GEOLOGICA CARPATHICA, 52, 2, BRATISLAVA, APRIL 2001
111—121
EVALUATION OF IDENTIFICATION METHODS FOR
CHRYSOCOLLA – A Cu-SMECTITE-LIKE HYDROUS SILICATE:
IMPLICATIONS FOR HEAP-LEACHING EXTRACTION OF COPPER
URSULA KELM
1
*, VILMA SANHUEZA
1
, JANA MADEJOVÁ
2
,
VLADIMÍR ŠUCHA
3
and FRANÇOISE ELSASS
4
1
Instituto de Geología Económica Aplicada, Universidad de Concepción, Casilla 160-C, Concepción 3, Chile
2
Institute of Inorganic Chemistry, Slovak Akademy of Sciences, 842 36 Bratislava, Slovak Republic
3
Department of
Mineral Deposits and Geology,
Faculty of Science, Comenius University, Mlynská dolina,
842 15 Bratislava, Slovak Republic
4
Station de Science du Sol, INRA, Versailles, France
(Manuscript received June 2, 2000; accepted in revised form March 15, 2001)
Abstract: Chrysocolla has been characterized by XRD, FTIR, DTA-TG, TEM and EPMA in order to clarify proper-
ties of this phase and to identify a suitable routine identification method. Chrysocolla is X-ray amorphous, although in
some cases slightly sensitive to EG-treatment. IR spectra indicate a trioctahedral layer-silicate structure, resembling
in many features that of trioctahedral smectites. HRTEM images do not show a phyllosilicate-like repetitive layering
in accordance with XRD results. The SiO
2
/CuO ratio is considered responsible for the degree of crystal ordering
achieved. SiO
2
/CuO ratio near 1 results in an amorphous phase, while SiO
2
/CuO near 1.3 generates more smectite-
like characteristics. For routine analyses in mining operations, DTA-TG is considered feasible, once basic mineraliza-
tion types are known at a site.
Key words: Cu-leaching, mineral analysis, Cu-smectite, chrysocolla.
Introduction
Chrysocolla occurs in oxidation zones of porphyry copper de-
posits and has often been used as exploration guide (Meach
1981). With colour shades from turquoise-blue and emerald-
green to blackish green and most frequently developed as col-
loidal crusts and aggregates, it is a stable member of many
mineral collections. Hardness of only 2 to 4 on the Mohs scale
makes chrysocolla unsuitable for a widespread application in
jewelry or ornamental items (Ďu a & Rejl 1997). The mineral
is translucent to opaque, biaxial (-), with
α
= 1.575—1.858,
β
=
1.597,
γ
= 1.598—1.635 (Anthony et al. 1995). Fine inter-
growths with the host rock make optical identification diffi-
cult, in particular for frequent associations of chrysocolla with
azurite, malachite and atacamite and intergrowths with phyllo-
silicates or zeolites of the ore host rock. This has been mani-
fested by the need to use alternative methods of characteriza-
tion, in particular for metallurgical applications (US Patent
1980). Mineralogical reference databases (e.g. JCPDS – In-
ternational Centre for Diffraction Data) cite chrysocolla as
Cu
2—x
Si
2
O
5
(OH)
3
.
H
2
O (Van Oosterwyck-Gastuche 1970).
The original formula, CuSiO
3
. H
2
O (Foote & Bradley 1913),
is still used in mineral atlases (Ďu a & Rejl 1997). Newberg
(1967) proposed the decomposition of feldspar in the presence
of copper-rich solutions as a mechanism for the formation of
exotic (i.e. sulphide ore body-distant) chrysocolla ores. Ac-
cording to this author, conditions favouring this reaction are:
low copper concentrations, silica concentration below the
equilibrium level for amorphous silica, low concentration of
interfering anions and a moderate to high pH (superior to 9).
The latter point has been experimentally reassessed by Yates et
al. (1998), indicating a pH window between 5—9, with opti-
mum conditions at neutrality.
Chrysocolla is a widespread exotic copper ore, with Mina
Sur (Chuquicamata) and Dona Ines de Collahuasi as the prime
examples in Chile. Research concerning the conditions of
chrysocolla formation and its characterization has been gener-
ated as collateral information in the synthesis of trioctahedral
smectites (Kloprogge et al. 1999 and papers cited therein) as
well as by metallurgical studies aiming to predict the behav-
iour of chrysocolla in flotation or acid leaching processing
(Meach 1981 and Lindsay 1994). These process-oriented stud-
ies focus in particular on the physico-chemical characteristics
of the ore minerals. EXAFS studies (McKeown 1994) have
been used to clarify the Cu-O and Cu-Cu environments using
copper oxide, dioptase and metal reference materials.
Chrysocolla often impregnates altered host rocks containing
smectite, chlorite, and laumonite. The Chile-based authors are
frequently faced with the need to unequivocally identify
chrysocolla finely dispersed with the above phases and to pre-
dict possible causes for elevated acid consumption in leaching
processes. The smectite-group minerals have been identified
as the main acid consumer from copper solutions in heap
leaching extraction processes (Gomer et al. 1991). However,
routine methods such as microscopy and XRD do not always
allow a satisfactory identification of minor admixtures of ei-
ther phase. It is therefore considered important to understand
single mineral characteristics before attempting to predict the
leaching behaviour, for example, of chrysocolla and smectite
or zeolite mixtures in a future stage of study.
112 KELM et al.
Therefore, six apparently pure chrysocolla samples have
been selected, with the aim of elucidating the texture and com-
position of the mineral at TEM level and thus to contribute in-
directly to the understanding of its behaviour in leaching pro-
cesses. The parallel use of high resolution observation and
analyses as well as more rapid and cheaper methods is aimed
at recommending an analytical procedure that will ease char-
acterization on a more routine and larger scale basis.
Methods
X-ray diffraction analyses have been performed using a
Rigaku D max X-ray diffractometer with Ni-filtered Cu radia-
tion at the following slit settings: 1° divergence and antiscatter
slits, 1.5 mm receiving slit. Also a Philips PW 1710 (Ni-fil-
tered Cu-K
α
radiation) diffractometer has been used. Oriented
specimens have been recorded in an air-dry state and saturated
with ethylene glycol overnight at 70 °C.
The dehydration pattern of chrysocolla has been observed
using a differential thermal/gravimetric analyser (Rigaku
TAS 100) at heating speeds of 10 °C/min in an air atmo-
sphere with approximately 30 mg of sample compacted in a
Pt microcrucible.
Infrared spectra have been obtained on a Nicolet Magna 750
Fourier transform infrared spectrometer equipped with a
DTGS detector. For each sample, 128 scans were recorded in
the 4000—400 cm
—1
range. The samples had been prepared as
pressed KBr disks (1 mg of sample and 200 mg of KBr). The
disks were heated overnight at 150 °C to minimize the amount
of residual water in the KBr.
The HRTEM measurements were performed using a JEOL
1200 EXII and a Philips 420 STEM microscope operated at
120 kV. Rock chips and small portions of water-saturated
<2
µ
m fractions in Na-form were coated with agar, embedded
in Spurr resin and sectioned by ultramicrotome. Under such
conditions, any potential smectite layers intercalated by organ-
ic compounds produce interlayer distances of about 1.35 nm
(Tessier 1984).
Electron microprobe analyses (EPMA) were carried out on a
wavelength dispersive Jeol 8600 probe with a defocused beam
(40
µ
m), 15 kV acceleration voltage and 10
—8
A probe current.
Nitrogen adsorption at 77K measurements has been ob-
tained on a Gemini 2370 instrument after pre-treating samples
(0.153 g) in a nitrogen atmosphere for two hours at 350 °C, as
well as on a NAPLO 58301 equipment at 60 °C after degas-
sing under vacuum. Two samples (Chuqui 1 and 3) have been
selected for this analysis due to mineral purity and abundance
of sample material. Determinations have been carried out on
the < 200 Tyler mesh fraction.
Samples
For the present study, a total of six samples have been used.
Samples Chuquicamata 1 to 5 (Chuqui 1—5) are from the Mina
Sur exotic ore sector of the Chuquicamata porphyry copper de-
posit. Chrysocolla impregnates gravel and basement rock
along a 6 km long and 200 m to 1200 m wide paleochannel
(Ruiz & Peebles 1988). Sample Chuqui 1 corresponds to a col-
loidal dark green chrysocolla-gypsum vein of about 10 cm di-
ameter in the basement rock. Chuqui 2 is a conglomerate ce-
mented with green chrysocolla. Chuqui 3 is a pale blue
colloidal chrysocolla sample from a 10 cm thick vein from a
core in the basement rock. Chuqui 4 represents a colloidal blue
chrysocolla sample from an approximately 1 m thick vein.
Chuqui 5 is a mixture of black and light blue chrysocolla,
forming veins of 3—5 cm thickness. A green-blue sample (col-
lection specimen) of unspecified origin has been incorporated.
All samples represent collection specimens, where chrysocolla
has been separated by handpicking for all stages of analyses
(see below). The chrysocolla of samples Chuqui 1, 3 and 5 are
obvious fillings of void spaces, thus minimizing the admixture
of host rock minerals during separation.
Results
XRD analyses
Greenish transparent and bluish parts of the sample, which
were visually identified with chrysocolla, were handpicked
and X-rayed in both random and oriented specimens (Figs. 1,
2). Samples of random orientation show broad peaks as indi-
cated by the ICCD reference card No. 27-188 of chrysocolla
accompanied by minor minerals (except for Chuqui 2) of the
host rock and vein phase gypsum (Table 1).
All patterns obtained, except for sample Chuqui 3, were
without distinct peaks, displaying only elevated background
between 0.59 nm and 0.225 nm. Oriented specimens show
poorly resolved peaks in a low 2°
θ
region. The shape of the
poor peaks seems to be slightly sensitive to ethylene glycol
saturation (insert of Fig. 2). XRD patterns of Chuqui 3 sample
slightly differ from other patterns. Both air-dried and glycolat-
ed patterns show much better resolved peaks in the low two
theta region and at about 0.44—0.38 nm, 0.30 nm, and 0.256—
0.225 nm. This pattern is much more sensitive to glycolation
than other samples. The most intensive change after EG satu-
ration has been observed at about 1.7 nm, which may also in-
dicate traces of 1.7 nm smectite peak, and in the 0.44—0.38 nm
region. No EG sensitivity for chrysocolla has been reported so
far. Prosser et al. (1965) indicate a possible water uptake of 25
% without noticeable lattice expansion.
FTIR analyses
The IR spectra of all chrysocolla samples are similar in the
whole spectral range (Fig. 3). Spectral features resemble those
of trioctahedral 2:1 layer silicates (Farmer 1974). Xiao &
Villemure (1998) and Decarreau et al. (1992) reported similar
spectra for synthetic Cu smectites. A sharp band near 3620
cm
—1
is supposed to correspond to OH stretching vibrations of
structural OH group coordinated to three octahedral cations,
although a band at 3680 cm
—1
for trioctahedral smectites is
missing (Decarreau et al. 1992). Due to the high content of Cu
in the chrysocolla samples, Cu
3
OH is the most probable octa-
hedral arrangement. A broad absorption in the 3500—3000 cm
—1
range is attributed to the OH stretching vibrations of water
IDENTIFICATION METHODS FOR CHRYSOCOLLA – A Cu-SMECTITE-LIKE SILICATE 113
Fig. 1. Random X-ray diffractograms of chrysocolla.
Fig. 2. Oriented diffractograms of EG-saturated samples Chuqui 1—3. EG sensitivity is demonstrated on the inset Chuqui 5 sample.
114 KELM et al.
molecules. A maximum of this band occurs near 3475 cm
—1
for
all chrysocollas except Chuqui 3, where a 3442 cm
—1
position
is observed (Fig. 3). Differences in the water band position
may indicate a change in the environment of the bound water
molecules. A shoulder near 3698 cm
—1
may indicate traces of
kaolinite present as admixture in Chuqui 2 and 3; this trace has
not been detected by XRD. The Si-O stretching band observed
near 1026 cm
—1
in the IR spectra of Chuqui 1, 2, 4 and 5 is
shifted by about 10 cm
—1
towards higher frequencies for Chu-
qui 3 and collection sample. In accordance with Decarreau et
al. (1992) a band near 780 cm
—1
is due to Si-O vibration, while
Cu
3
OH bending vibration absorbs nearly 677 cm
—1
. In the re-
gion below 600 cm
—1
, two clearly resolved bands are present.
The absorption bands near 500 and 470 cm
—1
are related to Cu-
O-Si and Si-O-Si bending, respectively.
TEM observations
Although all samples have been apparently pure handpicked
chrysocolla, differences in hardness have been observed dur-
ing impregnation with Spurr resin and subsequent cutting of
the samples. This may be due to different states of hydration of
the samples (see also DTA analyses); colloidal light blue sam-
ples (sample 3 and collection specimen) tended to disintegrate
easily but required longer periods of impregnation, whereas
dark green-black samples did not pose any problem with cut-
ting (samples 1, 5).
A spotty appearance of irregular by contoured aggregates
are common features observed in TEM images (Fig. 4). Spots
are present from the initial incidence of beam, becoming more
focussed and darker during the first minute of observation and
remaining stable even under long-term observation (>30 min).
Spot diameters for dark green (possibly more hydrated, see
DTA data) samples are larger than those for light blue ones.
Photographs represent the “definite” state of the spots (Figs. 4
and 5). HRTEM permits the observation of elongated diffuse
filaments, similar to fringe images, but with no regular spacing
observed between them. The width of the filaments is approxi-
mately 1.3 nm (white) and 1.6 nm (dark). They may also occur
as individual laths in an otherwise undifferentiated ground-
mass (Fig. 6). Because of the size of the aggregates and their
dispersed nature, it has not been possible to obtain a usable
electron diffraction image. When trying to focus diffraction
images, diffuse spots are observed frequently to disappear dur-
ing adjustment of the image, leaving only “amorphous” rings.
Similar changes have been observed by Van Oosterwyck-Gas-
tuche (1970) and McKeown (1994).
Ultrathin sections of sample Chuqui 4 show at very low
magnification dominant lath-shaped particles (Fig. 7), which
display at higher magnification a homogeneous amorphous-
like image. The image is very similar to that observed by
Šucha et al. (1998) for very early stages of illite syntheses
from gel and glass.
DTA-TG analyses
Analyses were carried out at ambiental humidity of 45—50
% (Fig. 8A and B). The pure samples of chrysocolla (Chuqui 1
and 3) show a broad endothermic peak below 200 °C, fol-
lowed by a continuous endothermic drift between 250 and 650
°C (Fig. 8B). Weight losses up to 200 °C have been 17 % for
dark green Chuqui 1 and only 13 % for pale blue Chuqui 3
(Fig. 8A). A marked exothermic peak occurs at 695—705 °C.
At 1027 °C an endothermic peak is observed. Fusion starts
above 1100 °C.
Compared to DTA spectra of many smectite group minerals,
chrysocolla lacks the bifurcated low temperature peak of most
smectite samples (Patterson & Swaffield 1987) attributed to
the loss of externally adsorbed and interlayer-bound water. No
endothermic peaks have been observed at 700 °C, nor the shal-
low endothermic/exothermic signal between 800 and 900 °C
(Paterson & Swaffield 1987), except for a slight endothermic
indentation at 550 °C for Chuqui 3.
To observe the formation of potential new phases, samples
have been heated for one hour to 350 and subsequently to 850
°C with each step checked by XRD analyses. No phase change
has been found at 350 °C. Sample Chuqui 1 exhibits a faint
1nm peak on the diffractogram, which may possibly be attrib-
uted to a smectite collapse, although no such phase has been
detected by XRD. Tenorite formed at 700 °C, followed by oxi-
dation to cuprite at 1027 °C.
DTA/TG characteristics of chrysocolla reported by Meach
(1981) are similar to the samples analysed in this study. How-
ever, for the present samples, the initial endothermic peak is
observed at about 100 °C, that is 50° lower than the descrip-
tion by Meach. It is possible that this difference is due to using
a high temperature DTA/TG equipment, which does not yield
best results below 100 °C. The water uptake of up to 25 % re-
ported by Prosser et al. (1965) could also contribute to this
shift.
Chuqui 2 is a polymineralic sample with two dehydration
peaks at 80 and 130 °C, respectively (Fig. 8). It is doubted
that this endotherm doublet is attributable only to smectite.
Fig. 3. The FTIR spectra of chrysocolla samples from Chuquica-
mata (Ch1 to Ch 5) and collection specimen (Coll).
Ab
sor
b
ance
Wavenumbers / cm
-1
674
675
774
790
Coll
Ch 5
Ch 3
Ch 4
Ch 2
499
471
1026
673
1038
785
783
1037
3476
3623
3200
3600 3400
3442
K
3623
3475
1300
1100
3622
500
K
Ch 1
900
700
3620
IDENTIFICATION METHODS FOR CHRYSOCOLLA – A Cu-SMECTITE-LIKE SILICATE 115
Fig. 5. Large spots, which initially change their morphology under the electron beam.
Fig. 4. Spotty appearance of chrysocolla surfaces under electron beam (TEM).
116 KELM et al.
Fig. 7. General lath-shaped aggregates at low TEM magnification.
Fig. 6. Short-range, non-repetitive fibres, most noticeable grey sector along the lower image margin.
IDENTIFICATION METHODS FOR CHRYSOCOLLA – A Cu-SMECTITE-LIKE SILICATE 117
For the second peak, the loss of adsorbed water from feldspar
surfaces is suspected. For the collection sample with mala-
chite, the second endotherm corresponds to the loss of CO
2
from the copper carbonate. The shift of tenorite formation to
a temperature of about 730 °C could be due to the presence
of the CO
2
generated in a static atmosphere (Bish & Duffy
1990). Sample Chuqui 5 shows a faint indentation of the
principal peak of water desorption at about 130 °C. This
could be due to the presence of a 1.4 nm, 0.7 nm or smectite-
type phyllosilicate not identifiable by XRD due to superposi-
tion with chrysocolla, although oriented XRD mounts do not
show clear smectite peaks. Chrysocolla is characterized by a
broad dehydration endotherm, lack of dehydroxylation en-
dotherms typical of smectite and a conversion to tenorite.
The latter two aspects mark a clear difference to discrete
smectite. Higher temperatures of tenorite formation in Chu-
qui 5 could correspond to the presence of Mn in the sample
(Table 2, black chrysocolla).
Specific surface area
Nitrogen adsorption indicated similar specific surface ar-
eas for samples 1 and 3 (purest and abundant samples avail-
able, Table 1), 465 and 457 m
2
/g respectively for <200 mesh
Tyler samples. However, it should be taken into consider-
ation that the samples have been heated to 350 °C, implying
a measurement post-dehydration. BET values drop to 276
and 229 m
2
/g respectively, if this analysis is carried out post
degasification in vacuum at only 60 °C. These values agree
with specific surface areas of 200 m
2
/g calculated by Wright
& Prosser (1965). Prosser et al. (1965) further described
chrysocolla to have high porosity, permitting a water uptake
up to 25 % without noticable expansion of the lattice. How-
ever, leaching or sintering drastically reduces specific sur-
face areas to < 50 m
2
/g (Pohlman & Olsen 1976; Raghaven
& Fuerstenau 1977).
Chemical analyses
Results are presented in Table 2. Due to the reduced amount
of sample available, electron microprobe analyses have been
chosen as the most adequate method to overcome this restric-
tion. Despite careful handpicking, traces of quartz, feldspar,
gypsum and possibly kaolinite have been detected in three
samples (Table 1). These are reflected as minor contents of
Al
2
O
3
, Na
2
O, SO
3
and CaO (Table 1, samples 1, 4, 5, collec-
tion sample). However, only sample Chuqui 2 proved to be a
very fine impregnation of feldspar by chrysocolla, making
EPMA single phase analysis impossible.
With respect to total and soluble copper analyses, it is inter-
esting to note that two very pure chrysocolla samples (Chuqui
1 – dark green, Chuqui 3 – light blue) correspond to the
highest and lowest Cu solubilities respectively (Table 2).
Discussion and conclusions
Research on chrysocolla has developed along three lines: (1)
metallurgy with a focus on methods to characterize bulk ore
for flotation and leaching for process design and quality con-
trol, (2) synthesis of transition metal smectites and (3) crystal-
lographic work at TEM scale. Unfortunately, cooperation be-
tween these closely related subjects has been restricted. The
present discussion will first focus on the texture and structure
of the mineral and then routine methods best suited for mineral
identification will be examined.
Texture and structure
With respect to crystallographic characterization Van Oost-
erwyck-Gastuche (1970) has pointed out the presence of fi-
brous phases on low magnification TEM images. That author
proposed a palygorskite-like structure and rejected smectite-
type structures as well as a completely amorphous character.
She proposes for the c-dimension a distorted kaolinite lattice;
the distortion being due to the incorporation of Cu
2+
in Al
3+
sites thus resulting in a chain structure perpendicular to the c-
axis. The present TEM observations confirm the presence of
fibrous aggregates without regular stacking (Fig. 6), appearing
as individual strings in an apparently amorphous groundmass.
XRD traces show the lack of long range ordering, whereas
FTIR suggests the presence of trioctahedral layer-silicate
phases. Fibres have thicknesses of approximately 1.3—1.5 nm,
but no HRTEM stacks or packages could be observed. EXAFS
studies (McKeown 1994) have explored the Cu-Cu and Cu-O
environment of chrysocolla, suggesting possibly linked CuO
4
units similar to tenorite and dioptase; however Cu-Si correla-
tions remain unknown.
The presence of poorly crystallized material is corroborated
by (1) the difficulty of focus for diffraction images and (2)
Table 1: Mineralogy of chrysocolla samples.
Chuqui 1
Chrysocolla***
Quartz*
Chuqui 2
Chrysocolla**
Plagioclase***
Quartz**
Gypsum**, Smectite or 1.4 o 7 nm Phyllosilicate?
Chuqui 3
Chrysocolla***
Chuqui 4
Chrysocolla***
Gypsum*
Chuqui 5
Chrysocolla***
Gypsum*
Quartz*
1.4 or 0.7 nm Phyllosilicate?
Collection Sample
Chrysocolla**
Malachite**
Quartz*
***
major phase,
**phase present, *phase at trace level, ? phase difficult to identify due to very low peak intensities and/or
superposition of peaks
118 KELM et al.
blister like TEM images (Figs. 4, 5) adjusting under the elec-
tron beam.
Electron microprobe defocussed spot analyses indicate the
presence of 1—1.5 % MgO+CaO+K
2
O+Na
2
O (Table 2). These
oxides are attributed to “contaminating” matrix minerals such
a feldspar and traces of chlorite, although the latter has not
been clearly identified in the present samples; its presence is
known from the study of the host rocks impregnated by
chrysocolla. The respective Al
2
O
3
contents for monomineralic
samples Chuqui 1 and 3 are 3.7 % and 3.1 %. Ratios of SiO
2
/
CuO are 1.19 and 1.39. Results are in good agreement with
wet chemical analyses compiled by Van Oosterwyck-Gastuche
(1970), although differences in the analytical method (AAS)
and the date of the cited analyses have to be taken into consid-
eration. Given the obvious partially amorphous character of
chrysocolla, it is felt that at present no compositional formula
based on crystalline smectite can be given. The uncertainty of
proposing a definite mineral formula has also been observed
for natural Cu incorporating smectites from Peru (Plötze &
Wolf 1999). The specification of Cu content in the crystalline
phase has only been possible for synthetic smectites of transi-
tion elements. Conditions of synthesis permitting the incorpo-
ration of Cu in the octahedral sheet require Cu/(Cu + Mg) ra-
tios <0.5 (Kloprogge et al. 1999), otherwise formation of
chrysocolla occurs (Güven 1988). The element contents deter-
mined for the present study (Table 2) do not indicate condi-
tions favourable for the formation of smectite phases. Xiao &
Villemure (1998) report synthesis experiments in which a non-
expanding chrysocolla phase results in the SiO
2
/CuO ratio
close to 1. At ratios around 1.3, an expandable phase forms.
Chuqui 1, 4, 5 and the collection sample have ratios close to 1,
Chuqui 3 approaches a ratio of 1.3 (Table 2). The presence of
minor quartz in the samples may affect the ratio. For most
samples no clear shift upon EG treatment has been observed,
although they seem to be sensitive to this treatment. The only
sample clearly sensitive to EG treatment is Chuqui 3, although
devoid of a discrete peak, which would support the above
mentioned observation by Xiao & Villemure (1998). The col-
lection sample should be considered with caution due to the
presence of malachite, although this mineral can be texturally
distinguished from chrysocolla under EPMA.
The large difference in the total and soluble copper analyses
between Chuqui 1 and 3 may point to the bonding of Cu in a
possibly better developed phyllosilicate structure, which is
more resistant against an acid attack. However, the increased
hydration of Chuqui 1 could have promoted leaching, although
further samples are needed to afirm or reject this suspicion.
Furthermore, chrysocolla is a high surface area mineral. Val-
ues of >400 m
2
/g (heated 350 °C) or 250 m
2
/g (60 °C) are
comparable with smectites, thus making acid attack easy.
FTIR (Si-O stretching band and OH-stretching vibration) and
XRD data show clear differences between Chuqui 3 and other
samples. FTIR seems to be more sensitive to the structuring of
the starting medium for crystallization (natural or synthetic),
Fig. 8A. TG curves of chrysocolla separates.
IDENTIFICATION METHODS FOR CHRYSOCOLLA – A Cu-SMECTITE-LIKE SILICATE 119
as it was observed also for synthetic ammonium illite (Šucha
et al. 1998) where XRD has shown no reflection, but FTIR
clearly indicated evolution of the structure towards dioctahe-
dral ammonium mica. As each layer of a layer silicate absorbs
infrared radiation almost independently of its neighbours, the
recognition of their structures by their absorption pattern does
not depend on the presence of the packets of layers required to
give a distinct X-ray diffraction (Farmer 1974). Chrysocolla
samples, except Chuqui 3, show no clear XRD peaks, but
FTIR indicate several trioctahedral layer silicate features, most
Fig. 8B. DTA curves of chrysocolla separates.
Table 2: Chemical analyses of chrysocolla samples by electron microprobe (EPMA) and atomic absorption spectroscopy (AAS). Elec-
tron microprobe analyses represent averages of 4-5 determinations. A complete set of analyses is available from the authors. Sample
Chuqui 2 represents a mixture of chrysocolla and feldspar (Table 1), no single phase analyses could be obtained.
Sample
Method
Chuqui 1
Chuqui 2
Chuqui 3
Chuqui 4
Chuqui 5
blue-green
Chuqui 5
black
Collection
Sample
Total Cu % (CuO %)
AAS
27.4 (34.3) 10.6 (13.25) 26.1 (38.75) 31.0 (32.25)
25.8 (32.25)
31.95 (39.9)
Soluble Cu % (CuO %)* AAS
27.0
9.5
18.1
28.9
24.8
29.38
SiO
2
%
EPMA
51.28
55.45
49.84
52.47
45.86
50.00
TiO
2
%
EPMA
0.02
0.02
0.01
0.00
0.00
0.02
Al
2
O
3
%
EPMA
3.68
3.06
0.84
4.00
3.99
4.21
FeO %
EPMA
0.02
0.01
0.00
0.16
0.03
0.00
MnO %
EPMA
0.02
0.02
0.07
0.01
6.94
0.01
MgO %
EPMA
0.44
0.28
0.11
0.39
0.39
0.49
CaO %
EPMA
1.05
0.59
0.24
0.94
0.86
1.44
Na
2
O %
EPMA
0.24
0.26
0.01
0.39
0.57
0.03
K
2
O %
EPMA
0.05
0.11
0.00
0.12
0.10
0.02
CuO %
EPMA
43.07
39.87
48.72
41.07
41.24
43.71
SO
3
%
EPMA
0.07
0.32
0.17
0.22
0.06
0.01
SiO
2
/CuO
1.39
1.02
1.28
1.11
0.95**
1.14
* Leaching by 1 molar citric acid for 1.5 hours. ** SiO
2
/ (CuO + MnO)
120 KELM et al.
probably without any periodicity. Chuqui 3 seems to be a more
evolved stage towards a phyllosilicate with a periodical struc-
ture. The differences in perfection of the sample structure
could ultimately be an explanation of the differences in the
Cu-leaching experiments, but further verification with more
samples is required. Conclusions explaining the <0.5 substitu-
tion of Mg by Cu in synthesis experiments of smectites may
also support the above ideas. Güven (1988) considers impor-
tant the similarity of the Si/metal ratio of the initial precipitate
to that of the resulting smectite (a feature also observed in zeo-
lite synthesis, Sanhueza et al. 1999). Under natural conditions,
environments favourable for formation of chrysocolla or
smectite-like phases are likely to fluctuate on a small scale un-
der natural conditions. The dominance of Cu over other metals
(see qualitative EDAX and EPMA analyses) also may prevent
the formation of smectite as observed by Mosser et al. (1990),
who found chrysocolla at Cu/Mg ratios >0.5. These authors
explain this phenomenon (using EXAFS) by a string distortion
of Cu octahedra due to the Jahn-Teller effect.
Methods of routine analysis
In determining differences among the chrysocolla-phyllosil-
icate materials, both DTA-TG and FTIR appear as powerful
tools, which could be applied also for technological purposes.
The only concern is that samples have to be picked very care-
fully in order to avoid any admixture which can affect the final
result of detailed XRD and FTIR measurement. XRD is poten-
tially useful, provided samples are of high purity and careful
EG treatment is carried out.
For routine analyses of chrysocolla ores, availability of the
method(s), easy use and time requirements have to be consid-
ered. For purposes of production, time available for analyses
may be only several hours. Unequivocal identification of
chrysocolla by X-ray diffraction is only possible when sam-
ples have high contents of the mineral and EG treatment is car-
ried out. The simultaneous presence of chrysocolla and dis-
crete smectite is problematic with respect to identification of
the expanding phase and its potentially elevated acid con-
sumption during leaching. IR spectra show chrysocolla with
several features of a trioctahedral smectite-like phase. Differ-
entiation of the mineral from a separate smectite phase will re-
quire considerable interpretation skills, not easily found in
quality control laboratories on site or in mining contract ser-
vices. For the samples analysed, a difference has been ob-
served in the DTA-TG curves for chrysocolla and chrysocolla-
smectite samples. The routine analyses of total and soluble
copper in bulk samples do not help the identification of
chrysocolla and especially chrysocolla-smectite mixtures, un-
less performed first on carefully picked monomineralic speci-
mens. On the basis of the present study, despite the reduced
data set, the authors propose identification of chrysocolla-con-
taining mineralization types by combining X-ray diffraction
and infrared spectroscopy at the stage of mineralization type
study. For a later fast identification of chrysocolla and chryso-
colla—smectite mixtures in a production process, where miner-
alization types and general rock mineralogy are already
known, thermal analysis represents a potential alternative, due
to its ease of sample preparation and interpretation (once cali-
brated) and the comparative low cost of a thermobalance.
Acknowledgements: The financial support of Fondecyt
Project 1980503 (Chilean Science Foundation) and Slovak
Grant Agency (Grant 2/7202) is appreciated. The authors
thank Andrés Reghezza, Superindentente de Ingeniería Met-
alúrgica, Subgerencia Mina Sur, CODELCO, Chuquicamata
for the supply of chrysocolla reference samples and staff in-
volved at Concepción and Bratislava Universities for their
help at all stages of the study. The comments and suggestions
by Sabine Petit and Istvan Viczian have greatly helped to im-
prove the manuscript.
References
Anthony J.W., Bideaux R.A., Bladh K.W. & Nichols M.C. 1995:
Handbook of mineralogy. Volume II, Silica, Silicates. Mineral
Data Publishing, Tucson, Arizona, USA, 1—446.
Bish D.L. & Duffy D.L. 1990: The thermogravimetric analysis of
minerals. In: Stucki J.W., Bish D.L. & Mumpton F.A. (Eds.):
Thermal Analysis in Clay Science. Clay Mineral Society Work-
shop Lectures 3, The Clay Mineral Society, Aurora, Colorado,
USA, 95—189.
Decarreau A. 1980: Cristallogenése expérimentale de smectites
magnésiennes: Hectorite, stévensite. Bull. Mineralogie 103,
579—590.
Decarreau A., Grauby O. & Petit S. 1992: The actual distribution of
octahedral cations in 2:1 clay minerals: Results from clay syn-
thesis. Appl. Clay Sci. 7, 147—167.
Ďu a R. & Rejl L. 1997: Edelsteine. Handbuch und Führer für den
Sammler. Aventinum Nakladatelství s r.o., Prague, 1—407.
Farmer V.C. 1974: Layer silicates. In: Farmer V.C. (Ed.): Infrared
Spectra of Minerals. Mineral. Soc. 331—363.
Foote H.W. & Bradley W.M. 1913: The composition of amourphous
minerals as illustrated by chrysocolla. Amer. J. Sci., 4th Ser. 36,
180—184.
Gomer J.S., Yopps S.W., Sandoval S.P. & Clark A.E. 1991: Copper
exchange capacity of clays and their potential effect in in situ
copper leaching. United States Department of Interior, Bureau
of Mines, Rep. Invest. 9396, 10.
Güven N. 1988: Smectites. In: Bailey S.W. (Ed.): Hydrous Phyllo-
silicates (Exclusive of Micas). Reviews in Mineralogy, Volume
19, Mineral. Soc. America 497—559.
Kloprogge J.T., Komarneni S. & Amonette J.E. 1999: Synthesis of
smectite clay minerals: a critical review. Clays and Clay Miner.
47, 5, 529—554.
Lindsay N.M. 1994: Process mineralogy of Manto Verde, Chile.
Austr. IMM Proc. 299, 2, 29—34.
McKeown D.1994: X-ray absorption spectroscopic study of copper
in an amorphous copper silicate: chrysocolla. J. Non-Crystal-
line Solids 180, 1—10.
Meach J.A. 1981: Chrysocolla ores. A technical and economic eval-
uation. International Academic Services Limited, Kingston,
Ontario, Canada, 205.
Newberg D.W. 1967: Geochemical implications of chrysocolla
bearing alluvial gravels. Econ. Geol. 62, 932—956.
Mosser C., Mestdagh M., Decarreau A. & Herbillon A. 1990: Spec-
troscopic (ESR, EXAFS) evidence of Cu for (Al-Mg) substitu-
tion in octahedral sheets of smectite. Clay Miner. 25, 271—282.
Patterson E. & Swaffield R. 1987: Thermal analyses. In: Wilson
M.J. (Ed.): A handbook of determinative methods in clay min-
IDENTIFICATION METHODS FOR CHRYSOCOLLA – A Cu-SMECTITE-LIKE SILICATE 121
eralogy. Blackie, Glasgow and London, 99—132.
Plötze M. & Wolf D. 1999: Incorporation of Cu in smectite from
Cerro Verde-Santa Rosa (Peru). Euroclay, 1999, Krakow, Po-
land, Abstracts.
Pohlman S.L. & Olson F.A. 1976: Characteristics of chrysocolla to
acid leaching. Proceedings of International Symposium on
Copper Extraction and Refining, Las Vegas, USA, February
1976, 943—959.
Prosser A.P., Wright A.G. & Stephens J.D. 1965: Physical and
chemical properties of natural copper silicates which resemble
chrysocolla. Trans. IMM 74, 233—258.
Raghaven S. & Fuerstenau D.W. 1977: Characteristics and pure
strucutre analysis of a copper ore containing chrysocolla. Int. J.
Miner. Proc. 4, 381—394.
Ruiz C. & Peebles F. 1988: Geología, distribución y génesis de los
yacimientos metalíferos Chilenos. Editorial Universitaria S.A.,
Santiago de Chile, 334.
Sanhueza V., Kelm U. & Cid R. 1999: Synthesis of molecular sieves
from Chilean kaolinites: 1. Synthesis of NaA type zeolite. J.
Chem. Tech. Biotech. 74, 356—363.
Šucha V., Elsass F. & Komadel P. 1998: Hydrothermal synthesis of
ammonium illite. Amer. Mineralogist 83, 1/3, 58.
Tessier D. 1984: Étude experimental de l‘organisation des materi-
aux argileux. Dr. Science Thesis, Univ. Paris VII, INRA publ.
361.
United States Patent 1980: Recovery of copper from copper oxide
minerales. United States Patent 4,212,983, Rule, T.W. (Inven-
tor), The Hanna Mining Company, Cleveland, OH, USA.
Van Oosterwyck-Gastuche M.C. 1970: La structure de la chrysocol-
le. Comptes Rendus de l’Académie des Sciences, Paris, 271,
1837—1840.
Wright A.J. & Prosser A.P. 1965: Study of the reactions and flota-
tion of chrysocolla with alkali-metal Xanthates and sulphides.
Trans. IMM 74, 259—279.
Xiao J. & Villemure G. 1998: Preparation, characterization and
electrochemistry of synthetic copper clays. Clays and Clay
Miner. 46, 195—203.
Yates D.M., Joyce K.J. & Heanney P.J. 1998: Complexation of cop-
per with polymeric silica in aqueous solution. Appl. Geochem.
13, 235—341.