01
Case 2b 1.03 0.28 0.29 99.3 91.2 1.02
Case 3a 0.68 0.41 0.28 0.07 – 99.4 57.9 4.6 0.80
Case 3b 0.61 0.46 0.28 0.06 99.6 63.3 5.09 0.85
Case 4a 0.82 0.34 0.29 0.30 – 99.1 88.3 4.18 0.91
Case 4b 0.91 0.32 0.29 0.29 99.3 89.7 3.20 1.00
Case 5a 0.90 0.30 0.27 0.27 0.27 99.3 91.2 5.78 2.98 0.96
Case 5b 0.91 0.32 0.29 0.29 0.30 98.8 90.0 6.06 4.05 1.02
510 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518

Table 2
Composition of precipitated solids at the end of the reaction period (12.5 h).

Test Initial solution Fe/As molar ratio As (%) Fe (%) Cu (%) Mg (%) S (%) Solid S/Fe molar ratio X Solid Fe/As molar ratio Phase
a
Case 1a 1.87 28.8 24.1 – – 1.6 0.11 0.16 1.12 SD-SO4
Case 1b 1.86 29.1 23.7 1.4 0.11 0.16 1.09
a
Case 2a 0.96 31.6 23.9 – – 0.2 0.01 0.01 1.01 SD-SO4
Case 2b 1.03 31.2 24.1 0.4 0.01 0.01 1.02
a
Case 3a 0.68 35.1 21.1 0.4 – 0.6 0.04 0.04 0.80 SD-SO4
Case 3b 0.56 34.2 21.8 0.3 0.9 0.07 0.07 0.85
a
Case 4a 0.82 32.5 22.0 1.32 – 0.1 0.01 0 0.91 SD-SO4
Case 4b 0.91 31.9 23.9 0.3 0.5 0.02 0 1.00
a
Case 5a 0.90 29.5 21.2 0.8 0.6 1.2 0.10 0.11 0.96 SD-SO4
Case 5b 0.96 29.7 22.6 0.6 0.7 1.6 0.12 0.13 1.02
a
SD-SO4: sulfate-containing scorodite: Fe(AsO4)1–0.67X(SO4)X
2H2O.

reflect near equilibrium conditions. Essentially nearly complete ar-
senic and iron precipitation (>90%) was observed in all tests (Cases
2, 4 and 5) with a solid Fe/As molar ratio close to one (correspond-
ing to the stoichiometry of scorodite). In contrast to Monhemius
and Swash [17], the presence of copper (or Mg; compare Case 2
to Case 4 or 5 in Tables 1 and 2) did not appear to significantly
influence the arsenic precipitation yield.
With reference to iron and arsenic precipitation as a function
of the Fe/As molar ratio, we observe the following: in Case 1, cor-
responding to nominal Fe/As ratio  2, the highest% As precipita-
tion efficiency was observed (97–98%) but also the lowest% Fe
efficiency (60%). Exactly the opposite happens in Case 3, where
a Fe/As ratio <1 (0.7) was used. This suggests that the iron ‘‘fol-
lows’’ the arsenic, i.e. the two elements precipitate as a single
compound (scorodite) as evident from a number of factors. First,
the molar ratio of iron and arsenic precipitated was found to be
around one (with some deviation in Case 3, Table 1). Similarly,
the molar Fe/As ratio close to one was obtained from the analysis
of the solids (Tables 1 and 2) which corresponds closely to the
theoretical scorodite Fe/As molar ratio. Diffraction and spectro-
scopic analysis shown later further confirm that all the products
produced once 150 °C is reached (2.5–12.5 h) are crystalline
scorodite.
3.1.2. Kinetic results
According to the data plotted in Fig. 1, the precipitation of ar-
senic and iron was essentially complete once the target tempera-
ture of 150 °C was reached (2.5 h), independent of the absence or
presence of copper and magnesium.
It is worthy to note that in Case 1 (Fe/As > 1), the iron precip-
itation efficiency was the lowest in comparison to all other cases
(Fe/As 6 1), reflecting the stoichiometric precipitation (Fe/As mo-
lar ratio = 1) of scorodite as evident by the solid product compo-
sition results (Fig. 1a–d). Similar behavior of iron in precipitation
of scorodite as a result of the presence of elevated Fe/As molar
ratio (>1) was also observed previously by Monhemius and
Swash for similar experiments [17]. This agreement with the
work of Monhemius and Swash [17] at 150 °C was unexpected
since their experimental procedure differed from ours. In their
work [17], it was reported that ‘‘the experiments were carried
out by heating the arsenic bearing solutions (without iron) in the
autoclave until the desired temperature was reached, then a solution
containing the required amount of iron was injected into the auto-
clave in order to start the precipitation reactions’’, these tests are
referred here as ‘‘at-temperature’’ tests. These ‘‘at-temperature’’
experiments are quite different than the present work where
the iron and arsenic are dissolved at 25 °C, then put into the
reactor and heated to the target temperature; these test are re-
ferred to as ‘‘heat-up’’ tests. Upon consultation of the dissertation
of Swash [21] – the source of the data reported in [17] – it was
discovered that indeed their test data published in [17] for the
Fe/As effect was mislabeled as it had been collected with ‘‘heat-
ing-up’’ (as in this work) and not with ‘‘at-temperature’’ experi-
ments. The ‘‘at-temperature’’ experiments exhibited much faster
kinetics than the corresponding ‘‘heating-up’’ tests as clearly
described in Swash’s thesis [21].
In terms of copper and magnesium (Fig. 1c and d), there was no
significant variation of their solution concentration with time sug-
gesting little precipitation tendency for them. For example, the% Cu
precipitated from the high copper concentration (0.3 M) tests
(Cases 4 and 5) was 3.7 ± 0.5 and 5.9 ± 0.1 while the %Cu content
of the solids was 0.8 ± 0.5 and 0.7 ± 0.1, respectively (refer to Ta-
bles 1 and 2). However, more careful analysis of samples collected
during the heat-up period of Case 4 test revealed the temporary co-
precipitation of copper along with iron and arsenic to be more sig-
nificant. This is discussed later in Section 3.2.
3.1.3. Characterization of scorodite
The chemical composition of all precipitated scorodite products
is reported in Table 2. Essentially their ferric and arsenic content
fell between 21.1–24.1% and 28.8–35.1% respectively, which on
average are 22.84 ± 1.52 and 31.36 ± 2.15 slightly lower than the
theoretical values (Fe 24% and As 32.3%) reflecting the small incor-
poration of minor elements, such as S (as SO4), Cu and Mg. All sco-
rodite products were found via chemical analysis to contain certain
amount of sulfate (from 0.3% to 4.8%, Table 2) as a result of the
SO4 M AsO4 substitution: Fe(AsO4)1 0.67x(SO4)x
2H2O where x = 0–
0.16 [6].
The presence of sulfate in the scorodite structure was further
verified via infrared spectroscopy (Fig. 2a and b), detailed assign-
ments and group analysis may be found in our previous work
[7]. This type of sulfate incorporation has also been previously ob-
served in the case of atmospherically produced scorodite from sul-
fate media [15,22].
Concerning Cu and Mg, there were only minor amounts that
were found to report as impurities in the scorodite solids (1%),
which is in contrast to the high copper content (3.4%) reported by
Monhemius and Swash [17]. The corresponding % Cu content of
atmospherically produced scorodite similarly has been reported to
be 1% and 2% [15,16]. The low (1%) Cu content in our work likely
reflects structural incorporation within scorodite, while the higher
amounts reported by Monhemius and Swash [17] and Fujita et al.
[16] reflect the formation of separate copper arsenate co-precipi-
tated phase(s). No such copper arsenate or copper sulfate phases
were detected in the present work via either XRD or vibrational
spectroscopy. The incorporation of the divalent cations into the sco-
rodite [9–11] crystal structure (Cu2+ or Mg2+ for Fe3+) is not favor-
able unlike the trivalent cations (Fe3+, Al3+, Ga3+, and In3+)
incorporations [23,24] due to ion charge mismatch. Alternatively,
the lower affinity of the M2+ cation (vs. M3+) towards the arsenate
ions may also arise from electronegativity differences of the
metal cations which can coordinate to the arsenate ligands.
 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518 511

18 24

16 (a) Fe test 1
22 (b) Fe test 1
As test 1 20 As test 1
14 Fe test 2 Fe test 2
18
As test 2 As test 2
12 16
Concentration (g/L)

Concentration (g/L)
14
10
12
8 10

6 8

6
4
4

2 2

0
0
-2
0 2 4 6 8 10 12 0 2 4 6 8 10 12 14
Reaction time (hrs) Reaction time (hrs)

35
Fe test 1
35
(c) Fe test 1 (d) As test 1
As test 1 30 Cu test 1
30 Cu test 1 Mg test 1
Fe test 2 As test 2
25
25 As test 2 Cu test 2
Concentration (g/L)
Concentration (g/L)

Cu test 2 Mg test 2
20 Fe test 2
20

15
15

10
10

5 5

0 0

0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Reaction time (hrs) Reaction time (hrs)

Fig. 1. Concentration profiles as a function of reaction time for tests in Case 1 (a), Case 2 (b), Case 3 (c) and Case 5 (d). The reaction time of 0 h corresponds to the beginning of
the experiment at 25 °C and 2.5 h corresponds to the time when the target temperature of 150 °C was reached.

weak weak
(a) (b)
Transmission Intensity

Transmission Intensity

H-bond H-bond
strong
H-bond
strong
H-bond SO4 SO4
H2O H2O

AsO4 OH/H2O
AsO4

Case 1 @ 12.5hrs , 150ºC

Case 1 @ 2.5hrs, 150ºC
Case 2
Case 2
Case 3
Case 3
Case 4
Case 4
OH/H2O Case 5
Case 5

3550 3050 2550 2050 1550 1050 550 3500 3000 2500 2000 1500 1000
Wavenumbers/cm-1 Wavenumbers/cm-1

Fig. 2. ATR-IR of the solids produced at 150 °C once the target temperature was reached at (a) 2.5 h and (b) at the end of the reaction period (12.5 h).

The electronegativities (v) of Fe3+, Mg2+, and Cu2+, Co2+, Zn2+, Ni2+ Positive identification of the formation of crystalline scorodite
are reported to be 1.68, 1.20, 1.51, 1.46, 1.50, and 1.42 respectively was made via XRD (Fig. S1a and b) and molecular analysis [ATR-
[25], i.e. ferric iron exhibits much higher electronegativity relative IR (Fig. 2a and b) and Micro-Raman (Fig. S1c and d)]. According
to the divalent cations and should therefore be more stable. to the XRD patterns, all precipitates formed at 150 °C consisted
512 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518
Case 1 Fe/As =1.9 at 12.5 hrs and 150 C
Solid: 23.9 % Fe, 30.1% As, 1.5 % S
(a)
H2O
SO4
AsO4
Weak
H-bond
Strong
H-bond
Case 1 Fe/As =1.9 at 3.5 hrs and 150 C
Solid: 23.8 % Fe, 28.1% As, 1.8% S
Case 1 Fe/As =1.9 at 2.5 hrs and 150 C
Solid: 23.5 % Fe, 27.5% As, 2.0% S
S1008 Fe/As =1.6 at 2.0 hrs and 175 C
Solid: 24.9 % Fe, 29.1 % As, 1.9 %S
3550 3050 2550 2050 1550 1050
-1
Wavenumbers (cm )
Fig. 3. (a) ATR-IR spectra and (b) PXRD analysis (k = 1.78897 Å) of Case 1 at 2.5, 3.5 and 12.5 h at 150 °C, as well as at 2.5 h and 175 °C.
of crystalline scorodite independent of the time of production:
2.5 h (end of heat-up period) or 12.5 h. However, molecularly-sen-
sitive ATR-IR analysis (Fig. 2a) showed that in Case 1 (Fe/As = 1.9)
at 2.5 h (150 °C), not all the H-bonding in the system had been fully
developed as evident by the broad OH/H2O band (3000–
4000 cm 1). This behavior of not fully resolved H-bonding environ-
ment was also observed at similar Fe/As ratio and reaction time,
but also at higher temperature (175 °C) Fig. 3a.
Upon extension of time to 3.5 h in Case 1, the H-bonding
environment became fully ordered and typical of crystalline sco-
rodite [7,24]. It is interesting to note that in the products where
the H-bonding environment was not well ordered (2.5 h), the
complementary XRD analysis (Fig. 3b) indicated that these sam-
ples were composed of crystalline scorodite similar to those pro-
duced at longer times (P3.5 h). In scorodite, two distinct types
of intramolecular hydrogen bonding occur between the molecu-
lar crystal water molecules (H-bond donors) of the metal octahe-
dra and the arsenate groups (H-acceptor) [7,10,11]. These
electrostatic interactions (H-bond) redistribute electronic charges
and stabilize (give order) the scorodite structure [11]. Both of
these are manifested as two IR OH/H2O bands assigned to
distinct H-bonding types occurring in scorodite [7,24]. Thus, it
appears that in cases where Fe/As P 1.5, the H-bonding environ-
ment (as observed via the IR spectra) is found to develop more
slowly (63.5 h) than in cases where the Fe/As 6 1(in the pres-
ence or absence of copper). These results also indicate that it
is not only temperature but time that affect the H-bonding
structure and crystallinity (order) of the scorodite structure,
something that has never been reported in the extensive scoro-
dite literature [6,9–11,17,21,26].
The particle morphology of the scorodite precipitates was
evaluated by SEM (Figs. 4, S2 and S3) and was found to be similar
in all cases. Nevertheless, some small differences are observed in
the morphology of particular samples (Fig. S3); for example after
12.5 h in Case 1 (Fe/As > 1.5) the particle morphology appears as
oval shaped which differs from that found in Case 2 (Fe/As 6 1)
where the particle morphology is much closer to that of ideal
orthorhombic (Pbca) crystals (Fig. S3). The particle morphology
in the presence copper (such as Case 4) was unlike that of Case
1 which was characterized by rounded-cauliflower morphology
made up of small well defined orthorhombic crystallites similar
to those observed in Case 2 (Figs. 4g, S2 and S3) and those re-
ported by Fujita et al. [27]. Thus, the presence of copper seems
to exercise a small positive influence on scorodite morphology
(b)
Case 1 Fe/As =1.9 at 12.5 hrs and 150 C
Fe/As =1.6 at 2.0 hrs and 175 C
Arb. Intensity
Case 1 Fe/As =1.9 at 2.5 hrs and 150 C
Case 1 Fe/As =1.9 at 3.5 hrs and 150 C
Standard
550 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
-1
q space (A )
without reducing the crystallinity of the final product (Figs.
S1–S3).
3.2. Formation of a Cu(II)–Fe(III)–AsO4–SO4 intermediate
3.2.1. Solution changes
A benefit of using a glass autoclave is that it allows to visually
monitor the solution as the reaction goes through the various pre-
cipitation (nucleation, growth, transformation etc.) stages (Fig. S4).
Thus during the heat up period of Case 4 tests (Fe/As = 0.82–0.91,
Cu = 0.29–0.30 M (18.6 g/L)) a gelatinous solid was observed to
form, which disappeared upon reaching the target temperature
(150 °C) where scorodite particles appeared. This prompted us to
have a closer look of this phenomenon1 via the collection of solu-
tion (and gel2) samples during the heat up period (before 2.5 h and
150 °C). It is noted that in the other cases (1–3 and 5), no precipitates
other then scorodite were observed to form during the heat up per-
iod and as such no additional data was collected.
The collected data for the intermediate phase in Case 4 are sum-
marized in Table 3 and shown in Fig. 5. In general, the arsenic and
iron concentration kinetic profiles (Fig. 5) during the heat up period
and beyond followed the same trend of precipitation as the other
tests (Fig. 1) and in agreement with previous works [17,21]. Iron
and arsenic (99% Fe and 90% As) precipitation was completed once
the target temperature of 150 °C was reached (2.5 h) (Fig. 5).
The copper, and sulfur (as sulfate) profiles for Case 4 were un-
like the other cases (Fig. 1) or any other reported in literature
[15,17,21]. In particular, a significant drop in the copper and sulfur
(sulfate) concentrations was observed during the heat up period
along with the visual formation of the gel already mentioned
(Figs. 5 and S4). Once 150 °C was reached only scorodite was found
to exist. This implies that any copper (and sulfur) containing pre-
cipitate that had formed re-dissolved before the 2.5 h and 150 °C
point. Of interest is also the acid concentration profile as the pre-
cipitation of scorodite leads to acid generation [6]. The acidity pro-
file similarly (as Cu and S) decreased before the 150 °C point, after
which point acid generation was observed indicative of scorodite
formation.
1
It should be noted that this experiment was repeated three times to ensure that
what is described here is a reproducible and standard feature of the process.
2
The gel/solid phase was separated from the solution by centrifugation for 3 min at
1000 rpm.
 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518 513

Fig. 4. SEM micrographs of scorodite particles of Case 1 (a and f), Case 2 (b and g), Case 3 (c and h), Case 4 (d and i) and Case 5 (e and j) at 2.5 h (top) and 12.5 h (bottom)
respectively. The scale of the images is 5 lm. The inset in figures b and i, display a graphic description of what an ideal orthorhombic–dipyramidal crystal looks like
superficially.

Table 3
Analysis (wt.%) of solid intermediate phase during the heat up period for Case 4 test
40
via ICP-AES.

Temp and time Solid/gel

As (%) Fe (%) Cu (%) S (%)
30
Concentration (g/L)
90 °C and 30 minb 25.6 16.1 4.68 or 11.4 5.75
(3.94 or 11.2)a,b
101 °C and 40 min 23.77 15.09 4.38 6.03
(3.64)a 20
115 °C and 50 min 23.3 15.1 4.26 5.89
Fe
(3.56)a
As
125 °C and 60 min 23.1 14.9 4.10 5.80
10 S
(2.82)a
130 °C and 70 min 22.91 14.9 4.00 5.75
Cu
(2.80)a
Acid
132 °C and 85 min 26.35 17.35 3.91 4.52
0
(2.50)a
138 °C and 96 min 29.65 20.24 2.16 0.5
(1.63)a
0 2 4 6 8 10 12
150 °C and 150 min 28.46 21.44 1.40 0.5
Reaction time (hrs)
(1.02)a
150 °C and 750 min 32.51 22.0 1.32 0.5
Fig. 5. Concentration profiles as a function of reaction time for Case 4 including the
Final (0.98)a
heat up period from 25 °C to 150 °C; the reaction time of 0 h corresponds to the
Standard scorodite [7] 31.0 24.0 0 0.2
beginning of the experiment at 25 °C and 2.5 h corresponds to the time the target
a temperature of 150 °C was reached.
Indicates a.t. % determined via XPS analysis.
b
As mentioned in the text this sample was found to be inhomogeneous (via XPS
analysis Fig. S5) and rather unstable under ambient conditions so these results for
the 30 min sample are considered tentative.

observed copper incorporation with all these other elements (Fe,

3.2.2. Characterization of the intermediate phase
The intermediate phase (gel-like and solid) was analyzed via
ICP-AES and XPS (Table 3 and Fig. 6). The elemental analyses at
30 min and 90 °C should be considered as tentative because of
the inherent complexity involved in sampling and analyzing such
highly hydrated3 unstable substance. The sample was found not to
be elementally homogeneously distributed as evident from our
XPS spot analysis (Fig. S5) which showed the copper concentrations
to significantly differ in two distinct locations (400 lm beam). In the
other samples (40–750 min) such element distribution variation was
not observed.
More importantly, XPS analysis of the intermediate phase (gel-
like or solids) showed (Fig. 6a and b) that indeed these solids con-
tained copper (along with Fe, As (as AsO4) and S (as SO4)) with the
largest amount occurring in the gel component of the intermediate
phase, in agreement with the ICP-AES results (Table 3). The
3
The amount of hydration according to TGA analysis done was at least 30–40%
water. This high degree of hydration is evident from the water band at 1633 cm 1 but
also from the broad water/hydroxyl IR band (3000–3500 cm 1) shown later.
AsO4, SO4) has never been reported in literature [15–17,21].
Visually we can see (Fig. 7) that initially a precursor dark gel
(30 min, 90 °C) formed which evolved to a mixture of gel and solid
light yellow powder precipitate (40 min, 101 °C), to a more solid
light yellow powder with a small surrounding gelatinous layer
(60 min, 125 °C) to the final pale green color of synthetic scorodite
(96 min, 138 °C), with all copper being rejected back into solution.
The solution was also observed to change color throughout the
reaction (Fig. S6). It is worth mentioning that the gel-like material
became physically hard (glass like) and as such had to be ground
for PXRD analysis (Fig. 8).
Fig. 8 shows that from 40 to 60 min (101–125 °C) an X-ray
amorphous type of phase formed, while at 85 min (132 °C) a mix-
ture of a crystalline scorodite component with an amorphous com-
ponent is observed and at 96 min (138 °C) crystalline scorodite is
formed and remained the dominant phase until the end of the
reaction. Since the intermediate sampled phase was X-ray amor-
phous (Fig. 8), we decided to investigate their molecular structure
using ATR-IR and Raman spectroscopies (Fig. 9), which are inde-
pendent of the state of the material (amorphous, crystalline or
dsorbed) [28–33].
514 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518

Fig. 6. XPS full (a) and close up (b) spectrum of samples collected between 30 and 96 min (90–150 °C) showing the major elements in the samples to be Fe, As, O, S, and Cu. A
standard scorodite [6] produced in the absence of copper (as that formed in Case 2) is also included for comparison.

Fig. 7. Photographs of products formed at (a) 30 min, 90 °C; (b) 40 min, 101 °C; (c) 60 min, 125 °C and (d) 96 min, 138 °C during the heat up period of Case 4 test.

the same at the molecular level as they exhibit the same AsO4
(771–801 cm 1), SO4 (114–1064 cm 1), OH/H2O (1625 cm 1 and
3052–3350 cm 1) bonding environments, which differ from that
40 minutes 101ºC
of scorodite [6,7,9–11]. At 85 min (132 °C), the IR spectra show
the m3 (F2) arsenate band becoming well defined (no splitting ob-
60 minutes 125ºC served) and very similar to that observed in scorodite but still
exhibiting a strong sulfate component; the hydroxyl region also
appears to have some OH structure similar to that found in scoro-
85 minutes 132ºC dite as highlighted by the arrows in Fig. 9a but largely disordered
as in Fig. 3a. The Raman spectra at 85 min (132 °C) similarly show
a mixture of both intermediate and scorodite in agreement with IR
96 minutes 138ºC
and XRD results (Figs. 8 and 9). Thus all of this indicates that the
150 minutes 150ºC sample at 85 min (132 °C) is a mixture of both components (scoro-
dite + intermediate). At 96 min (138 °C) only arsenate vibrations
Standard from scorodite are observed (IR and Raman) indicating that by that
time the intermediate gelatinous component had converted to the
scorodite stable phase (Figs. 8 and 9). It is noted that most but not
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 all the sulfate is expelled from the scorodite structure at 96 min –
q space ( -1) 12.5 h (138–150 °C) as shown in the IR and Raman analysis (Figs. 2,
9 and S1).
Fig. 8. Powder XRD spectra (k = 1.78897 Å) of solids produced during the heat up
period. A standard scorodite [6,7] sample is also included for easier comparison. From the above observations, we can infer that the product
formed during the heat up period up to 60 min and 125 °C is not
scorodite but rather a hydrated arsenate sulfate intermediate
According to the spectra of Fig. 9a and b, the intermediate phase which contains various amounts of copper as shown via ICP-AES
samples obtained from 30 to 60 min (90–125 °C) are essentially all and XPS analysis (Table 3 and Figs. 6, 9a and b). Similar vibrational
 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518
Scorodite
150 min 150C
96 min 138C
85 min 132C
60 min 125C
40 min 101C
30 min 90C
H2O/OH
SO4
3500 3000 2500 2000 1500
Wavenumbers (cm )
-1
Fig. 9. ATR-IR and Raman spectra of intermediates formed during the heat up period from 25 to 135 °C of Case 4 test. All spectra have been offset vertically for easier
comparison. Reference spectra for scorodite and BFAS prepared via hydrothermal synthesis [6,7] are also shown for easier comparison.
monitoring has been reported for synthetic aluminum phosphate
[34] but never in the case of scorodite formation in the presence
of foreign ions.
With the view of identifying this intermediate gelatinous phase,
we decided to examine if it resembled one of the other ferric arse-
nate phases (FAsH and/or BFAS) that are known to form under
hydrothermal conditions besides scorodite [5–7]. This examination
revealed that the arsenate and sulfate bonding environment as ex-
pressed in the IR and Raman spectra (Fig. 9a and b) exhibited a
similar symmetry and profile as the basic ferric arsenate sulfate
(BFAS) described elsewhere [7]. In particular it can be observed
that the samples of interest (30–60 min and 90–125 °C) exhibited
the characteristic BFAS vibrational features, where some degree
of degeneracy of the triply degenerate m3 (F2) arsenate (730–
806 cm 1) and sulfate (1051–1108 cm 1) modes are removed as
a result of a decrease in symmetry as evident from the IR and Ra-
man spectra (Fig. 9a and b). In particular for this reaction we have
the following symmetry change to the XO4 (X = As or S) groups
along the reaction: Td (solution) ? C2,2v(colloidal intermediate
gel) ? C1(crystalline scorodite) The vibrational symmetry of the
arsenate and sulfate molecules indicates that a bidentate binuclear
or bidentate mononuclear metal bonding coordination [7,32,33]
exists in the intermediate phase formed, which are similar to those
found in the BFAS phase. Yet the broadness of the Raman XO4
bands are indicative of a diffuse/disordered type of environment
that is expected in a colloidal gel-like phase [35]-this feature
becoming less pronounced as scorodite formation dominates and
the ions become either better ordered (AsO4) or expelled (SO4) in
the structure.
The intermediate green gel-like product (40 min, 101 °C) was
quite unusual and unexpected, thus it was further probed with
polarized microscopy (PM), selected area XRD (with Co Ka1,
k = 1.78897 Å), and TEM–SAED. The 40 min sample at 101 °C was
chosen to be looked at in more detail as it was more stable then
the 30 min sample.
Polarized microscopy was selected in this context as it can
determine whether solid or liquid crystal materials are isotropic
or anisotropic (based on difference of their refraction index) on a
mm scale [36–38]. For our gel-like intermediate (Fig. 10) there
are regions where texture (expressed as different colors) and
515
Scorodite
150 min 150C
96 min 138C
85 min 132C
60 min 120C
40 min 101C
30 min 90C
M-OH/OH2
BFAS
AsO4
1000 200 400 600 800 1000 1200 1400 1600
Wavenumbers (cm-1)
orientation of ordered domains (likely as nano-crystals) are
observed, similar to those encountered in birefringent anisotropic
liquid crystals [36,37]. However, in our gel-like intermediate, no
low angle diffraction peaks were observed and as such it cannot
be classified as a liquid crystal but it does exhibit texture and ori-
entation of nano-ordered domains similar to that of anisotropic
particles [37]. In addition to the texture, there exists an also isotro-
pic component (pink) in the sample.
Selected area XRD analysis (Fig. S7) on two different spots
(200 lm) of the gel-like intermediate showed an amorphous sig-
nal, typical of a material with poor long range order as in the bulk
XRD analysis (Fig. 8). However, a diffraction pattern indicative of a
semi-crystalline phase was observed in the second region indicat-
ing that this gel-like intermediate was not completely amorphous
but contained some ordered component (in agreement with the
polarized microscopy results) pointing to a transition process in
progress. Calculated d spacings of the XRD diffraction peaks
(k = 1.78897 Å) for the amorphous and semi-crystalline compo-
nents of the gel material are reported in Table S1. The selected area
XRD pattern of the amorphous or semi-crystalline components of
the gel-like intermediate at (Fig. S7) did not match scorodite nor
any other in the JCPS database.
TEM and SAED images of the gel-like intermediates, shown in
Figs. 10, S8 and S9, provided further evidence of the presence of
both amorphous and polycrystalline features indicative of the evo-
lution of a colloidal gel intermediate towards increasing atomic or-
der (Fig. 10). EDS analysis confirmed the intermediate products
(Figs. 10 and S9) were composed of Fe, As, S, O and Cu (not from
grid); in agreement with the ICP-AES (Table 3) and XPS results
(Fig. 6).4 To observe the atomic order in the gel-like intermediate,
higher magnification TEM images were taken (Figs. 11, S8 and S10).
From these images, the lattice fringes of these particles which
make up the gel-like intermediate were found to have spacings
of 3.27 Å, close to the 3.25 Å spacing of BFAS [6]. A closer look at
the end of the particles (Figs. 11, S8 and S10) revealed the nano-
scale organization of disordered particles to the formation of lattice
4
It is important to note that the presence of copper in the solid was verified via
several methods and is emphasized throughout the text since such a phenomenon has
never been observed/reported until now [15–17].
516 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518

Fig. 10. TEM, EDS, SAED and polarized microscopy images of the intermediate gelatinous product obtained at 40 min and 101 °C. The SAED was taken in different locations on
the sample showing both an amorphous and a polycrystalline type of material while the polarized images show different colors where texture and orientation occur. The
scale on the TEM–SAED and polarized microscopy images are 500 nm and 100 lm respectively. (For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)

Fig. 11. TEM of the semi-crystalline component of the gel product (40 min and 101 °C) showing the formation of ordered atomic planes from atoms in the unordered phase to
the lattice fringes.

fringes. TEM images (Fig. S11a and b) further showed this gel-like
substance to exhibit different lattice fringe directions5, which may
5
In addition the gel-like substance appeared to be highly porous (Fig. S11b and d).
explain the texture observed in the polarized microscopy images
(Fig. 10). Alternatively, this colloidal gel-like phase may be com-
posed of nanocrystals which exhibit gel-like properties and exhibit
anisotropic particle character [37]. This is not believed to be the case,
however, as the individual particles are not believed themselves to
 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518
have their own lattice fringes. At this point is worth noting that the
gel-like products (Fig. 7) were stable at ambient conditions after
1 week (Fig. S12) but the yellow type of solid powder eventually
decomposed.
From the above analysis, it is apparent that this intermediate
phase formed in Case 4 (30–85 min, 90–132 °C as shown in
Fig. 7) is a copper-carrying ferric arsenate sulfate hydrated species,
which converts (via a dissolution–recrystallization mechanism) at
longer times (96 min – 12.5 h) and higher temperatures (138–
150 °C) to scorodite via the exclusion of copper and sulfate [39].
The formation of this type of short-lived intermediate phase has
never been observed in scorodite precipitation under hydrother-
mal [17] or atmospheric pressure conditions [15,16] and as such
this work demonstrated that colloidal metastable phases can
incorporate foreign metal ions and interfere with the scorodite
crystallization process if in adequate times and temperatures are
employed.
3.3. Characterization of industrial copper pressure leach residue
In order to correlate the findings of our laboratory study to
industrial practice as exemplified by Teck’s CESL copper pressure
leaching process [3,4,40], an industrial arsenic containing residue
(Fig. S13) was analyzed. The residue studied here was principally
composed of iron (29.3%) and sulfur (36.3%) with the bulk being
in elemental form (32.1%) and the remaining as sulfate (4%). The
residue contained only 0.5% As (1% AsO4), which explains the diffi-
culty associated with identifying the arsenate form.
According to previous studies [5–7], scorodite is expected to be
the most stable phase at 150 °C (the temperature at which the CESL
process operates). However, in these studies, the Fe/As molar ratio
was much lower (Fe/As 6 4) than the Fe/As ratio (80) in the cop-
per concentrate (24%Fe, 0.4% As) and with longer reaction times
(P2 h). Hence it was important to verify if scorodite was indeed
present in the residue or another phase like our BFAS-resembling
intermediate phase is formed.
In order to determine the arsenate form in this industrial resi-
due, PXRD, Micro-Raman and ATR-IR analysis was conducted. Pow-
der XRD analysis (Figs. S14a and S15) gave us low matches (645%)
for hematite and elemental sulfur but failed to detect any type of
arsenate phase (such as scorodite), even when the appropriate ele-
ment restrictions were imposed (Fe, As, S, O and H). Therefore
vibrational spectroscopies (ATR-IR and Raman) were employed in-
stead as they offer the low detection limit appropriate for the
industrial material (1 wt.%). The Raman spectrum of the ar-
senic-containing residue (Fig. S14b in SI section) showed no arse-
nate vibrational modes at low (100–400 cm 1) or high
wavenumbers (700–900 cm 1) before and after elemental sulfur
removal [41; see Fig. S16a]. Since no arsenate phase was detected
via Raman spectroscopy, the use of ATR-IR spectroscopy was
employed.
Figs. S14c and S16b show that the ATR-IR spectrum does not
change much after sulfur removal (in agreement with PXRD
analysis). In general, the full IR spectra (Fig. S14c) show strong sul-
fate vibrations (1000–1200 cm 1), the water bend mode
(1630 cm 1) and a strong broad absorption due to hydroxyl
modes (3378 cm 1). Small vibrations of the arsenate m3 (F2)
antisymetric stretch (786 cm 1) were also observed, but due to
the large signals of the other molecular groups, they could not be
clearly deciphered. Thus we decided to focus (‘‘tune’’) only on
the arsenate vibrations by running multiple scans over the 700–
900 cm 1 region (Fig. S16b).
It can be observed (Fig. 12) that upon comparison to our previ-
ous vibrational work [6,7] on high temperature Fe(III)–AsO4–SO4
phases that the arsenate form in the Teck residue is not like
517
Fig. 12. ATR-IR spectra of the m3 (AsO4) mode, for the industrial pressure leaching
arsenic-containing residue compared to other relevant high temperature ferric
arsenate phases such as scorodite, ferric arsenate sub-hydrate (FAsH) and basic
ferric arsenate sulfate (BFAS) [6,7].
scorodite but rather resembles that of BFAS phase and the short
lived intermediate phase described earlier (Fig. 9a and b).
Another possibility is for arsenate to be covalently adsorbed
onto the hematite (major component of the reside) via bidentate
bridging surface complexes [42,43] where the arsenate (and sul-
fate) molecules have a lower symmetry (C2,2v) and thus give rise
to same type of IR signal. However, the striking resemblance of
the arsenate (sulfate and OH) IR footprint with that of BFAS and
the short time (60 min) employed in the CESL process make it most
probable to have the BFAS-like phase in the industrial residue.
4. Conclusions
Scorodite was found to be the most stable phase (at 150 °C) un-
der all test conditions investigated (0.7 < Fe(III)/As(V) < 1.9, 0.3 M
Fe(SO4)1.5, 0–0.3 M CuSO4, 0–0.3 M MgSO4 and >2 h reaction time).
The produced scorodite [Fe(AsO4)1–2/3X(SO4)X
2H2O] was found to
incorporate up to 5% SO4 in agreement with other recent reports
on hydrothermal precipitation of scorodite [6]. The presence of for-
eign cations (Cu2+ or Mg2+) did not affect the precipitation yield or
kinetics of scorodite precipitation with <1% incorporated in the fi-
nal scorodite solid product. This finding agrees with scorodite pre-
cipitation studies carried out under atmospheric pressure
conditions (at 95 °C) [15,16] but contradicts results from Monhem-
ius and Swash [17] obtained at 190 °C. This difference (lower yield
and higher Cu incorporation) may reflect the formation of ferric
arsenate sub-hydrate (FAsH:FeAsO4
3/4H2O) at the expense of sco-
rodite at 190 °C as reported by these authors [6].
Excess ferric iron (Fe/As = 2) in the initial solution was found to
slow down the development of H-bond structure in comparison to
other cases (Fe/As = 1) in the presence or absence of third ions. The
negative effect of excess ferric iron (Fe/As > 1.5) on the precipita-
tion kinetics of scorodite have been reported [15] but till now
the mode by which iron (as ferric sulfate) interfered with the sco-
rodite crystallization process was not understood.
Slow heating up tests involving equimolar composition acidic
sulfate solutions (0.3 M Fe, As, Cu) revealed the formation of a
short lived hydrated copper-containing ferric arsenate–sulfate
intermediate gelatinous phase whose molecular structure is simi-
lar to that of basic ferric arsenate sulfate (BFAS: Fe[(A-
sO4)1 x(SO4)x(OH)x]
wH2O). Upon further phase transformation
induced by continuous temperature elevation and extending heat-
ing time, the intermediate gelatinous phase was observed (using a
variety of characterization tools) to transform not to crystalline
BFAS but rather to scorodite via the exclusion of copper and sulfate
518 M.A. Gomez et al. / Journal of Colloid and Interface Science 360 (2011) 508–518
from its structure. The formation of this intermediate (metastable)
phase during hydrothermal precipitation of scorodite from multi-
component solutions typical to those encountered in industry con-
stitutes a discovery that can have implications in terms of the type
of arsenate phase forming in an actual industrial process. In the
field of industrial hydrometallurgy, the formation of such a copper
containing intermediate is highly undesirable as this may result in
valuable copper loss and production of potentially less stable (sol-
uble) arsenate phases (unlike scorodite) posing environmental
risks. Analysis of an industrial low arsenate phase (1% AsO4) via
ATR-IR revealed its arsenate profile to be molecularly similar to
the BFAS form reflecting the complications arising from high Fe/
As ratios and the presence of copper that according to the present
work both interfere with the crystallization kinetics of scorodite.
Longer reaction times (>60 min) and lower Fe/As ratios are ex-
pected to lead to scorodite as final product.
The discovery, isolation, detailed characterization and monitor-
ing of the gelatinous Fe(III)–Cu(II)–AsO4–SO4 intermediate phase
and its conversion towards an ordered crystalline scorodite phase
that is reported herein may lead to a better understanding of
nucleation and phase transition phenomena [19] for industrially
and environmentally important precipitation systems as is the for-
mation/crystallization of various iron arsenate compounds.
Acknowledgments
The support of this project by NSERC and Teck Metals is
acknowledged. The authors would like to thank Professors Derek
Gray for use of polarized microscope, Xue Dong Li for taking TEM
images at McGill University’s Chemistry and Physics departments.
Dr. Samir Elouatik and Petr Fiurasek from the Materials Surface
Analysis Center and the Center for Self Assembled Chemical Struc-
tures are acknowledged for their technical support and the use of
the Raman Microscope, TGA and ATR-IR instruments. The great
help received from Kee Eun Lee is also acknowledged.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcis.2011.05.010.
References
[1] P. Miretzky, A. Cirelli-Fernandez, Crit. Rev. Environ. Sci. Technol. 40 (2010) 93.
[2] O.A. Riveros, J.E. Dutrizac, P. Spencer, Can. Metall. Q. 40 (2001) 395.
[3] J. Defreyne, W. Grieve, D.L. Jones, K. Mayhew, in: J.E. Dutrizac, P.A. Riveros
(Eds.), The Role of Iron in the CESL Process. In Iron Control Technologies.
Canadian Institute of Mining, Metallurgy, and Petroleum Conference
Proceedings, Montreal, 2006, p. 205. <http://www.cesl.com>.
[4] K. Mayhew, P. Parhar, H. Salomon-de-Friedberg, CESL process as applied to
enargite-rich copper concentrates, in: Copper 2010 Conference Proceedings,
Hamburg, 2010, vol. 5, p. 1983 (ISBN 978-3-940276-29-2).
[5] J.E. Dutrizac, J.L. Jambor, Hydrometallurgy 86 (2007) 147.
[6] M.A. Gomez, L. Becze, J.N. Cutler, G.P. Demopoulos, Hydrometallurgy 107
(2011) 74.
[7] M.A. Gomez, H. Assaaoudi, L. Becze, J.N. Cutler, G.P. Demopoulos, J. Raman
Spectrosc. 41 (2010) 212.
[8] D. Filippou, G.P. Demopoulos, JOM 49 (1997) 52.
[9] K. Kitahama, R. Kiriyama, Y. Bala, Acta Crystallogr., Sect. B: Struct. Sci. 31
(1975) 322.
[10] F.C. Hawthorne, Acta Crystallogr., Sect. B: Struct. Sci. 32 (1976) 2891.
[11] Y. Xu, G. P Zhou, X.F. Zheng, Solid State Acta Crystallogr., Sect. E: Cryst. Struct.
Commun. 63 (2007) 67.
[12] E. Krause, V.A. Ettel, Am. Mineral. 73 (1988) 850.
[13] D. Langmuir, J. Mahoney, J. Rowson, Geochim. Cosmochim. Acta 70 (2006)
2942.
[14] M.C. Bluteau, G.P. Demopoulos, Hydrometallurgy 87 (2007) 163.
[15] S. Singhania, Q. Wang, D. Filippou, G.P. Demopoulos, Metall. Mater. Trans. B 37
(2006) 189.
[16] T. Fujita, R. Taguchi, M. Abumiya, M. Matsumoto, E. Shibata, T. Nakamura,
Hydrometallurgy 93 (2008) 30.
[17] A.J. Monhemius, P.M. Swash, JOM 51 (1999) 30.
[18] J.F. Le Berre, R. Gauvin, G.P. Demopoulos, Colloid Surf. A 315 (2008) 117.
[19] M.E. Blesa, E. Matijevic, Adv. Colloid Interface Sci. 29 (1989) 173.
[20] L. Wang, Z. Yan, S. Qiao, G.Q. Max Lu, Y. Huang, J. Colloid Interface Sci. 316
(2007) 954.
[21] P.M. Swash, The Hydrothermal Precipitation of Arsenic Solids in the Ca–Fe–
AsO4–SO4 System at Elevated Temperature, Ph.D. Dissertation, Imperial
College of Science, Technology and Medicine, University of London, 1996.
[22] S. Singhania, Q. Wang, D. Filippou, G.P. Demopoulos, Metall. Mater. Trans. B 36
(2005) 327.
[23] J.F. Le Berre, R. Gauvin, G.P. Demopoulos, Ind. Eng. Chem. Res. 46 (2007) 7875.
[24] M.A. Gomez, J.F. Le Berre, H. Assaaoudi, G.P. Demopoulos, J. Raman Spectrosc.
42 (2011) 62.
[25] Y. Zhang, Inorg. Chem. 21 (1982) 3886.
[26] G.P. Demopoulos, Hydrometallurgy 96 (2009) 199.
[27] T. Fujita, R. Taguchi, M. Abumiya, M. Matsumoto, E. Shibata, T. Nakamura,
Hydrometallurgy 96 (2009) 189.
[28] K.R. Liedl, T. Loerting, E. Mayer, Angew. Chem. Int. Ed. 48 (2009) 2690.
[29] S.K. Deb, M. Wilding, M. Somayazulu, P.F. McMillan, Nature 414 (2001) 528.
[30] A.B. Horn, K.J. Sully, Phys. Chem. Chem. Phys. 1 (1999) 3801.
[31] X.C. Tang, M.J. Pikal, L.S. Taylor, Pharm. Res. 19 (2002) 477.
[32] D. Peak, R.G. Ford, D.L. Sparks, J. Colloid Interface Sci. 218 (1999) 289.
[33] G.Y. Zhang, D. Peak, Geochim. Cosmochim. Acta 71 (2007) 2158.
[34] F. Fan, Z. Feng, K. Sun, M. Guo, Q. Guo, Y. Song, W. Li, C. Li, Angew. Chem. Int.
Ed. 48 (2009) 8743.
[35] J. Loun, J. Cejka, J. Sejkora, J. Plasil, M. Novak, R.L. Frost, S.J. Palmer, E.C. Keeffe, J.
Raman Spectrosc. 41 (2011), doi:10.1002/jrs.2900.
[36] J.R. Bellare, H.T. Davis, W.G. Miller, L.E. Scriven, J. Colloid Interface Sci. 136
(1990) 305.
[37] S.N. Fejer, D. Chakrabarti, D.J. Wales, Soft Matter 7 (2011) 3553.
[38] I.W. Hanley, Soft Matter 6 (2010) 1863.
[39] G.R. Desiraju, Angew. Chem. Int. Ed. 46 (2007) 8342.
[40] J. Defreyne, T. Cabra, Early copper production results from Vale’s
hydrometallurgical CESL refinery, in: ALTA Conference Proceedings, Utah,
2009. <http://www.cesl.com>.
[41] L. Bezce, M.A. Gomez, J.F. Le Berre, B. Pierre, G.P. Demopoulos, Can. Metall. Q.
48 (2009) 99.
[42] J.G. Catalano, Z. Zhang, C. Park, P. Fenter, M.J. Bedzyk, Geochim. Cosmochim.
Acta 71 (2007) 1883.
[43] G. Lefevre, M. Fedoroff, Phys. Chem. Earth 31 (2006) 499.