Journal of South American Earth Sciences 25 (2008) 359–376

www.elsevier.com/locate/jsames

Morphology and composition of gold in a lateritic profile,

Fazenda Pison ‘‘Garimpo’’, Amazon, Brazil
a,*
J.H. Larizzatti , S.M.B. Oliveira b, C.R.M. Butt c

a
CPRM/Geological Survey of Brazil, Av. Pasteur, 404, Rio de Janeiro, RJ 22290-240, Brazil
b
Universidade de São Paulo, Rua do Lago, 562, São Paulo, SP 05508-900, Brazil
c
CRC LEME, CSIRO Exploration and Mining, PO Box 1130, Bentley, WA 6102, Australia

Abstract

This study describes the morphological evolution of gold grains in a lateritic weathering profile in an equatorial rainforest climate.
Primary sources of gold are quartz veins associated with shallow granophyric intrusion. Gold grains were found in fresh ore, saprolite,
transition zones, ferruginous duricrust, red latosol, and yellow latosol. Irregularly shaped grains predominate, with smaller proportions
of dendritic and prismatic forms. Gold grains are weathered in the uppermost 10 m of the regolith. Mean gold grain size is maximum in
the duricrust (>125 lm) and decreases progressively upward into the yellow latosol (<90 lm). Voids and corrosion pits appear on grain
surfaces, and progressive rounding is observed from the bottom of the profile to the top. Gold grains can be classified as either homo-
geneous or zoned with respect to their chemical composition. Homogeneous grains contain 2–15% Ag (mean 8.3%). Zoned grains have
more variable Ag contents; grain cores have means of approximately 10% or 23% Ag, with Ag-poor zones of approximately 3.7% Ag
along internal discontinuities and/or outer rims. Formation of Ag-poor rims is due to preferential depletion of silver. Processes responsible
for duricrust formation may preserve some grains as large aggregates, but subsequent transformation into latosol further modifies them.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Gold grains; Morphology; Composition; Weathering; Laterites; Amazon

1. Introduction and the presence of corrosion features on the surface of pri-

mary crystals (Colin et al., 1989a,b; Colin and Vieillard,
The supergene mobility of gold in weathering profiles 1991; Porto and Hale, 1996). Gold precipitation is demon-
has been well documented in different parts of the world strated by the formation of secondary crystals of high fine-
(e.g., Boyle, 1979; Mann, 1984; Wilson, 1984; Webster ness and characteristic morphology (Angélica et al., 1995;
and Mann, 1984; Davy and El-Ansary, 1986; Michel, Freyssinet et al., 1987; Grimm and Friedrich, 1990; Bowell,
1987; Freyssinet and Butt, 1988; Colin et al., 1989a,b; 1992; Lawrance and Griffin, 1994) and gold nuggets enclos-
Freyssinet et al., 1987; Grimm and Friedrich, 1990; Colin ing neoformed goethite (Machairas, 1963; Wilson, 1984;
and Vieillard, 1991; de Oliveira and Campos, 1991; San- Vasconcelos and Kyle, 1991; Santosh and Omana, 1991;
tosh and Omana, 1991; Vasconcelos and Kyle, 1991; Edou de Oliveira and Campos, 1991; Zang and Fyfe, 1993). Fur-
Minko et al., 1992; Bowell, 1992; Freyssinet, 1993; Law- ther data supporting gold mobility in the supergene envi-
rance and Griffin, 1994; Sergeev et al., 1994; Porto and ronment include elevated gold concentrations in stream
Hale, 1996; de Oliveira and de Oliveira, 2000). Evidence waters draining auriferous zones (Benedetti and Boulège,
of gold dissolution includes the decreasing size of gold 1991; Andrade et al., 1991) and grain growth in stream sed-
grains from the bottom to the top of weathering profiles iments (Eyles, 1990; Groen et al., 1990; Craw, 1992;
Youngson and Craw, 1995; McCready et al., 2003). The
*
Corresponding author. Tel.: +55 21 2542 0586; fax: +55 21 2295 4292. objective of this study is to provide some data on the
E-mail address: [email protected] (J.H. Larizzatti). morphology and composition of gold in a well-preserved

0895-9811/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsames.2007.06.002
360 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

lateritic profile derived from a quartz (+pyrite) vein system by late sinistral brittle faults (ENE). A hydrothermal
in the Tapajós region. alteration envelope follows the host structure and is
expressed by mineralogical zoning around the mineralized
quartz veins:
2. Description of the area
Central zone: gold, quartz, white mica, epidote, pyrite,
hematite, and magnetite;
2.1. Climate
Intermediate zone: minor gold, quartz, white mica, epi-
dote, fluorite, and magnetite; and
The Fazenda Pison ‘‘Garimpo’’ (artisinal mine working)
Distal zone: minor gold, quartz, white mica, chlorite,
is located in northern Brazil, near the Tapajós River, about
and magnetite.
600 km SE of Manaus, Amazonas State (Fig. 1). The area
At the top of the saprolite, a meter-wide stockwork zone
has a warm to hot humid rainforest climate, with mean
is visible adjacent to the quartz veins. The lode is subverti-
minimum temperature of 17 C (February), mean maxi-
cal, with an overall width, including the alteration zone, of
mum temperature of 38 C (July), and mean annual rainfall
20 m. Rounded, decimeter-sized, quartz-rich zones can be
of 2750 mm. The relief is of the ‘‘half-orange’’ type, at ele-
observed within the stockwork zone.
vations of 140–200 m above sea level.
The Amazon region is characterized by a deep lateritic
regolith, and at Fazenda Pison, the weathering profile is
2.2. Geology 20–30 m deep. The studied profile is located in a hillslope.
Six main units can be recognized in a fully preserved later-
The Garimpo is located in the Tapajós Province itic profile (Fig. 2):
(Almeida et al., 1981), Tapajós Mineral Province (Santos
et al., 2001). Felsic volcanic rocks (Iriri Group, Andrade 1. Saprolite: >20 m thick, light red, with well-preserved
et al., 1978; CPRM, 1999) occur where shallow granitic primary fabrics and structures. Primary white micas can
intrusions (Maloquinha Intrusive Suite; Santos et al., be recognized, but feldspars, chlorite, fluorite, and sulp-
1997) are emplaced along brittle shear zones. hides cannot (boxwork texture is present). Quartz, white
Primary mineralization at Fazenda Pison follows a
third-order brittle shear zone (N20E) hosted by fine
granophyric rock. The mineralized veins are dislocated

Fig. 1. Area localization and geological sketch modified from RTDM

(1995). Fig. 2. Weathering profile and sampling sites.
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 361

mica, kaolinite, and hematite (+little goethite) are the most 5. Red latosol: clay-rich matrix (minor Fe) with nodules
abundant minerals present. The contact with lower transi- and pisoliths derived from the duricrust. Thickness var-
tion zoneisgradationaland centimetric; saprolitefragments ies from less than 1 m at the base of slopes to 6 m on hill
decrease in size and are immersed in a clay-rich matrix. crests. From duricrust to red latosol occurs an increase
2. Lower transition zone: <0.5 m thick, red, with numerous in quartz and kaolinite contents and important decrease
weathered lithic fragments that preserve primary rock in iron oxi-hydroxides (especially hematite). The transi-
fabrics immersed in a kaolinite-rich matrix containing tion to yellow latosol is gradational; at the top horizon,
iron oxi-hydroxides (hematite/goethite). From saprolite iron-rich pisoliths are not observed.
to the lower transition zone, there is an increase in 6. Yellow latosol: similar to red latosol, with lower con-
quartz content, decrease in white-mica content, and no tents of iron oxi-hydroxides. From red to yellow latosol,
change in iron oxi-hydroxides and kaolinite contents. minor variations in quartz, kaolinite, and iron oxi-
The transition to the upper transition zone is grada- hydroxides contents are observed. At the top of this
tional; iron-rich pisoliths are formed and increase in size, horizon rests a centimetric layer of a dark-brown
forming nodules in the top of the horizon. organic matter-rich material.
3. Upper transition zone: 0.2 m thick, with ferruginous
pisoliths and nodules embedded in a ferruginous kaoli-
nite matrix. No primary fabric can be observed in this 3. Methods
material. From lower to the upper transition zone,
quartz and white-mica contents decrease, and an impor- A schematic cross-section of the pit at the Fazenda Pison
tant increase in iron oxi-hydroxides contents (specially Garimpo and the sampling sites are shown in Fig. 2. Samples
hematite) and kaolinite can be noted. The transition to were collected from the mineralized zone and different hori-
duricrust is gradational; iron-rich nodules become bigger zons of the regolith developed above it, as follows: One sam-
and more numerous and form an indurated duricrust. ple (0.94 kg) of primary ore was collected from a diamond
4. Duricrust: (discontinuous ferruginous horizon) of vari- drill core and jaw-crushed to 5 mm, then ground in a disc mill
able thickness (usually <1 m), with a high concentration to 1.5 mm. It is not shown in Figs. 2 and 3.
of hematite. From upper transition zone to duricrust, an Bulk samples of 50 kg each where taken from the rego-
important decrease in quartz content and important lith as follows: (1) six samples of saprolite, specifically, four
increase in hematite/goethite and kaolinite contents samples from granophyre saprolite, at 10, 8, 7, and 4 m;
occur. Contact with red latosol is sharp, but at the base, one sample from the stockwork zone; and one sample from
it is possible to observe iron-rich pisoliths with yellow a quartz-rich zone; (2) two samples from the transition
cortexes immersed in the clay-rich matrix. zone, one from the upper and one from the lower; (3)

Fig. 3. Characteristics of gold particles according to regolith unit (YL, yellow latosol; RL, red latosol; D, duricrust; TZt, transition zone-top; TZb,
transition zone-base; G4, saprolite, granophyre 4 m deep; QZ, saprolite, quartz-rich zone; SZ, saprolite stockwork zone; G7, saprolite; granophyre 7 m
deep; G8, saprolite, granophyre 8 m deep; G10, saprolite, granophyre 10 m deep). (A) Number of particles; (B) particle size (lm); (C) particle shape (%),
where Prism, prismatic; Dendr., dendritic; Irreg., irregular; and (D) edge roundness (%), where V. Rounded, very rounded; S. Rounded, sub-rounded.
362 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

one sample from the duricrust; (4) one sample from the red Table 2
latosol; and (5) one sample from the yellow latosol. Number of gold grains analysed and number of analyses according to
sampling medium (EMPA)
Samples were processed as follows:
Sampling Homogeneous Zoned grains
medium grains
1. Field: samples collected at the lateritic profile were
panned and dried in the field. Samples from primary Number Number Number Number of Number of
of of of analyses – analyses –
ore and duricrust were hand crushed before panning. grains analyses grains core rim
2. Laboratory preparation: samples were oven dried at 70 C.
Yellow 3 11 7 11 9
Magnetic minerals were separated using a hand magnet latosol
and Frantz magnetic separator. Heavy minerals were sepa- Red latosol 4 15 1 1 2
rated using heavy liquid (methylene iodide). Gold grains Duricrust 6 21 8 38 20
were handpicked under a binocular microscope. Transition None None 5 19 4
3. Optical examination: gold grains were mounted on glass zone
Stockwork None None 5 10 7
slides with double-sided adhesive tape and examined using Granophyre None None 9 19 14
a binocular microscope. Grains were classified according Fresh ore 4 9 2 2 2
to their morphology; some were selected for further study.
4. Scanning electron microscopy (SEM): grain morphology
was examined by SEM, and the Ag content was determined
semi-quantitatively using an energy dispersive spectrometer. in Fig. 3. The grains were classified in terms of location
Instruments used were a JEOL JSM-6400 and Electroscan of regolith material (Fig. 2); size, as (length · width)1/2
ESEM(S) E-3, Centre of Microscopy and Microanalysis, (Fig. 3B); habit, as to prismatic, when gold grains show
University of Western Australia (CMM/UWA), operating similar length in three dimensions and smooth/near
at 20–30 kV accelerating voltage, 3 · 10 12 A primary beam smooth faces (e.g., Figs. 4B, 7E, and 8B) or dendritic, when
current, and 9–20 mm working distance. one dimension is much more developed than the other two
5. Polished section: selected grains were examined in pol- (e.g., Figs. 6A, B, and 7F) or irregular, when gold grains do
ished section by SEM. not show smooth faces or one distinctive dimension (e.g.,
6. Electron microprobe (EPMA): fifty-four gold grains Figs. 5A, B, E, F, 7A–D, and 8A, E, F); and edges, in terms
(polished sections) were analysed for Au, Ag, As, Bi, of whether they are sharp (e.g., Figs. 4A, B, and 5A), sub-
Cu, Fe, Hg, Pt, Sb, and Zn using a JEOL 6400 instru-
ment at CMM/UWA. Operating conditions and detec-
tion limits are given in Table 1.

Approximately 1850 gold grains were examined by opti-

cal microscopy, 300 representative gold grains by SEM,
and 54 grains by EPMA (Table 2).
The data discussed in this paper refer only to the grains
recovered and examined by these procedures. A high pro-
portion of grains smaller than 10 lm were lost, with mod-
erate losses of those between 10 and 20 lm.

4. Results

4.1. Size and morphology of gold grains

4.1.1. Classification
The number and morphological characteristics of gold
grains in the different regolith horizons are summarized
Table 1
Operating conditions and detection limits for EMPA
Instrument conditions
Accelerating voltage = 25 kV Primary beam
12
current = 2.8 · 10 A
Au Ag As Bi Cu Fe
Detection limit (wt%) 0.131 0.033 0.031 0.092 0.018 0.017
Counting time (s) 25 25 25 25 25 25
Fig. 4. Backscattered electron images (SEM) of primary gold grains.
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 363

Fig. 5. Backscattered electron images (SEM) of gold grains recovered from saprolite. (A–D) Granophyre; (E) stockwork zone; and (F) quartz-rich zone.

rounded (e.g., Figs. 5C, E, F, and 6A), rounded (e.g., Figs. iable, as reflected by the number of grains recovered at dif-
5B and 6B), or very rounded (e.g., Fig. 8A and F). ferent horizons (Fig. 3). The mean grain size is 92 lm
(r = 46). The majority of grains (Fig. 3C) are irregular in
4.1.2. Gold in primary mineralization shape (83%), with about 10% of the grains dendritic and
Primary gold is associated with pyritic quartz veins. the remainder (7%) prismatic. Most grains have slightly
Only 12 gold grains were recovered from the fresh ore, with rounded edges (94%), with the remainder strongly rounded
a mean size of 56 lm. All grains are xenomorphic, with or sharp.
very smooth faces that show the impression of the sur- More than 90% of the grains recovered from both stock-
rounding quartz. There are two habits: irregular work zones and quartz-rich zone are irregular (Fig. 5E and
(Fig. 4A) and prismatic (Fig. 4B). F); the remainder are dendritic. Most grains from quartz-
rich and stockwork zones show subrounded edges (85%
4.2. Saprolite and 92%, respectively), and the rest show rounded edges.
Corrosion pits occur on most grains observed in the sapro-
Grains separated from saprolite are shown in Fig. 5. The lite materials (Fig. 5A–F). Some grains show scratches
gold distribution in the granophyre saprolite is highly var- (tool marks).
364 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

upper transition zone; and the remainder are dendritic

(Fig. 3C). Nearly 90% of the grains have very rounded
edges, all show numerous corrosion pits (Figs. 3D and
8A, B), and the mean grain size (105 lm, r = 63; Fig. 3B)
is smaller than the underlying duricrust.
In the yellow latosol (Figs. 3C, 8C–F), the proportion of
prismatic grains is approximately 6%, similar to that in the
saprolite; most grains (88%) are irregular, and 6% are den-
dritic. The mean grain size is less than in the red latosol
(86 lm, r = 28). Most grains (65%) have very rounded
edges with well-developed corrosion pits; nevertheless, it
is possible to recognize the primary habit in many of the
gold particles.
The progressive decrease of mean grain size from the
duricrust to the yellow latosol, together with the increasing
number of corrosion pits and degree of rounding, implies
dissolution of gold. Gold is preferably dissolved from
prominent parts of grains, and dissolution seems more
active on dendritic and irregular grains than prismatic
grains. Some grains recovered from both latosols show tool
marks.

4.6. Chemical composition of gold grains

Chemical analysis of 54 gold particles was performed

by EPMA (Tables 1–4). Only Au, Ag, and Cu are present
Fig. 6. Backscattered electron images (SEM) of gold grains recovered in detectable amount in most grains; some also contain
from transition zone. (A) Bottom of transition zone showing subrounded As, Bi, and Fe. Silver is the only element, other than gold,
edges; (B) top of transition zone showing rounded edges. present in high concentrations. According to the silver
distribution, there are two groups of gold grains: homoge-
4.3. Transition zone neous and zoned (Fig. 9). The 17 analysed homogeneous
grains (56 analyses) show a similar grey hue throughout
Mean grain size is 108 lm (r = 76) at the base and the hole-polished section under SEM backscattered mode
81 lm (r = 23) at the top of transition zone. Most grains (Fig. 9A) and have Ag contents of 2.40–17.34% (mean
from the transition zone are irregular (86%), and the rest 8.28%; Tables 3 and 4). The 37 analysed zoned grains
are dendritic (8%) or prismatic (6%). There are more pris- (54 rim and 104 core analyses) show variable silver con-
matic grains (12%) in the upper part of the zone than all tent throughout the polished section under SEM back-
underlying horizons. Edges and faces of grains are cor- scattered mode (Fig. 9B), lower silver contents in the
roded and pitted (Fig. 6A and B). Some grains show outer rims, and internal discontinuities (0.02–14.09%,
scratches (tool marks) and folded rims. mean 3.64% Ag) compared with the cores (3.84–31.10%,
mean 17.30% Ag; Tables 3 and 4). These internal varia-
tions are evident using SEM backscatter imaging and
4.4. Duricrust can be quantified by EPMA analysis. Some homogeneous
and zoned grains show compositional variation within a
Gold grains in the duricrust have the greatest mean size single grain, core, or rim. In zoned grains, the contact
(124 lm, r = 72; Fig. 3B). Nearly 80% of the grains are between grain cores and silver-poor rims is commonly
irregular (Figs. 3C and 7A–D), with nearly equal propor- sharp. Zoned gold grains from the fresh ore do not show
tions of prismatic (Figs. 3C and 7E) and dendritic (Figs. silver-poor rims; they show internal variations in Au and
3C and 7F). Most grains have slightly rounded edges Ag contents that cannot be characterized as rims or cores,
(96%) and numerous corrosion pits; some grains show and they were not studied together with regolith zoned
scratches. gold grains.

4.5. Latosol 4.6.1. Homogeneous grains

Homogeneous gold grains were recovered from primary
In the red latosol, only 53% of the grains are irregular in mineralization (fresh ore), duricrust, and latosols. The
habit; some 40% are prismatic, a much higher proportion compositions of homogeneous particles vary between
than all underlying horizons, including the duricrust and 82.25% and 98.20% Au and 2.40% and 17.34% Ag
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 365

Fig. 7. Backscattered electron images (SEM) of gold grains recovered from duricrust. (A–D) Irregular particles; (E) prismatic particle; and (F) dendritic
particle.

(Fig. 10, Table 4), with minor variation in Au–Ag contents for all zoned gold grains; the composition of rims and
within the same grain (exception for one gold grain col- cores can be observed in black. Rims are composed of sil-
lected in primary ore that shows variation of 6.94% Au). ver-poor gold (mean 3.64% Ag) and tend to be thinner in
The highest Cu contents were observed in gold grains the latosol horizons. Composition of cores is bimodal,
recovered from the fresh ore; high Ag contents are concen- with cores containing less silver (mean 9.76% Ag) or more
trated in yellow latosol grains; and high gold content is silver (mean 22.60% Ag). Due to the small number of
concentrated in duricrust grains (Fig. 11). These differences grains analysed, it is not possible to correlate core and
in composition between horizons may be a function of low, rim composition. Copper contents are similar to those in
non-representative number of samples. homogeneous grains. Grain cores contain <0.02–0.1%
Cu in most horizons, possibly richer (>0.04–0.09% Cu)
4.6.2. Zoned grains in the granophyre saprolite and poorer (<0.04% Cu) in
Zoned grains are present throughout the regolith, and the transition zone. Rims have a similar range of compo-
37 of them were analysed by EMPA (Tables 2 and 3). sitions (<0.02–0.15% Cu) in all horizons, except the stock-
Fig. 12A and B show the histograms of EMPA analyses work (0.23–0.30% Cu).
366 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

Fig. 8. Backscattered electron images (SEM) of gold grains recovered from latosols. (A–B) Particles from the red latosol, very rounded and subrounded,
respectively; (C–F) particles from yellow latosol, (D) is a detail from (C).

Some zoned grains show a thin, silver-enriched zone Gold grain sizes range from <20 to >120 lm. Primary
between the core and rim and in discontinuities within a grains are xenomorphic and irregularly distributed in the
single grain (Fig. 13). mineralized zone; gold grains are irregular, dendritic or
prismatic in shape, and show smooth faces and sharp
5. Discussion and conclusions edges. The effects of weathering are visible from the base
of the saprolite upward, with grains edges becoming
Particulate gold in the Fazenda Pison Garimpo is asso- rounded and corrosion pits evident on grain surfaces.
ciated with quartz veins, situated in voids between quartz Rounding increases slightly in the horizons of the base of
grains, and enclosed by or filling fractures in pyrite. Gold the profile and abruptly from duricrust to latosol. Tool
is also heterogeneously disseminated in the granophyre, marks and folded edges likely are products of crushing
stockwork zone, and quartz-rich zone. A lateritic profile and panning processes; these features are found only in
composed of saprolite, transition zone, duricrust, red lato- some of the studied gold grains and not as a general rule.
sol, and yellow latosol results from the weathering evolu- The degrees of rounding and pitting are not homoge-
tion of underlying mineralized granophyric rock. neous in gold grains, even those collected in the same
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 367

Table 3 Table 3 (continued)

Results of EMPA analyses of gold grains in polished sections
Particle Composition Au (%) Ag (%) Cu (%)
Particle Composition Au (%) Ag (%) Cu (%)
D4 Homogeous 95.68 4.50 0.06
YL1 Homogeous 88.52 11.95 0.03 D4 Homogeous 95.50 4.49 0.06
YL1 Homogeous 89.00 11.80 0.01 D5 Homogeous 90.47 9.66 0.08
YL1 Homogeous 88.10 12.20 0.05 D6 Homogeous 95.63 4.18 0.13
YL1 Homogeous 87.37 11.83 0.02 D6 Homogeous 93.11 5.95 0.08
YL1 Homogeous 88.25 12.27 0.03 D6 Homogeous 95.19 5.55 0.10
YL2 Homogeous 82.90 17.34 0.06 D6 Homogeous 94.88 5.22 0.08
YL3 Homogeous 86.44 13.96 0.07 D6 Homogeous 95.52 5.23 0.09
YL3 Homogeous 87.46 12.82 0.04 D7 Zoned – rim 96.30 3.75 0.08
YL3 Homogeous 86.80 13.07 0.04 D7 Zoned – rim 96.50 3.64 0.02
YL3 Homogeous 86.50 14.10 0.03 D7 Zoned – rim 96.30 3.61 0.09
YL3 Homogeous 86.64 13.04 0.08 D8 Zoned – rim 93.58 6.92 0.09
YL4 Zoned – rim 93.70 5.16 0.13 D8 Zoned – rim 93.52 7.14 0.13
YL5 Zoned – rim 88.60 12.06 0.06 D9 Zoned – rim 94.12 5.92 0.09
YL6 Zoned – rim 100.00 0.35 0.04 D9 Zoned – rim 95.20 4.84 0.02
YL6 Zoned – rim 99.70 1.02 0.06 D9 Zoned – rim 95.31 5.18 0.21
YL6 Zoned – rim 99.00 0.18 0.04 D10 Zoned – rim 96.40 3.73 0.02
YL7 Zoned – rim 99.60 0.85 0.01 D10 Zoned – rim 96.96 3.84 0.03
YL8 Zoned – rim 97.15 2.56 0.06 D11 Zoned – rim 97.60 2.37 0.04
YL9 Zoned – rim 97.60 2.51 0.05 D11 Zoned – rim 90.28 10.19 0.05
YL10 Zoned – rim 85.75 14.09 0.03 D12 Zoned – rim 99.70 0.69 0.04
YL4 Zoned – core 92.36 7.40 0.10 D12 Zoned – rim 100.00 0.16 0.03
YL5 Zoned – core 85.75 14.09 0.03 D13 Zoned – rim 100.00 0.29 0.03
YL6 Zoned – core 77.94 21.21 0.05 D13 Zoned – rim 100.20 0.22 0.05
YL7 Zoned – core 75.06 25.57 0.01 D14 Zoned – rim 100.30 0.15 0.07
YL8 Zoned – core 73.34 26.87 0.04 D14 Zoned – rim 100.30 0.07 0.03
YL8 Zoned – core 73.47 26.11 0.05 D7 Zoned – core 89.42 9.74 0.06
YL9 Zoned – core 74.36 24.42 0.03 D7 Zoned – core 90.60 9.29 0.05
YL9 Zoned – core 73.94 25.23 0.05 D7 Zoned – core 85.52 14.31 0.08
YL9 Zoned – core 74.56 24.68 0.03 D7 Zoned – core 87.07 13.14 0.07
YL9 Zoned – core 73.96 24.61 0.05 D8 Zoned – core 89.95 9.89 0.07
YL9 Zoned – core 74.67 24.49 0.04 D8 Zoned – core 90.34 10.15 0.04
YL9 Zoned – core 73.47 26.11 0.05 D8 Zoned – core 90.27 9.74 0.05
YL10 Zoned – core 71.25 28.64 0.07 D8 Zoned – core 90.82 9.78 0.04
RL1 Homogeous 91.99 7.52 0.08 D8 Zoned – core 77.66 23.10 0.03
RL1 Homogeous 92.90 7.40 0.05 D8 Zoned – core 76.88 23.49 0.03
RL1 Homogeous 91.66 7.60 0.09 D8 Zoned – core 77.51 23.35 0.04
RL1 Homogeous 92.07 7.61 0.08 D9 Zoned – core 75.52 23.58 0.01
RL2 Homogeous 91.76 8.20 0.03 D9 Zoned – core 78.63 21.66 0.05
RL2 Homogeous 91.72 8.21 0.05 D9 Zoned – core 79.59 21.04 0.03
RL3 Homogeous 90.36 9.74 0.07 D9 Zoned – core 77.03 23.15 0.06
RL3 Homogeous 90.07 9.56 0.13 D9 Zoned – core 77.56 22.56 0.04
RL3 Homogeous 90.17 9.24 0.07 D10 Zoned – core 87.54 12.98 0.05
RL3 Homogeous 91.20 8.78 0.05 D10 Zoned – core 86.96 13.68 0.13
RL3 Homogeous 91.10 8.09 0.06 D10 Zoned – core 88.84 11.60 0.05
RL3 Homogeous 90.96 8.53 0.05 D10 Zoned – core 88.38 12.21 0.07
RL4 Homogeous 89.53 9.77 0.09 D11 Zoned – core 78.77 21.61 0.02
RL4 Homogeous 88.83 10.80 0.09 D11 Zoned – core 78.10 20.74 0.05
RL4 Homogeous 88.03 11.79 0.10 D11 Zoned – core 80.73 20.11 0.05
RL5 Zoned – rim 92.18 7.76 0.05 D12 Zoned – core 76.59 24.08 0.05
RL5 Zoned – rim 98.40 0.32 0.09 D12 Zoned – core 90.46 9.33 0.04
RL5 Zoned – core 75.25 25.10 0.07 D12 Zoned – core 88.95 11.02 0.05
D1 Homogeous 97.80 2.69 0.14 D13 Zoned – core 78.18 21.70 0.01
D1 Homogeous 98.20 2.40 0.12 D13 Zoned – core 78.66 21.89 0.01
D1 Homogeous 97.65 2.43 0.12 D13 Zoned – core 78.72 21.93 0.01
D1 Homogeous 97.50 2.43 0.13 D13 Zoned – core 78.84 21.20 0.02
D2 Homogeous 96.90 3.32 0.12 D13 Zoned – core 70.15 30.33 0.02
D2 Homogeous 97.00 3.26 0.09 D13 Zoned – core 88.92 11.76 0.08
D2 Homogeous 96.48 3.22 0.14 D13 Zoned – core 87.36 12.96 0.09
D2 Homogeous 97.00 3.29 0.12 D14 Zoned – core 89.88 10.19 0.10
D2 Homogeous 96.90 3.26 0.12 D14 Zoned – core 91.91 8.23 0.07
D3 Homogeous 96.69 3.82 0.05 D14 Zoned – core 91.98 7.76 0.05
D3 Homogeous 96.13 2.59 0.20 D14 Zoned – core 90.35 9.71 0.06
D4 Homogeous 95.69 4.63 0.15 TZ1 Zoned – rim 92.01 8.62 0.02
D4 Homogeous 95.95 4.36 0.07 TZ2 Zoned – rim 91.48 9.08 0.05
(continued on next page)
368 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

Table 3 (continued) Table 3 (continued)

Particle Composition Au (%) Ag (%) Cu (%) Particle Composition Au (%) Ag (%) Cu (%)
D14 Zoned – core 87.38 12.61 0.06 G6 Zoned – core 92.54 7.62 0.08
TZ3 Zoned – rim 92.29 8.36 0.10 G6 Zoned – core 92.19 7.34 0.08
TZ4 Zoned – rim 91.05 8.78 0.04 G6 Zoned – core 92.21 7.66 0.09
TZ1 Zoned – core 87.72 13.00 0.01 G6 Zoned – core 92.56 7.27 0.07
TZ2 Zoned – core 78.60 21.23 0.02 G7 Zoned – core 94.76 5.28 0.07
TZ2 Zoned – core 79.08 19.90 0.01 G7 Zoned – core 94.98 4.94 0.04
TZ2 Zoned – core 79.15 20.81 0.04 G7 Zoned – core 94.70 5.34 0.04
TZ3 Zoned – core 80.07 20.63 0.03 G8 Zoned – core 95.68 4.79 0.09
TZ3 Zoned – core 80.62 19.24 0.02 G8 Zoned – core 96.28 4.29 0.07
TZ4 Zoned – core 79.23 20.34 0.01 G8 Zoned – core 96.08 3.84 0.08
TZ4 Zoned – core 79.15 20.32 0.01 G9 Zoned – core 93.68 7.05 0.05
TZ4 Zoned – core 79.30 20.69 0.01 FO1 Homogeous 91.80 7.65 0.05
TZ4 Zoned – core 79.11 20.76 0.03 FO1 Homogeous 91.27 8.16 0.08
TZ4 Zoned – core 79.19 20.46 0.04 FO1 Homogeous 91.44 8.15 0.20
TZ4 Zoned – core 79.02 20.64 0.01 FO2 Homogeous 88.29 10.64 0.17
TZ4 Zoned – core 79.37 20.45 0.03 FO2 Homogeous 89.19 10.53 0.87
TZ4 Zoned – core 79.46 20.20 0.05 FO2 Homogeous 88.88 10.71 0.73
TZ4 Zoned – core 79.30 21.00 0.01 FO3 Homogeous 82.25 17.05 0.02
TZ5 Zoned – core 79.97 20.74 0.01 FO4 Homogeous 85.68 13.63 0.35
TZ5 Zoned – core 79.68 20.29 0.08 FO4 Homogeous 86.00 13.50 0.21
TZ5 Zoned – core 79.70 20.84 0.01 FO5 Zoned 96.30 3.49 0.19
TZ5 Zoned – core 79.67 20.70 0.01 FO5 Zoned 72.85 26.92 0.07
TZ5 Zoned – core 79.02 20.64 0.01 FO6 Zoned 96.69 3.62 0.17
TZ5 Zoned – core 79.37 20.45 0.03 FO6 Zoned 75.96 24.16 0.06
TZ5 Zoned – core 79.46 20.20 0.05
TZ5 Zoned – core 79.30 21.00 0.01
SZ1 Zoned – rim 93.43 5.71 0.29 Table 4
SZ1 Zoned – rim 94.22 5.61 0.29 Summary statistics report for EMPA analyses of gold grains
SZ2 Zoned – rim 93.94 5.55 0.24
SZ2 Zoned – rim 93.71 6.00 0.23 Variable Count Mean StdDev Min Max Range
SZ3 Zoned – rim 93.12 5.70 0.24 Homogeneous
SZ4 Zoned – rim 94.92 5.43 0.23 Au (%) 56 91.63 4.11 82.25 98.20 15.95
SZ5 Zoned – rim 92.40 7.52 0.05 Ag (%) 56 8.28 4.00 2.40 17.34 14.94
SZ1 Zoned – core 74.99 24.12 0.01 Cu (%) 56 0.12 0.15 0.01 0.87 0.86
SZ1 Zoned – core 76.34 23.91 0.02
SZ1 Zoned – core 76.26 23.50 0.04 Zoned – rim
SZ2 Zoned – core 76.07 23.97 0.03 Au (%) 54 96.48 3.63 85.75 100.74 14.99
SZ2 Zoned – core 75.65 23.83 0.03 Ag (%) 54 3.64 3.60 0.02 14.09 14.07
SZ3 Zoned – core 76.78 23.76 0.02 Cu (%) 54 0.08 0.07 0.01 0.29 0.28
SZ3 Zoned – core 77.47 23.16 0.04 Zoned – core
SZ4 Zoned – core 82.95 16.70 0.12 Au (%) 104 82.71 7.00 68.35 96.28 27.93
SZ5 Zoned – core 89.39 11.04 0.02 Ag (%) 104 17.30 6.91 3.84 31.10 27.26
SZ5 Zoned – core 87.98 12.00 0.05 Cu (%) 104 0.05 0.03 0.01 0.15 0.14
G1 Zoned – rim 94.85 5.82 0.04
G3 Zoned – rim 100.26 0.02 0.04
G4 Zoned – rim 100.74 0.02 0.05 regolith material. This difference is caused by differential
G4 Zoned – rim 100.48 0.02 0.07 exposition to weathering solutions. Whereas some grains
G5 Zoned – rim 100.45 0.02 0.08
G5 Zoned – rim 100.01 0.82 0.09 were totally free of surrounding primary minerals, others
G6 Zoned – rim 98.96 1.02 0.15 were partially or totally encapsulated by quartz or other
G6 Zoned – rim 99.53 0.61 0.07 minerals. Therefore, (1) in a single grain, it is possible to
G7 Zoned – rim 99.89 0.16 0.09 observe smooth surfaces and sharp edges in one area and
G7 Zoned – rim 100.02 0.06 0.08
rounded edges and corrosion pits in another; (2) gold
G7 Zoned – rim 100.04 0.02 0.04
G8 Zoned – rim 98.92 1.30 0.04 grains show different degrees of weathering in the same reg-
G8 Zoned – rim 99.04 0.44 0.05 olith horizon; and (3) gold grains in upper horizons show
G9 Zoned – rim 99.85 0.23 0.01 less weathering than some of those found below in the lat-
G1 Zoned – core 68.35 31.10 0.07 eritic profile. It is also possible to observe that, as a general
G2 Zoned – core 75.07 24.70 0.09
rule, rounding and pitting increase upward in the profile,
G3 Zoned – core 86.07 13.03 0.15
G4 Zoned – core 88.03 11.79 0.07 and the mean grain size decreases progressively from duri-
G4 Zoned – core 88.48 10.55 0.09 crust upward as corrosion and dissolution continue.
G4 Zoned – core 89.59 10.86 0.08 In the duricrust, the presence of gold particles with cores
G5 Zoned – core 90.30 9.99 0.08 with different Ag content indicate that some aggregate
G5 Zoned – core 90.15 10.58 0.09 grains comprising several smaller particles formed (note
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 369

alized profile; (3) the presence of relict pisoliths of iron oxi-

hydroxides (fragments of duricrust) in the latosol (Larizz-
atti, 2002); and (4) the recognition, despite weathering pro-
cesses, of gold grain primary habits, similar to those
observed in the saprolite, in the latosols.
Grains of weathered primary gold with Ag-poor rims
and Ag-rich zones on internal grain boundaries have also
been observed at Gentio do Ouro in the semiarid northeast
of Brazil (Grimm and Friedrich, 1990) and have several
features similar to those seen at Fazenda Pison. Silver
enrichment always occurs at the contact between the pri-
mary core of the grain and the Ag-depleted rim or zone,
and there is an abrupt contact between silver-rich zones
and silver-poor rims. However, at Fazenda Pison, the con-
tacts between the Ag-rich zones and the core of primary
gold grains tend to be gradational, as shown in Fig. 14A
and B. Grimm and Friedrich (1990) suggest Ag-rich zones
‘‘result from mobilization, separation, and rapid precipita-
tion of electrum by thiosulphate complex formation’’ dur-
ing weathering.
Webster and Mann (1984) stress the influence of climate,
rock composition, and geomorphological factors in gold
mobility. The supergene conditions prevailing in the Ama-
zon (one bar atmospheric pressure; temperatures of 20–
30 C; excess free water), the presence of pyrite in the pri-
mary ore, the abundance of organic matter in the surface
Fig. 9. Backscattered electron image (SEM) of polished sections of gold soil, and the mushroom-shaped surface dispersion pattern
particles. (A) Homogeneous gold grain; (B) zoned gold grain. of gold at Fazenda Pison (Larizzatti, 2002) suggest the
metal could be complexed by a wide range of ligands and
different tones of grey in the particle shown in Fig. 14). The is relatively soluble (Goleva, 1970; Roslyakov et al., 1971;
cores of these individual small particles have different Au/ Baker, 1978; Plyusnin et al., 1981; Stoffregen, 1986; Web-
Ag compositions and were joined by zones of Ag-poor gold ster, 1986; Vlassopoulos and Wood, 1990; Bowell, 1993).
between them. The largest gold grains (mean 124 lm) were Evidence of gold dissolution includes the progressive
found in the duricrust; only smaller grains were observed in decrease in grain size upward in the lateritic profile (except
the materials present below the duricrust, which suggests in the duricrust, where grains are aggregates of separate
that a process of gold grain development was active during individual small grains) and the presence of corrosion fea-
iron duricrust formation. Gold nugget formation in later- tures on grain surfaces, as observed by other authors (Colin
itic profiles was reported previously, for example, by et al., 1989a,b; Colin and Vieillard, 1991; Porto and Hale,
Machairas (1963), Mann (1984), and Wilson (1984). 1996). Rounding and pitting increase upward in the profile,
These aggregates were not found in the overlying red indicating dissolution of gold due to exposure to weather-
and yellow latosols, which suggests that the process of duri- ing solutions. There is, however, no unequivocal evidence
crust transformation into latosol was accompanied by gold of reprecipitation of secondary gold. In most weathering
dissolution and particle division. Silver-poor rims are thin- environments, secondary gold is of very high fineness, con-
ner in the latosols than in the duricrust, suggesting a pro- taining less than 1.0% Ag (e.g., Mann, 1984; Butt, 1989;
gressive upward weathering in the profile. Gold grades Freyssinet et al., 1987; Santosh and Omana, 1991; Gray
are higher in the latosol than in the duricrust (Larizzatti, et al., 1992). All gold recovered at Fazenda Pison contain
2002), despite the smaller size and number of gold particles silver, either throughout in homogeneous grains or in the
recovered from latosol, which also suggests gold was dis- core in zoned grains; even the rims of most zoned grains
solved during the transformation of duricrust into latosol. contain >1.0% Ag. No grains of pure gold were recovered,
All gold grains observed in the regolith are interpreted though this may be a function of sample preparation, in
as residual relicts that survived weathering of the primary that few grains <10 lm were recovered. The secondary
mineralization. To support this hypothesis, we note (1) gold noted by Freyssinet et al. (1987) in Mali, for example,
the mushroom-type gold anomaly observed just above is very fine grained, mostly <1 lm, and fragile dendrites,
mineralized quartz veins (Larizzatti, 2002); (2) that an readily destroyed by panning. Freyssinet et al. (1987) and
exploration survey performed in the area (geophysics, geo- Santosh and Omana (1991) also observe fine secondary
chemistry, drilling; RTDM, 1995) was unable to find other grains in cavities in gold grains, but these were not found
known mineralization in the vicinities of the studied miner- at Fazenda Pison.
370 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

Fig. 10. Frequency diagrams for ‘‘N’’ analyses of Au, Ag, and Cu contents in 17 homogeneous gold particles and 37 zoned gold particles. All data shown
in outline, with results for each phase highlighted in black. (EPMA analyses.)

Silver (0.1–25%) and trace copper (0.02–0.9%) are the work saprolite differs from that in the granophyre sapro-
only elements other than gold present in significant abun- lite, in that both cores and rims in grains from the
dance. Compositions generally vary from grain to grain stockwork appear to be more silver rich, and rims are rel-
or even within the same grain, as noted by other authors atively rich in copper.
(e.g., Boyle, 1979; Desborough et al., 1971; Antweiler and Silver-poor rims have been observed on the surfaces of
Campbell, 1977; Morrison et al., 1991; Knight et al., primary gold grains in lateritic profiles and placer deposits
1999). The grains are either homogeneous (with minor vari- (Desborough, 1970; Mann, 1984; Giusti, 1986; Michel,
ations in Au–Ag contents) or zoned, with silver-rich cores 1987; Freyssinet et al., 1987; Freyssinet and Butt, 1988).
and thin, silver-poor rims. Some of these silver-poor rims The presence of zoned grains in the regolith at Fazenda
are composed of essentially pure gold. Copper contents Pison, silver-poor zones adjacent to internal fractures and
show much less variation, with similar abundances in as rims to the grains, sharp internal contact between the
homogeneous grains, cores, and rims. Gold from the stock- core and these zones, and high fineness of the rims are
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 371

Fig. 11. Frequency diagrams for ‘‘N’’ analyses of Au, Ag, and Cu contents in 17 homogeneous gold particles. All data shown in outline, with results for
each phase highlighted in black. (EPMA analyses.)

consistent with the rims being secondary features. The quence or drainage, and precipitates on the surface of the
results from Fazenda Pison provide more quantitative data grain. Precipitation can be caused by changes in solution
for silver and copper, which indicate a general relationship chemistry, reaction with metal ions such as Fe and Mn that
between the silver contents of core and rim and that copper break down complexing ligands (Conklin and Hoffmann,
contents are seemingly unchanged. These latter observa- 1988), or catalysis by solids (Boyle, 1979), of which free
tions differ from other reports, which do not note any com- gold is one of the most effective. Plating by reprecipitated
positional relationships between core and rim, rather the gold is considered to have occurred in some placer deposits,
absence of both silver and copper. The presence of zones which may contain pristine, commonly fragile, discrete
of silver enrichment between the rim and core confirms grains of secondary gold, and zoned grains in which relict
similar findings by Grimm and Friedrich (1990). Three primary gold has silver-poor rims exhibiting delicate struc-
principal theories can be advanced regarding the formation tures and clear crystalline habits (e.g., Webster and Mann,
of silver-poor rims: 1984; Giusti, 1986; McCready et al., 2003). Such structures
1. Precipitation (plating) of gold. The Ag-poor gold is have not been reported from weathering profiles; crystal-
derived from higher levels in the regolith profile, topose- line and other forms of secondary gold are described as dis-
372 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

Fig. 12. Frequency diagrams for ‘‘N’’ analyses of Au, Ag, and Cu contents in (A) rims and (B) cores of 37 zoned particles. All data shown in outline, with
results for each phase highlighted in black. (EPMA.)

crete grains, not continuous overgrowths on relict primary 2. Self-electrorefining. In this process, an Au–Ag alloy
grains. Furthermore, precipitation along internal disconti- dissolves electrochemically at the interface with a weather-
nuities is difficult to explain by this mechanism. ing solution, and gold, the more inert metal, reprecipitates
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 373

Fig 12. (continued )

onto the surface of the grain (Groen et al., 1990). The pro- mary gold–silver alloy. However, overgrowths that com-
cess results in the formation of lobate overgrowths of rep- pletely cover the primary grain core would restrict access
recipitated Au, and silver is leached by the weathering to the primary gold–silver alloy and interrupt the process.
solution. The reaction is maintained by the difference in 3. Selective depletion of silver. Preferential silver deple-
electrical potential between the gold-rich areas and the pri- tion from the primary grain by interaction with soil,
374 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

Gold would reprecipitate immediately, whereas Ag would

be removed by the leaching solution. Retention of Cu in
the rims and zones of silver enrichment, noted in this study,
add weight to this conclusion.
The variations in morphology and composition of gold
grains in the regolith at the Fazenda Pison ‘‘Garimpo’’
are products of processes that have caused the weathering
of the mineralized rocks to form saprolite, the formation of
the lateritic duricrust, the destruction of the duricrust, and
its transformation to latosol. Costa (1997) describes mature
and immature laterite profiles in the Amazon region as a
result of two distinct periods of laterite formation in differ-
ent climatic conditions. The observation that large, aggre-
gate grains occur only in the duricrust may be significant,
assuming it is not due to poor sampling statistics or a var-
iation in the primary mineralization. It is possible these are
remnants of primary polycrystalline grains, preserved by
the cementation of the duricrust. Similar grains would have
been present in the saprolite, but ongoing exposure to
weathering may have caused them to disaggregate or be
sufficiently weakened not to survive panning. Similarly,
aggregates released from duricrust during latosol forma-
tion may have been further weathered, so that none sur-
vived. The corrosion of silver-poor rims on grains in the
latosol is evidence for this latter weathering.

Fig. 13. SEM images (backscattered electrons) of gold grains recovered Acknowledgments
from saprolite (granophyre) showing silver-enriched zones (dark grey),
following discontinuities in gold grains filled by silver-poor gold (light This research was undertaken as part of a Ph.D.
grey). research project by JHL at the University of Sao Paulo
and during a study visit to the CRC for Landscape Evolu-
tion and Mineral Exploration, CSIRO Division of Explo-
ration and Mining, and the University of Western
Australia in Perth, Western Australia. FAPESP (State of
São Paulo Research Foundation) is acknowledged for re-
search Grant 1999/1537-9. CAPES Foundation (Coordina-
tion of Training of Higher Education Graduate) is
acknowledged for scholarship BEX0575/00-3. Prof.
Dr. Rômulo Angélica is thanked for discussions on gold
grain geochemistry and morphology and for his suggestions
and comments as a reviewer. Dr. Erik Van Noort is thanked
for discussions during EMPA analysis. Prof. Dr. Brendan J.
Griffin and Ms. Sharon Platten are thanked for SEM assis-
tance. Dr. Robert Hough is thanked for comments on a pre-
Fig. 14. Backscattered electron image (SEM) of a zoned aggregate grain vious version of the manuscript. Constructive reviews by
from duricrust, with silver-poor zones at the contact between individual
Dr. J.C. Parisot helped improve the manuscript.
particles. Note different tones of grey in core zones.

ground, or drainage water is the most probable process in References

the regolith, with the presence of silver-poor zones along
de Almeida, F.F.M., Hasui, Y., de Brito Neves, B.B., Fuck (sic), R.A.A.,
internal fractures as the most compelling evidence (Colin 1981. Brazilian structural provinces: an introduction. Earth Science
and Vieillard, 1991; Gray et al., 1992). Silver depletion Review 17, 1–29.
develops from the outer rims and discontinuities (probably Andrade, A.F., Santiago, A.F., Melo, C.F., Bizinella, G.A., Moreira,
crystal boundaries) within the grains, allowing weathering H.L., Santos, J.O.S., Oliveira, J.R., Moura, P.A., Lopes, R.C., Rosa
Filho, S.F., Neves, S.A.V., 1978. Projeto Tapajós – Sucunduri:
solutions to interact from the surface to the centre of the
relatório de integração geolóaus, CPRM/DNPM. 3v. (internal report).
grains. According to Gray et al. (1992), the rims could be Andrade, W.O., Mackesky, M.L., Rose, A.W., 1991. Gold distribution
a consequence of galvanic reactions on the grain surface and mobility in the surficial environment, Carajás region, Brazil.
involving simultaneous solution of Ag+ and Au+ ions. Journal of Geochemical Exploration 40, 95–114.
 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376 375

Angélica, R.S., Costa, M.L., Pöllmann, H., 1995. Mineralogia e geoquı́- Freyssinet, P., 1993. Gold dispersion related to ferricrete pedogenesis in
mica das partı́culas de ouro nos perfis laterito-gossânicos das áreas South Mali: application to geochemical exploration. Chronique de.
Igarapé Bahia e Águas Claras, Carajás, Pará. In: V Congr. Bras. Recherche des Mines 510, 25–40.
Geoq., Anais, Niterói, RJ. Giusti, L., 1986. The morphology, mineralogy, and behaviour of ‘‘fine
Antweiler, J.C., Campbell, W.L., 1977. Application of gold compositional grained’’ gold from placer deposits of Alberta: sampling and implica-
analyses to mineral exploration in the United States. Journal of tions for mineral exploration. Canadian Journal of Earth Sciences 23,
Geochemical Exploration 8, 17–29. 1662–1672.
Baker, W.E., 1978. The role of humic acid in the transport of gold. Goleva, G.A., 1970. Geochemical trends in the occurrence and migration
Geochimica et Cosmochimica Acta 42, 645–649. forms of gold in natural waters. Geochemistry International 7, 518–
Bowell, R.J., 1992. Supergene gold mineralogy at Ashanti, Ghana: 529.
implications for the supergene behavior of gold. Mineralogical Gray, D.J., Butt, C.R.M., Lawrance, L.M., 1992. The geochemistry of
Magazine 56, 545–560. gold in lateritic terrains. In: Butt, C.R.M., Zeegers, H. (Eds.), Regolith
Bowell, R.J., 1993. The role of fulvic acid in the supergene migration of Exploration Geochemistry in Tropical and Subtropical Terrains.
gold in tropical rain forest soils. Geochimica et Cosmochimica Acta Handbook of Exploration Geochemistry, 4. Elsevier, Amsterdam,
57, 4179–4190. pp. 461–482.
Boyle, R.W., 1979. The geochemistry of gold and its deposits. Geological Grimm, B., Friedrich, G., 1990. Weathering effects on supergene gold in
Survey of Canada Bulletin 280, 564p. soils of a semiarid environment, Gentio do Ouro, Brazil. In: Noack,
Benedetti, M., Boulège, J., 1991. Mechanism of gold transfer and Y., Nahon, D. (Eds.), Geochemistry of the Earth’s Surface and of
deposition in a supergene environment. Geochimica et Cosmochimica Mineral Formation. Chemical Geology, 84, 70–73.
Acta 55, 1539–1547. Groen, J.C., Craig, J.R., Rimstidt, J.D., 1990. Gold-rich rim formation on
Butt, C.R.M., 1989. Genesis of supergene gold deposits in the lateritic electrum grains in placers. Canadian Mineralogist 28, 207–228.
regolith of the yilgarn block, Western Australia. In: Keays, R.R., Knight, J.B., Mortensen, J.K., Morison, S.R., 1999. Lode and placer gold
Ramsay W.R.H., Groves, D.I. (Eds.), The Geology of Gold Deposits: composition in the Klondike district, Yukon Territory Canada:
The Perspective in 1988. Economic Geology Monograph 8. Proceed- implications for the nature and genesis of Klondike placer and lode
ings of Bicentennial Gold 88, 460–470. gold deposits. Economic Geology 94, 649–664.
Colin, F., Edou-Minko, A., Nahon, D., 1989a. Lór particulaire dans le Lawrance, L.M., Griffin, B.J., 1994. Crystal features of supergene gold at
profils latéritiques: altérations géochimiques et disprsion superficielles Hannan South, Western Australia. Mineralium Deposita 29, 391–
em cditions équatorialles. Comptes Rendues de ĺAcadémie des 398.
Sciences, Paris 309 (serie II), 553–560. Larizzatti, J.H., 2002. Gold and pathfinder elements in the regolith of
Colin, F., Lecomte, P., Boulangé, B., 1989b. Dissolution features of gold Fazenda Pison garimpo – dispersion processes and implications for
grains in a lateritic profile at Dondo Mobi, Gabon. Geoderma 45, 241– exploration. Unpublished PhD. thesis. University of São Paulo, 204p.
250. (in Portuguese).
Colin, F., Vieillard, P., 1991. Behavior of gold in the lateritic equatorial Machairas, G., 1963. Étude des phénomènes de migration chimique de
environment: weathering and surface dispersion of residual gold l’or. Cas de la Guyane Française et d’Ity en Côte-d’Ivoire. Bull. Soc.
grains, at Dondo Mobi, Gabon. Applied Geochemistry 6, 279–290. Franç. Minér. Crist. 86, 78–80.
Conklin, M.H., Hoffmann, M.R., 1988. Metal ion–sulfur (IV) chemistry. Mann, A.W., 1984. Mobility of gold and silver in lateritic weathering
3. Thermodynamics and kinetics of transient iron (III)–sulfur (IV) profiles, some observations from Western Australia. Economic Geol-
complexes. Environmental Science & Technology 22, 899–907. ogy 79, 38–49.
Costa, M.L., 1997. Lateritization as a major process of ore deposit Michel, D., 1987. Concentration of gold in situ laterite from Mato Grosso.
formation in the Amazon region. Exploration and Mining Geology 6, Mineralium Deposita 22, 185–189.
79–104. McCready, A.J., Parnell, J., Castro, L., 2003. Crystalline placer gold from
CPRM, 1999. Geology and Mineral Resources of Vila Mamãe Anã Sheet the Rio Neuquén, Argentina: implications for the gold budget in placer
(SB.21-V-D). 1:250.000 scale. CPRM, Manaus, 60p, unpublished (in gold formation. Economic Geology 98, 623–633.
Portuguese). Morrison, G.W., Rose, W.J., Jaireth, S., 1991. Geological and geochem-
Craw, D., 1992. Growth of alluvial gold grains by chemical accretion and ical controls on the silver content (fineness) of gold in gold–silver
reprecipitation, Waimumu, New Zealand. New Zealand Journal of deposits. Ore Geology Reviews 6, 333–364.
Geology and Geophysics 35, 157–164. de Oliveira, S.M.B., Campos, E.G., 1991. Gold-bearing iron duricrust in
Davy, R., El-Ansary, M., 1986. Geochemical patterns in the laterite profile Central Brazil. Journal of Geochemical Exploration 41, 233–244.
at Boddington gold deposit, Western Australia. Journal of Geochem- de Oliveira, S.M.B., de Oliveira, N.M., 2000. The morphology of gold
ical Exploration 26, 119–144. grains associated with oxidation of sulphide-bearing quartz veins at
Desborough, G.A., 1970. Silver depletion indicated by microanalysis of São Bartolomeu, central Brazil. Journal of South American Earth
gold from placer occurrence, Western United States. Economic Sciences 13, 217–224.
Geology 65, 304–311. Plyusnin, A.M., Pogrelniak, Y.F., Mironov, A.G., Zhmodik, S.M., 1981.
Desborough, G.A., Heidel, R.H., Raymond, W.H., Tripp, J., 1971. Primary The behavior of gold in the oxidation of gold bearing sulphides.
distribution of silver and copper in native gold from six deposits in the Geochemistry International 18, 116–123.
Western United States. Mineralium Deposita 6, 321–334. Porto, C.G., Hale, M., 1996. Mineralogy, morphology and chemistry of
Edou Minko, A., Colin, F., Trescases, J.J., Lecomte, P., 1992. Altération gold in stone line lateritic profile of the Posse deposit, Central Brazil.
latéritique du gı̂te aurifère d’Ovala (Gabon), et formation d’une Journal of Geochemical Exploration 57, 115–125.
anomalie superficielle de dispersion. Mineralium Deposita 27, 90–100. RTDM, 1995. Relatório de avaliação do potencial aurı́fero da área
Eyles, N., 1990. Post-depositional gold nugget accretion in Late Cenozoic Fazenda Pison (AM). Rio Tinto Denvolvimentos Minerais. Confiden-
glacial placer deposits of Western Canada. Sediments 1990 – 13th tial report, 29pp. (in Portuguese).
International. Sedimentological Congress, pp. 74–75. Roslyakov, N.A., Peshevitsky, B.I., Nepeyna, L.A., Timbalist, V.G., 1971.
Freyssinet, P., Butt, C.R.M., 1988. Morphology and geochemistry of gold Geochemistry of gold in processes of weathering crust formation.
in a lateritic profile, Reedy mine, Western Australia. CSIRO Australia International Geological Congress, Moscow, Abstracts of reports. II,
Division of Exploration Geoscience Restricted Report MG 58R. 15pp. 179–180.
Freyssinet, P., Zeegers, H., Tardy, Y., 1987. Néoformation d’or dans les Santos, J.O.S., Hartman, L.A., Gaudetti, H.E., 1997. Reconnaissance U–
cuirrasses latéritiques, dissolution, migration, précipitation. Comptes Pb in zircon, Pb–Pb in sulphides and review of Rb–Sr geochronology
Rendues de l’ Académie des Sciences, Paris 305 (série II), 867–874. in the Tapajós Gold Province, Pará/Amazonas States, Brazil. South
376 J.H. Larizzatti et al. / Journal of South American Earth Sciences 25 (2008) 359–376

American Symposium on Isotope Geology – Brazil (SSAGI), Campos Vlassopoulos, D., Wood, S.A., 1990. Gold speciation in natural waters. I.
do jordão, SP. Proceedings. 280–282. Solubility and hydrolisis reactions of gold in aqueous solution.
Santos, J.O.S., Groves, D.I., Hartmann, L.A., Moura, M.A., McNaugh- Geochimica et Cosmochimica Acta 54, 3–12.
ton, N.J., 2001. Alta Floresta Domains, Tapajós-Parima orogenic belt, Webster, J.G., Mann, A.W., 1984. The influence of climate, geomorphol-
Amazon Craton, Brazil. Mineralium Deposita 36, 278–299. ogy and primary geology on the supergene migration of gold and
Santosh, M., Omana, P.K., 1991. Very high purity gold from lateritic silver. Journal of Geochemical Exploration 22, 21–42.
weathering profiles of Nilambur, southern India. Geology 19, 746–749. Webster, J.G., 1986. The solubility of gold and silver in the system Au–
Sergeev, N.B., Zaikov, V.V., Laputina, L.P., Trofimov, O.V., 1994. Gold Ag–S–O2–H2 at 25 C and 1 atm. Geochimica et Cosmochimica Acta
and silver in the supergene zone of the pyritic lode off the Gai deposit, 50, 1837–1846.
the southern Urals. Geology of Ore Deposits 36, 152–164. Wilson, A.F., 1984. Origin of quartz-free gold nuggets and supergene gold
Stoffregen, R., 1986. Observations on the behavior of gold during found in laterites and soils – a review and some new observations.
supergene oxidation at Summitville, Colorado, U.S.A., and implica- Australian Journal of Earth Sciences 31, 303–316.
tions for electrum stability in the weathering environment. Applied Youngson, J.H., Craw, D., 1995. Evolution of placer gold deposits during
Geochemistry 1, 549–558. regional uplift, Central Ontago, New Zealand. Economic Geology 90,
Vasconcelos, P.M., Kyle, J.R., 1991. Supergene geochemistry and 731–745.
crystal morphology of gold in a semiarid weathering environment, Zang, W., Fyfe, W.S., 1993. A three-stage genetic model for the Igarapé
application to gold exploration. Journal of Geochemical Exploration Bahia lateritic gold deposit, Carajás, Brazil. Economic Geology 88,
40, 115–132. 1768–1779.