OCTOBER 1999

SEG
NEWSLETTER
HSOCIETY OF ECONOMIC GEOLOGISTSH NUMBER 39

Alteration Mapping in Exploration:

Application of Short-Wave Infrared
(SWIR) Spectroscopy
ANNE J. B. THOMPSON (SEG 1990) PHOEBE L. HAUFF AUDREY J. ROBITAILLE
PETRASCIENCE CONSULTANTS INC. SPECTRAL INTERNATIONAL INC. PETRASCIENCE CONSULTANTS INC.
3995 W. 24TH AVENUE P.O. BOX 1027 3995 W. 24TH AVENUE
VANCOUVER, B.C. • CANADA V6S 1M1 ARVADA, COLORADO 80001 VANCOUVER, B.C. • CANADA V6S 1M1
EMAIL, [email protected]

ABSTRACT ore. In addition to its early use in remote sensing, development of

Alteration mineral assemblages are important to the the PIMA-II in 1991 allowed direct use of SWIR on rocks, greatly
understanding of and exploration for hydrothermal ore deposits. enhancing its practical application to exploration. SWIR
Conventional mapping tools may not identify fine-grained minerals spectroscopy detects minerals such as phyllosilicates, clays,
or define important compositional variations. Field portable short- carbonates, and selected sulfates and is also sensitive to variations in
wave infrared (SWIR) spectrometers solve some of these problems individual mineral species.
and provide a valuable tool for evaluating the distribution of SWIR field spectrometers are used in numerous deposit
alteration assemblages. Spectrometers such as the PIMA-II allow environments, including high- and low-sulfidation epithermal,
rapid identification of minerals and mineral-specific variations at a porphyry, mesothermal, sediment-hosted gold and copper, uranium,
field base. Mineral assemblages, integrated with other exploration VMS, and kimberlite deposits (Table 1, page 16). In addition,
data, are then used to target drill holes and guide regional spectrometers aid regolith mapping, both for determination of
exploration programs. Data collection must be systematically bedrock composition and for differentiation of residual and
organized and carried out by a trained operator. Analysis of data transported regolith. Publications on the results of SWIR
sets requires the use of spectral reference libraries from different spectroscopy are sparse, reflecting the confidential nature of most
geological environments and may be aided in some cases by companies’ programs and the lack of academic work applied to field
computer data processing packages. Integration of results with field mapping. A selection of recent papers and abstracts, however,
observations, petrography, and X-ray diffraction analysis is necessary highlight the work that is currently underway: Stewart and
for complete evaluation. The PIMA (portable infrared mineral Kamprad, 1997, and Shen et al., 1999 (regolith mapping); Zhang et
analyzer) has been used successfully in the high-sulfidation al., 1998 (uranium); Passos and de Souza Filho, 1999 (Archean
epithermal, low-sulfidation epithermal, volcanogenic massive sulfide greenstone); Denniss et al., 1999, and Huston et al., 1999 (VMS);
(VMS) and intrusion-related environments. Case studies from these Martinez-Alonso et al., 1999, and Kruse and Hauff, 1991 (epithermal
systems demonstrate the ability to rapidly acquire and process SWIR clays); and Crowley, 1996 and 1999 (evaporites).
data and produce drill logs and maps. The resulting information is SWIR field spectrometers fill an important gap in exploration data
critical for targeting. by helping to map alteration consistently throughout a mineralized
system. Determining alteration mineralogy routinely during an
INTRODUCTION exploration program aids rapid evaluation and therefore increases
efficiency.
Field portable SWIR spectrometers are becoming increasingly
important to exploration. Spectrometers typically are employed to Alteration Mapping
determine the mineralogy of altered rocks and hence assist in Determination of the type and distribution of
classifying ore systems, identifying alteration patterns, and locating alteration minerals is a routine part of exploration to page 16 . . .
16 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 1 S W I R S P E C T R O S C O P Y , C O N T .
4. Use the above data to establish the relationships among the
main minerals;
5. Outline the distribution at the map scale;
for hydrothermal mineral deposits and is useful in the assessment of
6. Use petrography of selected samples to further define
exploration properties and the construction of deposit models.
relationships of minerals;
Typically, alteration maps are based on macroscopic field
7. Augment with X-ray diffraction (XRD) analysis if necessary;
observations, supported by limited petrographic or X-ray diffraction
8. Use scanning electron microscope (SEM) with energy dispersive
studies. Alteration studies at the deposit scale are limited, or reliant system (EDS) to determine variations in individual minerals and
on detailed but necessarily restricted sample suites. assist with interpretation in fine-grained material;
Lithogeochemistry is used in some environments to evaluate 9. Refine and reevaluate, continually, the interpretation and
alteration but only works well where lithologies and their petrology integrate the results with other geologic, geochemical, and
are well understood. Lithogeochemistry is difficult to apply to geophysical data sets.
extensive areas of clay alteration, where protoliths are hard to
identify during the exploration program. SWIR analysis aids exploration from regional to property scales.
Fine-grained alteration minerals commonly are grouped as For example, in complex zoned intrusive systems, alteration
“argillic” or “phyllic” (Thompson and Thompson, 1996). Such mineralogy determined routinely during mapping helps to define
descriptions ignore the mineralogy and lose valuable information vertical and horizontal zoning and related ore environments. Within
regarding the nature of the alteration. The importance of using each environment, alteration mineralogy can define local zoning,
minerals and mineral assemblages was noted by Rose and Burt providing vectors to mineralization. SWIR spectroscopy is most
(1979) and subsequent authors, but this approach is not always helpful where alteration mineralogy is not easily identified in hand
applied during exploration. Classification of alteration by mineralogy specimen because of grain size or weathering. Even where field
involves field observations that may be aided by SWIR spectroscopy mapping of alteration minerals is effective, SWIR spectroscopy
(Table 2). The use of SWIR spectrometers at a field base allows allows recognition of subtle mineralogic and compositional
mineralogy to be mapped or placed on cross sections. The resultant variations; these can be important in locating ore.
interpretation can be applied in real time to guide drilling, and
ultimately can be integrated with other data to develop targets, REFLECTANCE SPECTROSCOPY
models, and regional guides. Reflectance spectroscopy is an analytical technique used by
Field observations must be made in a careful and systematic chemists and mineralogists since the early 1900s, with infrared data
manner. Care is needed in determining the relationship among on minerals published between 1905 and 1910 by W.W. Coblentz of
minerals prior to assigning them to a single assemblage or the U.S. Bureau of Standards. Commercially available infrared
interpreting their relationship to other types of alteration. A series of spectrophotometers in the mid-1940s led to increased use of the
steps should be followed in order to make realistic interpretations of technique for mineralogy. Early reviews of mineral spectra were
the observed hydrothermal alteration. These steps are: published by Lyon (1962) and Moenke (1962). Farmer (1974)
1. Determine the minerals present, based on field observations; published a comprehensive book on theoretical and practical
2. Determine their distribution at the outcrop and hand specimen aspects and Marel and Beutelspacher (1976) compiled clay minerals.
scales; Kodama (1985) published spectra of minerals typically found in
3. Employ SWIR analysis carefully, analyzing a variety of locations soils, including numerous hydroxides, oxides, phyllosilicates,
on each sample and using systematic sampling techniques; carbonates, and sulfates.

Table 1. Examples of the Use of SWIR in Exploration

Mineral Identification Alteration Interpretation Exploration Application

Alunite Advanced argillic — High-sulfidation environment, and zoning around high sulfidation
— Steam-heated zones in low sulfidation
Dickite Advanced argillic — Zoning around high sulfidation
— Sediment-hosted Au, with mineralization
Kaolinite Advanced argillic and — High-sulfidation
weathered rock — Sediment-hosted Au, zoning
Dickite, pyrophyllite, diaspore Advanced argillic — Depth estimation
Chlorite Propylitic, chloritic — VMS zoning
— Uranium zoning
Illite/smectite Argillic — High and low sulfidation, zoning
— Uranium, zoning
Carbonate Carbonate — Mesothermal, zoning
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 17

Mineral spectra in the SWIR were first compiled by Hunt and the ability to field check Landsat Thematic Mapper imagery. Several
Salisbury (1970, 1971), and Hunt et al. (1971a, b, c). Hunt’s database field spectrometers are available. These include the Geophysical
provided a basic reference for infrared active minerals in the SWIR Environmental Research, Inc. (GER-IRIS), Analytical Spectral Devices
region that is still in active use. The work was expanded on by (ASD-FieldSpec) and Integrated Spectronics Pty. Ltd. (PIMA). The
Clark et al. (1990). Hauff (1993) published a commercial reference GER and ASD instruments provide data in the visible, near, and
library. Workers at the Jet Propulsion Laboratory also added to the shortwave infrared wavelengths. The instruments are field portable,
references available (Grove et al., 1992). but require use of solar illumination. Early published papers include
Field Portable Spectrometers documentation of the GER instrument (Marsh and McKeon, 1983).
Geologists in the remote-sensing community drove the initial The instrument was used in field checking of airborne spectroradio-
development of field portable SWIR spectrometers that were meter data in the Oatman district (epithermal veins), Arizona. Hunt
particularly useful to the mineral exploration industry. The ability to and Ashley (1979) and Crowley (1984) linked
to page 18 . . .
easily produce laboratory-quality data in field situations enhanced SWIR spectroscopy to alteration mapping.

Table 2. Summary of Infrared-Active Minerals, with Distinctive Spectra in the SWIR

Environment of formation Standard terminology SWIR active mineral assemblage (key minerals are in bold)
Intrusion-related Potassic (biotite-rich), K silicate, Biotite (phlogopite), actinolite, sericite, chlorite, epidote,
biotitic muscovite, anhydrite
Sodic, sodic-calcic Actinolite, clinopyroxene (diopside), chlorite, epidote, scapolite
Phyllic, sericitic Sericite (muscovite-illite), chlorite, anhydrite
Intermediate argillic, sericite- Sericite (illite-smectite), chlorite, kaolinite (dickite),
chlorite-clay (SCC), argillic montmorillonite, calcite, epidote
Advanced argillic Pyrophyllite, sericite, diaspore, alunite, topaz, tourmaline,
dumortierite, zunyite
Greisen Topaz, muscovite, tourmaline
Skarn Clinopyroxene, wollastonite, actinolite-tremolite, vesuvianite,
epidote, serpentinite-talc, calcite, chlorite, illite-smectite, nontronite
Propylitic Chlorite, epidote, calcite, actinolite, sericite, clay
High-sulfidation epithermal Advanced argillic—acid sulphate Kaolinite, dickite, alunite, diaspore, pyrophyllite, zunyite
Argillic, intermediate argillic Kaolinite, dickite, montmorillonite, illite-smectite
Propylitic Calcite, chlorite, epidote, sericite, clay
Low-sulfidation epithermal “Adularia” — sericite, sericitic, Sericite, illite-smectite, kaolinite, chalcedony, opal,
argillic montmorillonite, calcite, dolomite
Advanced argillic— Kaolinite, alunite, cristobalite
acid-sulphate (steam-heated) (opal, chalcedony), jarosite
Propylitic, zeolitic Calcite, epidote, wairakite, chlorite,
illite-smectite, montmorillonite
Mesothermal Carbonate Calcite, ankerite, dolomite, muscovite (Cr-/V-rich), chlorite
Chloritic Chlorite, muscovite, actinolite
Biotitic Biotite, chlorite
Sediment-hosted gold Argillic Kaolinite, dickite, illite
Volcanogenic massive sulfide Sericitic Sericite, chlorite, chloritoid
Chloritic Chlorite, sericite, biotite
Carbonate Dolomite, siderite, ankerite, calcite, sericite, chlorite
Sediment-hosted massive sulfide Tourmalinite Tourmaline, muscovite
Carbonate Ankerite, siderite, calcite, muscovite
Sericitic Sericite, chlorite
Albitic Chlorite, muscovite, biotite
Minerals are grouped by assemblages of alteration minerals, and keyed to commonly used terminology; Complete assemblages are in
Thompson and Thompson (1996)
18 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 17 S W I R S P E C T R O S C O P Y , C O N T . Short-wave infrared spectroscopy detects the energy generated

by vibrations within molecular bonds. These bonds have bending
and stretching modes within the 1,300- to 2,500-nm region of the
The most widely used portable instrument in exploration is the
electromagnetic spectrum. The observed absorption features are
PIMA, which collects data only in the SWIR region. The PIMA-II is a
manifestations of first and second overtones and combination tones
commercial field instrument manufactured by Integrated Spectronics
of fundamental modes that occur in the mid-infrared region. SWIR is
Pty. Ltd. in Australia. The instrument has an internal light source, particularly sensitive to certain molecules and radicals, including
allowing collection of laboratory-quality data in the field by direct OH, H2O, NH4, CO3, and cation-OH bonds such as Al-OH, Mg-OH,
illumination of the rock sample. In addition, internal calibration and Fe-OH. The positions of the features in the spectrum and their
results in reliable spectra not subject to variability due to the characteristic shapes are a function of the molecular bonds present
conditions under which they were measured. The instrument is in the mineral. Variations in chemical composition may be detected
capable of measuring a variety of sample types, including rocks, as the wavelength positions of features shift consistently with
chips, core, powders, and liquids. An analysis typically takes less elemental substitution. SWIR spectroscopy is partly sensitive to
than 30 seconds. PIMA dominates current usage in the industry for crystallinity variations, but may not detect primary changes in the
the purpose of alteration mapping. Several PIMA-II instruments lattice structure. A typical spectrum consists of several absorption
were used in the collection of the data discussed in this paper. features. Figure 1 illustrates the various aspects of an absorption
All instruments require training for effective use, both in the feature, including wavelength position, depth and width (full-height,
interpretation of results and in instrument operation. Lack of training half-width maximum). The outline of the hull or continuum is also
can result in broken and malfunctioning instruments or worse, shown.
misinterpretation of data. The limitations of the technique must be
understood in order to utilize fully a powerful tool. Integration of
spectral data with geologic, geochemical, and geophysical
information is also critical.
Field use of short-wave infrared spectrometers has increased
dramatically in the last five years. The increased application of the
tool is the result of several developments during the twentieth
century. The milestones in the development of infrared
spectroscopy for minerals include:
• Early documentation (1905–1910);
• Laboratory use, expansion of mineral reference databases
(1940–1985);
• Development of field portable instruments (1978–1991);
• Real time processing of data (late 1980s);
• Commercial availability of portable instrument with internal light
source, PIMA (1991);
• Continued expansion of mineral reference data sets (1990s);
• Dissemination of case histories and examples of the application
of SWIR spectroscopy to mineral exploration (1990s);
• Use of PCs in the field, allowing rapid data interpretation
(1995–present);
• Heightened interest due to use of airborne hyperspectral scanners
and increasingly sophisticated data processing software
(1998–present).
SWIR spectrometers also are now employed in numerous other
capacities beyond exploration. In particular, they are useful in
mineral processing control procedures and evaluation of leach piles
and tailings dumps. Continued development of applications will
lead to uses in other environmental applications and the
geotechnical fields.
SWIR Spectroscopy
Remote-sensing geologists use a variety of bands within the
electromagnetic spectrum, including the visible-near infrared (VNIR),
short-wave infrared (SWIR), and mid-infrared (MIR). Field portable
instruments detect in the SWIR region, which is sensitive to
molecular changes, and also in the VNIR, where color variations and
Figure 1. Detail of feature in kaolinite SWIR spectrum collected with PIMA-II
changes in elemental oxidation states (e.g., iron and chromium) are spectrometer. The hull, feature depths, position and the full width half-maximum
observed. VNIR, however, does not relate directly to composition. (FWHM) are shown.
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 19

Minerals can be distin-

guished not only on the
basis of distinctive features
and wavelength positions,
but also by the character of
the profile (without hull
subtraction). Examples of
common alteration minerals
are shown in Figure 2.
Mineral identification is
based on wavelength posi-
tions, intensity and shape of
absorption troughs, and the
overall shape of the entire
spectrum.
The short-wave infrared
wavelength region is not
suitable for most anhydrous
silicates. In addition, it is
difficult to identify minerals
present in amounts less than
5%, unless the sample is a
simple mixture with quartz
and the mineral is highly
reflective. Infrared reflec-
tivity varies between mineral
species. In mixtures of
infrared-active minerals, the
dominant and typically most Figure 2. Stack plot with examples of spectra
reflective mineral is easily characteristic of individual minerals. Reflectance values
are offset for clarity. Examples are from the SPECMIN™
identified; however, as a database.
general rule, 10% or more of
a mineral must be present
for positive identification.
Where low reflectance min-
erals are present, recognition
may require 20% or more of
the mineral in the sample
(e.g., carbonate, chlorite).
Mineral chemistry: Vari-
ations in mineral chemistry
are typically detected by
shifts in wavelength posi-
Figure 4. Examples of K-, Na-, Ca-, and NH 4-bearing alunite group
tions or changes to the hull minerals. An inset shows the positions of distinctive features for K and
shape. The presence of iron Figure 3. Example of Fe- and Mg-bearing clinochlore. Ca in the Ca-dominant sample. Elements present in this sample were
Note the steep slope on the Fe-rich sample from 1,300 confirmed by EDS analysis (scanning electron microscope). Examples
in most minerals results in a nm to 1,900 nm. are from the SPECMIN™ database.
strong positive slope from
1,300 to 1,900 nm. A com-
parison of the spectra from Fe-rich and Mg-rich clinochlore is shown may be required in order to assign an observed variation to a
in Figure 3. Chemical variation in the carbonate group of minerals is change in chemical composition. Mineral composition variations are
gauged by a shift in the position of the major feature as a function best evaluated from monomineralic samples; however, it may be
of the cation present. The dominant feature varies widely, including possible to define variations in some mineral mixtures.
magnesite (Mg) at 2,300 nm, dolomite (Mg, Ca) at 2,320 nm, calcite
(Ca) at 2,330 nm and rhodochrosite (Mn) at 2,360 nm. Variations in DATA COLLECTION AND ANALYSIS
alunite-group mineral chemistry are manifested by shifts in the Data Collection
1,480-nm position, with values ranging from ~1,461 (NH4), to ~1,478 Understanding the variables that affect spectra is critical to the
nm (pure K) to ~1,496 (Na) to 1,510 nm (Ca). Examples of all four interpretation of spectral data sets. These variables include grain
to page 20 . . .
spectra are shown in Figure 4. Depending on the quality of size, transparency, sulfide content, overall
reference spectra, petrographic, SEM or electron microprobe data reflectivity, water content, heavy element
20 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 19 S W I R S P E C T R O S C O P Y , C O N T .
able to identify many spectra characteristic of minerals by eye, as
data collection is carried out. Even novice users can quickly learn
content, contaminants (e.g., oil, organic material), orientation of the basic minerals important in their study area. Only a few minerals
minerals (e.g., micas), and mineral mixtures. Good data collection typically are required to characterize an area, which also realistically
procedures that minimize the effect of these variables should be allows for the tracking of variability within a single mineral species.
adopted. A minimum of 2 spectra per sample is necessary, both for Reference databases are extremely useful in refining the ability of
reproducibility of data and to test for heterogeneity of the samples. data processing software to provide automatic identification.
By analyzing the groundmass, veins, phenocrysts, vug infill, fracture Automatic identification may be helpful when working with large
coatings, and weathered surfaces, the data gatherer can identify data sets on well-defined areas. In order to achieve high-quality
several minerals in one sample. Clear descriptions of basic results, variations at the deposit scale must be observed and
observations — e.g., color, texture, veining or fractures, vein recorded by the user, using reference data sets created for that
envelopes, and weathering state — are important for high-quality deposit. Deposit or region specific data sets appear to be critical in
spectral interpretation. obtaining reliable results from attempts at unmixing (identifying
Sample types and sample processing may also affect the spectra. mineral mixtures) using algorithms. Identification of complex
Hand specimens, powders, rock chips, liquids, and reject samples mixtures will require geologic context, user experience, and
can all be analyzed, with some minor variations observed in the establishment of reference samples with additional information, e.g.,
spectra. Samples that have been pulverized (e.g., analytical pulps) petrography, XRD, and SEM analysis, but may be difficult with
commonly yield extremely degraded patterns, where many of the currently available algorithms.
spectra are similar in appearance. Pulps are typically made with ring Data processing software allows the subtraction of the hull (see
mills that generate heat during the crushing process. Clay analysis, Fig. 1), typically followed by extraction of feature positions,
in particular, will be inaccurate if the structure of the mineral intensities, and widths. There are various methods to extract
changes with heating. Standard XRD analysis for clays is also done diagnostic derivative spectra from those that are measured. Two
without the use of ring mills for the same reason. common methods are the hull quotient, which is a “rubber band”
Instrument stability must be considered in evaluating spectra, method of removing the effects of variable background slopes, and
particularly in variable field conditions. Without calibration, the first derivative that removes the effects of background by
wavelength positions will shift as the instrument heats up or is emphasizing changes in response. A variety of software packages
moved. Analysis of a standard kaolinite with the PIMA-II shows a are available commercially, with the most flexible being those that
systematic wavelength shift of 2 nm downward as the instrument allow the importing of data from a variety of sources (spectrometers
heats up from 22° to 44°C. Although small (within the spectral or scanner data). Care must be taken when extracting data to ensure
resolution of the PIMA spectrometer), this shift highlights the need that the data is treated in the same manner, e.g., feature positions all
for consistent recalibration of the instrument. Good laboratory based on the hull quotient, or on the first derivative.
practice also includes use of standards and the saving of calibration In some cases, use of a single feature position, depth, or ratio of
files as references for future checks on instrument drift. Some depths of two features may provide broad outlines of alteration
workers have reported 2 nm variations in as few as 10 zones. This style of data extraction with a computer program can be
measurements. done extremely quickly; however, the data must be carefully
SWIR spectroscopy is a useful tool for identifying minerals in evaluated by an experienced user to confirm that the comparison is
individual samples; however, its greatest power comes from of similar material, with comparable mineral assemblages and
consistent collection of data in a systematic manner. Sample features. Since wavelength positions for various minerals overlap,
intervals may need to be as small as 1 to 2 m to evaluate gradients faulty results which are based on a single feature may be produced.
in alteration mineralogy and define boundaries when used for drill For example, the Al-OH feature at ± 2,200 nm may represent alunite,
core logging or detailed traverses. Once the basic variations are pyrophyllite, kaolinite, dickite, illite, mixed-layer illite/smectite,
described, spacing can be widened, depending on the area to be smectite, or muscovite, which all obviously have different
covered and the goal of the survey. For example, core logging may implications in terms of deposit modeling. Contouring of such data
be done on 5- to 10-m spacing, whereas mapping may be widened must also be carefully carried out. Individual analyses may falsely
to 50 or even 100 m. Closely spaced sampling typically produces weight the data, resulting in spurious features.
the most useful information. For mapping, the sample locations may Mineral percentages: A common objective of SWIR analysis is to
be laid out on a grid pattern and include soils or may be tied to determine not only the minerals present in a sample but also their
outcrop patterns. Data processed and evaluated concurrent with relative abundance. Many software programs attempt to provide
mapping can have a direct impact on an exploration program. mineral unmixing as part of the package. This task is challenging
Data Analysis and Processing due to the lack of knowledge regarding the absorption coefficients
Mineral identification is based on the use of reference data sets, for the molecular bonds detectable in the SWIR range of the
which are empirical records of each mineral’s characteristic spectra. electromagnetic spectrum. The spectral data indicate that minerals
Visual observation of a group of mineral spectra will quickly show apparently are not present in linear mixed configurations, but rather
variations based on numerous factors, including mineral chemistry, as a function of these unknown absorption coefficients. The
temperature, and mode of formation (reflected in crystallinity), and intensities of the absorption features, therefore, cannot be used as a
other subtle changes. The user’s greatest asset, then, is a well- 1:1 correlation to relate directly with the amount of mineral present.
defined and large reference collection or spectral database created For instance, an iron chlorite will absorb more energy at its
from samples representative of a wide variety of deposit diagnostic wavelengths and reflect less back to the detector than an
environments and occurrences. Experienced users are ultimately aluminum-bearing mineral, a muscovite, which more accurately
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 21

reflects the absorbed energies. In addition, there is a matrix effect Alteration was defined based on mineral assemblages, and the
problem when non-infrared active minerals are also present and alteration patterns determined by spectral analysis were used to help
absorbing, but not reflecting the excitation energy. predict zones of mineralization. It was found that dickite increased
In cases where the absorption coefficients are likely to be similar in areas associated with gold mineralization, thus providing a
for the minerals in question, the resulting spectrum can be treated mineralogical pointer to areas favorable for gold. Figure 5 is a drill
essentially as a linear mixture. The key to accurate results lies in log from the Alumbre zone, showing the relationship of the
building calibration files from the sample suite under investigation. alteration to mineralization. Spectral analyses were carried out every
Monomineralic end members must be chosen for the models to 2 to 3 m down hole. This hole was subsequently deepened and
additional mineralization associated with
to page 22 . . .
succeed. The accuracy of the mixing algorithm may be as good as
4%; however, this will vary depending on the software (algorithm), dickite, alunite, and vuggy quartz was found.
the materials in the mixture, and their relative abundance.
Instrumentation is also a limiting factor in producing accurate
results. PIMA-II has approximately a 5- to 6-nm resolution and
samples are collected in 2-nm steps. This over-sampling is done to
improve the reproducibility of the method; however, it does not
necessarily improve the accuracy and leads to an artificial perception
of a 2-nm resolution. Therefore, the limit of the method for
resolving the wavelength positions remains between 5 and 6 nm.

CASE STUDIES
The following case studies illustrate the use of SWIR in
exploration. All of the analyses were collected with PIMA-II
spectrometers. These examples emphasize the use of the technique
as a mapping tool, for integration with other data types. The
examples include alteration maps, detailed drill logs, and integration
of geochemistry, petrography and spectral data. The mineral
identifications and variations in feature positions are shown to
define alteration zones and provide vectors toward mineralization.
Blind application of digital data, however, may lead to false results
based on inadequate sampling or misinterpretation of spectral
variations.
High-Sulfidation Epithermal Gold Deposits
The general characteristics of high-sulfidation deposits are well
known and are summarized in Arribas (1995), who includes
numerous examples of deposits around the world. These deposits
are known for their extremely fine grained alteration minerals and
typically homogeneous appearance. Minerals that are infrared active
and form in these environments are shown in Table 2. Alteration
mapped in the field typically relies first on varying degrees of silicic
alteration ranging from leached, vuggy quartz to zones of
replacement quartz. Beyond the quartz-dominated areas, however,
the alteration assemblages are commonly mapped simply as
advanced argillic and argillic alteration during exploration programs.
Identification of individual minerals, some of which are critical to
zoning patterns, is extremely difficult. Use of SWIR spectroscopy
allows the major alteration minerals to be easily and rapidly
identified.
Virgen: A comprehensive alteration study of the Virgen property
was completed for Gitennes Exploration Inc., Vancouver. The
Virgen property is a gold prospect located 180 km east of Trujillo,
Peru. Cretaceous sedimentary rocks and Tertiary andesites host
mineralization. The aim of the work was to determine alteration
zone patterns with respect to mineralization using the PIMA-II
spectrometer. Data were collected from available drill holes, hand
samples, and road cuts. A total of 22 drill holes were analyzed and
over 900 spectra were obtained across the property. Alteration
Figure 5. Drill log from the Alumbre zone, Virgen deposit, Peru, showing variations in
minerals observed using spectral analysis include alunite, dickite, alteration mineral assemblage, lithology, and mineralization. Dickite in the advanced
pyrophyllite, diaspore, kaolinite, smectite, illite, and quartz. argillic alteration was used as an indicator for mineralization.
22 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 21 S W I R S P E C T R O S C O P Y , C O N T .
zone contain two to three minerals in a single spectrum. Examples
of these spectra are shown in Figure 7. The alteration and style of
mineralization is consistent with a high-sulfidation epithermal to
Pamel: An alteration study of the Pamel prospect in the Western
magmatic-hydrothermal environment.
Cordillera of Peru was conducted for Candente Resource Corp. The
property is at an early stage of exploration. Geologists employed at
the property selected hand and reject samples from the geochemical
sampling program and submitted these to Vancouver for analysis
with a PIMA-II spectrometer. The results from approximately 128
samples were integrated with geologic information and outcrop
patterns. Based on three days work, combined with the previous
geologic data, an alteration map was created for the property. The
results clearly showed distinct alteration zones and helped delineate
zones of interest (Figure 6). The alteration varies, from silicification
to alunite-dickite, to alunite-kaolinite, to kaolinite dominant, and
outward to sericite, illite, and chlorite. Small amounts of diaspore,
topaz, and tourmaline were also noted locally. A detailed study was
conducted in the western portion of the map area. Samples from the

1000metres
Figure 7. Examples of spectra for mineral mixtures, from the Pamel prospect,
Candente Resource Corp.

Low-Sulfidation Epithermal Gold Deposits

Alteration in low-sulfidation deposits is characterized by adularia
and calcite-replacement textures within quartz veins that grade
outward in the host rock to illite, illite-smectite, and illite/smectite-
chlorite zones. Calcite may also occur within the alteration
envelope. The variation in clay alteration outward from
mineralization is typically difficult to define in the field, but may be
detected with SWIR. Zoning patterns are well described by
numerous authors, including White and Hedenquist (1990) and
Sillitoe (1993). The widths of alteration envelopes vary from
Figure 6. Alteration map based on outcrop patterns for the Pamel prospect,
centimeters to meters. Zones of steam-heated, advanced argillic
Candente Resource Corp. alteration may also cap or develop laterally from low-sulfidation
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 23

mineralization. Differentiation of these zones from high-sulfidation Volcanogenic Massive Sulfide Deposits
systems is critical to exploration in this environment. Table 2 lists Alteration mapping is an important aspect of VMS exploration in
minerals found in the low-sulfidation and steam-heated areas where metal distribution may provide limited information.
environment. Typically, distribution of Fe and Mg chlorite and sericite (muscovite)
Patagonia, southern Argentina: SWIR spectroscopy was carried zones are used as a vector toward ore lenses. Clays may also be an
out systematically on RC (reverse circulation) chips from a property important part of some systems. The style of alteration varies,
in Patagonia, southern Argentina. The area contains zones of high- depending on the setting of the deposit; Franklin (1993) gives
grade mineralization associated with illite-dominant alteration. The several models. Lithogeochemistry is commonly employed to
use of spectral analysis allowed mapping of alteration patterns. differentiate alteration types, but use of SWIR spectroscopy may also
Figure 8 is a drill log showing the distribution of illite, illite-gypsum, provide valuable, direct information regarding the alteration
illite-smectite, and illite-chlorite. The illite-gypsum zones clearly mineralogy.
flank the mineralization. Kidd Creek: Distribution of chlorite and sericite is outlined by
Koopman et al. (1999) for the Kidd Creek deposit, western Abitibi
Subprovince, Canada. Variations in proportions of the two minerals
appear to reflect proximity to ore and help to outline mineralized
zones. The initial work was based on field observations, X-ray
diffraction and petrographic techniques. Seventy-six of the XRD
samples were analyzed with a PIMA-II spectrometer. A variety of
processing techniques were applied to the data set, including
automatic mineral unmixing using commercially available software;
comparison of ratios of feature depths in the Al-OH region to the Fe-
OH (as described by Huston, 1999); and simple comparison of the
data to a set of artificially (linear) mixed spectra. The linear mixed
spectra used site-specific end members, and 10% increments (Figure
9, page 24). Results from this comparison were the most consistent
and correlate well with the estimated mineral abundance based on
peak intensities from the XRD analysis (Table 3, page 26). The
attempts at semiquantification using the other techniques, however,
did not produce reliable results. The effect of different reflectivity
and of mixed mineral assemblages may hinder the use of the
automatic techniques in determining mineral percentages in this
environment until better data processing is possible.
Intrusion-Related Deposits
Many deposit types occur within the intrusion-related
environment. Mapping of alteration has application to both broad
and local zoning. As shown in Table 2, a wide variety of alteration
minerals may be present in this environment and can be used to
focus exploration on a specific target type.
Red Mountain, British Columbia: A detailed study of the
relationship of alteration and mineralization was completed at the
Red Mountain project in northern British Columbia and is described
by Rhys et al., 1995. Gold mineralization in the area is spatially
related to a porphyry Cu-Mo stockwork.
Hydrothermal alteration is pervasive throughout the pre-Tertiary
rocks on Red Mountain, including all phases of the main intrusions.
Several shallow-dipping alteration zones are stacked above a quartz
stockwork/molybdenum zone with pervasive sericite alteration.
These zones include: (1) sericite-quartz-pyrite alteration (pyrite-
dominant), (2) chlorite-K feldspar-sericite-titanite alteration with
disseminated and vein pyrrhotite, (3) brown to black tourmaline
veins, and (4) K feldspar-titanite-actinolite alteration. Anomalous
gold values are associated with the transition from pyrite to
pyrrhotite, and sericite to K feldspar alteration. High-grade zones are
focused below areas of abundant tourmaline, in pervasive sericite
alteration. Details of the alteration zones are given in Table 4 (page
to page 24 . . .
Figure 8. Drill log shows the distribution of alteration with respect to gold assays.
26), along with a summary of major element
Example of low-sulfidation system, Patagonia, southern Argentina. geochemistry. The deposit is structurally
24 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 23 S W I R S P E C T R O S C O P Y , C O N T .
provides a framework for application of SWIR spectroscopy to the
deposit area and exploration within the district.
disrupted by several faults. The distinct alteration zones provide a CONCLUSIONS
way to reconstruct relative locations within the system where fault SWIR spectroscopy is a tool to assist field mapping in mineral
boundaries are crossed in drilling. exploration. The ability to rapidly differentiate fine-grained
Data gathered by the PIMA clearly outlines the major alteration alteration minerals in the field allows for an enhanced understanding
zones, based on the presence of key minerals for each assemblage. of the property under investigation, and the results can be applied
The results of SWIR analysis on drill core samples are compared to immediately to the exploration program. Further refinement of the
geochemistry and petrography in Table 4, and representative spectra alteration assemblages, including the use of other analytical
from the deposit are included in Figure 10. The complete log for a techniques, will yield data important for the development of deposit
drill hole from the Marc zone is shown in Figure 11. Integration models and regional exploration programs. Several components are
with the previous petrographic, geologic, and geochemical data critical to a successful survey. These are:

Figure 9. Artificial linear mineral mixtures determined from actual chlorite and muscovite end members for the Kidd Creek footwall rhyolite. The percentage of chlorite
increases down the plot. Examples of the comparisons of real spectra with selected mineral mixtures are shown on the right. Mineral mixtures were generated with
SPECWIN©, Spectral International Inc.
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 25

• A systematic, well-planned sampling program, with consistent ······················· ACKNOWLEDGMENTS ·······················

sample spacing depending on the purpose of the survey; We would like to thank the many geologists and consultants in
• Alteration mapping concurrent with a mapping or a drilling exploration for their contributions to the use of SWIR spectroscopy.
program, to allow rapid incorporation of information and The successful application of the technique would not have been
effective application of data; possible without them. Data in this paper was published with the
• SWIR data collection by a trained operator with a geologic permission of Jerry Blackwell (Gitennes Exploration Inc.), Joey
background; Freeze (Candente Resource Corp.), Mark Hannington (Geological
• Use of mineral-dominant assemblages for preliminary mapping; Survey of Canada) and Jacques Houle (Royal Oak Mines) who are
• Subsequent data processing of wavelength positions or other all thanked for their support. The paper benefited from reviews by
spectral characteristics, to further evaluate alteration; Noel White and Charles Tarnocai. In particular,
• Selected petrography and XRD for complete alteration
to page 26 . . .
Noel White is thanked for his enthusiasm, well-
assemblages that can be related to the spectral data. timed reminders, and critical review.
With the increased use of SWIR spectrometers in the field,
distribution patterns over large areas can be delineated. The use of
field spectrometers has provided a vast new database that not only
is aiding exploration, but also will contribute ultimately to the
understanding of these systems.

Figure 11. Drill log from the Marc zone, Red Mountain, British Columbia. Log shows
major alteration zones, lithologies, and the location of Au-Ag mineralization. The
SWIR active minerals identified in each alteration zone are: A. actinolite dominant:
actinolite + chlorite (Fe > Mg) + datolite + prehnite + axinite, B. tourmaline
dominant: chlorite (Mg > Fe) + muscovite + schorl + axinite + calcite, C. Tourmaline
stockwork: schorl + Mg chlorite + muscovite, D. auriferous pyrite-pyrrhotite
Figure 10. Representative spectra used to determine alteration zones at Red stockwork: muscovite + chlorite, and E. pyrite dominant: muscovite + chlorite +
Mountain, British Columbia. clinozoisite.
26 SEG NEWSLETTER Nº 39 • OCTOBER ’99

... from 25 S W I R S P E C T R O S C O P Y , C O N T .

Table 3. Comparison of SWIR and XRD Results for Selected Samples from the Kidd Creek Footwall Felsic Units

Spectra Lithology SWIR X-ray diffraction

ah09667b Rhyolite Chlorite >> muscovite Quartz >> clinochlore > muscovite
ah09682b Quartz porphyry Chlorite > muscovite Quartz >> clinochlore, muscovite
ai00510b Quartz porphyry tuff Chlorite >> muscovite Quartz >> albite, clinochlore, calcite > muscovite
ai00557a Felsic tuff Chlorite > muscovite Quartz >> clinochlore, albite, calcite > muscovite
ai00671b Rhyolite Chlorite > muscovite Quartz >> clinochlore > muscovite
ai00724b Rhyolite Muscovite > chlorite Quartz >> albite, clinochlore > muscovite
ai01511b Quartz porphyry tuff Chlorite > muscovite Quartz >> albite > clinochlore, muscovite
ai01532b Quartz porphyry tuff Chlorite > muscovite Quartz >> clinochlore, calcite > muscovite
ai01538a Quartz porphyry tuff Chlorite = muscovite > kaolinite? Quartz >> clinochlore > muscovite
ai01545a Rhyolite Muscovite >> chlorite > kaolinite? Quartz >> albite, muscovite > clinochlore
ai01550b Felsic tuff Muscovite = chlorite Quartz >> albite > clinochlore, muscovite

SWIR analysis identified the chlorite as Fe-bearing

Table 4. Distribution of Alteration Zones at Red Mountain, B.C.

Alteration Thickness Geochemistry Veins Petrography SWIR

Zone
Actinolite >150m Na2O > 3.3%,; K2O < 0.5%; Chlorite + pyrite + K feldspar + actinolite + Actinolite + chlorite
CaO > 2.8%; Sr > 400 ppm actinolite + calcite chlorite + titanite + albite (Fe>Mg) ± axinite* ±
+ pyrite ± datolite ± muscovite ±
prehnite
Tourmaline 100-300m Na2O > 3.3%; K2O < 0.5%; Tourmaline + pyrite + K feldspar + chlorite + Schorl (tourmaline) +
stockwork CaO > 2.8%; Sr > 400 ppm chlorite + pyrrhotite titanite + pyrite + chlorite (Mg) ± muscovite
tourmaline + pyrrhotite ± carbonate ± axinite*
Pyrrhotite 100–200m Na2O > 3.3%; K2O < 0.5%; Pyrrhotite + pyrite ± K feldspar + sericite + Muscovite + chlorite (Mg)
CaO >2.8%; Sr > 400 ppm chalcopyrite ± chlorite ± pyrrhotite + pyrite + axinite ± carbonate ±
calcite ± quartz ± chlorite ± tourmaline prehnite
sphalerite ± galena
Auriferous 10-50m Na2O < 1.5%; K2O > 5%; Pyrite ± pyrrhotite ± Intense sericite + pyrite; Muscovite + chlorite (Mg)
Py-Po high values in Au (>0.5 quartz ± chlorite mantled by disseminated
stockwork ppm); Ag, As, Sb and and veinlet sphalerite +
locally Cu, Zn correspond pyrrhotite + pyrite
with ore zones
Pyrite 100-200m Na2O < 3.5%; K2O > 4%; Pyrite ± calcite ± quartz ± Sericite, pyrite ± calcite ± Muscovite + chlorite
CaO < 3.3%; Sr < 100 ppm chlorite chlorite ± tourmaline
Gypsum <5–100m Similar to pyrite-dominant Gypsum + pyrite + calcite Sericite + pyrite ± quartz Gypsum*
stockwork alteration ± quartz ± K feldspar
Quartz >200m Cu > 300 ppm; Mo > 30 Quartz + pyrite ± chlorite Sericite + quartz + pyrite Chlorite (Mg), muscovite ±
stockwork, ppm; SiO2> 55% and ± epidote ± magnetite ± + chlorite + K feldspar ± clinozoisite*
Mo-Cu similar values to pyrite molybdenite ± chalcopyrite epidote ± tourmaline ±
dominant alteration for magnetite ± hematite
Na2O, K2O, CaO, and Sr
* Denotes vein mineralogy
Zones are listed from the top down, with their associated geochemistry, vein mineralogy and alteration mineralogy (petrography and SWIR
analysis); after Rhys et al (1995)
OCTOBER ’99 • Nº 39 SEG NEWSLETTER 27

····························· REFERENCES ····························· alteration at Kidd Creek, Economic Geology Monograph 10, in
Arribas, A., Jr., 1995, Characteristics of high-sulfidation epithermal press.
deposits, and their relation to magmatic fluids: Mineralogical Kruse, F.A. and Hauff, P. L., 1991, Identification of illite polytype
Association of Canada Short Course v. 23, p. 419–454. zoning in disseminated gold deposits using reflectance
Clark, R.N., King, T., Klefwa, M., Swayze, G.A., and Vergo, N., spectroscopy and X-ray diffraction — potential for mapping with
1990, High spectral resolution reflectance spectroscopy of imaging spectrometers: IEEE Transactions on Geoscience and
minerals: Journal of Geophysical Research, v. 95, no. B-8, p. Remote Sensing, v. 29, p. 1101–1104.
12,653–12,680. Lyon, R.J.P., 1962, Minerals in the Infrared — A critical bibliography:
Crowley, J.K., 1984, Near infrared reflectance of zunyite: Stanford, California, Stanford Research Institute, 76p.
Implications for field mapping and remote-sensing detection of Marel, H.W. and Beutelspacher, H., 1976, Atlas of infrared
hydrothermally altered high alumina rocks: Economic Geology, spectroscopy of clay minerals and their admixtures: Amsterdam,
v. 79, p. 553–557. Elsevier, 396 p.
——1996, Mg- and K- bearing borates and associated evaporites at Marsh, S.E. and McKeon, J.B., 1983, Integrated analysis of high-
Eagle Borax Spring, Death Valley, Califonia: A spectroscopic resolution field and airborne spectroradiometer data for alteration
exploration: Economic Geology, v. 91, p. 622–635. mapping: Economic Geology, v. 78, p. 618–632.
——1999, Association of magnesium-rich clay minerals and borate Martinez-Alonso, S.E., Goetz, A.F.H., Atkinson, W.W., Kruse, F.A.
deposits: A potential remote sensing guide for borate mineral and Eberl, D.D., 1999, Multiscale study of infrared spectra of
exploration?: International Conference on Applied Geologic clay minerals: from ab initio quantum calculations to
Remote Sensing, 13th, Vancouver, B.C., March 1 – 3, 1999, ERIM hyperspectral remote sensing: International Conference on
International Inc., Proceedings, p. I-255–262. Applied Geologic Remote Sensing, 13th, Vancouver, B.C., March
Denniss A.M., Colman, T. B., Cooper, D.C., Hatton, W.A., and 1 – 3, 1999, ERIM International Inc., Proceedings, p. I-174–181.
Shaw, M.H., 1999, The combined use of PIMA and VULCAN Moenke, H., 1962, Spektralanalyse von mineralien und gesteinen;
technology for mineral deposit evaluation at Parys Mountain eine anleitung zur emissions- und absorptionsspektroskopie:
mine, Anglesey, UK: International Conference on Applied Leipzig, Akademie Verlagsges. (in German).
Geologic Remote Sensing, 13th, Vancouver, B.C., March 1 – 3, Passos, R.V., and de Souza Filho, C.R., 1999, Spectro-mineral
1999, ERIM International Inc., Proceedings, p. I-25–32. mapping using field spectroscopy and the geometry of
Farmer, V.C., 1974, The infrared spectra of minerals: Mineralogical hydrothermal alteration zones associated with mesothermal gold
Society of London, Darking, Surrey, Adlard and Son Ltd., mineralization: A case study in the Brumal Deposit, Quadrilatero,
Monograph 4, 539 p. Ferrifero, Brazil: International Conference on Applied Geologic
Franklin, J.M., 1993, Volcanic-associated massive sulphide Remote Sensing, 13th, ERIM International Inc., Proceedings, p. I-
deposits: Geological Association of Canada Special Paper 40, p. 73–80.
403–417. Rhys, D.A., Sieb, M., Frostad, S.R., Swanson, C.L., Prefontaine,
Grove, C.I., Hook, S., and Paylor, E.D., 1992, Laboratory M.A., Mortensen, J.K. and Smit, H.Q., 1995, Geology and
reflectance spectra of 160 minerals, 0.4 to 2.5 micrometers: Pilot setting of the Red Mountain gold-silver deposits, northwestern
Land Data System, Jet Propulsion Laboratory, Pasadena, British Columbia: Canadian Institute of Mining Metallurgy,
California, JPL Publication 92-2. Special Volume 46, p. 811–828.
Hauff, P.L., 1993, SPECMIN™ Mineral Identification System and Rose, A.W. and Burt, 1979, Hydrothermal alteration, in Barnes,
Spectral Library, v. 1 and 2.: Arvada, Colorado, Spectral H.L., ed., Geochemistry of hydrothermal ore deposits: New York,
International, Inc., 600 p. Wiley-Interscience, p. 173–235.
Hunt, G.R. and Ashley, R.P., 1979, Spectra of altered rocks in the
Shen, X.C., Dunlop, A.C., and Cohen, D.R., 1999, Characterization
visible and near infrared: Economic Geology, v. 74, p.
of residual and transported regolith profiles using the PIMA-II
1613–1629.
(abs): The Association of Exploration Geochemists, International
Hunt, G.R., and Salisbury, J.W., 1970, Visible and near-infrared
Geochemical Exploration Symposium, 19th,Vancouver, B.C., April
spectra of minerals and rocks: I. Silicate minerals: Modern
1999, Program and Abstracts, p. 137–138.
Geology, v. 1, p. 23–30.
Sillitoe, R.H., 1993, Epithermal models: genetic types, geometrical
——1971, Visible and near-infrared spectra of minerals and rocks:
controls and shallow features: Geological Association of Canada
II. Carbonates: Modern Geology, v. 2, p. 283–300.
Special Paper 40, p. 403–417.
Hunt, G.R., Salisbury, J.W., and Lenhoff, C.J., 1971a, Visible and
Stewart, A. and Kamprad, J., 1997, Piercing the regolith veil;
near-infrared spectra of minerals and rocks: III. Oxides and
identifying parent rocks from weathered equivalents, Eastern
hydroxides: Modern Geology, v. 2, p. 195–205.
——1971b, Visible and near-infrared spectra of minerals and rocks: Goldfields, Western Australia: Australian Geological Survey
IV. Sulphides and sulphates: Modern Geology, v. 3, p. 1–14. Organisation Research Newsletter 26, May 1997, 6 p.
——1971c, Visible and near-infrared spectra of minerals and rocks: Thompson, A.J.B., and Thompson, J.F.H., 1996, Atlas of
VI. Additional silicates: Modern Geology, v. 4, p. 85–106. alteration: A field and petrographic guide to hydrothermal
Huston, D.L., Kamprad, J., and Brauhart, C., 1999, Definition of alteration minerals: Geological Association of Canada, Mineral
high-temperature alteration zones with PIMA: An example from Deposits Division, 119 p.
the Panorama VHMS district, central Pilbara Craton: Australian White, N.C., and Hedenquist, J.W., 1990, Epithermal environments
Geological Survey Organisation Research Newsletter, v. 30, p. and styles of mineralization: Variations and their causes, and
10–12. guidelines for exploration: Journal of Geochemical Exploration,
Kodama, H., 1985, Infrared spectra of minerals; reference guide to v. 36, p. 445–474.
identification and characterization of minerals for the study of Zhang, G., Pan, Y. and Chen, N., 1998, Short wave infrared (SWIR)
soils: Research Branch Agriculture Canada, Technical Bulletin reflectance spectroscopic study of kaolinite-group minerals at the
1985-1E, 197 p. Athabasca basin, Saskatchewan [abs]: Geological Association
Koopman, E., Hannington, M., Santaguida, F., and Cameron, B., of Canada and Mineralogical Association of Canada Abstract
1999, Petrology and geochemistry of proximal hydrothermal Volume, Quebec 1998, Abstracts, p. A-203. 1