Heavy Minerals of Late Eocene Sandstones of Southwestern North Dakota

1
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Heavy Minerals of Late Eocene Sandstones of
Southwestern North Dakota
John Webster - Minot State University
INTRODUCTION Ongoing faculty-student research
projects at Minot State University (MSU) have
focused on the nature, age, and provenance of
Eocene sandstones of southwestern North Dakota.
The sandstones being studied are the Chalky
Buttes Member of the Chadron Formation and the
Medicine Pole Hills sandstone. The Chalky Buttes
Member sandstone is exposed at a number of buttes
in southwestern North Dakota (Figure 1). The
Medicine Pole Hills sandstone is exposed
southwest of the buttes (Figure 1). Both
sandstones are conglomeratic in part. Murphy et
al. (1993) correlated the Medicine Pole Hills
sandstone with the Chalky Buttes Member. This was
a lithostratigraphic correlation based on the
fact that both the Chalky Buttes Member and
Medicine Pole Hills sandstones overlie
(unconformably) various formations of the
Paleocene Fort Union Group (Figure 2). The
research effort at MSU has involved a variety of
studies designed to characterize and compare the
Medicine Pole Hills and Chalky Buttes Member
sandstones. The studies have included comparison
of textures (Frederick 2004), major mineralogy
(Arneson 2005), and pebble lithologies (Whitlow
2003). Several studies have focused on heavy
mineral analysis of the Medicine Pole Hills
sandstone and the Chalky Buttes Member
Figure 1 Map of southwestern North Dakota
showing the locations of buttes that expose White
River Group or Arikaree Formation strata. Red
labels show locations that have been sampled for
heavy mineral analysis. MPH Medicine Pole Hills,
RSB Rattlesnake Butte, CB Chalky Butte, LB Little
Badlands, SB Square Butte.
Figure 2 Generalized stratigraphic column for
Tertiary strata of North Dakota (from Murphy et
al. 1993).
1
2
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
sandstone. Kight (2002) and Kuhnhenn (2005)
completed heavy mineral analyses of the Medicine
Pole Hills sandstone. Hicks (2007) carried out a
trace element study of diopside, the dominant
heavy mineral in the Medicine Pole Hills
sandstone. The goal was to determine whether the
diopside was igneous or metamorphic in origin.
Heavy mineral analysis of the Chalky Buttes
Member sandstone has been completed for the
sample from Rattlesnake Butte (Nelson 2008), and
a sample from Square Butte. This report provides
a review of the heavy mineral research of Denson
and co-workers on Tertiary strata of the central
Rocky Mountains and northern Great Plains, which
included samples from southwestern North Dakota
(Denson and Gill 1965, Sato and Denson 1967,
Denson and Chisholm 1971). It also provides a
summary of the results of MSU heavy mineral
research, comparing the Medicine Pole Hills and
Chalky Buttes Member sandstones to one another,
and to the results of Denson and co-workers.
Finally, some needs for future research are
discussed. HEAVY MINERALS OF TERTIARY
STRATA Denson and Chisholm (1971) reported the
general heavy mineral compositions of Tertiary
strata of the middle Rocky Mountains and northern
Great Plains based on the study of 3,000 heavy
mineral separates from 200 stratigraphic
sections. For most samples,
heavy minerals were separated from the fine sand
fraction (0.063-0.125 mm) for a limited number
of samples, the coarse silt fraction (0.031-0.063
mm) was used. Non-opaque grains (100 or more per
sample) were identified by optical microscopy.
The changes in heavy minerals throughout
Tertiary strata reveal changing sources that
correspond to tectonic and volcanic episodes in
the middle Rocky Mountains and northern Great
Plains. Denson and Chisholm reported six major
lithogenetic units (Table 1).
Table 1 Summary of six lithogenetic Tertiary
units of Denson and Chisholm (1971)
2
3
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Heavy Mineral Sources Denson and Chisholm (1971)
reported that the three basic sources for heavy
minerals in Tertiary strata were (1) recycled
sediment derived from Paleozoic and Mesozoic
sedimentary rocks, (2) Precambrian plutonic and
metamorphic rocks exposed in the cores of
mountain ranges, and (3) Tertiary volcanic rocks.
A summary of these three sources, as given by
Denson and Chisholm is as follows 1.
Recycled tourmaline zircon (pinkish-violet
plutonic) 2. Plutonic and Metamorphic blue-green
hornblende garnet epidote zircon tourmaline
gray-green biotite 3. Volcanic green-brown
hornblende red oxyhornblende augite hypersthene
red-brown biotite
Sato and Denson (1967) studied the heavy mineral
compositions of sediment in streams draining
Precambrian plutonic and metamorphic mountains
and of Tertiary volcanics (ash deposits). A
summary of their results is given in Table 2.
These results
show the dominance of blue-green hornblende,
garnet, and epidote for Precambrian plutonic and
metamorphic sources, and the dominance of
green-brown hornblende and pyroxene (mostly
augite) in volcanic sources.
Table 2 Non-opaque heavy mineral compositions of
Precambrian sources and volcanic (ash) sources
number of samples analyzed shown in parentheses
3
4
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Heavy Minerals of the Williston Basin Denson and
Gill (1965) reported heavy mineral compositions
of Cretaceous and Tertiary strata exposed at
buttes in southwestern North Dakota, southeastern
Montana, and northwestern South Dakota (Table 3).
Of significant result was the general dominance
of opaque heavy minerals in most Fort Union Group
samples and the White River Group Chadron
Formation. The opaque minerals consisted largely
of magnetite (up to 50 in some samples),
ferruginous and clay aggregates, leucoxene, and
pyrite. Non-opaque minerals dominated in the
White River Group Brule Formation, in the
Arikaree Formation, and to a lesser extent in the
Fort Union Group Ludlow Formation. Among the
non-opaque heavy minerals, some significant
variations included (1) a decreasing abundance
of garnet with time, (2) a close similarity
between the Sentinel Butte and Chadron
Formations, and (3) the appearance of abundant
hornblende and pyroxene in the Brule and Arikaree
Formations.
Table 3 Heavy mineral compositions of Tertiary
strata of the southwestern Williston Basin area
number of samples analyzed shown in
parentheses individual mineral data are
normalized percentages of non-opaque minerals
4
5
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Heavy Minerals of the White River Group The
heavy mineral compositions of White River Group
samples reported by Sato and Denson (1967) are
summarized in Table 4. Their results indicate a
mixture of Precambrian and volcanic sources. The
appearance of augite and rd-br hornblende and the
significant increase in gr-br hornblende indicate
an influx of volcanic minerals during the latest
Eocene. Early Eocene strata were dominated by
Precambrian source heavy minerals. Based on
comparison with data reported by Denson and Gill
(1965), shown in Table 3, it seems that the
Williston Basin data in Table 4 primarily reflect
the Brule Formation rather than the Chadron
Formation (especially the high percentages of
hornblende and biotite). Thus, while the Brule
Formation reflects a mixture of Precambrian
plutonic - metamorphic and volcanic sources, the
Chadron Formation data of Table 3 suggest a
different mixture consisting of multicycle
minerals (e.g., zircon and tourmaline) and
metamorphic minerals (staurolite, kyanite,
andalusite, garnet, and epidote). A significant
difference between the metamorphic minerals and
those typical of Precambrian sources in general
is the lack of hornblende in the Chadron
Formation (and the Sentinel Butte Formation).
METHODS Field Work Allen Kihm, John Webster,
and several students carried out fieldwork in
southwestern North Dakota over a period of
several days during the summer of 2002.
The goal of the fieldwork was to collect
sandstone and pebble samples from both the
Medicine Pole Hills and Chalky Buttes Member
sandstones. For heavy mineral analysis,
approximately 4 kg of sandstone were collected
from each location in canvas bags. The Medicine
Pole Hills sandstone
Table 4 Non-opaque heavy minerals of the White
River Group a
a numbers in parentheses indicate the number of
samples analyzed b 13 green biotite 3 rd-br
biotite
5
6
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
was sampled from a pit previously dug by Dean
Pearson (Pioneer Trails Regional Museum) and
coworkers. Chalky Buttes Member sandstone samples
were collected at Rattlesnake Butte, Chalky
Butte, Little Badlands, and Square Butte (see
Figure 1). Sample Preparation Sample processing
techniques have evolved throughout the several
heavy mineral studies. Initially, disaggregation
involved running sample through a blender. This
was abandoned because of the possible effects on
more fragile grains. The first two heavy mineral
studies of the Medicine Pole Hills sandstone
(Kight 2002, Kuhnhenn 2005) utilized a separatory
funnel for heavy mineral separation. Later
studies made use of a high-speed centrifuge. The
procedures outlined here are based on the more
recent studies. Disaggregation and Grain-Size
Separation Each sample was initially
disaggregated as much as possible in a large,
heavy plastic tub (80 x 60 x 10 cm) using gentle
hammering with a rock hammer. Distilled water was
added to fill the tub approximately half full,
and three small spoonfuls of Calgon were added.
The sample was allowed to soak for one day, with
occasional agitation using a large wooden spoon
and plastic potato masher. To remove as much silt
and clay as possible, the liquid was decanted and
rinsed in a 0.063-mm brass
sieve (8-inch) resting in an underneath spray
system developed by Jason Hicks during his study
of diopside grains from the Medicine Pole Hills
sandstone (Hicks 2007). More water and Calgon
were added to the tub, and this process repeated
until the majority of the clay in particular was
removed. Then the entire sample was washed
through the 0.063-mm sieve wash system in small
batches. The cleaned sand from each small batch
was transferred to smaller plastic tubs (four
total) using distilled water. With the cleaned
sand evenly distributed among the four small
tubs, the sand was stirred, allowed to settle
briefly, and the water decanted through a small
(3-inch) 0.063-mm brass sieve. Sand in each tub
was rinsed in the same way with distilled water
twice, with as much water as possible decanted
after the final rinse. Grains collected in the
0.063-mm sieve were washed back into one of the
tubs using a small amount of distilled water. The
tubs were then placed in a large drying oven at
80 C until they were completely dried. The sand
was then combined in one tub (or two if
necessary). Each sample was sorted into 1/4 ?
size fractions from 2.00 mm (-1 ?) to 0.063 mm (4
?) using 8-inch sieves (21 total) and a sieve
shaker. The sieves were divided into three sets,
and the entire sample was run through a set
before the next set was used. Approximately 100 g
of sample at a time were run through a set of
sieves, with the sieve shaking running 10-12
minutes each time. The sieves were emptied onto
labeled
sheets of butcher paper after each two 100-g
batches. Each grain size fraction was stored in
an appropriate labeled plastic bag, and the
weight of each fraction was recorded. Heavy
Mineral Separation Heavy minerals were separated
from the selected grain size fraction(s) using a
heteropolytungstate solution (LST) with a density
of 2.85 g/cm3. Separation was carried out in
500-mL centrifuge bottles, each filled with
approximately 200 mL of LST. The number of
bottles used was dependent on the number of size
fractions used, and the amount of sand in each
fraction. The amount of sand in each bottle was
kept to 75 g or less. The centrifuge was run at
2000 rpm for 30 minutes. Heavy minerals were
recovered from the bottom of the bottles using a
vacuum extraction tube (Figure 3). A hollow glass
tube with a cone-shaped plastic cap on the bottom
end was slowly inserted into the centrifuge
bottle though the layer of light minerals. The
plastic cap was used so none of the light
minerals at the top of the bottle would enter the
tube. Once the glass tube was through the top
minerals, a thin glass rod was used to push out
the cap. A smaller glass tube attached to a
vacuum system was then inserted through the first
tube so its end was just above the heavy minerals
that collected at the bottom of the centrifuge
bottle. Heavy minerals were transferred by vacuum
suction into an Erlenmeyer trap flask. The heavy
minerals that were
6
7
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Grain Mount Preparation Polished grain mount
sections of heavy minerals were prepared on
one-inch round glass slides. A large drop of
epoxy was placed on a frosted glass slide, and
heavy mineral grains were sprinkled into the
epoxy. This was covered with a thin glass cover
slip, which was moved around under light pressure
to spread out the epoxy and grains. After at
least 24 hours cure time, a thin section machine
was used to grind away the cover slip, and
partially grind into the layer of grains.
Hand-lapping with 600-grit silicon carbide
abrasive on a glass plate was used to slowly
grind the grains some more, exposing as many
grains as possible. The grain mounts were then
polished on a Buehler Metaserve 2000 using a
succession of diamond (6-, 3-, and 1-micron) and
alumina (0.3- and 0.05-micron) polishes. Each
polished section was coated with carbon using the
carbon accessory on a Denton Vacuum Desk 2
sputter coater. Grain maps were prepared for
each section from high-resolution
transmitted-light scanned images. For each, the
scanned image was imported into a computer
drawing program. The program was used to draw
lines separating the section into quarters,
designated with capital letters A-D. Each quarter
section was divided with lines (polygons) into
subsections of 35-40 grains each these were
labeled with lower case letters. Within each
subsection, each grain was numbered sequentially,
providing a system for reference to individual
grains
(e.g., Ba32 refers to grain number 32 in
subsection a of quarter B). Optical Microscopy
and SEM-EDS Analysis A polarized-light
petrographic microscope was used for optical
identification of heavy mineral grains. The
identifications were recorded on a table (one for
each subsection) along with any qualifying
remarks. In particular, grains of uncertain
identification were noted. Chemical analysis of
grains was carried out using energy dispersive
spectroscopy (EDS) microanalysis using a LEO
435VP scanning electron microscope (SEM) equipped
with an EDS detector and a Thermo Electron Noran
System Six (NSS) analysis system. With the SEM in
spot mode, SEM-EDS microanalysis was used to
analyze individual spots within grains. For
heterogeneous grains, multiple spots were
typically analyzed. Microanalysis was used to
obtain a number of representative analyses of all
types of grains, and more importantly to help in
the characterization of grains of uncertain
optical identification. Mineral formulae were
recalculated from the analyses using a
spreadsheet written by the author. Optical
identifications were corrected as necessary based
on SEM-EDS analyses. They were then entered into
a spreadsheet for calculation of the percentage
of each heavy mineral. The percentages obtained
are grain-count percentages. In the early
Figure 3 Apparatus for recovery of heavy
minerals from centrifuge bottles.
recovered in the flask were filtered, washed with
distilled water, and then dried in a drying oven.
The dried heavy mineral grains were brushed onto
weighing paper and then transferred to a small
vial for storage. The light mineral grains
remaining in the centrifuge bottles were also
filtered, washed, dried, and stored for possible
future research.
7
8
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
(Medicine Pole Hills sandstone) studies, the size
(area) of each grain was also tabulated in the
spreadsheet, and weight percentages were
calculated from total areas of each mineral and
appropriate densities determined from average
chemical analyses. Weight percentages were needed
for comparison with heavy mineral analyses
carried out using X-ray diffraction (XRD). These
early optical-SEM studies (Kight 2002, Kuhnhenn
2005) showed minimal difference between weight
percentages and grain-count percentages, so the
practice of measuring grain areas was
discontinued for subsequent studies. Rietveld
Quantitative X-Ray Diffraction Analysis Two
studies involved heavy mineral analysis using
Rietveld quantitative XRD (RQXRD) analysis.
Grains not used for optical-SEM work were put in
a McCrone micronizing mill with agate cylinders
and hexane to reduce the particle size to less
than 5 µm. The micronizing mill was run for 10
minutes and the contents were transferred to an
evaporating dish. Additional hexane was added to
the micronizing bottle and the bottle was placed
back onto the micronizer for 30 seconds. The
purpose of this was to remove particles adhered
to the bottle and milling cylinders this was
repeated twice. A small portion of the sample was
analyzed by XRD using a slurry mount on a zero
background silicon disk. Jade 6.0 (Materials Data
Incorporated) was used for
identification of phases present, in particular
to determine if rutile is present. Rutile is
typically used as the internal standard for RQXRD
analysis. If rutile is present in the sample, ZnO
is used as the internal standard. For the RQXRD
analysis, 0.9000 g of the heavy mineral sample
was combined with 0.1000 g of the internal
standard using a mortar and pestle. The
homogenized mixture was packed in an XRD sample
holder using standard rear packing techniques.
The sample was analyzed using a PANalytical
XPert Pro MPD system. The scan covered a 2?
range of 20 to 80 with a step size of 0.0167
2?. Fixed slits were used in order to ensure a
constant irradiated volume throughout the scan,
and a 10-mm mask was used in order to ensure the
beam was constrained within the sample boundary.
The sample stage was spun at a rate of one cycle
per 30 seconds in order to increase counting
statistics. A scan length of 20205 (hms) with
a copper tube yielded sufficient scan intensity.
These parameters have been shown to yield
sufficient counting statistics for routine
analysis (Brayko 2007). The XRD scan was
examined using Jade 6.0 for identification of
phases present. Rietveld refinements were carried
out to quantify the minerals that were
identified. Rietveld analysis was done using
General Structure Analysis System (GSAS) (Larson
and Von Dreele 1994) and EXPGUI (Toby 2001) to
determine weight percent for each phase. The
refinement parameters included background (6th
order Chevyshev
polynomial), scaling, specimen displacement
(shft), lattice parameters, Lorentzian broadening
(LX), and preferred orientation using spherical
harmonics. The order of refinements was varied
from phase to phase as needed to maintain
refinement stability (i.e., maintaining non-zero
weight fractions of each material in refinement
when possible). The refinements were continued
until GSAS showed the refinement to be
converged. RESULTS Results of the optical-SEM
heavy mineral analyses carried out at MSU are
shown in Table 5. Work on the Medicine Pole Hills
sandstone was carried out by Kight (2002) and
Kuhnhenn (2005). Work on the Chalky Buttes Member
sandstone from Rattlesnake Butte was carried out
by Webster (reported in Nelson 2008), and from
Square Butte by Webster and Fogarty (2008).
Results of RQXRD heavy mineral analyses are not
reported here, but are consistent with the
optical-SEM results (Webster et al. 2002, Cool
2006, Nelson 2008). The optical-SEM analyses
show that the Medicine Pole Hills sandstone heavy
minerals are dominated by diopside and
hornblende, with modest abundances of
epidote-clinozoisite, augite, and almandine
garnet. The two Chalky Buttes Member sandstones
yielded rather different results from the
Medicine Pole Hills sandstone, and from one
another. The Rattlesnake Butte sample was
dominated by epidote-clinozoisite, biotite, and
almandine garnet,
8
9
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
with modest abundances of ilmenite and
pseudobrookite. The Square Butte sample was
dominated by rutile, with modest abundances of
epidote-clinozoisite, pseudo--brookite, ilmenite,
staurolite, aluminosilicate (probably kyanite),
apatite, and tourmaline. DISCUSSION Comparison
of MSU Results Abundances of the non-opaque
heavy minerals (normalized to 100) from the MSU
studies are compared in Table 6. The two Medicine
Pole Hills samples compared very well with one
another, which is not surprising because they
were collected from the same location and
stratigraphic horizon. The two Chalky Buttes
Member samples do not compare very well with one
another. The Rattlesnake Butte non-opaque heavy
minerals were dominated by epidote-clinozoisite
(47.4), biotite (24.3), and almandine garnet
(23.3). The Square Butte sample did not contain
biotite, and contained considerably less
epidote-clinozoisite and almandine garnet. It was
dominated by rutile (48.2), and contained modest
(gt10) amounts of epidote-clinozoisite (13.5),
staurolite (10.7), and apatite (9.7), with
minor amounts of aluminosilicate (4.7),
almandine garnet (3.5), and tourmaline (3.5).
With the exception of epidote-clinozoisite, these
minerals of the Square Butte sample were found in
trace amounts (if at all) in the Rattlesnake
Butte sample.
Table 5. Heavy mineral compositions of Eocene
sandstones MSU studies
a weighted average of three grain sizes (1.9
from 0.180-0.212 mm, 30.0 from 0.212-0.25 mm,
and 68.1 from 0.25-0.30 mm) b weighted
average of two grain sizes (60.9 from 0.212-0.25
mm, and 39.1 from 0.25-0.30 mm)
9
10
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Comparison of MSU Results with Chadron Formation
Results Results of the MSU studies are compared
with the Chadron Formation data of Denson and
Gill (1965) in Table 6. In this table, to allow
direct comparison with the data reported by
Denson and Gill, the non-opaque minerals were
normalized to 100 (excluding quartz and
feldspar). While there are a few similarities
between data from the Chadron Formation and the
MSU studies, there are a number of significant
differences. These differences are summarized
here. Opaques The Chadron Formation contained a
very high percentage of opaque heavy minerals,
averaging 78. The Chalky Buttes Member samples
had much lower abundances of opaques. The Square
Butte and Rattlesnake Butte samples contained
16.5 and 8.5 opaques respectively, both
dominated by pseudobrookite and ilmenite. The
Medicine Pole Hills sandstone contained less than
1 opaque heavy minerals. Diopside,
Pyroxenes The Chadron Formation contained no
pyroxenes, and in this sense is similar to the
MSU samples of the Chalky Buttes Member (trace
amounts at most). The Medicine Pole Hills
sandstone contains abundant pyroxenes, with
59-62 diopside, 2-6 augite, and a trace amount
of hypersthene.
Table 6. Heavy mineral compositions of Tertiary
strata of the southwestern Williston Basin area
10
11
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Hornblende The Chadron Formation averaged 3
hornblende. The Chalky Buttes Member sandstone
samples also had low hornblende (0.3 to 1.6),
but the Medicine Pole Hills sandstone contained
23-26. Aluminosilicates The Chadron Formation
averages 8 kyanite and 6 andalusite.
Aluminosilicates (undifferentiated) are only
found in trace amounts in the MSU samples, with
the exception of the Square Butte sample
(4.7). Staurolite The Chadron Formation
contained 22 staurolite. Of the MSU samples,
only the Square Butte sample contained a modest
amount of staurolite (10.7). Ultra-stable
Minerals The Chadron Formation contained
relatively abundant ultra-stable grains,
averaging 28, 10, and 3 zircon, tourmaline,
and rutile respectively. The Medicine Pole Hills
sandstone contains no ultra-stable grains. The
Square Butte sample has a high percentage of
ultra-stable grains, but they are dominantly
rutile, with a modest amount of tourmaline and a
low abundance of zircon. The Rattlesnake Butte
sample contains only trace amounts of the
ultra-stable grains.
Significance of Heavy Mineral Comparisons The
significance of the comparisons between the MSU
data and Chadron Formation data must be viewed in
light of some differences that existed in the
nature of the samples and the methods used for
analysis. The most significant difference between
the samples is in grain size. The MSU studies
focused on fine to medium sand (overall,
0.180-0.30 mm). Denson and co-workers, including
Denson and Gill (1965) reported data on very fine
sand (0.063-0.125 mm), and in some cases coarse
silt. However, the data of Kight (2002) for three
separate grain size ranges do not indicate any
significant trends in mineral percentages as a
function of grain size. Also, although no formal
results are available for very fine sand
fractions of MSU samples, preliminary XRD
analysis of the very fine sand fraction heavy
minerals of the Medicine Pole Hills sample shows
a diffractogram that is nearly identical to that
of the 0.25-0.30 mm fraction. One aspect of the
heavy mineral abundance that is perhaps more
likely to reflect differences in grain sizes used
and/or sample processing techniques relates to
opaque grains. It is possible that the high
percentage of opaque grains in the Chadron
Formation samples of Denson and Gill (1965) was
due to small aggregate grains that included iron
oxides and hydroxides (ferruginous and clay
aggregates, and possibly leucoxene). These
aggregate grains may have been much more abundant
in the very fine sand and silt grain
size range than in fine to medium sand. Sample
processing used for MSU samples (processing in a
blender and/or ultrasonic bath) was designed to
break down aggregate grains. A final factor that
must be considered when comparing MSU data with
the Chadron Formation data is that Denson and
Gill (1965) reported an average of 23 samples,
and that they included some samples from the
upper part of the Chadron Formation (i.e., the
mudrocks of the South Heart Member). However, a
review of the stratigraphic sections (with sample
locations) presented by Denson and Gill (1965) in
their Plates 3 and 4 show that most samples (13
of the 17 for which stratigraphic sections were
shown) did come from the lower part of the
formation (i.e., the Chalky Buttes Member).
Comparison of the average non-opaque heavy
mineral compositions of the 13 Chalky Buttes
Member and four South Heart Member samples showed
they were very similar. The South Heart Member
samples had only slightly lower staurolite (14.8
versus 22.2) and slightly higher zircon (33.8
versus 28.3) abundances. Thus, it seems unlikely
that their average Chadron Formation heavy
mineral abundances were significantly affected by
South Heart Member samples. Implications for
Correlations The heavy minerals found in the MSU
studies and work of Denson and Gill (1965) were
divided into four groups based on
11
12
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
source type. Three groups correspond to the three
general sources identified by Denson and Chisholm
(1971) Tertiary volcanics, Precambrian
metamorphic and plutonic rocks, and recycled
grains, with the recognition that the
ultra-stable grains of the latter could also be
derived from Precambrian plutonic sources. The
fourth group consist of minerals that could be
derived from Tertiary volcanics or Precambrian
rocks. Minerals interpreted to be from a
Tertiary volcanic source were biotite and the
pyroxenes. Biotite was interpreted to be from a
Teriary volcanic source because of the rather
Mg-rich composition of biotite in the Rattlesnake
Butte sample (Webster, unpublished data). While
biotite can be derived from Precambrian sources,
this interpretation would have little effect on
other samples because only the Chadron Formation
contains biotite, and only a small amount. A
significant question regarding pyroxenes was the
nature of the source of the very abundant
diopside in the Medicine Pole Hills sample. Hicks
(2007) separated diopside grains for trace
element analysis. Based on the concentrations of
compatible trace elements (Cr, Ni, Sc, and V), he
concluded that the diopside was derived from an
igneous source rather than a metamorphic
(meta-carbonate) source. The diopside was similar
to phenocrysts known to exist in Tertiary shallow
mafic intrusive rock (basaltic trachyandesite)
exposed at Sugarloaf, in the southern part of the
Bear Lodge uplift, northwest of the Black Hills
(Webster and Bell 1996). Whatever the source of
the diopside, it
is significant that little diopside is reported
in Tertiary strata by Denson and co-workers.
Denson and co-workers published nine 30 x 60
geologic maps (eight in northeastern Wyoming, one
in southeastern Montana) that show heavy mineral
sample locations and list data for individual
samples. Most samples were from Paleocene strata
(Fort Union Group) and the Eocene Wasatch
Formation. A very few Wasatch samples contained
up to 4 diopside, but there was one sample
listed with 60 diopside (near the center of the
Buffalo 30 x 60 quadrangle Denson et al.,
1990). This is the only sample (other than the
Medicine Pole Hills) found to date with abundant
diopside. Minerals interpreted to be from
Precambrian metamorphic/plutonic sources included
the metamorphic minerals (garnet,
epidote-clinozoisite, talc, aluminosilicates, and
staurolilte). They also included sphene,
monazite, allanite, and corundum, which are
accessory minerals that could be metamorphic or
plutonic in origin. The ultra-stable minerals
zircon, tourmaline, and rutile, were placed in
the recycled/plutonic group. These could have
been recycled, derived from Paleozoic or Mesozoic
strata eroded around Laramide uplifts, or from
Precambrian plutonic rocks exposed in the
uplifts. Two minerals, hornblende and apatite
were placed in a fourth category recognizing they
could be derived from Tertiary volcanics or
Precambrian rocks, although it should be noted
that apatite grains could also be bone fragments.
The interpretation of hornblende is
perhaps most significant because of its abundance
in the Medicine Pole Hills sandstone. Denson and
Chisholm (1971) recognized three color varieties
of hornblende, interpreting blue-green hornblende
as metamorphic (Precambrian) in origin, and both
green-brown and red-brown varieties as having a
Tertiary volcanic source. All of these colors
were found in the Medicine Pole Hills sandstone,
but only two varieties were categorized
green-brown and red-brown. Some grains were more
of a blue-green color, but were placed in the
green-brown category. SEM-EDS analyses of the
hornblendes showed there was a compositional
distinction between the two varieties, but that
they all formed a continuum. The analyses did not
indicate that there was a third compositional
type of hornblende. It might be reasonable to
interpret the Medicine Pole Hills hornblende as
derived from Tertiary volcanic sources, but
because of the uncertainty, it is placed in this
fourth category. When the heavy minerals are
grouped as described, the analyses indicate that
each of the samples had distinct sources, or at
least distinct proportions of multiple sources.
The generalized source contributions are
illustrated in the pie diagrams of Figure 4. All
four are rather different from one another.
Perhaps the most similar are the Chadron
Formation and Square Butte analyses, both
dominated by Precambrian metamorphic/plutonic and
recycled/plutonic sources. However, the
similarity breaks down when considering
individual minerals. While
12
13
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
the assemblages of Precambrian metamorphic
minerals are rather similar, the ultra-stable
grains are not. The Chadron Formation is
dominated by zircon, and the Square Butte sample
by rutile. The Medicine Pole Hills and
Rattlesnake Butte samples are similar in that
they are dominated by Tertiary volcanic and
Precambrian metamorphic/ plutonic source
minerals, but in different proportions. Again,
the modest similarity breaks down when individual
minerals are considered. While the Precambrian
metamorphic minerals are similar in that
epidote-clinozoisite gt garnet, the individual
Tertiary volcanic minerals are quite distinct.
The Medicine Pole Hills sample is dominated by
diopside, while the Rattlesnake Butte sample is
dominated by biotite. Thus, it seems it was not
simply a matter of different proportions of the
same two sources. In short, none of the three
samples studied at MSU are similar to Chadron
Formation as described by Denson and Gill (1965).
Also, the Medicine Pole Hills heavy minerals are
distinct from those found in the two Chalky
Buttes Member samples. Overall, the results to
date suggest that the Medicine Pole Hills
sandstone should not be correlated with the
Chalky Buttes Member of the Chadron Formation as
suggested by Murphy et al. (1993). The
dissimilarities among the two Chalky Buttes
Member samples and the data of Denson and Gill
suggest that correlation, in the strict sense, of
sandstones mapped as the Chalky Buttes Member
should be viewed with suspicion. Additional work
on Chalky Butte Member samples may well
Figure 4 Comparison of Eocene sandstone sources
as interpreted from heavy mineral analyses.
Chadron Fm Chadron Formation data from Denson
and Gill (1965) MPH Medicine Poles Hills data
from Kight (2002) and Kuhnhenn (2005) RSB
Rattlesnake Butte data from Webster, reported in
Nelson (2008) SQB Square Butte data from
Fogarty and Webster (unpublished).
13
14
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
reveal a complex depositional setting consisting
of multiple depositional systems receiving
sediment from multiple, changing sources through
time. FUTURE RESEARCH To continue to address
the question of correlation of Eocene sandstones
of North Dakota, research is needed along several
lines. First, heavy mineral analysis of
additional Chalky Buttes Member sandstone samples
will be important. Understanding the geographic
variability will be essential for a better basin
analysis of this sandstone, and to further test
whether the Medicine Pole Hills sandstone could
possibly fit with this variability. Initially,
samples from three more locations will be
studied. Samples from the Little Badlands and
Chalky Butte (Figure 1) have been processed and
are ready for heavy mineral analysis. Samples
from the Slim Buttes area (northwestern South
Dakota) are also available for study. Two other
lines of research will also contribute to the
overall study. Pebbles collected during the 2002
fieldwork need to be studied in more detail.
Whitlow (2003) characterized (grouped) pebble
types from some samples, but there are others
that need to be studied. Also, more detailed work
on the pebbles is needed. Thin section
petrography and whole-rock chemical analysis will
allow more detailed comparison of the igneous
pebbles in particular. A class project in GEOL
305 at MSU started this work on the Rattlesnake
Butte sample, and one
Denson NM, Macke DL, Schumann RR. 1990. Geologic
map and distribution of heavy minerals in
Tertiary and uppermost Cretaceous rocks of the
Buffalo 30 x 60 Quadrangle, Johnson and
Campbell Counties, Wyoming. USGS Misc. Inv.
Series, Map I-2022. Frederick K. 2004. Grain size
comparison of sands from the Chalky Buttes
Member, Chadron Formation, southwestern North
Dakota. Senior Seminar Paper, Minot State
University. 47 p. Hicks J. 2007. A geochemical
study of diopside found in the Eocene sandstone
of the Medicine Pole Hills in southwestern North
Dakota Senior Seminar Paper. Minot (ND) Minot
State University. 26 p. Larson AC, Von Dreele RB.
1994. Generalized Structure Analysis System
GSAS. Los Alamos National Laboratory Report LAUR
86-748. Kight R. 2002. Heavy mineral analysis of
sand-stones using Rietveld X-ray diffraction.
Senior Seminar Paper, Minot State University. 17
p. Kuhnhenn S. 2005. Heavy mineral analysis and
provenance of a sandstone from the Medicine Pole
Hills of North Dakota Senior Seminar Paper.
Minot (ND) Minot State University. 21 p. Murphy
EC, Hoganson JW, Forsman NF. 1993. The Chadron,
Brule and Arikaree Formations in North Dakota.
North Dakota Geologic Survey Report of
Investigation 96. 144 p. Sato Y, Denson NM.
1967. Volcanism and tectonism as reflected by the
distribution of nonopaque heavy minerals in some
Tertiary rocks of Wyoming and adjacent states.
USGS Prof. Paper 575-C, p. C42-C54. Schumaker K.
2003. Marsupials, multituberculates, and
insectivores from the Medicine Pole Hills Local
Fauna, Bowman County, North Dakota. Senior
Seminar Paper, Minot State University. 71 p.
student is continuing that study, but work on
igneous pebbles from other locations is
needed. Finally, additional work on the major
mineralogy of the sandstones is needed. Arneson
(2005) carried out a comparison of major grains
by optical microscopy. Additional work could
characterize the major minerals in more detail.
Work on the light mineral grains recovered during
the heavy mineral studies is needed. A
combination of optical microscopy and SEM-EDS
microanalysis would provide more detailed
information on the abundances and compositional
ranges of feldspar grains.
REFERENCES Arneson C. 2005. Lithologic comparison
of sand grains of the Chadron Formation at three
localities in southwestern North Dakota. Senior
Seminar Paper, Minot State University. 37
p. Brayko A. 2007. Analysis of errors due to
microabsorption and Bindley correction within
complex mixtures. Senior Seminar Paper. Minot
(ND) Minot State University. 17 p. Cool C.
2006. Heavy mineral analysis of sandstone using
X-ray diffraction and the Rietveld method Senior
Seminar Paper. Minot (ND) Minot State
University. 11 p. Denson NM, Gill JR. 1965.
Uranium-bearing lignite and carbonaceous shale in
the southwestern part of the Williston Basin - a
regional study. USGS Prof. Paper 463, 75
p. Denson NM, Chisholm WA. 1971. Summary of
mineralogic and lithologic characteristics of
Tertiary sedimentary rocks in the Middle Rocky
Mountains and the northern Great Plains. USGS
Prof. Paper 750-C, p. C117-C126.
14
15
Heavy Minerals of Tertiary Strata
Introduction
Methods
Results
Discussion
Future Research
References
Toby BH. 2001. EXPGUI, a graphical user
interface for GSAS, J. Appl. Cryst. 34,
210-213. Webster JR, Bell KJ. 1996. Tertiary
volcanics of Sugarloaf Mountain, southeastern
Bear Lodge Mountains, Crook County, Wyoming
(abs) Geological Society of America Abstracts
with Programs, v. 28 Webster JR, Kight RP,
Winburn RS, Cool CA. 2002. Heavy mineral
analysis of sandstones by Rietveld analysis.
Advances in X-Ray Analysis 46198-203. Whitlow T.
2003. Lithology of pebbles collected in Eocene
sandstones of southwest North Dakota. Senior
Seminar Paper, Minot State University. 16 p.
15