&EPA
EPA910-R-01-002
    United States Environmental Protection Agency
    Region 10,1200 Sixth Avenue, Seattle, WA 98101-1128
        Feasibility for Identifying Mineralogical and Geochemical
                              Tracers for Vermiculite Ore Deposits
                                                          February, 2001
                                                            Prepared By
                                            David Frank and Lorraine Edmond
                                         U.S. Environmental Protection Agency
                                          Office of Environmental Assessment
                                                              Region 10

-------
ACKNOWLEDGMENTS

       Extensive library assistance by Alison Keyes and Joanne Meyer, ASRC Aerospace
Company, greatly aided compilation of this literature review. We thank Alfred Bush, Gregory
Meeker, and Gregg Swayze of the U.S. Geological Survey and several colleagues within the EPA
for their critical comments.
                                          11

-------
CONTENTS

Acknowledgments 	ii
Contents	iii
Abstract	1
Introduction	1
Background 	2
       Methods of Vermiculite Formation	2
       Ore Minerals and Parent Minerals 	2
       Product Minerals 	3
Overview of Vermiculite Ore Deposits 	4
       Type 1 Examples 	5
             Libby, Montana 	5
             Palabora, South Africa 	7
       Type 2 Examples 	9
             North Carolina Deposits, Blue Ridge Mountains Province  	9
       Type 3 Examples 	9
             South Carolina Deposits 	9
             Louisa, Virginia 	10
             Dillon, Montana 	11
       Unclassified Examples 	11
             Hafafit, Egypt 	11
             Russian Deposits 	12
             Chinese Deposits 	12
Potential Tracers	12
       Mineralogy  	13
             Vermiculite and Parent Minerals 	13
             Accessory Minerals 	14
       Chemical Composition 	15
             Major and Minor Elements 	15
             Trace Elements	16
             Isotope Ratios	17
       Likelihood of Deposits being Chemically Distinctive 	17
       Limitations of the Data 	17
Recommendations for Further Study 	18
       Analytical Methods  	18
       Study Design	20
Conclusions	22
References	23
Glossary 	28
                                         in

-------
FIGURES

1.  Index map of locations ofvermiculite deposits discussed in this report 	33
2.  Graph of mineral composition ofvermiculite ore (data from Atkinson and others, 1982)
       A. Mine atLibby, Montana 	34
       B. Mines at Enoree, South Carolina 	35
3.  Graph of elemental composition reported for vermiculite and associated minerals 	36
4.  Chart showing analytical approach for screening phase of tracer analysis 	37
TABLES

1. Vermiculite and associated minerals	39
2. Locations of mines that produced vermiculite for consumer products 	40
3. Minerals reported at selected vermiculite deposits  	41
4.  Chemical composition ofvermiculite and associated minerals
       A. Oxide concentration in weight percent 	42
       B. Element concentration in parts per million  	43
5. Description of methods for identifying tracer constituents in vermiculite ore 	44
                                           IV

-------
ABSTRACT

       A review of the geological, industrial, and health literature on vermiculite ore deposits
indicates that mineralogical or chemical fingerprints may exist that would allow linking
vermiculite in consumer products to a particular source ore deposit.  Two characteristics of
vermiculite suggest that pursuit of a set of tracers could be successful. First, vermiculite deposits
are formed in geochemically distinct, ultramafic environments that may concentrate a unique set
of minor or trace elements.  Second, vermiculite is a mineral with particularly high cation
exchange capacity and is amenable to sequestering and retaining trace elements. A two-phase
approach toward identifying suitable tracers is recommended. The first phase would include a
screening study of a small set of samples from the few major sources that serve the U.S. market.
The screening study would determine whether enough variation exists among ore deposits to
yield measurable differences in diagnostic minerals or elements.  The screening study would also
evaluate whether diagnostic characteristics can be detected in a few selected vermiculite
products. If a screening study successfully identifies potential tracers, a second-phase detailed
study would use a large set of samples to determine whether tracer variability among ore deposits
is greater than variability within a deposit.  The second phase would also determine a range of
vermiculite products in which characteristics can be measured.

INTRODUCTION

       Vermiculite is used as a component of many consumer products sold in the northwestern
states within the jurisdiction of Region 10 of the U.S. Environmental Protection Agency.  These
states include Washington, Oregon, Idaho, and Alaska. EPA investigations at a major U.S.
source of vermiculite at Libby, Montana, have shown that asbestiform amphibole minerals occur
as accessory minerals in vermiculite ore (Atkinson and others, 1982). Asbestiform minerals have
also been reported at other commercial vermiculite deposits.  Some of these occurrences are
reported at concentrations less than those found at Libby, but most are at poorly known or
unknown concentrations. The asbestiform component of vermiculite ores may not be completely
removed in the preparation of consumer products, as shown by recent testing of vermiculite
garden products (USEPA, 2000). In order to evaluate risk of exposure to asbestos in vermiculite-
containing materials, it is important to know which materials contain asbestiform accessory
minerals.

       As used in this report, asbestos refers to varieties of serpentine and amphibole minerals
that have asbestiform characteristics (see Glossary).  Asbestiform is used to describe minerals
that have a fibrous form with relatively small fiber thickness, large fiber length, flexibility, easy
separability, and a parallel fiber arrangement.

       Direct analysis of asbestos content can be straightforward when the mineral is
concentrated enough to allow easy identification and quantification.  However, analysis of very
low, but potentially hazardous, concentrations of asbestos in consumer products may require
multiple, expensive techniques in order to achieve both certainty in mineral identification and
high confidence in reported concentrations.  If the presence of unacceptable concentrations of
asbestos in commercial vermiculite is largely confined to that derived from only one or a few

-------
discrete sources or mines, then some low-cost means of determining the source of vermiculite in
consumer products would be advantageous as a tool in screening for products that likely contain
asbestos. An efficient screening tool could thereby allow more focused use of expensive
asbestos-specific analytical methods on those products from problematic sources.

       The goal of this report is to determine if it is a reasonable assumption that a chemical or
mineralogical characteristic of vermiculite ore can be used as a tracer to identify the source of
vermiculite in consumer products. The goal is not to identify a specific tracer, but rather to
determine if pursuit of a tracer might be successful.  Specific objectives include the following:

       - Conduct a literature search of the occurrence and characteristics of vermiculite, with
       emphasis on diagnostic features of ore deposits,

       - Evaluate the variability of the compositional characteristics of vermiculite ore in light of
       potential mineralogical or geochemical tracers, and

       - Recommend whether additional work or analytical techniques might be successful in
       identifying a tracer for the source of vermiculite in consumer products.

BACKGROUND

Methods of Vermiculite Formation

       Vermiculite is  a secondary mineral formed primarily by the alteration of mica, and less
commonly by alteration of pyroxene, amphibole, olivine, chlorite, or other clay minerals (de la
Calle and Suquet, 1988, p. 455; Bush, 1976, p. 150). Weathering, alteration by ground water,
and hydrothermal processes have been proposed as mechanisms for vermiculite formation.
Industrial grade vermiculite deposits are believed to be formed both by weathering and ground
water alteration of macroscopic mica particles of biotite or iron-bearing phlogopite (Hindman,
1994).

       Geochemical environments in which vermiculite forms are generally associated with
ultramafic rocks rich in magnesium silicate minerals. Three categories of deposits are common
and are described in more detail below (see Overview of Vermiculite Ore Deposits, page 4).

Ore Minerals and Parent Minerals

       Vermiculite and its precursors (parent minerals) are part of a broad class of platy minerals
called sheet silicates (Table 1). Vermiculite can display a range of compositions depending on
the composition of the original parent minerals and the progress of chemical changes during
weathering.  Table 1 provides a comparative list of the ideal compositions of vermiculite, its
micaceous parent minerals, and a selection of other associated minerals that can be found in
vermiculite ore deposits.  The ideal composition of vermiculite is a magnesium iron aluminum

-------
silicate hydroxide hydrate. The hydrate refers to structural water that occurs between silicate
layers. In addition to structural water, vermiculite also contains non-structural water in an
amount depending on porosity and relative humidity.

       The most common parent minerals of industrial vermiculite are phlogopite and biotite
mica (Table 1). These two mica minerals are part of a solid solution series of potassium
magnesium iron aluminum silicate hydroxides that ranges from a magnesium-rich end member
of phlogopite to more iron-rich biotite. Besides vermiculite, another common mineral that forms
from mica is mixed-layer mica/vermiculite. In the mixed-layer minerals, interstratified mica and
vermiculite can occur with variable amounts of either component in either a random or regular
sequence of mixed layers. If the interstratification of biotite and vermiculite is regular the
mineral is called hydrobiotite (Brindley and others, 1983). By analogy with hydrobiotite, mixed-
layer phlogopite/vermiculite has been referred to as hydrophlogopite by Schoeman (1989, p. 7).

       Some disagreement occurs over whether mixed-layer mica/vermiculite can be formed by
hydrothermal processes as well as by lower temperature weathering or ground water alteration.
Roy and Romo (1957) pointed out that vermiculite was unlikely to form above 200° C.
Boettcher (1968) suggested that although vermiculite appeared to be a weathering product at the
large commercial deposit at Libby, Montana, hydrobiotite may have formed by hydrothermal
alteration. Libby (1975), on the other hand, demonstrated that both hydrobiotite and vermiculite
formed by weathering in large commercial deposits in the Enoree district, South Carolina.

       The industrial or commercial use of the term "vermiculite" is broader in meaning than the
mineralogical definition. Industrial vermiculite refers not only to ideal vermiculite but also to the
mixed-layer phases that are interstratified with biotite and phlogopite; that is, hydrobiotite and
hydrophlogopite.  This report retains the mineralogical meaning of vermiculite as the unstratified
mineral and will refer to it as distinct from hydrobiotite and hydrophlogopite (see Glossary).

Product Minerals

       Prior to industrial and consumer use, vermiculite is concentrated (beneficiated) by any
number of methods that can include screening, floating, or grading. Beneficiated ore is then
thermally or chemically treated to spread apart the macroscopic platy layers like an accordion in
order to make a lightweight, porous product. This process is called exfoliation or expansion.
The commonly used thermal treatment for exfoliation physically expands the layers by flashing,
or very rapidly boiling, the interlayer water to  steam. The characteristics of the product minerals
after the exfoliation process depend on the parent minerals and the length of time the product is
held at high temperature.

       Industrially exfoliated vermiculite may therefore include partly or completely dehydrated
vermiculite or mixed-layer mica/vermiculite. Industrial vermiculite may also include the parent
minerals, biotite or phlogopite, which may have been ineffectively removed during beneficiation.
Consequently, the mineralogical characteristics of vermiculite in the product materials may

-------
depend as much on the beneficiation and exfoliation processes as on the original ore.  For
example, the relative amount of vermiculite, mixed-layer mica/vermiculite (hydrobiotite or
hydrophlogopite), and parent mica (biotite or phlogopite) strongly depends on the efficiency of
beneficiation, and on the temperature and duration of heating during exfoliation.

       Several other major, minor, and trace minerals in the original ore may also occur as minor
or trace minerals in the exfoliated vermiculite product. Minerals that are mined along with ore
but are generally unwanted in the final ore product, are known as gangue. Table 1 includes a
selection of minerals in addition to vermiculite, that may be common in vermiculite ore.  The
presence of gangue in vermiculite products depends on the efficiency of the beneficiation
process.  Examples of studies that have documented the presence of non-vermiculite minerals in
beneficiated or exfoliated vermiculite products are discussed below in Accessory Minerals (see p.
14).

OVERVIEW OF VERMICULITE ORE DEPOSITS

       For ease of reference in discussion of ore deposits, the usage of several terms is noted
here; see the Glossary  for additional terms. In this report,  a mineral occurrence refers to a
concentration of a particular mineral or minerals in earth materials, regardless of the degree of
concentration. A mineral deposit refers to a mineral occurrence that might be concentrated
enough to be economic; an ore deposit is economic. Economic deposits are those that are
concentrated enough to be profitable to mine. Mines occur at deposits that have undergone some
extractive development of ore.  Ore is the naturally occurring earth material from which valuable
minerals can be extracted.  Ore minerals are that part of ore that is of economic value, whereas
gangue is that part of ore without value.

       This report reviews the mineralogical and geochemical characteristics of vermiculite and
some of the associated minerals by focusing primarily on deposits that are either within the
United States, or which are likely to be the source of imports to the U.S. (Figure 1, Table 2).
Within the United States, vermiculite deposits occur in twenty of the fifty states (Bush, 1976),
but ore has been mined in only six states, and in the last quarter century, only from Montana,
South Carolina, and Virginia.  As of 1998, the major operating vermiculite mines in the U.S.
included the W.R. Grace mines near Enoree, South Carolina,  and the Virginia Vermiculite Ltd.
mines near Woodruff,  South Carolina, and Louisa County, Virginia (Potter, 1998).  In addition,
the Elk Gulch deposit operated by Stansbury Holdings Corporation and Dillon Vermiculite
L.L.C. near Dillon, Montana, has produced small amounts of ore sporadically since 1990 (Berg,
1997).

       For many years, the W.R. Grace mine (Rainy Creek) at Libby,  Montana was the world's
largest producer of vermiculite, but since the mine's closure in 1990, the Palabora deposit near
Phalaborwa, Transvaal, Republic of South Africa has become the world's largest producer
(Hindman, 1994). In 1998, the United States producers sold approximately 170,000 tons of
vermiculite while the U.S. imported approximately 68,000 tons.  Approximately 60% of the

-------
imports came from Palabora in South Africa and 35% from Xinjiang Province in China (Potter,
1998). The Libby closure eliminated the only domestic source of coarse-sized concentrates with
flakes greater than 2 mm across, and by the end of 1991 essentially all such coarse vermiculite
was obtained from foreign sources (Hindman, 1994).

       Economic vermiculite deposits  commonly occur in ultramafic intrusive rocks, such as
coarse-grained pyroxenites,  and metamorphic bodies containing biotite schists and gneisses. The
ore is generally 20-35% vermiculite (Hindman,  1994). Most vermiculite deposits are associated
with ultramafic igneous or metamorphic rocks cut by igneous intrusions of silicic, alkalic, and
carbonatitic rocks.  While the deposits  can be grouped into three categories following the
classification system described by Bush (1976), below, there are strong similarities among all
three.

       Type 1. The largest deposits form in large ultramafic intrusive bodies, some of which
       are zoned, and are cut by intrusions of syenite, carbonatite, or both. Examples are Libby,
       Montana and Palabora,  South Africa.

       Type 2. Deposits occur in small to large unzoned dunite, pyroxenite, or peridotite
       intrusive bodies that are cut by pegmatite, syenite, or granite.  These deposits vary  in
       size, but most are small. They are characteristic of the Blue Ridge region of the
       southeastern U.S.  Examples  are the North Carolina deposits. None of the currently
       active mines in the U.S. are found in Type 2 deposits.

       Type 3. The most common deposits of this type are found in layered ultramafic schist
       that has been cut by pegmatite,  less commonly by pyroxenite or peridotite. This type
       includes the second largest productive district in the U.S., in the Piedmont of South
       Carolina. Examples are the Enoree district, South Carolina; Louisa County, Virginia;
       and most of the  deposits in Colorado, Wyoming and Texas.

Type 1 Examples

                                    Libby, Montana

       The Rainy Creek mine in the vermiculite deposit at Libby, Montana (Figure 1) was the
major commercial source of vermiculite in the world for many years. Detailed mapping by
Boettcher (1966) shows that the vermiculite is found within a concentrically zoned complex of
igneous intrusions that include biotite pyroxenite and biotitite, a rock made up of almost pure
biotite.  Both of these ultramafic intrusions are cut by alkaline pegmatites. In addition, Boettcher
(1966) describes late-stage hydrothermal veins that are widely distributed in the igneous complex
and adjacent rocks.  Most of the biotite has been altered to hydrobiotite and vermiculite. The
pegmatites have altered the pyroxenite  to fibrous soda-rich amphiboles along their contacts with
the pyroxenite. Alteration zones are  generally less than a foot wide and are reported by Boettcher
(1966, p. 82) to be composed mainly of "mass-fiber tremolite," an amphibole in asbestiform

-------
habit.

       The pyroxenite surrounding the biotitite contains the vermiculite ore zone, which consists
of mostly diopside, hydrobiotite, and apatite. Fibrous amphibole is found both in veins and
disseminated throughout the intrusive rock in thin layers along cleavage planes of the pyroxene.
Thin veins (approximately one inch) are common throughout the deposit, while thicker veins
with quartz cores are less common (Bassett,1959). Boettcher (1966, p. 103) notes for late-stage
hydrothermal veins that, "Most of the tremolite in the small stringers is cross-fiber material, but
the bulk of this mineral is a compact, mass-fiber variety..." The fibrous minerals were
sufficiently abundant that early in the history of the Libby deposit, several companies were active
in the development of tremolite asbestos deposits along with the vermiculite in the Rainy Creek
complex (Boettcher, 1966, p. 7).

       The mineralogical characteristics of the varieties of amphibole asbestos at Libby are
currently under study (G. Meeker, USGS, personal communication, 2000). The  widespread
amphibole was described as tremolite and "soda-rich tremolite" by Boettcher (1966, p.  102).
Boettcher also described another amphibole, asbestiform richterite, as  occurring in an unusual
tabular body within the complex that also contains vanadium-rich pyroxene, plus strontianite,
barite, and copper and zinc sulfides. Larsen (1942) reported an analysis for richterite that was
included in the dissertation by Boettcher (1966) as well as in the compilation by Deer, Howie,
and Zussman (1997). Wylie and Verkouteren (2000) reviewed the amphibole nomenclature with
respect to asbestiform amphiboles from Libby, and found reference to "tremolite, actinolite, soda
tremolite, richterite, and winchite."  In their own analyses of two samples from Libby, Wylie and
Verkouteren (2000) identified asbestiform winchite as defined under the most current amphibole
nomenclature of Leake and others (1997), but stated that the mineral could have been called a
soda tremolite, sub-calcic actinolite, or richterite under older naming conventions. Because of
the variability in chemical composition that may be common among amphibole minerals,
determination of the full range of fibrous amphibole varieties that exist at Libby awaits a more
complete study of many more samples representative of the whole mine, and the use of current
nomenclature.

       With respect to other characteristic minerals at Libby, the unusual vanadium-rich
pyroxenes are also associated with some of the hydrothermal veins in the complex, as are some
rocks high in titanium. Fluorapatite is a common accessory mineral throughout the Libby  deposit
and is found enclosed in the pyroxene grains and in the books of vermiculite, hydrobiotite, and
biotite. A magnetite-rich zone surrounds the vermiculite zone, but the ore zone itself is  described
as being relatively low in iron (Boettcher, 1966).

       A few studies have specifically  examined fibrous minerals in raw, beneficiated, and
exfoliated ore from Libby. Rohl and Langer (1977) reported the presence of chrysotile asbestos
in addition to amphibole asbestos in Libby ore. A study by the EPA (USEPA, 1980) reported
asbestiform tremolite as a contaminant  in the exfoliated vermiculite product from Libby at a
concentration of at least 1%.  A subsequent detailed mineralogical study commissioned by the

-------
EPA looked at various grades of beneficiated ore and found fibrous tremolite-actinolite at a
concentration of 2-7%, with the higher concentrations being in the finer-sized grades of
unexfoliated vermiculite (Atkinson and others, 1982).  A sample of "head feed" from the Libby
mill contained 21-26% fibrous tremolite-actinolite (Atkinson and others, 1982). Moatamed and
others (1986) later reported fibrous amphibole identified as actinolite at a maximum
concentration of 2% in unexfoliated vermiculite ore from Libby and 0.6% in the exfoliated
product, both collected at an exfoliation plant in Utah.

                                 Palabora, South Africa

       Potter (1998) reported that 65% of vermiculite imported into the U.S. in 1996 came from
South Africa. The Palabora mine, near the town of Phalaborwa in the Republic of South Africa
(Figure 1), is the source of the imported vermiculite.  The Palabora deposit has  many features in
common with the Libby deposit.  The Palabora Igneous Complex is also a zoned deposit with
ultramafic rocks (pyroxenite) at its core. Here the ultramafic  deposits were also intruded by
alkalic rocks, most of which are syenitic in composition.

       One significant difference between the Palabora deposit and the Libby deposit is that the
primary mica at Palabora is phlogopite rather than biotite, and the mixed-layer mica/vermiculite
alteration product that forms the vermiculite ore is hydrophlogopite rather than  hydrobiotite
(Palabora Mining Company, 1976; Schoeman, 1989; Evans, 1993). Minor hydrobiotite is
present at Palabora but is not mined, as it is finer-grained, and thus of a lower economic grade
(Schoeman,  1989).

       There are three separate open-pit mines in the Palabora Complex. Along with the
vermiculite mine, there is a copper mine (also known as Loolekop) which has byproducts of
magnetite, apatite, gold, silver, zirconium, uranium, and nickel as well as a pit in the world's
largest igneous phosphate (apatite) deposit.  The relationship  among these deposits is not well
understood, but some work has shown that copper sulfide liquid was present early in the
crystallization sequence (Evans,  1993).

       The vermiculite is mined from a coarse-grained zoned ultramafic body that consists of
phlogopite-serpentinite rock enveloped by phlogopite-diopside rock.  Apatite at Palabora is so
abundant in some of the rock types that it is found in economic concentrations over large areas,
but apatite content in the ultramafic rocks is quite variable (Palabora Mining Company,  1976).

       Carbonatites are a source of rare earth elements (REE), and at Palabora, they have been
evaluated for potential economic extraction.  It is not clear whether the rare earth elements are
found specifically in the vermiculite orebody or are restricted to the separate Loolekop orebody.
The Loolekop copper orebody contains uranothorianite (a variable oxide of uranium and
thorium) and baddeleyite (zirconium oxide with a trace of hafnium) (Palabora Mining Company,
1976).  While these minerals are present in small amounts, both are considered economically
recoverable. It was not determined if these unusual minerals  exist within the vermiculite ore

-------
deposit.

       No mention of fibrous minerals in the Palabora deposit was noted in the geological
literature, but some reports were found in the industrial and health literature. In a report
contracted by Mandoval, Ltd., Chatfield and Lewis (1979) examined vermiculite ore from
Palabora for chrysotile asbestos fibers. They suggest that chrysotile-like fibers (about 2 ppm) in
South African ore are instead rolled-up scrolls of vermiculite.  These scrolls are described as
forming at the cracks in vermiculite plates, becoming tubular in appearance, and forming bundles
similar to chrysotile (Chatfield and Lewis, 1979, 1980).

       Moatamed and others (1986) also found what they described as "rolled vermiculite"
fibers from Palabora ore sampled from a rail car at a plant in Utah that used vermiculite in
chemical processing. In addition, they reported the presence of actinolite and rare anthophyllite
fibers with a low length-to-width ratio. According to Moatamed and others (1986, p. 214), the
unexfoliated ore they analyzed contained 0.4% fibrous amphibole. However, elsewhere the
report describes the  South Africa sample as being  distinguished by "a near absence of fibers" (p.
215).  The method used for quantification was listed as X-ray diffraction, which leaves some
uncertainty as to whether the percentage reported includes solely fibrous, or both fibrous and
non-fibrous, amphibole.

       In an article evaluating the health of South African vermiculite workers, baghouse dust at
Palabora is described as containing small amounts of asbestos fibers (Hessel and Sluis-Cremer,
1989). The mineralogy of the fibers was not determined, but the authors state that it "may be
tremolite" (p. 22). Both the statements regarding asbestos in baghouse dust and its likely
mineralogy are cited as personal communications, with no additional documentation. Fiber
counts in the mill air filter samples were done in 1987, and eight samples varied from 0.27 to
0.80 f/ml, which, according to Hessel  and Sluis-Cremer (1989, p. 22),  are similar to counts
measured in Libby around  1980. It is  important to note that the measurement of fibers per
milliliter of air sampled is a measure of potential exposure at the mill and not an estimate of the
amount of fibrous minerals present in  the ore or in consumer products. Exposure measurements
are affected by the milling  methods, indoor air control, personnel movements, and other factors
as well as the abundance of fibers in the ore.

       The Palabora ore is described in an EPA report as being "essentially free of asbestiform
fibers" (USEPA, 1980, p. 4), a statement based on the work of Chatfield and Lewis (1979).
Analyses of the ore for asbestiform minerals have  been commissioned by Mandoval, the
European distributor of Palabora ore.  A recent such report by the Institute of Occupational
Medicine in Edinburgh (IOM, 2000) includes analyses from samples of six grades of commercial
vermiculite from Palabora. Samples were analyzed by polarized light  microscopy, and reported
to have a detection limit of 1 ppm. No amphibole or chrysotile asbestos  fibers were reported in
any of the samples, and further quantification efforts using electron microscopy and X-ray
diffraction were not conducted.

-------
Type 2 Examples

                North Carolina Deposits, Blue Ridge Mountains Province

       This type of vermiculite occurrence is common, but has not produced any of the large
economic deposits. Numerous bodies of dunite and peridotite have intruded metamorphic
country rock. Vermiculite is found in veins and lenses along the serpentinized contact between
the two rock types. Many of the vermiculite deposits are associated with pegmatites  (Bush,
1976).

       Unusual accessory minerals include nickel silicates and corundum.  Talc is also a
common accessory. Anthophyllite is mentioned as a "byproduct" in some deposits and has been
mined locally. The anthophyllite is not specifically described as fibrous, but elsewhere
"asbestos" (mineralogy unknown) is said to have been mined along with corundum in a
vermiculite-containing deposit (Murdock and Hunter, 1946).  Chrysotile is reported to occur in
one of the deposits in veins and in clusters of fibers.  In the Day Book dunite (Figure  1),
vermiculite is present as hydrophlogopite outside a talc zone and within a zone of "asbestos"
(Kulp and Brobst, 1954).

Type 3 Examples

                                South Carolina Deposits

       The most numerous vermiculite deposits of the United States are in potassic ultramafic
intrusions that have been regionally metamorphosed and cut by pegmatites. These deposits are
found in several of the belts within the Piedmont  physiographic province of the southeastern U.S.
The cores of potassic ultramafic bodies are "biotitites," composed almost wholly of biotite. They
were intruded as small plutons into biotite  gneisses, and they host occurrences of similar
vermiculite-bearing rocks as scattered occurrences from Georgia to Virginia (Libby, 1975).

       The economic deposits contain primarily hydrobiotite, which is found in the upper
portion of the biotite intrusions (Maybin and others, 1990). In his dissertation on the Enoree
district (Figure 1), Libby (1975) distinguished two types of vermiculite deposits, which he
denoted by  the prefixes "V-" and "HB-" to distinguish those that were dominated by vermiculite
from those that were dominated by hydrobiotite.  The hydrobiotite-bearing deposits have been
preferentially mined because of superior commercial properties compared to vermiculite (Libby,
1975, p. 30).

       Fluorapatite is a common accessory mineral in the South Carolina deposits. Sphene is an
ubiquitous minor phase in the  "HB" deposits at Enoree, occurring as wedges between the flakes.
Zircon is widely dispersed throughout the plutons (Libby,  1975). Other accessory minerals
include talc, chlorite, chromite, rutile, zircon, titanite, corundum, anatase, and amphibole
asbestos (Hunter, 1950).

-------
       At some of the deposits in the Enoree district, amphibole asbestos has been found
associated with the vermiculite.  The amphibole, identified by Libby in his dissertation as
tremolite, is described as occurring in two forms in the HB-biotitites, as a fibrous variety
mantling pyroxene and as stubby euhedral to subhedral crystals.  In the V-biotitites, the tremolite
is found in the same two forms, but a larger proportion of it is described as being of the fibrous
variety (Libby, 1975).

       A study commissioned by the EPA (Atkinson and others, 1982) examined samples from
two mills that processed ore from some of the mines in the Enoree district. This study, which
looked at the fibers using X-ray diffraction, identified some of the tremolite from Enoree as being
sodium-bearing.  The same study also described the South Carolina samples from the two mills
as varying somewhat depending on source location.  Tremolite-actinolite and anthophyllite were
found as mixed bundles of fibers from the W.R. Grace mill at Enoree and separately at the
Patterson mill (Table 2). Bulk samples from both the Grace and Patterson mills were reported as
containing "<1%" fibrous minerals (Atkinson and others, 1982). At the Grace mill, some of the
vermiculite was intergrown with the amphibole fibers, and some occurred separately in a fibrous
habit.

                                    Louisa, Virginia

       There is minimal information available in the geological literature describing the Virginia
Vermiculite mine in Louisa County, Virginia (Figure 1).  The deposits were briefly described
long before the mine completed its first year of production in 1979 (Meisinger, 1979).  Gooch
(1957) described deposits in Louisa County as mafic rocks intruded by a series of small
pegmatites. In the same area, he described some lenses of vermiculite that were 20 feet thick and
more than 100 feet long. Not long before the opening of the mine, Bush (1976) classified the
Louisa deposit as Type 3 and suggested that it may be similar to the South Carolina deposits.

       Limited information on asbestos content at the Louisa deposit occurs in the health
literature.  Rohl and  Langer (1977) found both amphibole fibers and chrysotile fibers in all six of
the ore samples from the Virginia vermiculite deposit that they examined. The chrysotile
asbestos was found both as individual fibers and in bundles.  The amphibole was described as
varying widely in composition, but with more than half being near the range of compositions for
actinolite.

       Moatamed and others (1986)  sampled Virginia ore collected from a rail car at a plant in
Utah that used vermiculite in chemical processing. An ungraded vermiculite sample was
reported to contain traces of fibrous amphibole (1986, p. 214), with actinolite "mostly as
cleavage fragments with a low length to width ratio (p. 217)." Moatamed and others (1986)
reported 1.3%  amphibole in both unexfoliated  and exfoliated ore, although the report is unclear
as to whether these concentration measurements represent total, rather than fibrous amphiboles.
                                           10

-------
                                    Dillon, Montana

       The Elk Gulch mine, near Dillon, Montana (Figure 1), has produced small quantities of
ore intermittently since 1990 (Berg, 1997).  The vermiculite occurs with hydrobiotite in pods and
stringers along the contact between an ultramafic body and the gneisses it intrudes.  A biotite
schist developed at the contact of the ultramafic bodies with the enclosing gneiss, and has
weathered to form vermiculite (Desmarais,  1976).

       The ultramafic body is composed of hypersthene, amphibole (actinolite-tremolite,
anthophyllite, cummingtonite), spinel, and olivine. The ultramafic rocks have been prospected
for nickel, which has been found, but not in economic concentrations. The deposit contains a
distinctive cinnamon-brown anthophyllite (non-fibrous) which may be present in the ore zone as
well (R. Berg, personal communication, 2000).

       The ultramafic rock has been serpentinized in places and chrysotile asbestos veinlets a
few millimeters thick surrounded by massive serpentine are conspicuous locally (Berg, 1995).
Analyses performed on drill core samples showed detectable amounts of asbestiform actinolite in
one sample, and anthophyllite in another. Additional samples from the ultramafic zone bordering
the deposit showed low but detectable amounts (<0.75%) of fibrous actinolite/tremolite and
chrysotile asbestos.  Transmission electron  microscopy (TEM) results indicated asbestiform
fibers made up less than 0.1 % of the sample (MDEQ 1999).

       A health risk summary report (Behre Dolbear and Company, 2000) conducted for the
company developing the mine reported fibrous talc and fibrous biotite as well as chrysotile and
asbestiform anthophyllite and tremolite-actinolite. The report estimates that amphibole asbestos
minerals make up less than 0.001% (10 ppm) of the ore prior to processing.

Unclassified Examples

       Insufficient information is available to place the following deposits into one of the three
types described above.  They are noted here to  indicate the degree to which other deposits are
found in similar geologic environments and contain distinctive accessory minerals, such as
fibrous amphiboles.

                                     Hafafit, Egypt

       In a group of small deposits near Hafafit, Egypt (Figure  1), asbestiform minerals and
vermiculite are found where ultramafic rocks which have been altered to serpentinite are cut by
pegmatites (El Shazly, 1975a,  1975b). Both asbestos and vermiculite have been mined from the
area, but neither are known to be imported into the United States. Phlogopite, hydrophlogopite
and vermiculite are found in association with asbestiform anthophyllite and lesser amounts of
asbestiform actinolite and tremolite.  Other minerals associated with the deposit include
serpentine, talc, and apatite.
                                           11

-------
                                    Russian Deposits

       Several types of vermiculite deposits are found in Russia (Vorovikov, 1973). Deposits
are found in the Ural Mountains, the Kola Peninsula, Kazakhstan, and other areas (Figure 1).
Among these, Vorovikov (1973, p. 172) reported Kovdarsk to be the largest deposit on the Kola
Peninsula.  Commercial vermiculite deposits are associated with ultramafic intrusions of dunite
and pyroxenite as well as alkaline intrusions.  The vermiculite minerals are described as
alteration products of both biotite and phlogopite (Vorovikov, 1973). One type of deposit is
described as being associated with pegmatites, corundum, talc, asbestos and other deposits as
well as with veins in serpentinites. The type and concentration of asbestos found in association
with the vermiculite is not specified. Vermiculite from the Russian deposits is not known to be
imported into the U.S.

                                    Chinese Deposits

       Potter (1998) reported that approximately 35% of the vermiculite imported into the US in
1996 came from China, with the remaining 65% coming from South Africa.  Using the estimated
domestic production for 1996, this would mean that the Chinese imports made up approximately
10% (24,000 metric tons,) of the total vermiculite used in the United States.  An industry
estimate of Chinese imports for 1998 (Moeller, 2000) reported 40,000 tons (22% of imports) as
coming from the Qieganbulake deposit in the Xinjiang Province (Figure 1).

       Little information is available in English on the geology or mineralogy of the Chinese
deposits. Tongjiang and others (1996a, 1996b) describe phlogopite, vermiculite, and mixed-layer
phlogopite/vermiculite in the Weili deposit, Xinjiang, which appears to be one of the few major
commercial deposits other than Palabora with phlogopite as the main mica mineral associated
with vermiculite. Rongqi and Junchen (1993) describe the same group of minerals at
Tseganbrark, Xinjiang.

POTENTIAL TRACERS

       A review of the mineralogy and geochemical variability among the major vermiculite
deposits may provide a basis for anticipating potentially significant  compositional differences.  If
sufficiently unique, such differences may then provide a fingerprint with which vermiculite in a
commercial product may be traced back to a particular ore deposit.  Table 3 lists a selected group
of minerals, including vermiculite, mixed-layer mica/vermiculite, parent mica, amphiboles,
serpentine, talc and apatite found in most if not all vermiculite ore deposits which serve the U.S.
market. The relative abundance of these minerals varies somewhat  with locality, but they
generally are all present.  Some pertinent characteristics of the major commercial deposits are
discussed first for vermiculite and parent minerals, and then for accessory minerals.
                                           12

-------
Mineralogy

                             Vermiculite and Parent Minerals

       All of the deposits producing commercially important vermiculite are derived from
alteration of a parent mica, either biotite or phlogopite. The presence of either biotite or
phlogopite in a vermiculite product would suggest that the ore came from a deposit rich in that
particular parent mineral. Phlogopite is reported as the main parent mineral for vermiculite at
only one of the major producers for the U.S. market, that at Palabora (Table 3). However, the
Weili and other deposits in China, a small but steadily increasing contributor to U.S. vermiculite
consumption, apparently also have phlogopite as the main mica parent mineral (Tongjiang and
others, 1996b; Rongqi and Junchen, 1993).  Phlogopite has been reported as the parent mica for
vermiculite deposits in North Carolina, Texas, and Egypt, but these are not current producers for
the U.S. market.

       The phlogopite/biotite ratio in a vermiculite product may therefore provide a potential
signature for the Palabora and at least some of the Chinese ore deposits, provided no other
sources of a phlogopite-bearing vermiculite have entered the U.S. market.  Color can be a means
of distinguishing large flakes of darker biotite from somewhat lighter phlogopite. Since the two
minerals are part of a solid solution series that varies mainly in iron content accompanied by only
subtle structural  change (Brindley and Brown, 1980), the discrimination of phlogopite from
biotite can be achieved with chemical analysis. Reflectance spectroscopy can also be a useful
method of discrimination (G. Swayze, personal communication, 2000).

       Mixed-layer mica/vermiculite,  either hydrobiotite or hydrophlogopite, has been reported
at all major commercial deposits (Table 3).  In fact, there is some evidence that the presence of
mixed-layer mica/vermiculite is what makes a vermiculite ore particularly amenable to producing
a durable, exfoliated  material.  For example, Libby (1975) reported that when non-mixed-layer
vermiculite is exfoliated,  the resulting material tends to be more brittle  and subject to
pulverization.

       Hydrobiotite at some deposits,  such as Enoree, is reported to occur in more or less
concentrated pods within the orebody (Libby,  1975), potentially leading to distinctive but
temporally variable vermiculite/hydrobiotite ratios in the product.  The  beneficiation process,
however, should tend to mix different qualities of vermiculite ore so that a more uniform product
results with a less variable vermiculite/hydrobiotite ratio than found in the raw ore.  No report
has been found for any of the ore producers that would indicate mining or milling procedures
specifically attempt to segregate products that have widely different ratios of vermiculite to
mixed-layer mica/vermiculite.  Therefore a unique ratio of these minerals in a vermiculite
product may not be found to be a useful signature.

       As previously noted,  the process of exfoliation alters the mineralogical characteristics of
both vermiculite and mixed-layer mica/vermiculite by removal of water. The degree of
                                            13

-------
dehydration that accompanies exfoliation measurably alters the mineral structure providing a
potential useful identifier.  The exfoliation process, however, may occur at any of a large number
of exfoliation plants across the country rather than at the original mine sites.  As many as
nineteen exfoliation plants were active in 1999 (Potter, 2000).  Since exfoliation plants may
receive both domestic and imported ore from a variety of mines, the potentially unique
mineralogical effects of exfoliation of a particular ore material may be confounded by different
operating conditions at different plants and by mixing with ore from more than one source.
Consequently, though the mineral ogical structure of vermiculite may be distinctively different in
various consumer products, evaluation would be needed to be certain the difference is not due
more to the characteristics of the exfoliation process rather than the ore deposit. Other than
changes in water content, the chemical compositions of individual mica and vermiculite phases
would not be expected to change during exfoliation and could remain potentially unique
identifiers of the individual mine sites, as discussed below in the section on Chemical
Composition (see p. 15).

                                   Accessory Minerals

       The classification of vermiculite ore deposits (see p.  5) emphasizes the zoned nature and
distinctive rock types associated with vermiculite deposits, for example ultramafic rocks intruded
by alkaline magmas. Zoning derived from magmatic, metamorphic, hydrothermal, or weathering
processes would tend to provide a wide variety of minerals within a single ore deposit. On a
broader scale, the vermiculite-forming processes as described in the literature for the different
commercial ore deposits have a common ultramafic thread, tending to make mineral suites
similar from deposit to deposit.  Under such conditions, the variability in minerals may likely be
as great within a deposit as among deposits.

       Table 3 notes that amphibole, serpentine, talc, and apatite accessory minerals have been
reported for most ore deposits for which information is available.  Perhaps the best potential
mineralogical tracers among the accessory minerals are the amphiboles, which are relatively
abundant at all sites and which have a wide range of compositions. Additionally, fibrous
amphiboles (asbestos), which are the impetus for this evaluation, are reported for all sites though
in highly variable and debatable amounts.

       Few studies in the literature have sought to contrast in detail both vermiculite and
accessory minerals at vermiculite ore deposits. Perhaps the most detailed to date is the study
contracted by the EPA that examined material from the W.R. Grace Rainy Creek mine at Libby,
Montana, and the W.R. Grace and Patterson mills in the Enoree district, South Carolina
(Atkinson and others,  1982). The variability of vermiculite and accessory minerals from raw ore
and beneficiated ore, as reported by Atkinson and others (1982), is shown in Figure 2A for Libby
and Figure 2B for Enoree samples representing one point in  time.  The diagrams indicate the
change in relative abundance of the minerals as raw ore with 20-40% vermiculite is concentrated
into beneficiated ore with 65-95% vermiculite. The data from both localities indicate continued
presence of the same accessory minerals in the beneficiated ores, notably amphiboles in the range
                                            14

-------
of 2-16%. Differences in beneficiated ore among localities include a relatively greater proportion
of talc and apatite at Enoree, and biotite, pyroxene, and calcite at Libby. Data from Atkinson and
others (1982) show that the accessory mineral patterns are carried through to exfoliated ore from
the Enoree site (Figure 2B). Though the total amphiboles remain high in beneficiated ore from
both localities, the Libby site retains a higher component of fibrous amphibole of up to 7%
compared to less than 1% at Enoree (arrows in Figures 2A and 2B).  The relative amounts of
accessory minerals in beneficiated ore may therefore provide a signature for at least one point in
time for ore from a particular site.  Accessory minerals can change significantly with location
within a deposit, however, leading to variation in the ore as different parts of a mine are
developed.  Furthermore, dilution with minerals from other material in a mixed consumer
product may mask such an accessory mineral signature.

Chemical Composition

                               Major and Minor Elements

       Table 4 lists oxide and element concentrations for vermiculite ore minerals (vermiculite,
mixed-layer mica/vermiculite, and mica) from deposits that serve the U.S. market. Values for
oxide weight percent in Table 4A are taken from the literature; element concentrations in Table
4B are recalculated from the oxide values. Sources of data include vermiculite, hydrobiotite, and
biotite in mine samples from Libby, (Boettcher, 1966); vermiculite, hydrobiotite, and biotite in
mine samples from several Enoree deposits (Libby, 1975); vermiculite, hydrobiotite, and
phlogopite in orebody samples from Palabora (Schoeman, 1989);  hydrophlogopite and
phlogopite in mine samples from Tseganbrark, Xinjiang (Rongqi and Junchen, 1993);
vermiculite concentrates from Libby (Bassett, 1959); and vermiculite concentrate from Louisa,
Enoree, Palabora, and Qieganbulake, Xinjiang (Hindman, 1984).

       The limited data set suggests that some gross discrimination  among biotite-derived versus
phlogopite-derived deposits may be made on the basis of major elements associated with the
phlogopite-biotite solid solution series. For example, Figure 3 shows the relationship between
the ratio of magnesium to magnesium plus iron [Mg/(Mg+Fe)] versus ferric iron concentration.
Ferric iron allows discrimination of the parent mica minerals, biotite (EB and LB in Figure 3)
and phlogopite (XP and PP) from the more oxidized vermiculite and mixed-layer
mica/vermiculite (uncircled symbols). The Mg/(Mg+Fe) ratio in turn allows some discrimination
of the more magnesium-rich phlogopite-derived vermiculite from Palabora and Xinjiang (dashed
tie-lines on the right side of Figure 3) from the biotite-derived material at Libby, Louisa, and
Enoree (dotted tie-lines).  Of commercial ore deposits that serve the U.S., only the imported
deposits are largely phlogopite-derived, and hence more enriched in magnesium. These data,
therefore, suggest a potentially useful method for discriminating the more magnesium-rich,
imported materials from domestic sources.

       Not all of this sparse data set would support a general rule for discriminating sources on
the basis of major elements, however. For example, one sample of Enoree vermiculite (EV)
                                           15

-------
plots in the magnesium-poor, ferric iron-rich region of Figure 3 as would be expected
considering the biotite parent mica, but another sample of Enoree vermiculite plots in the
magnesium-rich, ferric iron-poor region more suitable for a phlogopite parent mica. Likewise,
the few data points for the mixed-layer mica/vermiculite samples do not all support a general rule
on discriminating imports.  For example, the only chemical data found for a Palabora mixed-
layer mica/vermiculite were for a hydrobiotite (PH in Figure 3) rather than for hydrophlogopite
which is described in the literature as the dominant mixed-layer mica/vermiculite at that mine
(Schoeman, 1989). Therefore, the utility of major-element relationships among deposits would
need more extensive samples and better documentation of samples and analytical methods to
support a consistency in patterns.

       Data for minor elements may also show potentially diagnostic differences among the
major deposits. For example, inspection of Table 4 shows that Palabora vermiculite has
relatively elevated values for fluoride, whereas Enoree vermiculite may be associated with
relatively high aluminum and chromium. Louisa vermiculite has high sodium, and Louisa and
Libby vermiculite have high potassium.  As with major elements, the significance of minor
element relationships for vermiculite among the different deposits cannot be fully determined
from these few data, but the patterns suggest that additional analyses  could be informative.

                                     Trace Elements

       Although major and minor element analyses of vermiculite have  been found for some of
the ore deposits, to date little information has been found on the trace element content.  In
general, vermiculite has a high ion exchange capacity, perhaps the highest among layer-silicates,
and a mineral structure suitable for sequestering trace elements (McBride, 1994). At least two
features of the environment of vermiculite deposits foster the occurrence of elevated trace metals.
First, the ultramafic character of the geologic setting in which vermiculite deposits form is
commonly associated with a distinctive set of elevated trace elements including titanium,
chromium,  nickel, and cobalt.  Chromium is shown in Table 4B to be elevated in vermiculite at
levels greater than  1000 ppm in one or more samples from all of the deposits which have data,
with Enoree vermiculite particularly high in chromium (3400 ppm). Maybin and others (1980)
found that the South Carolina vermiculite deposits have associated stream sediment and soil
anomalies of chromium,  nickel  and phosphorous and possibly cobalt  and copper that are high
enough to have possible exploration potential as indicators of vermiculite ore. Murdock and
Hunter (1946) reported that North  Carolina vermiculite deposits are enriched in nickel  and even
have nickel silicate minerals associated with them. El Shazly and others (1975a) noted nickel,
chromium,  and cobalt enrichment in Egyptian vermiculite deposits.

       The second trace element feature of potential significance is that vermiculite has been
found to have the capability to preferentially concentrate the group of trace elements known as
rare earth elements (Pastor and others, 1988). The rare earth elements, including in particular the
lanthanide series from atomic number 57 (lanthanum) to 71 (lutetium), with the addition of
elements 21 (scandium) and 39 (yttrium) have been found in geochemical studies to be
                                           16

-------
particularly useful in fingerprinting applications.  A useful characteristic of rare earth elements is
their limited reactivity under conditions at the earth's surface, so that once incorporated into a
deposit they are likely to remain present and retain similar relative proportions during weathering
processes.

       Therefore, the abundance of rare earth and other elements that tend to be less mobile at
ambient temperatures may potentially provide a diagnostic signature for vermiculite ore deposits.
Fortuitously, carbonatites which are associated with the Palabora deposit, and other alkaline
rocks that are in general of the type associated with other vermiculite deposits, are commonly
elevated in rare earth elements (Korotev, 1996).  Indeed for Palabora, Evans (1993) reported that
rare earth elements are sufficiently abundant as to be considered potentially extractable.

                                     Isotope Ratios

       No data have been found on stable or radiogenic isotopes for vermiculite deposits.
Isotopes have been found to be of use in fingerprinting geologic deposits and may have
application with vermiculite ores.  No further evaluation of isotopes has been made for the
purposes of this report.

Likelihood of Deposits being Chemically Distinctive

       Several lines of reasoning argue for vermiculite deposits from widespread locations being
chemically distinctive.  While the association of vermiculite with ultramafic rocks is a common
characteristic that sets the local geologic environment apart from the surrounding regional rock,
the deposits  differ in detail because of the regional influence. The rock type surrounding each
vermiculite orebody is  one important regional variable, as is the composition of crosscutting
intrusions and the resulting alteration.

       As the ultramafic magmas move upward through the earth's crust, many variables can
influence their composition and mineralogy.  Potential influences include the composition of the
rocks the magmas move through and the degree to which those are assimilated, the sequence and
timing of associated fluids of various compositions, and the pressure and temperature conditions
under which the rocks crystallize.

Limitations of the Data

       While it seems more likely that vermiculite deposits are chemically distinctive rather than
identical, the differences could be difficult to detect.  Little is known at present about the amount
of variability within the deposits.  Without this knowledge, it is difficult to determine the number
of samples necessary to measure the differences among deposits with enough confidence to
provide a predictive tool.   Some deposits, such as Libby and Palabora, are zoned on a large scale,
but smaller scale differences that would affect trace element chemistry may also exist.
Additionally, efforts to link vermiculite consumer products with particular ore deposits may be


                                            17

-------
stymied by the practice of exfoliation plants or distributors mixing ores from different locations
and thereby masking or confounding any possibly unique chemical signature.

       The limitations suggest that a phased approach to tracer identification is appropriate,
whereby an initial study is designed to verify that gross compositional differences occur between
deposits. A subsequent more detailed study would determine whether within-deposit variability
masks or exceeds between-deposit variability.

RECOMMENDATIONS FOR FURTHER STUDY

       This literature review indicates that unique signatures for different vermiculite deposits
are not unreasonable, considering the distinctive geologic environments in which vermiculite
deposits are formed as well as the geochemical reactivity of vermiculite.  Mineralogical and
geochemical data are sparse but suggest that a set of chemical signatures may be found,
especially among trace element content of either vermiculite minerals or the bulk ore.
Furthermore, limited mobility for some of the more likely elevated trace elements suggests that
their signature may be retained through the vermiculite milling and exfoliation process and
ultimately remain measurable in consumer products.

       The mineralogy of a deposit and the major element composition may also have a unique
character, though possibly not as distinctive as the trace element content. Nonetheless, any
fingerprinting effort should include mineralogy and major element analysis in order to provide a
framework in which to place and interpret trace element data.

Analytical Methods

       Analytical methods for a tracer analysis should include both mineralogical and chemical
methods that can be applied to bulk samples as well as mineral grains.  Mineral grains can be
analyzed as either concentrates derived from the bulk samples or as individual  particles.  The
reason for emphasis on mineral grains is that vermiculite in some consumer products would be
highly diluted by material from sources other than the vermiculite ore deposit.  Conceivably, the
vermiculite grains themselves, or a particularly abundant accessory mineral such as amphibole,
would hold the key to geochemical discrimination.

       Table 5 lists a limited selection of methods potentially useful in tracer analysis. A variety
of other methods exist, some of which are pointed out below as also particularly suited to the
problem. The limited list in Table 5 is not meant to be exclusive, but serves to highlight readily
available methods capable of yielding a low-cost approach. Although some methods provide
mineral identification and compositional information with the same instrument, generally no
single instrument is sufficient for mineral characterization. Therefore a complement of different
instrumental techniques would be most appropriate allowing both low and high magnification,
and measurement of optical properties, structure, and composition.
                                           18

-------
       A selective set of methods (Table 5) includes optical microscopy (OM) consisting of both
stereomicroscopy for sample description and polarized light microscopy for mineral
identification, quantitation and textural description; X-ray diffraction (XRD) for mineral
identification and qualitative or quantitative abundance; scanning electron microscopy with
energy dispersive X-ray spectroscopy (SEM/EDS) for mineral identification, texture, and semi-
quantitative chemical composition;  electron probe microanalysis with wavelength dispersive X-
ray spectroscopy (EPMA) as well as energy dispersive X-ray spectroscopy for mineral
identification, texture, and quantitative chemical analysis; and transmission electron microscopy
with energy dispersive X-ray spectroscopy and electron diffraction (TEM/EDS/ED) for
identification and quantitative chemical composition of fine-grained fibrous minerals.

       Two additional microanalytical techniques not listed in Table 5, but which can achieve
low detection limits for rare earth elements in solid samples should be considered; these are
secondary ion mass spectrometry (SIMS) and laser ablation - mass spectrometry (LA-ICPMS)
(Williams, 1996, p. 344). A rapidly developing method, also not listed on Table 5, that has the
advantage of discriminating potentially unique characteristics in bulk samples without having to
conduct microanalysis on every sample is high resolution reflectance spectroscopy (Clark, 1999).
The method can be applied with lab- or field-based spectrometers or airborne imaging
spectrometers.  The method relies on identification of distinctive absorption spectra, particularly
in the visible and near-infrared part of the energy spectrum. By comparison and calibration with
samples previously evaluated by other mineralogical methods, reflectance spectra can be used to
characterize unknown material and  should be considered as a potentially efficient and quick
method for fingerprinting either bulk samples or components of vermiculite ore.

       Several  methods listed on Table 5  are available for trace element analysis of bulk
material, either as ore, mineral concentrates, or consumer products.  The most suitable multi-
element methods (Table 5) that are  commonly used for bulk solid samples include inductively
coupled plasma atomic emission spectroscopy and mass spectrometry (ICP-AES, ICP-MS), X-
ray fluorescence spectroscopy (XRF), and instrumental neutron activation analysis (INAA). A
combination of analytical requirements, such as low-level detection of trace metals including rare
earth elements,  non-destructive analysis of whole solid samples, minimal specimen preparation,
and a throughput of many samples could perhaps be best met with INAA techniques (Marfunin,
1995; Williams, 1996) provided costs are  competitive with the other techniques. A disadvantage
of INAA would be the inability to analyze some potentially important elements including some
of the major species. An advantage of ICP methods over INAA is a much more complete list of
available analytes, whereas a disadvantage is the preparation requirement for whole-solid
analysis to completely dissolve the  specimen. ICP methods using non-whole rock acid digestion
preparation could be adapted for tracer analysis, provided a preliminary sensitivity study is
conducted to establish minerals left unanalyzed by incomplete acid digestion. Traditional whole-
rock XRF  analysis of thick specimens also has an advantage of a broad analyte list but
preparation requires pulverization and homogenization of the sample and possibly fusion prior to
analysis. Thin-film XRF analysis is a technique developed for air-filter samples that could be
adapted to other granular material and avoid destructive dissolution preparation required by
                                           19

-------
traditional XRF analysis. In all techniques the detection limits will vary depending on the
analyte of interest.

Study Design

       The lack of information on compositional variability within vermiculite deposits as well
as among different deposits argues for a phased approach in search of useful tracers. Initially, a
first-phase screening study should be undertaken to confirm or further evaluate some of the
patterns among vermiculite deposits that have been described in this report based on the
literature. Some of the assumptions on trace element concentrations, particularly that of elevated
chromium, nickel, and cobalt, and the rare earth elements, also need elaboration.

       The goal of the first phase would be to verify on the basis of a small set of samples that
differences in potentially diagnostic mineralogical or geochemical characteristics among sites are
measurable.  The  first phase would analyze samples from known sources, recognizing that the
small number of samples will not fully characterize the variability within any particular source.

       The goal of the second phase would be to verify on the basis of a large set of samples that
the spatial variability of a potential  fingerprint within a source is less than the variability among
sources.  The second phase would require a statistically significant set of samples that would
more fully characterize variability within and among ore deposits and their commercial products.

       The design for conducting the second phase would be based on evaluation of variability
of first-phase data. If the first-phase variability among all samples is minimal, the study would
end.  If there are apparent differences among samples, an evaluation would be attempted to
determine whether differences could be attributed to different sources. If variability can be
attributed more to different sources than to differences within a source, then a more robust
second phase would be undertaken.

       Recommended objectives for a screening study are to  examine a limited set of samples
(approximately 35) that would include the following (Figure 4):

       - Beneficiated ore from up to three batches each from  at least five major deposits that are
       spatially distant from each other. A batch of beneficiated ore refers here to material
       produced during a discrete time period.  Major deposits are those in ore districts that
       supply products for use in the U.S. Spatially distant deposits are those in  different ore
       districts.

       - Exfoliated products known to have originated from the sampled ore deposits.

       - A few consumer products made from vermiculite originating from sampled ore deposits.

       - One site  each for beneficiated ore and exfoliated product should be duplicated to
                                           20

-------
       estimate the combination of sampling and analytical variability.

       - Consumer products should be analyzed both as bulk samples and as separated
       vermiculite concentrate.

       Sample volumes for most samples should be sufficient for both the screening analysis and
for a detailed second-phase study, if undertaken, in order to minimize repetitive field sampling.
Each batch of beneficiated ore should be derived from different mixes of raw ore from spatially
different parts of the mine.  Where a variety of mixes for beneficiated ore are not available for an
individual mine, raw ore samples should be collected to allow lab generation of beneficiated ore
using different mixes of raw ore. The purpose of multiple mixes is to account for some degree of
within-mine variability in the results.

       Recommended analyses for the screening effort should include mineral identification and
qualitative estimate of abundance for major, minor and trace phases, and chemical analysis of
major, minor and trace elements for both bulk material and for separated vermiculite
concentrates. As indicated in Figure 4, mineral identification should proceed from optical
microscopy to X-ray diffraction, with feedback of information between the two methods to
characterize the mineral content. Mineral separations by size, shape, density, or magnetic
methods will aid in identification as well as estimation of abundance.

       Once the general mineral content is known, electron microprobe or scanning electron
microscope methods should be used to verify mineral species by microchemical and textural
analysis of individual grains (Figure 4). Transmission electron microscope analysis could be
used to more fully characterize fibrous minerals, but only where such minerals are initially
evident from the other observations. Chemical characterization of bulk material should begin
after the preliminary identification of the mineral content by microscopy and X-ray diffraction
(Figure 4), in order to allow results on the mineral content to be used as a guide in setting up
element-specific data objectives. Major and trace element analysis by X-ray fluorescence and
additional trace element analysis by instrumental neutron activation would be a suitable
approach.  ICP methods could be considered if they were accompanied by a separate study to
determine the comparative influence of digestion procedures relative to whole-sample
composition. Additional innovative microanalytical and reflectance spectroscopic methods
should also be considered depending on availability through cooperating investigators.

       For the full fingerprinting effort, recommended analyses should include mineral
identification and quantitative abundance, and chemical composition of both bulk samples and
vermiculite grains or other selected minerals such as amphiboles.  Results of the screening effort
should provide guidance on how to structure the approach toward full fingerprinting, such as
selection of sample preparation, additional analytical methods, target minerals, and target
elements.
                                           21

-------
CONCLUSIONS

       A review of the literature indicates that many, if not most, commercial vermiculite
deposits contain asbestiform accessory minerals.  The amount of asbestiform minerals appears to
vary substantially from source to source, though documentation in the open literature remains far
from adequate to fully evaluate asbestos content at mines.  Furthermore, the efficiency of any
asbestos removal that might occur during the course of ore beneficiation or other mineral
processing activities likely varies from source to source. Without knowing the characteristics of
the vermiculite source, the risk from asbestiform minerals in a vermiculite product can only be
determined by relatively expensive asbestos analysis of each product. To help anticipate
potential risk in vermiculite products, a low-cost mineralogical or geochemical means of
determining the source of vermiculite would be helpful as a tracer to identify those products
more likely to contain asbestiform minerals.

       A two-phase approach toward identifying  suitable mineralogical or geochemical tracers is
recommended using the general study design outlined in this report. The approach
accommodates consideration that the success of a tracer study would depend on being able to
measure variations in mineralogical or geochemical characteristics  among different ore deposits
beyond those variations that normally occur within each ore deposit.  A first phase is
recommended to examine a limited set of samples to verify that enough variability exists among
ore deposits to measure differences in diagnostic minerals or elements.  If the first phase
successfully identifies potential tracers, a second-phase detailed study should use a large set of
samples to determine whether tracer variability among ore deposits is greater than variability
within a deposit. Both phases should examine the degree to which  ore deposit characteristics
remain identifiable in samples of vermiculite products.
                                           22

-------
REFERENCES

Atkinson, G.R., Rose, D., Thomas, K., Jones, D., Chatfield, E.J., and Going, I.E., 1982,
       Collection, analysis, and characterization of vermiculite samples for fiber content and
       asbestos contamination: Midwest Research Institute (MRI), Task 32, Final Report
       prepared for U.S. Environmental Protection Agency, Project 4901-A32 under EPA
       Contract No. 68-01-5915, Washington, D.C., EPA 0717, 69 p., 5 app. [also cited as MRI
       (1982)]

Bassett, W.A., 1959, The origin of the vermiculite deposit at Libby, Montana:  American
       Mineralogist, v. 44, p. 282-298.

Behre, Dolbear & Co., 2000, Health risk summary report for the Dillon, Montana, vermiculite
       mine and plant:  unpublished report prepared for Stansbury Holdings Corporation, Behre
       Dolbear Project JOO-034, June 2000, 43 p.

Berg, R. B., 1995, Geology of the Elk Creek vermiculite deposit, Madison and Beaverhead
       Counties, Montana:  Montana Bureau of Mines and Geology Open-File Report, MBMG
       335, 9 p.

Berg, R.B., 1997, Montana's changing industrial minerals industry in Jones, R.W., and Harris,
       W. E., eds., Annual Forum on the Geology of Industrial Minerals, 32nd, Laramie, 1996,
       Proceedings: Wyoming State Geological Survey, Public Information Circular, no. 38, p.
       167-175.

Boettcher, A. L., 1966, The Rainy Creek Igneous Complex near Libby,  Montana: University
       Park, Pennsylvania, Pennsylvania State University, unpublished Ph.D. thesis, 155 p.

Boettcher, A. L., 1968, Origin of vermiculite and hydrobiotite near Libby, Montana [abs.]:
       Abstracts for 1966, Geological Society of America Special Paper, n. 101, p. 20-21.

Brindley, G. W., and Brown, G., eds.,  1980, Crystal structure of clay minerals and their X-ray
       identification: London,  Mineralogical Society, Monograph, 495 p.

Brindley, G. W., Zalba, P. E., and Bethke, C. M., 1983, Hydrobiotite, a  regular 1:1
       interstratification of biotite and vermiculite layers: American Mineralogist, v. 68, p. 420-
       425.

Bush, A. L., 1976, Vermiculite  in the United States in Montana Bureau  of Mines and Geology,
       Eleventh Industrial Minerals Forum:  Montana Bureau of Mines and Geology, Special
       Publication 74, p. 146-155.

Chatfield, E. J., and Lewis, G.M., 1979, Examination of vermiculite for the presence of asbestos


                                          23

-------
       fibers:  Mississauga, Ontario, Canada, Ontario Research Foundation Report No. 22056-2,
       22 p., 2 app.

Chatfield, E. J., and Lewis, G.M., 1980, Development and application of an analytical technique
       for measurement of asbestos fibers in vermiculite:  Scanning Electron Microscopy, p.
       329-340.

Clark, R.N., 1999, Spectroscopy of rocks and minerals, and principles of spectroscopy in Rencz,
       A.N. and Ryerson, R.A., eds., Manual of remote sensing, volume 3, Remote sensing for
       the earth: New York, John Wiley and Sons.

de la Calle, C.  and Suquet, H., 1988, Vermiculite, chapter 12 in Bailey S. W., ed., Hydrous
       phyllosilicates: Mineralogical Society of America, Reviews in Mineralogy, v. 19, p. 455-
       496.

Deer, W.A., Howie, R.A., and Zussman, J.,  1997, Rock-forming minerals, Volume 2B, Double-
       chain silicates (second edition): London, The Geological Society, 764 p.

Desmarais, N.R., 1976, Structural and petrologic study of Precambrian ultramafic rocks, Ruby
       Range, southwestern Montana:  Missoula, Montana, University of Montana, Missoula,
       unpublished MS thesis, 88 p.

El Shazly, E. M., Saleeb-Roufaiel,  G. S., andRasmy, A. H., 1975a, Geology, petrogenesis, and
       mode of formation of asbestos-vermiculite deposits at Hafafit, Egypt:  Egyptian Journal
       of Geology, v.19, no.2, p. 87-104.

El Shazly, E. M., Saleeb-Roufaiel,  G. S., El Ramly, M. F., and Rasmy, A. H., 1975b,
       Mineralogical study of anthophyllite, phlogopite, and vermiculite from Hafafit, Egypt:
       Egyptian Journal of Geology, v. 19,  no.2, p. 105-124.

Evans, A.M., 1993, Ore geology and industrial minerals, An introduction, (third ed.): Oxford,
       Blackwell Scientific Publications, 389 p.

Gaines, R.V., Skinner, H.C.W., Foord, E.E., Mason, B., Rosenzweig, A., King, V.T., and Dowty,
       E., 1997, Dana's new mineralogy: New York, John Wiley and Sons.

Gooch, E.O., 1957, Vermiculite: Virginia Minerals, v. 3, no.l, 5p.

Guthrie, G.D. and Mossman, B.T., eds, 1993, Health effects of mineral dusts:  Mineralogical
       Society of America, Reviews in Mineralogy, v. 28, 584 p.

Hessel, P. A. and Sluis-Cremer, G. K., 1989, X-ray findings, lung  function, and respiratory
       symptoms in black South African vermiculite workers:  American Journal of Industrial
                                           24

-------
       Medicine, v. 15, p. 21-29.

Hindman, J. R., 1994, Vermiculite, in Carr, D.D., ed., Industrial minerals and rocks, (sixth ed.):
       Littleton, Co., Society of Mining, Metallurgy, and Exploration, p. 1103-1111.

Hunter, C.E., 1950, Vermiculite of the southeastern states, in Snyder, F.G., ed., Symposium on
       Mineral Resources of the Southeastern United States: Knoxville, Tn., University of
       Tennessee Press, p. 120-127.

Institute of Occupational Medicine (IOM), 2000, Sampling and analysis of crude vermiculite
       samples for possible asbestiform content: unpublished report prepared for Mandoval
       Limited, March 29, 2000, 14 p.

Jackson, J.A., ed., 1997,  Glossary of geology, (fourth edition):  Alexandria, Virginia, American
       Geological Institute, 769 p.

Korotev, R.L., 1996, A self-consistent compilation of elemental concentration data for 93
       geochemical reference samples: Geostandards Newsletter, v. 20, p. 217-245.

Kulp, J. L, and Brobst, D.A., 1954, Notes on the dunite and the geochemistry of vermiculite at
       the Day Book dunite deposit, Yancey County, North Carolina: Economic Geology, v. 49,
       p. 211-220.

Larsen, E.S., 1942, Alkalic rocks of Iron Hill, Gunnison County, Colorado:  U.S. Geological
       Survey Professional Paper  197-A. 64 p.

Leake, Bernard E.(chairman), Wooley, A.R., Arps, C.E.S., Birch, W.D., Gilbert, M.C., Grice,
       J.D., Hawthorne, F.C., Kato, A.,  Kisch, H.J., Krivovichev, V.G., Linthout, K., Laird, J.,
       Mandarine, J., Maresch, W.V., Nickel, E.H., Rock, N.M.S.,  Schumacher, J.C., Smith,
       D.C., Stephenson, N.C.N.,  Ungaretti, L., Whittaker, E.J.W.,  and Youzhi, G., 1997,
       Nomenclature of amphiboles: report of the subcommittee on amphiboles of the
       International Mineralogical Association Commission on New Minerals and Mineral
       Names: Mineralogical Magazine, v. 61, p. 295-321.

Libby, S. C., 1975, The origin of potassic ultramafic rocks in the Enoree "Vermiculite" District,
       South Carolina: University Park, Pennsylvania, Pennsylvania State University,
       unpublished thesis, 116 p.

Marfunin, A.S., ed., 1995, Methods and instrumentations: results  and recent developments:
       Berlin, Springer-Verlag, Advanced Mineralogy, v. 2, 441 p.

Maybin, A.H. Ill, and Carpenter, R. H, 1990, Geochemistry: a new approach for vermiculite
       exploration in South Carolina, in Zupan, Alan-Jon and Maybin, A.H.  Ill, eds., Forum on
                                           25

-------
       the Geology of Industrial Minerals, 24th, Greenville, 1988, Proceedings:  Columbia, South
       Carolina Geological Survey, p.57-69.

McBride, M. B., 1994, Environmental chemistry of soils: New York, Oxford University Press,
       406 p.

Meisinger, A.C., 1979, Vermiculite in Minerals yearbook, 1978-79, metals and minerals: Bureau
       of Mines, v. 1, p. 977-980.

MRI (Midwest Research Institute), 1982, [see Atkinson and others (1982)].

Moatamed, Farhad, Lockey, J. E., and Parry, W.T., 1986, Fiber contamination of vermiculites: a
       potential occupational and environmental health hazard:  Environmental Research, v. 41,
       p 207-218.

Montana Department of Environmental Quality (MDEQ), 1999, Final environmental assessment
       for Dillon vermiculite application for operating permit: Helena, Montana Department of
       Environmental Quality, and Dillon, Bureau of Land Management, 27 p.

Moeller, E., 2000, Vermiculite: Mining Engineering, p. 66-67'.

Murdock, T. C., and Hunter, C. E., 1946, The vermiculite deposits of North Carolina: Raleigh,
       North Carolina Division of Conservation and Development, Division of Mineral
       Resources, 44p.

Palabora Mining Company Limited Mine Geological and Mineralogical Staff, 1976, The geology
       and the economic deposits of copper, iron, and vermiculite in the Palabora Igneous
       Complex: a brief review:  Economic Geology, v. 71, p. 177-192.

Pastor, P. O. , Rodriquez-Castellon, Enrique, and Garcia, A. R.,  1988, Uptake of lanthanides by
       vermiculite:  Clays and Clay Minerals, v. 36,  no. 1, p. 68-72.

Potter, M.J., 1998, Vermiculite in Minerals yearbook, metals and minerals: U.S. Geological
       Survey, v. 1, p.

Potter, M.J., 2000, Vermiculite in Minerals commodity summaries: U.S. Geological Survey,
       February 2000, p. 186-187.

Rohl, A.N., and Langer, A.M., 1977, Mineral analysis of core samples from the Green Springs
       area, Virginia vermiculite deposit:  unpublished  letter report from Mt Sinai School of
       Medicine, 10 p.

Rongqi, Xu and Junchen, Cao, 1993, A mineral ogical study of vermiculite from Tseganbrark,
                                          26

-------
       Xinjiang: Acta Mineralogica Sinica, v. 13, n. 1, p. 37-45 [in Chinese, abstract in
       English].

Roy, Rustum and Romo, L.A., 1957, Weathering studies. 1. New data on vermiculite: Journal of
       Geology, v. 65, n. 6, p. 603-610.

Schoeman, J.J., 1989, Mica and vermiculite in South Africa:  Journal of South African Institute
       of Mining and Metallurgy, v. 89, no.  1, p. 1-12.

Skinner, H.C.W., Ross, Malcolm, and Frondel, Clifford, 1988, Asbestos and other fibrous
       minerals: New York, Oxford University Press, 197+p.

Tongjiang, Peng, Pu, Wan, Zhaolu, Pan, and Jianhong, Zhang, 1996a, A study of interstratified
       minerals of industrial vermiculite from Weili vermiculite mine, Xinjiang, China [abs.]:
       International Geological Congress, 30th, Beijing, China, 1996, Abstracts, v. 2, p.  446.

Tongjiang, Peng, Pu, Wan, Zhaolu, Pan, and Jianhong, Zhang, 1996b, A study of the
       interstratified structure of phlogopite and vermiculite from Weili mine, Xinjiang, China:
       Acta Petrologica et Mineralogica, v. 15, n. 3, tot. 59, p. 250-258 [in Chinese, abstract in
       English].

U.S. Environmental Protection Agency (USEPA), 1980, Priority review level 1— Asbestos-
       contaminated vermiculite: Washington, D.C., USEPA, Office of Toxic Substances, 39 p.

U.S. Environmental Protection Agency (USEPA), 2000, Sampling and analysis of consumer
       garden products that contain vermiculite:  EPA 744-R-00-010.

Vorovikov, P.P., 1973, Genetic types, conditions of formation and economic evaluation of
       vermiculite deposits, in Daragan, V. K., ed., Perlite and vermiculite (Geology, exploration
       and production technology): translated from Russian and published for the U.S.
       Department of the Interior, Geological Survey, and the National Science Foundation,
       Washington, D.C., by the Indian National Scientific Documentation Centre, New Delhi,
       p. 139-176. [Available in libraries of the U.S. Geological Survey.]

Williams, C.T., 1996, Analysis of rare earth minerals, Chapter 13 in Jones, A. P., Wall, Frances,
       and Williams, C. T., eds., Rare earth minerals, chemistry, origin and ore  deposits:
       London, Chapman and Hall, p. 326-348.

Wylie, A.G., and Verkouteren, J. R., 2000, Amphibole asbestos from Libby, Montana: aspects of
       nomenclature:  American Mineralogist, v. 85, p. 1540-1542.
                                           27

-------
GLOSSARY

The definitions of terms as used in this report are derived, where noted, from the numbered
references. See also Table 1 for mineral formulae.

[ 1 ]    Gaines and others (1997)
[2]    Guthrie and Mossman (1993)
[3]    Jackson (1997)
[4]    Leake and others (1997)
[5]    Skinner, Ross and Frondel (1988)

Acicular - "Said of a crystal that is needle-like in form." [2]

Actinolite - "An amphibole with the ideal composition  Ca2(Mg,Fe2+)5Si8O22(OH)2.  Actinolite is
       a species in the Mg-Fe2+ series, tremolite - ferro-actinolite, with 0.9 > Mg/(Mg+Fe2+) >
       0.5." [2]  Actinolite can occur with different habits, one of which can be asbestiform.

Amphibole - A group of minerals that are hydroxylated chain silicates with Mg, Fe, Ca, and Na
       as the dominant cation species. These, together with other cations, substitute at specific
       sites in the crystal structure producing a group of minerals closely related
       crystallographically and chemically.  Fibrous varieties of minerals in the amphibole group
       occur.  Not all members of the amphibole group occur in fibrous form, and those that do
       may also be found in other habits. [5]

Asbestiform - "An adjective describing inorganic materials that possess the form and appearance
       of asbestos. Asbestiform is a subset of fibrous, where asbestiform implies relatively
       small fiber thickness and large fiber length, flexibility, easy separability, and a parallel
       arrangement of the fibers in native (unprocessed) samples. Often, asbestos fibers occur in
       bundles, i.e., they are often polyfilamentous." [2]

Asbestos -  "A term applied to asbestiform varieties of serpentine  and amphibole, particularly
       chrysotile, 'crocidolite' [asbestiform  riebeckite], 'amosite' [asbestiform grunerite],
       asbestiform tremolite, asbestiform actinolite, and asbestiform anthophyllite. The asbestos
       minerals possess asbestiform characteristics." [2]

Aspect Ratio - "The ratio of length to width." [2]

Beneficiated Ore  - Ore that has been processed in a mill to concentrate the material of
       commercial interest.  In the case of vermiculite, beneficiated  ore has had most but not
       necessarily all  of the accessory minerals (gangue) removed.  See exfoliated ore.

Beneficiation - "Improvement of the grade of ore by milling, flotation,  sintering, gravity
       concentration,  or other processes. The resultant product is a concentrate." [3]


                                            28

-------
Biotite - "A series of 2:1 layer silicates of ideal composition K(Mg,Fe)3Si4O10(OH)2. Phlogopite
       is the magnesium end-member of the series; annite is the iron end-member." [2]  In this
       report, biotite refers to a mineral intermediate in the annite-phlogopite series, listed as
       K(Fe2+,Mg)3AlSi3O10(OH,F)2 in Table 1. "The biotite composition field is arbitrarily
       defined as that having Mg:Fe ratios between 2:1 and 1:4, but the division is obscured by
       substitution of other cations in the octahedral as well as tetrahedral sites." [1]

Carbonatite - "An igneous rock composed of at least 50% carbonate minerals." [3]

Chain Silicate - A mineral class consisting of silica tetrahedra linked in one direction.
       Amphiboles, pyroxenes, and pyroxenoids are included in this class. [2]

Chlorite - A group of platy sheet silicates with the general formula (R+2R+3)6AlSi3O10(OH)8,
       where R is usually  Fe or Mg. [3]

Chrysotile - A hydrated magnesium  silicate mineral that is a member of the serpentine group and
       is usually asbestiform. [2,4]

Cross Fiber - Aggregate of asbestos in a vein in which parallel fibers or bundles of fibers are
       oriented perpendicular to the margins of the vein.

Equant - "Said of a crystal having the same or nearly the same dimensions in all directions." [2]

Exfoliated Ore - Vermiculite ore that has undergone thermal, or less commonly chemical,
       processing to expand the platy layers  like an accordion.

Fiber - A mineral with a highly elongate morphology developed during growth. [2] "A long, thin
       thread or threadlike solid with distinctive elongate shape that may be natural or synthetic
       and organic or inorganic in composition." [5]

Fibrous - Said of a mineral that gives the appearance of being composed of fibers. [2]
       "Aggregates of any size of individual fibers may form relatively thick fibrous bundles,
       thus becoming visible to the naked eye." [5]

Framework Silicate - A mineral class having silicate tetrahedra polymerized in three dimensions.
       [2]  Quartz and feldspar are among minerals included in this class.

Gangue - "The valueless rock or mineral aggregates in an ore; that part of an ore that is not
       economically desirable but cannot be avoided in mining.  It is separated from the ore
       minerals during concentration." [1]

Habit - "The shape or morphology that a crystal or aggregate of crystals assumes during
       crystallization." [2]  Examples of habit are equant, prismatic, acicular, fibrous, and
                                            29

-------
       asbestiform.

Head Feed - Ore that has been mined and is at the point of entering the processing system in a
       mill.

Hydrobiotite - A sheet silicate mineral that is a regular mixed-layer mica/vermiculite. "A
       regularly ordered, 1:1 mix of biotite/vermiculite." [3]

Hydrophlogopite - A sheet silicate that is a regular mixed-layer mica/vermiculite. A regularly
       ordered, 1:1 mix of phlogopite/vermiculite, by analogy with hydrobiotite.

Layer Silicate - A sheet silicate mineral class having silica tetrahedra polymerized in two
       dimensions. [2]  Micas, vermiculite, mixed-layer mica/vermiculite, and the chlorite
       smectite, and serpentine groups are among minerals included in this class.

Mass Fiber - Aggregate of asbestos in which fibers or bundles of fibers have random orientation.

Mica - A group of sheet silicate minerals that have an ideal electrical charge of-1 per formula
       unit. [2]  The group includes biotite, phlogopite, muscovite,  and many others. Mica is
       the most common precursory or parent mineral that alters to form vermiculite.

Mixed-layer mica/vermiculite - A group of sheet silicate minerals that have a mica component
       and a vermiculite component mixed in either a regular or random interstratification of
       layers.

Ore - The naturally occurring material from which a mineral or minerals of economic value can
       be extracted at a reasonable profit. [3]

Ore Feed - Ore that has been mined, transported to a mill site, and manipulated perhaps by size
       reduction in preparation for beneficiation; ore just prior to entering a mill for processing.

Phlogopite - "A 2:1 layer silicate with an ideal composition KMg3(Si3 A1)O10(OH)2.  Phlogopite
       is the magnesium end-member of the biotite series." [2]  Phlogopite, listed as
       K(Fe2+,Mg)3AlSi3O10(OH,F)2 in Table 1, has relatively less iron than biotite. "Members
       of the [biotite] series with Mg:Fe ratios of > 2:1 are generally regarded as phlogopites, but
       the boundary is arbitrary largely because substitutions other than Fe for Mg may also
       occur." [1]

Prismatic  - "A term used to describe crystals exhibiting aspect ratios greater than one and having
       parallel sides." [2]

Raw Ore - Unprocessed ore.
                                            30

-------
Richterite - An amphibole mineral with an end-member composition Na(Ca,Na)Mg5Si8O22(OH)2.
       [4] Richterite can occur with different habits, one of which can be asbestiform.

Serpentine - A group of common rock-forming minerals that are always derived from alteration
       of magnesium-rich silicates, especially olivine; including minerals such as antigorite,
       lizardite, and chrysotile. [3]

Serpentinite - "A rock consisting almost wholly of serpentine-group minerals." [3]

Sheet Silicate - A platy mineral also called a layer silicate. A mineral class having silica
       tetrahedra polymerized in two dimensions.  Micas, the chlorite group, vermiculite, and
       the serpentine group are among minerals included in this class.

Tremolite - "A species of amphibole with the ideal composition Ca2(Mg,Fe2+)5Si8O22(OH)2.
       Tremolite  is the magnesium-rich end member of the Mg-Fe2+ series, tremolite - ferro-
       actinolite,  with Mg/(Mg+Fe2+) > 0.9." [2]  Tremolite can occur with different habits,
       one of which can be asbestiform.

Ultramafic - An igneous rock composed chiefly of mafic minerals, such as hypersthene, augite,
       or olivine. [3]

Vermiculite - A sheet silicate with a general formula:
       Mg035+/.(Mg,Fe3+)3Si3(Al,Fe3+)O10(OH)2 v?H2O. [1]  Vermiculite has high exchange
       capacity and exfoliates  like an accordion when quickly heated at high temperature.

Winchite - An amphibole mineral with an end-member composition
       (Ca,Na)Mg4(Al,Fe3+)Si8O22(OH)2.  [4] Winchite can occur with different habits, one of
       which can be asbestiform.
                                           31

-------
FIGURES

1.  Index map of locations of vermiculite deposits noted in this report.
2.  Graph of mineral composition of vermiculite ore (data from Atkinson and others, 1982).
       A. Libby, Montana.
       B. Enoree district, South Carolina.
3.  Graph of elemental composition reported for vermiculite and associated minerals.
4.  Chart showing analytical approach for screening phase of tracer analysis.
                                           32

-------
                                                                                Kola Peninsula
                                                                                  .—-3
Figure 1. Locations of vermiculite deposits discussed in this report.  Current major sources for the U.S. market are in bold, including
       Louisa, Virginia; the Enoree district, South Carolina; and Phalaborwa, South Africa.  A previous major source was at Libby,
       Montana.

                                                             33

-------
      100
                                                            Libby, Montana
 O)
'
-------
        100
                                              Enoree district, South Carolina
  O)
         80
                                                      Ore Type


                                                        head feed

                                                        beneficiated ore
  o>
  CJ
  cu
  Q.
60
  CU
  O
  c
  O
  O
  •a
  Q)
  "cc
  E

  "w
  LU
40
20
               CD

               "o
                              ^j
                              CD
£
CD
                                            CD
                                            X
                                            o
                 CD

                 i
                 ^
                 O
                 E
.CD
O


Q.
E
CD

"CD
"o
CD
"O
CD
.a
0






CD
0
.Q
Q.
E
CD
.0
"CD
E
CO
Q.
"o
CD
O
.Q
Q.
CD
CO
0

i^


                                                             CO
                                                             O
                                        to
                                        Q.
                                        CD
CD


E
Figure 2B. Mineral composition of vermiculite ore (data from Atkinson and others, 1982).

       Data include head feed (solid line) and three grades of beneficiated ore (dotted) from the

       W.R. Grace mill, and ungraded ore (dashed) from the Patterson mill in the Enoree

       district, South Carolina. The graph shows three types of amphiboles (outlined) that make

       up total amphiboles. Asterisk indicates fibrous amphiboles which are present, but at a

       level of less than 1%.
                                            35

-------
       8
  o
  o
 •^
  
-------
       Raw Ore
     0-3 samples
  (if beneficiated ore
     not available)
Lab beneficiation
 (if necessary)
Beneficiated Ore
   3 samples
                                  Lab exfoliation
                   (if necessary)
Exfoliated Ore
  2 samples
Consumer Product
     1 sample
                                                                           vermiculite
                                                                           separation
                                                                           subsample
                                             OM
                                     mineral identification
                                          7 samples
                                                          mineral
                                                         separation
                                            XRD
                                    mineral identification
      XRF or ICP
  major / trace chem
     INAA
  trace chem
                 EPMA or SEM/EDS   «
              mineral ID / particle chem
                  TEM/EDS/ED
                fibrous mineral ID
Figure 4. Analytical approach for the screening phase of a tracer analysis. Seven samples, not counting duplicates, are shown for
      materials derived from a single vermiculite ore deposit. Where beneficiated or exfoliated materials are not available in the
      field, raw ore would be collected for lab beneficiation. As many as 35 samples would be derived from 5 vermiculite mines.
      See Table 5 for methods.
                                                      37

-------
TABLES

1.  Vermiculite and associated minerals.
2.  Locations of mines that produced vermiculite for consumer products.
3.  Minerals reported at selected vermiculite deposits.
4.  Chemical composition of vermiculite and associated minerals.
       A. Oxide concentration in weight percent.
       B. Element concentration in parts per million.
5.  Description of methods for identifying tracer constituents in vermiculite ore.
                                           38

-------
 Table 1.  Vermiculite and some associated minerals.  Formulae are from Gaines and others
          (1997), except for those noted in footnotes and for amphiboles, which are taken from
          Leake and others (1997).
MINERAL
                        I DEAL FORMULA
                                                                    MINERAL
                                                                                            I DEAL FORMULA
SHEET SILICATES
Industrial Vermiculite
Vermiculite
Hydrobiotite *
[Hydrophlogopite] *
Mica
Biotite **
Phlogopite **
Annite
Muscovite
Serpentine
Antigorite
Chrysotile ***
Other Clay Minerals
Talc
Smectite group
Chlorite group
Kaolinite

CHAIN SILICATES
Pyroxene
Augite
Diopside
Rhodonite
Amphibole
Magnesiohorneblende
Anthophyllite ***
Tremolite ***
Actinolite ***
Richterite ***
Winchite ***
Grunerite [Amosite ***]
Riebeckite [Crocidolite ***]


Mg0.35+/.(Mg,Fe3+)3Si3(AI,Fe3+)010(OH)2-nH20
K(Mg,Fe)6(Si,AI)8020(OH)4 -nH2O
KMg6(Si,AI)8020(OH)4-nH20

K(Fe2+,Mg)3AISi3010(OH,F)2
KMg3AISi3010(F,OH)2
KFe2+3AISi3010(OH,F)2
KAI2AISi3010(OH,F)2

(Mg,Fe2+)3Si205(OH)4
Mg3Si205(OH)4

Mg3Si4010(OH)2
(Na,Ca,K)(AI,Mg,Fe)2(Si,AI)4O10(OH)2-rtH2O
(AI,Fe,Mg,Mn,Ni)5.6(AI,Si,Fe)4010(OH)8
AI2Si205(OH)4



(Ca,Na)(Mg,Fe,AI,Ti)(Si,AI)2O6
CaMgSi2O6
(Mn+2,Fe+2,Mg,Ca)SiO3

Ca2(Mg4(AI,Fe3+))Si7022(OH)2
Mg7Si8022(OH)2
Ca2Mg5Si8O22(OH)2
Ca2(Mg,Fe2+)5Si8022(OH)2
Na(Ca,Na)Mg5Si8022(OH)2
(Ca,Na)Mg4(AI,Fe3+)Si8O22(OH)2
Fe2+7Si8022(OH)2
Na2(Fe2+3,Fe3+2)Si8022(OH)2
FRAMEWORK SILICATES
Quartz
Feldspar
Alkali feldspar subgroup
Plagioclase series

OTHER SILICATES
Olivine group
Garnet group
Titanite [Sphene]
Zircon

OXIDES
Magnetite
Hematite
Chromite
Rutile
Anatase
Pyrolusite
Spinel
Corundum

CARBONATES
Calcite
Dolomite
Strontianite

SULFATE
Barite

PHOSPHATE
Apatite


SiO2

(K,Na)AISi308
(Na,Ca)(Si,AI)408


(Mg,Fe2+)2Si04
(Ca,Fe,Mg,Mn)3(AI,Fe)2(Si04)3
CaTiSiO5
ZrSi04


Fe2+Fe3+204
a-Fe2O3
Fe2+Cr2O4
Ti02
Ti02
MnO2
MgAI2O4
AI203


CaCO3
CaMg(CO3)2
SrC03


BaSO4


Ca5(PO4)3(F,OH,CI)

Notes:
 [Brackets] Mineral names in brackets are informal in that they are not recognized by the International Mineralogical Association (IMA),
  though they occur in the literature. The informal names are used here for ease of reference.
 *   Mixed-layer minerals. The formula for hydrobiotite is generalized from that in Gaines and others (1997):
    K(Mg,Fe)3Si3AI010(OH)2 / Mgo.35+/-(Mg,Fe)3Si3AIO10(OH)2 -nH2O.
    The formula for hydrophlogopite is also generalized by analogy with that for hydrobiotite.
 **  Biotite and phlogopite are common parent minerals of vemiculite and mixed-layer mica/vermiculite (hydrobiotite or hydrophlogopite).
    Biotite is also a series name extending from annite, which is iron-rich, to phlogopite, which ideally lacks iron. Members of the series
    with Mg:Fe ratios greater than 2:1 are generally regarded as phlogopites, though the boundary is arbitrary (Gaines and others, 1997).
 *** Mineral may commonly occur with an asbestiform habit.
                                                          39

-------
Table 2. Locations of mines that produced vermiculite for use in consumer products in the
       United States during the last 10 years. Major producers are listed in bold. Data are taken
       from Potter (1998 and earlier).
LOCATION OF MINES
MONTANA
Libby
Dillon

VIRGINIA
Louisa County

SOUTH CAROLINA
Enoree district*
 Enoree*
 Woodruff*
COMPANY
STATUS
W.R. Grace & Co.
Stansbury Holdings Corp.
Closed
Active, limited production
Virginia Vermiculite, Ltd.  Active
W.R. Grace & Co.
Patterson Vermiculite Co.
Virginia Vermiculite, Ltd.
  Carolina Vermiculite Co.
Active
(not determined)
Active
Active
SOUTH AFRICA
Phalaborwa

CHINA
Xinjiang Province*
 (several mines)
Other Provinces*
Palabora Mining Co., Ltd.  Active
China Xinjiang Metals &    Active
  Minerals Imp & Exp Corp
(not determined)            (not determined)
* Location may represent multiple mines.
                                         40

-------
 Table 3.  Minerals associated with  selected vermiculite deposits.
Vermiculite Deposits
                                              5
                                              I
                                                    -g.
                                                    c
                                                         1
                                                         0
                                                         8-
                                                         I
                                                                              o:
                                                                                      Other Minerals and Comments
Libby, Montana
Palabora, South Africa
North Carolina deposits
Enoree district,
South Carolina
Louisa, Virginia
                                                    **    *     *     *   **  1,4, 13    (Apatite is fluorapatite), vanadium-rich pyroxene, strontianite, barite, copper and zinc sulfides.
                                                                           14, 19, 21   Diopside, augite, hornblende, sphene, magnetite, hematite, quartz, .
                                                                              23      Locally, carbonate and feldspar are intergrown with vermiculite.
                                                                                      Antigorite reported in head feed (13); mention of chrysotile (19, 13), very low abundance..
                                                                                      Amphibole asbestos reported as tremolite-actinolite (13), sodium tremolite and richterite (4), winchite (23).

                                                    *    **   See  NR  **   5, 7, 8    See text for minerals assoc. with the Igneous Complex, but not necess. with the vermiculite deposit.
                                                             Notes         10, 14, 17   Some describe chrysotile, others say it is scrolls of vermiculite (5); rare anthophyllite fibers reported (14).
                                                                                      The mica at Palabora is phlogopite rather than biotite.

                                                   See   **     *     **  NR 9, 11,16   Nickel silicates, corundum, chlorite.
                                                  Notes                               Fibrous amphibole and chrysotile are reported as abundant in individual deposits.

                                                    *    *     *     *   **   12,13    (Apatite is fluorapatite), sphene, zircon, chlorite, chromite, rutile, titanite, corundum, anatase,
                                                                                      hornblende, magnetite, hematite, rhodonite, pyrolusite, quartz, feldspar.
                                                                                      Tremolite-actinolite and anthophyllite occur mainly in prismatic form.

                                                    *   NR    *    NR NR   14,19    No detailed geologic reports were found describing this deposit.
                                                                                      Traces of fibrous amphibole reported (14).
Dillon, Montana ** *
Xinjiang Province, China **
Hafafit, Egypt
Russia (Urals, Kola Peninsula) ** *
I I I
I See I
Notes
I
NR
I
* NR
I I
I
2,3,15
18,20
6
22
Magnetite, spinel, and olivine (in the ultramafic bodies), garnet and chlorite in vermiculite ore zone.
Fibrous tremolite, actinolite, anthophyllite, and fibrous talc and biotite (2).
No detailed geologic reports were found describing these deposits.
Amphibole asbestos minerals include primariliy anthophyllite with lesser tremolite and actinolite.
Several amphibole minerals of unspecified habit, including Na-rich species, reported at many deposits.
Unspecified type of asbestos reported to occur with some deposits.
Explanation
A vermiculite includes mixed layer mica / vermiculite.
**   mineral is abundant.
*   mineral is mentioned in literature as being associated with the vermiculite deposit.
                                                                                      NR  mineral is not reported, may or may not be present.
                                                                                      I     insufficent information available in literature.
                                                                                      Deposits in bold type are either producing mines in the US or major importers to the US.
References: 1. Bassett, 1959; 2. Behre Dolbear and Co., 2000; 3. Berg, 1995; 4. Boettcher, 1966; 5. Chatfield and Lewis, 1979, 1980; 6. El Shazly, 1975a; 7. Evans, 1993; 8. Hessel and Sluis-Cremer, 1989;
            9. Hunter, 1950; 10. IOM, 2000; 11. Kulp and Brobst, 1954; 12. Libby, 1975; 13. MRI, 1982; 14. Moatamed et al, 1986; 15. Montana DEQ, 1999; 16. Murdock and Hunter, 1946;
            17. Palabora Mining Co, 1976;  18. Potter, 1998; 19. Rohl and Langer, 1977; 20. Rongqi and Junchen, 1993; 21. USEPA, 1980; 22. Vorovikov, 1973; 23. Wylie and Verkouteren, 2000.
                                                                                         41

-------
Table 4. Chemical composition of vermiculite, mixed-layer mica/vermiculite, and mica from selected ore deposits.  Dashes indicate
       no analysis conducted.
       A.  Oxide weight percent from listed references.
LOCALITY AND MINERALS
SiO2 AI2O3
Libby, Montana
vermiculite 38.64 14.94
vermiculite 35.57 11.47
vermiculite 35.43 1 1 .30
hydrobiotite 35.60 11.85
hydrobiotite 36.77 11.57
biotite 38.63 13.08
biotite 39.10 13.30
Louisa, Virginia
vermiculite 38.34 12.85
Fe203

9.29
7.49
6.65
10.28
8.28
2.50
2.56

8.80
FeO

-
0.34
0.27
0.81
0.98
8.75
7.23

-
MgO

22.68
22.57
23.56
20.17
20.04
19.94
21.55

16.38
CaO

1.23
0.73
0.39
1.44
1.94
0.18
0.12

1.12
Na2O

-
-
-
0.16
0.12
0.26
0.23

1.72
K2O

7.84
0.96
0.14
3.17
3.84
10.00
10.05

6.63
OXIDES, wt percent
TiO2

-
1.06
0.91
1.13
1.02
1.55
1.21

1.66
Cr203

0.29
0.18
0.26
0.03
0.27
0.23
0.25

0.23
MnO NiO

..
0.06 0.02
0.05 0.01
0.08 --
0.08 --
0.14 --
0.10 --

0.14 --
Li2O Rb2O SrO Cs2O BaO

..
0.01 -- 0.10
0.01 -- 0.03
0.0129 0.005 0.00041 0.17
0.0158 0.01 0.00050 0.19
0.04 0.005 -- 0.45
0.03 <0.005 -- 0.35

0.01 	
F

-
-
-
0.21
0.30
0.30
0.35

--
Cl P205

0.28 -
0.06
0.05
0.07
0.06
0.06
-

-
H2O+

-
9.01
9.47
7.56
6.69
3.52
3.74

--
H2O-

-
10.11
11.26
7.20
7.80
0.30
0.06

--
REFERENCE
H2Otot

5.29
19.12
20.73
14.76
14.49
3.82
3.80

10.66
Oxide tot

100.59 1
99.74 2
99.79 2
99.95 2
99.98 2
99.94 2
100.23 2

98.54 3
Enoree district, South Carolina
vermiculite 38.66 17.36
vermiculite 39.79 10.96
hydrobiotite 38.63 13.43
biotite 40.30 12.43
biotite 40.23 12.95
Phalaborwa, South Africa
vermiculite 39.37 12.08
vermiculite 38.53 8.35
vermiculite 35.93 9.05
hydrobiotite 38.74 10.38
phlogopite 40.25 10.45
Xinjiang Province, China
vermiculite 41.20 12.68
hydrophlogopite 36.30 12.45
phlogopite 39.38 14.44
phlogopite 38.35 15.44
8.45
4.33
8.89
1.86
1.97

5.45
6.31
5.11
8.96
2.50

4.60
5.95
2.87
2.62
-
0.66
1.13
9.28
7.72

1.17
0.64
0.16
1.98
2.18

1.54
0.37
2.40
2.76
20.04
24.42
17.93
19.42
20.15

23.37
25.74
26.15
21.77
26.11

24.22
22.82
23.61
23.02
0.75
0.90
1.28
0.42
0.43

1.46
-
0.50
-
—

0.93
0.68
0.95
0.63
-
-
-
-
-

0.80
-
-
0.27
0.25

1.61
1.10
2.50
3.70
4.24
0.28
5.48
8.82
9.56

2.46
4.73
0.16
7.82
10.35

5.97
3.70
10.40
10.20
-
0.43
1.39
2.32
2.14

1.25
1.12
0.93
2.11
1.15

1.38
1.32
1.49
1.34
0.50
0.15
0.08
0.10
0.07

-
0.12
0.48
-
0.07

-
-
-
~~
0.07 --
0.07 0.02
0.19 0.02
0.17 0.07
0.17 0.03

0.30 --
0.03 --
0.02 --
0.04 --
0.02 --

0.05 --
0.04 --
0.03 --
0.03 --
0.12
0.003 0.01 -- 0.02
0.059 0.01 -- 0.08
0.049 0.01 -- 0.43
0.063 0.01 -- 0.17

0.03
..
..
..
—

..
..
..
__
-
0.52
0.78
0.63
0.86

-
0.92
0.67
0.53
0.62

-
--
--
—
..
0.17
0.09
0.03
0.02

..
0.01
0.05
..
—

..
..
..
__
-
6.39
4.17
2.66
2.76

-
14.58
11.32
7.50
6.48

-
6.62
1.11
1.06
-
10.25
5.92
0.30
0.34

-
-
10.36
--
—

-
8.08
0.34
0.35
8.71
16.64
10.09
2.96
3.10

11.20
14.58
21.68
7.50
6.48

5.82
14.70
1.45
1.41
98.90 3
99.37 4
99.56 4
99.29 4
99.64 4

98.94 3
101.08 6
100.89 6
100.10 6
100.43 6

100.00 3
99.43 5
99.52 5
99.50 5
 References  1 - Bassett (1959)
2-Boettcher(1966)
3-Hindman(1984)
4- Libby (1975)
5 - Rongqi and Junchen (1993)
6-Schoeman (1989)
                                                                 42

-------
Table 4 (continued).  Chemical composition of vermiculite, mixed-layer mica/vermiculite, and mica from selected ore deposits.
       Dashes indicate no analyses conducted.
       B. Element concentration in parts per million (ppm), calculated from oxide weight percent. Symbols listed are used in Figure 3.
 LOCALITY AND MINERALS
                    ELEMENTS, ppm
SYMBOL DESCRIPTION

Si
Libby, Montana
vermiculite 181000
vermiculite
vermiculite
hydrobiotite
hydrobiotite
biotite
biotite
166000
166000
166000
172000
181000
183000
Al Fe tot
79100 65000
60700 55000
59800 48600
62700 78200
61200 65500
69200 85500
70400 74100
Fe3+ Fe2+
65000 -
52400 2600
46500 2100
71900 6300
57900 7600
17500 68000
17900 56200
Mg Ca
137000 8800
136000 5200
142000 2800
122000 10300
121000 13900
120000 1300
130000 860
Na K
- 65100
8000
1200
1200 26300
890 31900
1900 83000
1700 83400
Ti Cr
- 2000
6400 1200
5500 1800
6800 210
6100 1800
9300 1600
7300 1700
Mn Ni Li
Rb Sr
Cs Ba F Cl
P
	 .. .. 2800 —
460 160 -
390 80 -
620 - -
620 - -
1100 - -
780 - -
- 80
- 80
120 40
140 80
370 40
270 <40
900 -
270 -
3.9 1500 2100 -
4.7 1700 3000 -
- 40000 3000 -
- 3100 3500 -
260
220
300
260
260
-

LV
LV
LV
LH
LH
LB
LB

concentrate
RCSa-59
RC-level 12
RCSAa-36
RCSp-49
RCB-12
RCSp-55
 Louisa, Virginia
 vermiculite     179000 68000 61600 61600

 Enoree district, South Carolina
98800  8000 12800 5500010000 1600 1100 -   46 -
vermiculite 181000 91900
vermiculite 186000 58000
hydrobiotite 181000 71100
biotite 188000 65800
biotite 188000 68500
Phalaborwa, South Africa
vermiculite 184000 63900
vermiculite 180000 44200
vermiculite 168000 47900
hydrobiotite 181000 54900
phlogopite 188000 55300
Xinjiang Province, China
vermiculite 193000 67100
hydrophlogopite 170000 65900
phlogopite 184000 76400
phlogopite 179000 81700
59100
35400
71000
85100
73800

47200
49100
37000
78100
34400

44100
44500
38700
39800
59100
30300
62200
13000
13800

38100
44100
35700
62700
17500

32200
41600
20100
18300
-
5100
8800
72100
60000

9100
5000
1200
15400
16900

12000
2900
18700
21500
121000
147000
108000
117000
122000

141000
155000
158000
131000
157000

146000
138000
142000
139000
5400
6400
9100
3000
3100

10400
-
3600
-
-

6600
4900
6800
4500
-
-
-
-
-

5900
-
-
2000
1900

12000
8200
19000
27000
35200
2300
45500
73200
79400

20400
39300
1300
64900
85900

49600
30700
86300
84700
-
2600
8300
13900
12800

7500
6700
5600
12600
6900

8300
7900
8900
8000
3400
1000
550
680
480

-
820
3300
-
480

-
-
-
-
540 	
540 160 - 30
1500 160 - 540
1300 550 - 450
1300 240 - 580

2300 - 140 -
230 	
150 	
310 	
150 	

390 	
310 	
230 	
230 	
-
80
80
80
80

-
-
-
-
-

-
-
-
-
                                                                                            1100  -
                                                                                            180 5200
                                                                                            720 7800
                                                                                            3900 6300
                                                                                            1500 8600
                                                                                                9200
                                                                                                6700
                                                                                                5300
                                                                                                6200
                                                                      VV
                                                                            concentrate
-
720
390
110
74

44
220
-
—
..
-
-
-
EV
EV
EH
EB
EB
PV
PV
PV
PH
PP
XV
XH
XP
XP
concentrate
A-24V
Y-26V
A-48B
Y-35B
concentrate
VOD orebody
VOD orebody
sw of Loolekop
VOD orebody
Qieganbulake
Tseganbrark v-5
Tseganbrark v-1
Tseganbrark v-3
                                                               43

-------
Table 5.  Description of methods useful for identifying tracer constituents in vermiculite ore.  See
        text for additional innovative methods not listed here.
     Method          Data Objectives                                            Material
                                                                                      Bulk*   Grain
        OM       Optical Microscopy **
                        Grain morphology, size, habit, color, surface and boundary texture
                        Refractive indices and other optical crystallographic properties
                        Other crystallographic properties
                        Mineral identification by optical properties
                        Qualitative mineral abundance and quantitative concentration

        XRD      X-ray Diffraction
                        Mineral identification by structure
                        Qualitative mineral abundance and quantitative concentration

     SEM/EDS    Scanning Electron Microscopy
                    with Energy Dispersive X-ray Spectroscopy
                        Grain morphology, size, habit, color, surface texture
                        Mineral identification by chemical composition
                        Qualitative mineral abundance and quantitative concentration

       EPMA      Electron Probe Microanalysis
                   with Wavelength and Energy Dispersive X-ray Spectroscopy
                        Grain size, habit, boundary texture
                        Mineral Identification by quantitative chemical composition

    TEM/EDS/ED  Transmission Electron Microscopy
                   with Energy Dispersive X-ray Spectroscopy and Electron Diffraction
                        Grain size, habit, internal texture
                        Mineral Identification by chemical composition and structure

        XRF      X-ray Fluorescence Spectroscopy
                        Chemical composition

        INAA      Instrumental Neutron Activation Analysis
                        Chemical composition

      ICP-AES     Inductively Coupled Plasma-Atomic Emission Spectroscopy
                        Chemical composition

      ICP-MS     Inductively Coupled Plasma-Mass Spectrometry
                        Chemical composition
                     Bulk material includes bulk samples of raw ore, beneficiated ore, exfoliated ore,
                      consumer products, and mineral concentrates from all of above.
                     Optical microscopy includes both low-power stereomicroscopy with reflected light,
                      and polarized light microscopy with transmitted light.
                                                  44

-------