-
A Thesis
Presented to
The Graduate Faculty of The University of Akron
In Partial Fulfillment
of the Requirements for the Degree
Master of Science
DEVELOPING A USEFUL SET OF PROXY ELEMENTS FOR THE TARGETING
AND EXPLORATION OF GOLD DEPOSITS, BLACK HILLS, SOUTH
DAKOTA
Michael T. Harp
December, 2010
-
Approved:
______________________________ Advisor
______________________________Faculty Reader
______________________________Faculty Reader
______________________________Department Chair
Accepted:
______________________________Dean of the College
______________________________ Dean of the Graduate School
______________________________Date
Thesis
ii
Dr. LaVerne M. Friberg
______________________________
Dr. John A. Peck
______________________________
Dr. John P. Szabo
______________________________
Dr. John P. Szabo
Dr. Chand K. Midha
Dr. George R. Newkome
DEVELOPING A USEFUL SET OF PROXY ELEMENTS FOR THE TARGETING
AND EXPLORATION OF GOLD DEPOSITS, BLACK HILLS, SOUTH
DAKOTA
Michael T. Harp
-
ABSTRACT
iii
X-Ray Fluorescence (XRF), petrography, and Energy-Dispersive
X-ray Spectrometry
(EDAX) have been used to determine the element concentration in
samples and their
distribution within minerals for 222 metamorphic rock samples
from the Black Hills, S.D.
Element concentrations in these samples are compared to sample
location and known
gold deposits in the Black Hills.
XRF data of rock chips from whole rock samples were collected
using a portable
XRF unit to determine major, minor and trace element abundances.
Statistical analyses
of the XRF data indicates a moderate to strong correlation
between gold and the elements
Mn (19 to 16,116 ppm), S (1,283 to 79,452 ppm), As (n.d. to 132
ppm), Pb (n.d. to 318
ppm), Cl (625 to 31,277 ppm), Ba (n.d. to 1,101 ppm), and Zn
(n.d. to 266 ppm) thus
indicating these elements may serve as proxy indicators of gold.
The integration of
elemental data with ArcGIS was used to test the spatial
relationship of proxy elements to
known gold deposits in the Precambrian core of the Black
Hills.
Sixteen samples having gold concentration greater than 18 ppm
were chosen for
more detailed analyses. EDAX raster scans of these samples
determined proxy element
variations within individual mineral grains. Petrographic
analyses were done to identify
minerals and their textural relationships.
Sample proximity to known gold deposits in the Black Hills can
be correlated with
increases in minor and trace proxy element concentrations.
-
iv
ACKNOWLEDGEMENTS
First and foremost I would like to thank my advisor Dr. LaVerne
Friberg. His
guidance and knowledge gave this project life as well as my
interest in this field.
Because of him and this research I have discovered my passion in
the vast field of
geology. I would like to thank the Department of Geology and
Environmental Science
at the University of Akron for the use of the departmental
equipment and facilities that
aided in this research, but most especially for allowing me to
become part of the graduate
program that allowed me to get this far. I would like to thank
Dr. John Szabo and Dr.
John Peck for their willingness to be part of my thesis
committee and their guidance
throughout my Masters program. XRF analysis was conducted
through an Academic
and Research Relations Grant to The Department of Geology and
Environmental Science
provided by Innov-X Systems.
I would like to extend special thanks to Mr. Tom Quick for his
expert knowledge,
willingness to help me succeed, his ability to fix all that goes
wrong, and for being ever
present throughout my research. To Dr. Kevin Butler for his
expert advice and assistance
with all things related to ArcGIS. To Ms. Elaine Butcher for her
guidance through the
processes and procedures that come with the Masters program, for
helping me to learn
the formatting and the finishing of my thesis, but most
importantly for her friendship
and for being there when things were at their best and their
worst. Finally I would like
to thank my fellow graduate and undergraduate students, who
provided me with their
support, friendship, and their help in keeping my eyes on the
horizon when my family
couldnt be there.
-
vI would like to thank my parents Steve and Cheryl, for giving
me a solid base, a good
work ethic, their interest in my research, and for believing in
me all the way. I would
also like to thank my grandmother, Dorothy Hoffman for being a
mentor and an ever
present figure in my life. I would like to thank my brother
Brian and my sister Amanda
for keeping my head up and their support through this project.
My son Ayden and my
daughter Delaney for giving me a reason to keep going and to
make myself better in
every way. Finally I would like to thank my wife Taryn. She is
my biggest fan, my
strongest supporter, and my best friend. Her interest in my life
kept the fire burning and
without her none of this could have been possible.
-
vi
Page
TABLE OF CONTENTS
LIST OF FIGURES viii
LIST OF TABLES x
CHAPTER
I. INTRODUCTION 1
Overview 1
Geologic Setting 2
Previous Works 10
II. METHODS 13
Sample Locations 13
Laboratory Methods 13
X-ray Fluorescence 13
Energy Dispersive X-ray Analysis 17
ArcGIS Analysis 19
Statistical Analysis 20
Petrographic Analysis 20
III. RESULTS 22
X-ray Fluorescence and ArcGIS 22
Statistical Analysis 33
Petrographic Analysis 37
Energy Dispersive X-ray Analysis 47
IV. DISCUSSION 49
-
vii
Chlorine as a Predictor for Gold 49
Barium as a Predictor of Gold 51
Arsenic as a Predictor of Gold 51
Manganese as a Predictor of Gold 54
Sulfur as a Predictor of Gold 54
Zinc as a Predictor of Gold 57
Lead as a Predictor of Gold 57
Spatial Analysis of the Proxy Elements 57
Petrographic and EDAX analysis 63
Manganese 64
Barium 69
Sulfur72
V. CONCLUSIONS79
REFERENCES 81
APPENDICES 84
APPENDIX A. LATITUDE AND LONGITUDE FOR SAMPLE LOCATIONS 85
APPENDIX B. STATISTICAL DATA FOR ROCK CHIP ORIENTATION 91
APPENDIX C. XRF DATA 95
APPENDIX D. DATA FOR EDAX AND MICROPROBE COMPARISON 136
APPENDIX E. STATISTICAL DATA BASED ON XRF BULK ELEMENT ANALYSIS
144
APPENDIX F. MINERAL ASSEMBLAGE 156
-
viii
FigurePage
LIST OF FIGURES
1 Generalized diagram showing the geology and geomorphology of
the Black Hills, SD 3
2 Geologic cross section of the Black Hills after the Laramide
Orogeny, SD (Carter, et al., 2003) 4
3 Geologic map of the Black Hills, SD (Modified after Dahl et
al., 2005a) 6
4 Precambrian area and Metamorphic Isograds of the Black Hills,
South Dakota 9
5 Precambrian area and Metamorphic Isograds of the Black Hills,
South Dakota 14
6 Elemental concentration variance based on sample orientation
16
7 EDAX dot map scan of sample BHMA-27a 18
8 Gold concentration (ppm) contour map of the Precambrian area
of the Black Hills, South Dakota using the kriging method 23
9 Arsenic concentration (ppm) contour map of the Precambrian
area of the Black Hills, South Dakota, using the kriging method
25
10 Barium concentration (ppm) contour map of the Precambrian
area of the Black Hills, South Dakota, using the kriging method
26
11 Chlorine concentration (ppm) contour map of the Precambrian
area of the Black Hills, South Dakota, using the kriging method
28
12 Manganese concentration (ppm) contour map of the Precambrian
area of the Black Hills, South Dakota, using the kriging method
29
13 Lead concentration (ppm) contour map of the Precambrian area
of the Black Hills, South Dakota, using the kriging method 31
14 Sulfur concentration (ppm) contour map of the Precambrian
area of the Black Hills, South Dakota, using the kriging method
32
15 Zinc concentration (ppm) contour map of the Precambrian area
of the Black Hills, South Dakota, using the kriging method 34
-
ix
16 Precambrian area and Metamorphic Isograds of the Black Hills,
South Dakota. 38
17 Cross plot graph of gold versus chlorine concentration 50
18 Cross plot graph of gold versus barium concentration 52
19 Cross plot graph of gold versus arsenic concentration 53
20 Cross plot graph of gold versus manganese concentration
55
21 Cross plot graph of gold versus sulfur concentration 56
22 Cross plot graph of gold versus zinc concentration 58
23 Cross plot graph of gold versus lead concentration 59
24 Dot map indicating manganese concentration in garnet 65
25 Photomicrograph of sample BHMA-57b 66
26 Dot map indicating iron concentration in garnet 67
27 Dot map indicating tellurium concentration in garnet 68
28 Peak data indicating occurrence and intensity of elements
within sample BH-19a 70
29 Photomicrograph of sample BH-19a 71
30 Dot map indicating barium concentration in titanite 73
31 Dot map indicating tellurium concentration in titanite 74
32 Peak data indicating occurrence and intensitiy of elements
within sample BH-4 75
33 Dot maps indicating sulfur and iron concentrations in pyrite
76
34 Photomicrograph of sample BH-4 78
-
xTable Page
LIST OF TABLES
1 Correlation coefficients between paired elements: r is
significant when P 0.05 36
-
1CHAPTER I
INTRODUCTION
Overview
According to Rambeloson (1999), Gold occurs in four main kinds
of deposits: 1)
as a diffuse component of crystalline basement rocks, 2) in
concordant quartz veins
within the metamorphic rocks of the Precambrian basement, 3) in
recent discordant
veins, and 4) in recent and ancient alluvial deposits. Gold in
the Black Hills occurs as
four types of deposits: 1) within Precambrian quartz veins that
have been injected into
the metamorphic basement rock, 2) in ancient placer deposits
within the Deadwood
Formation, 3) as hydrothermal deposits associated with Tertiary
igneous activity, and 4)
as modern placer deposits. This study will focus on gold
deposited along quartz veins and
as hydrothermal deposits in the Precambrian metamorphic
rocks.
In the Black Hills there are many locations where gold mining
occurred in the
past. The historic sites of mining are the Lead-Deadwood
District, Rochford-Hill City
District, and The Keystone District. The Lead-Deadwood District
is encompasses the
town of Deadwood in eastern Lawrence County and the city of Lead
in central Lawrence
County, which is also the central area of the mineralized zone.
This area also contains
the Homestake Mine. The Rochford-Hill City District is located
in the western portion of
Pennington County in the vicinity of Hill City, near the
headwaters of Spring Creek and
extending into the city of Rochford to the northwest. The
Keystone District is located in
-
2western Pennington County on the northeastern side of Harney
Peak near (Koschmann et
al., 1968).
Using a collection of 222 samples taken from the Black Hills,
the purpose of this
study is to assess if minor elements occur within silicate
minerals, and if they can be used
as proxies for the presence of gold. Due to the conditions at
which gold and these proxy
elements are mobilized and then deposited by metamorphic fluids,
it can be hypothesized
that the proxy elements will substitute into the crystal
lattices of metamorphic
silicate, oxide, and sulfide minerals in occurrence with gold.
Because of the similar
electro-chemical characteristics of the proxy elements and gold,
I will show that gold
concentrations are in areas of the Black Hills where Mn, S, As,
Ba, Pb, Zn, and Cl are in
higher concentration.
Geologic Setting
The Black Hills is an elliptically domed region in the
southwestern portion of South
Dakota that extends into the northeastern portion of Wyoming
(Figure 1). The area is 200
km long and about 105 km wide, with its highest point being
Harney Peak at an elevation
of 2207 m. The Black Hills is an area that has been subjected to
multiple geologic events,
including mountain-building episodes, igneous intrusions, and
polymetamorphism related
to the tectonic episodes of the area.
As the Black Hills were uplifted by the Laramide Orogeny and
eroded throughout the
late Precambrian and into the Paleozoic, the last rocks to be
deposited were sedimentary
and dip away from the granitic core. The Homestake mine, located
in the northern part
of the Black Hills, is the location of the first discovery of
gold in the Black Hills. The
area is composed of Precambrian schists that are surrounded by
steeply outward-dipping
Paleozoic and Mesozoic rocks (Noble, 1950) (Figure 2).
-
3Figure 1. Generalized diagram showing the geology and
geomorphology of the Black Hills, SD. (Modified from Strahler and
Strahler, 1987).
-
4Fig
ure
2. G
eolo
gic
cros
s se
ctio
n of
the
Bla
ck H
ills
aft
er th
e L
aram
ide
Oro
geny
, SD
(C
arte
r, et
al.,
200
3)
-
5Exposed Precambrian rock is believed to be the source of
paleo-placer deposits that
are present in the Deadwood Formation (Noble, 1950). The
Deadwood Formation, which
is mostly sandstone, was another source of gold discovered in
the Black Hills as placer
deposits in creeks where the gold was derived from Precambrian
rock (Rahn et al., 1996).
As erosion occurred, gold was transported by the river systems
and deposited in paleo-
channels within the Cambrian-aged Deadwood Formation.
During the Laramide Orogeny, which occurred in late Cretaceous
into the early
Cenozoic, uplift intensified as deformation of the rock
continued and another episode of
hydrothermal alteration associated with Tertiary igneous
activities occurred, depositing
many economic minerals, including gold (Figure 3). Tertiary
intrusive dikes have also
been known to carry high concentrations of Zn and Ba in the
northeastern portion of the
Black Hills where remobilization of gold is believed to have
occurred (Uzinlar, 2010).
According to Dewitt et al. (1996), XRF analysis indicated that
barium occurred in high
abundance within a range of 580-1,700 ppm to the southeast of
Deadwood and lower
concentrations centered on the Whitewood Peak pluton to the
northeast of Deadwood.
High abundance of barium can be correlated to intrusion of
Tertiary dikes that are
prevalent in the northern Black Hills.
Due to igneous, metamorphic and sedimentary processes that have
acted on the
Black Hills throughout time, the area has become a location
known for its rare minerals
and shows evidence of geological processes that acted upon these
rocks. Multiple
episodes of deformation, uplift, and hydrothermal fluid
activities have been preserved
in the rocks in the Black Hills area that spans geologic time
from the Precambrian to the
present.
The Proterozoic thermotectonic and magmatic history of the Black
Hills crystalline
core is associated with arc accretion and continental collision
(Redden et al. 1990; Dahl
et al., 2005a, b, 2006; Nabelek et al., 2006). In a study done
by Frei et al., (2009), The
-
6Figure 3. Geologic map of the Black Hills, SD (Modified after
Dahl et al., 2005a).
-
7mode of occurrence of gold at Rochford is strikingly similar to
that in the Homestake
Iron Formation of the Lead District (Slaughter, 1968; Bayley,
1972). However the two
intracratonic basins developed independently from each other, in
space and time, i.e, ~25
km and ~80-130 Myr apart. Results obtained from this study will
identify similarities in
petrogenic origins of the Rochford Iron Formation and deposition
of the Homestake Iron
Formation which is constrained within a 2012-1974 Ma time frame
(Frei, et al., 2009).
The structural history of this area includes major tectonic
rifting and convergence
during the Proterozoic that caused multiple metamorphic episodes
in preexisting
basement rock. As the area began to rift, a period of rapid
erosion of Archean basement
rock to the west from the Wyoming Craton resulted in deposition
of over 3000 m of
sedimentary rock. As rifting stopped, plates were forced back
together causing structural
deformation and metamorphism of sedimentary rock (Dahl &
Frei, 1998).
In the Precambrian Era, the granitic core of the Harney Peak
area of the Black Hills
was formed when magma forced its way into existing rocks during
the Trans-Hudsonian
Orogeny (Dahl & Frei, 1998). The Trans-Hudsonian Orogeny was
the collision of the
Wyoming and Superior Cratons and accretion of arcs along the
southeastern margin of
the Wyoming Craton (Van Boening & Nabelek, 2008).
Intrusion of the Crook Mountain and Harney Peak magmatic bodies
are associated
with the Black Hills dynamothermal metamorphic event as well as
localized contact
metamorphism. Following closely to emplacement of this large
buried pluton in the
northeast, there was continuation of the magmatic event in which
large pegmatite bodies
were emplaced around the Harney Peak granite core of the Black
Hills. The Harney
Peak granitic core has been dated to 1.75 Ga years with
surrounding sedimentary rocks
being dated from 1.8 to 1.9 Ga years. The maximum age for
mineralization is 1,746
+/- 10 Ma as indicated from step-leach Pb-Pb dating of garnet
from mineralized samples
in the Homestake Mine (Terry et al., 1998). This tectonic
episode uplifted, eroded, and
-
8tilted rocks in the area as the diapir of magma rose through
rock layers causing contact
metamorphism with metasedimentary layers giving the area a bulls
eye appearance
with all surrounding rock dipping away from the center (Dahl
& Frei, 1998).
This area can be broken into areas of equal grade of
metamorphism or metamorphic
zones (Figure 4). The highest grade metamorphic rocks are in the
sillimanite zone which
extends west of the Harney Peak Granite. Temperature dropped to
the north where a
staurolite zone occupies an area a few kilometers wide. Farther
north is a broad area
belonging to the garnet zone. Northeast of the garnet isograd,
the biotite zone extends
from west of Lead to the southeast (Redden et al., 1975).
Heat and pressure applied to the rocks decreased with distance
from the Harney Peak
Granitic intrusion. The area that is preferential to deposition
of epigenetic gold is within
the biotite zone of metamorphism. Temperatures in the biotite
zone were less than 350
degrees Celsius. Gold deposition is only found within the
biotite zone, and as the garnet
isograd is crossed, deposition of gold ceases. The Homestake and
Rochford mining
districts, where gold has been actively mined, lie within the
biotite zone of the northern
Black Hills.
As hydrothermal fluids move through rock and interact with
grains of biotite,
chlorite, and garnet, proxy-element exchange occurs between the
fluid and the minerals.
In the case of biotite and chlorite structures, most of the
proxy elements in hydrothermal
solution substitute into octahedral sites, whereas in the
garnets they will substitute into
cubic or octahedral sites (Klein, 2002). This exchange should
occur at the rims in the
highly refractory garnet and penetrate into the interior of the
less refractory micas and
opaques along grain boundaries and along fractures and cleavage
planes.
-
9Figure 4. Precambrian area and Metamorphic Isograds of the
Black Hills, South Dakota. Black line indicates the Precambrian
boundary. Red lines indicate metamorphic isograds boundaries
showing level of metamorphism of the area. Brown Area indicates the
Harney Peak Granite (HPG).
-
10
Previous Works
High manganese concentrations have been discovered within black
smokers, or
hydrothermal vents that form at spreading centers due to
hydrothermal circulation
(Zierenberg et al., 1993). Anomalous concentrations of manganese
commonly occur in
or near sulfide ore environments. These anomalies occur as
manganiferous limestone
horizons (Russell, 1974; Gwosdz and Krebs, 1977), as
manganiferous garnet lithologies
within, above, or beneath metamorphosed massive sulfide deposits
(Spry, 1978; Stumpfi,
1979), and as ferro-manganiferous sediments associated with
ancient and active mid-
ocean spreading centers (Robertson and Hudson, 1973; Alt et al.,
1987). Seafloor
manganiferous sediments can arise from a number of processes,
some of which are
not related to sulfide mineralization. These processes include
halmyrolysis and occurs
between basalt and sediment, low-temperature precipitation as
nodules and crusts, and
diagenetic enrichment in the sediment column. The anomalous
manganese and sulfide
occurrence of the Black Hills, as well as gold, may have been
deposited at an ancient
spreading center during rifting associated with the
Trans-Hudsonian Orogeny (Dahl &
Frei, 1998).
In a study done by Redden (1990), an imprecise upper-intercept
207Pb/206Pb age of
1,884 29 Ma was obtained for bulk zircons in felsic tuff
interlayered with the Montana
Mine Formation that underlies the Rochford Formation. Dahl et
al., (2008) improved
this age constraint to 1,887 7 Ma U-Pb SIMS age from the same
felsic tuff. This age
constrained a maximum depositional age of ~ 1.887 Ga for the
Rochford Iron Formation.
A tuffaceous layer within the Ellison Formation, which overlies
the Homestake
Formation, was dated at 1,974 8 Ma (Redden et al., 1990) and
constrains a minimum
deposition age of ~1.974 Ga for the Homestake Iron Formation.
Therefore, gold
-
11
deposition of this area is constrained by the ages at which the
formations were deposited
and gold mobilization began.
A possible mode of gold deposition favors an epigenetic origin
for the Homestake
gold deposit, while also inferring a strong genetic association
of the gold event with the
late stages of nearby granite magmatism (1.75 Ga) (Caddey et
al., 1991). According
to Frei et al. (2009), the timing of Homestake gold
mineralization has been estimated
at ~1,730 Ma from Re-Os dating of arsenopyrite (Morelli et al.,
2005), which falls
within the known ~1,780-1,715 Ma interval of regional
metamorphism and igneous
emplacements.
In a study done by Caddey et al. (1991), gold-sulfide
mineralization in the Homestake
Iron Formation and in the Rochford district (Bayley, 1972) was
found to be hosted by
quartz veins that were formed during retrograde shearing. Three
sequential stages of
quartz veins (stage I, II, and III), associated with ductile,
ductile-brittle, and semi-brittle
shear zones, respectively, have been recognized and described in
the Homestake Mine
area
In a study by Paige (1924), the timing of sulfide mineralization
and gold deposition
has been delineated. All ores of the Homestake lode carry
sulfides; and generally,
where sulfides are abundant, the best ore is found. Sulfides
that occur are arsenopyrite,
pyrrhotite, and pyrite. Gold is associated with each of these
minerals either as inclusions
within them or in gangue minerals that are close by. These
sulfides replaced portions
of the carbonate schist and conform to the schistose structure
of the rock. Based on
evidence in this study, sulfides were introduced before final
stages of metamorphism.
Arsenopyrite was introduced at late stage metamorphism of
schists and was partly
deformed; and then shortly after pyrrhotite and pyrite were
introduced as gold
mineralization occurred.
-
12
Previous studies at the Homestake underground mine in the
northern Black Hills
show that the manganese content of chlorite increases with
proximity to gold-bearing
quartz veins (Armstrong and Friberg, 1998). This study showed
that manganese variation
in chlorite was not directly correlated to metamorphic grade or
rock type. Manganese
content within the chlorite weakly to moderately correlates with
manganese content in
biotite and garnet within a sample, but more strongly correlates
with high concentrations
in close proximity to the gold mining districts.
In a similar study by Friberg et al. (1997), chlorite occurs in
the greenschist facies
(biotite grade) through the lower amphibolites facies
(staurolite grade) rocks having
a wide range of composition. This study showed that chlorite
formed during the
dynamothermal event associated with the emplacement of the
Harney Peak Granite,
pegmatites, and the late quartz veining and was re-equilibrated
with the associated
mineralizing fluids which introduced higher concentrations of
manganese along the rims
and cleavage planes in the chlorite.
The chemistry of chlorite appears to be controlled by bulk
composition of the host
rock, metamorphic intensity-related chemical exchanges within
the coexisting minerals,
as well as mineralizing fluids that are associated with quartz
veins (Friberg et al., 1997).
This study also showed that manganese contents of chlorite
generally increase with
metamorphic grade. As zones of higher grade metamorphism are
crossed, manganese
within chlorite increases. In addition, manganese content in
chlorite also increases
toward known gold deposits in the chlorite-biotite grade
rock.
-
13
CHAPTER II
METHODS
Sample Locations
Samples used in this research were collected for previous
research across the
Precambrian core of the Black Hills in 1977 by Dr. L.M Friberg
and 1998 by M.
Armstong. The 222 samples analyzed in this study are samples
collected from all grades
of metamorphism within the Precambrian core of the Black Hills.
Samples were taken
and locations of collection marked (Figure 5, Appendix A).
Laboratory Methods
Samples used in this research were both cut rock chips and
polished thin sections.
Samples were analyzed using X-ray fluorescence (XRF), energy
dispersive X-ray
analysis (EDAX) attached to an environmental scanning electron
microscope (ESEM),
ArcGIS; and full petrographic analysis was conducted on 85
samples.
X-ray Fluorescence
Bulk elemental composition was determined using a handheld
Innov-X XRF analyzer
(Model Alpha). Standard soil mode was used to obtain elemental
compositions present
samples. A standard was inserted in order to calibrate the
analyzer at the beginning of
-
14
Figure 5. Precambrian area and Metamorphic Isograds of the Black
Hills, South Dakota. Red dots indicate sample locations collected
by L.M. Friberg (1977) and M. Armstrong(1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds showing level of metamorphism of the area. Blue line
indicatesouter boundary of Harney Peak Granite. Green lines
indicate Rochford and Homestake Mining Districts.
!
!!
! !
!!!!
!! !
!!
!!
!!!
!!!
!
!!
!!
!! !
!! ! !!
! !!
! !!
!!
!!
! !
! !
! !! !
! !!
! !!
!
!!
!! !
!!
!!
! !! !
! ! !
! !!
!
! !
!!
!!
!
!
! ! ! ! !!
! !
!!
!
!
!!
!
!!!!!!!!
!!
!!!
!!
!!!
!!
!
!!!!
!!
!!
!
!
!
!
!
!!
!!
!!
!
!!!! !!
! !!!
!!!!
!!
!
!
!
!!
!
!
!
!!!
!!
!
!
!
!
Custer
Rapid City
Lead
Rochford
Deadwood
0 10 205Kilometers
Key_Cities_BH! Sample Locations
Harney Peak GraniteMetamorphic IsogradsMining DistrictsBlack
Hills Area
Figure x. Precambrian area and metamorphic isograds of the Black
Hills, South Dakota. Red dots indicate sample locations collected
by L.M. Friberg (1977) and M. Armstrong(1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds showing level of metamorphism of the area. Blue line
indicatesouter boundary of Harney Peak Granite.Green Lines indicate
Rochford and Homestake Mining Districts
-
15
analysis and was reinserted after 20 consecutive samples had
been run to insure that the
analyzer was still operating properly and the system remained
clean of foreign particles.
Samples used for the XRF analysis were unpolished rock chips
that are were cut into
1x3x1-inch sized chips. Samples were wrapped in plastic wrap in
order to keep particles
off the analyzer and the rock chips were placed in the apparatus
for 2-minute intervals.
For the first 23 samples, different orientations were used to
ascertain whether
placement of the sample into the apparatus affected the
elemental concentration
measurement. The orientations chosen were 90, 135, and 225 from
the vertical
position. Orientations were graphed using a scatter plot to test
for variance between
samples (Figure 6, Appendix B).
For each sample and orientation, concentrations of proxy
elements were summed to
obtain proxy element abundance. Proxy element abundance for each
of the 3 orientations
was plotted. How well the trend lines aligned with each other
determined variance.
Where trend lines were overlapped, variance is considered low;
and where they deviated
from one another, variance increased.
Based on the data, three trend lines, each one representing an
orientation, indicated
low variance for most samples; and areas where deviation
occurred have been labeled
A-F. At point A, all three lines do not coincide. Based on the
data, this deviation can
be attributed to changes in sulfur concentration, which vary
considerably between the
three orientations, possibly due to bedding planes and foliation
in the rock. At point B,
the 90 position trend line is not aligned to the other two
trends. This can be attributed
to localized areas of concentration for sulfur. At point C, the
90 position trend line
does not coincide with the other two trends. This deviation can
be attributed to a higher
concentration of sulfur. At point D, all lines are not
coincident. Based on the data, this
deviation can be attributed to changes in sulfur concentration,
which vary considerably
among the three orientations. At point E, the 90 position trend
line does not match the
-
16
Fig
ure
6. E
lem
enta
l con
cent
rati
on v
aria
nce
base
d on
sam
ple
orie
ntat
ion.
X-a
xis
indi
cate
s sa
mpl
e nu
mbe
r.
Y-a
xis
indi
cate
s su
m o
f th
e pr
oxy
elem
ent a
bund
ance
in p
pm.
Are
as w
here
dev
iati
on o
ccur
red
are
labe
led
A-F
. B
lue
line
indi
cate
s 90
or
ient
atio
n tr
end.
Red
line
indi
cate
s 13
5 o
rien
tati
on tr
end.
Gre
en li
ne
indi
cate
s 22
5 o
rien
tati
on tr
end.
-
17
other two trends. This deviation can be attributed to a higher
concentration of sulfur at
the 90 position. At point F, the 135 position trend line does
not align with the other two
trends. This can be attributed to a higher concentration of
sulfur at the 135 position.
Because sulfide deposition occurs within foliations present in
the rock (Paige, 1924),
it is believed that variance in sulfur concentrations is
attributed to orientation of the
foliation direction in which sulfides were deposited. In samples
where sulfur was highly
variable, orientation of foliation affected the concentration of
sulfur. Although there was
variance in these samples, orientation was not considered a
major factor in the analysis.
For consistency, samples were then placed into the apparatus at
90 from vertical.
Data obtained from XRF analysis are expressed as parts per
million (ppm) and are
reported completely in Appendix C.
Energy Dispersive X-ray Analysis
Using XRF data, 16 samples were chosen for detailed analysis
using EDAX
(Appendix D). Standards were run on EDAX using known samples
that had been
analyzed using an electron microprobe. After standard samples
were completed, it was
determined that the EDAX data closely matched data available
from microprobe analysis
and would be employed in this research.
Polished thin sections were inserted into the ESEM; and a
backscatter image was
taken so that the observed area could be matched up with the
slide for petrographic
analysis. Samples were run for an average of 64 frames or 32
minutes at a spot size of
3.5-4.2 m at 25.0 kV at a chamber pressure of 0.60 torr. Sample
magnification varied
depending on area being scanned. Bitmap raster scans were
created, which show areas
of high elemental concentrations across the thin section (Figure
7). All EDAX data is
stored on the accompanying CD.
-
18
Figure 7. EDAX dot map scan of sample BHMA-27a.
-
19
ArcGIS Analysis
Using ArcGIS software version 9.3, sample locations and
metamorphic isograds
were included as layers on a digitized map. XRF data of bulk
elemental composition
associated with each location was entered into an attribute
table. Using the kriging
method of analysis, chosen proxy elements were contoured onto
maps showing areas of
high and low concentration.
The kriging method is a technique for interpolating which honors
data points exactly.
An output point is calculated as a linear combination of known
data points. Kriging
attempts to produce the best linear unbiased estimate (Glossary
of Geology, 2005).
Using points that are in proximity to each other, data are
extrapolated, and in areas where
data were not present an estimation can be derived to reflect
that data more precisely.
Sampling errors, known as edge effects, occur near the edges of
an area where sampling
ceases or in areas where sampling coverage is sparse. The result
is data that may not
reflect the true concentration value in an area where data has
been extrapolated. Areas
where gold has been previously mined, such as the Homestake and
Rochford mining
districts, were marked and used to locate and compare proxy
element concentration maps
for Au, Mn, Cl, Zn, Ba, As, Pb, and S.
Keystone and Hill City Districts are other areas that will be
focal points of this study.
The Hill City District is an area of widely scattered gold
deposits in the vicinity of Hill
City, near the headwaters of Spring Creek and around Rochford,
northwest of Hill City.
The Keystone District extends 5.5 km northwest of Keystone to
2.5 km southeast and is
northeast of the Harney Peak intrusion (Koschman et al.,
1968).
-
20
Statistical Analysis
Statistical analysis was run to determine if correlations exist
between the occurrence
of gold and potential proxy elements. Pearson Product Moment
correlation coefficients
(r) were obtained and tested at a 5% significance level to
determine if a trend exists.
Variables having significant positive correlation coefficients
tend to increase together,
whereas variables having significant negative correlation
coefficients tend to decrease
while others increase.
A multiple linear regression was also employed to generate cross
plot graphs to see
if a trend emerged. Gold was plotted on the X-axis as the
dependant variable and the
proxy elements on the Y-axis as the independent variable. A
multiple linear regression
determines whether there is a positive or negative trend between
gold and the proxy
elements. Complete data set is included in Appendix E.
Petrographic Analysis
Of the 222 samples used in this project, 16 samples were chosen
for detailed
petrographic analysis based on their XRF elemental gold
concentrations. Detailed
petrographic analysis of these samples was completed, and
photomicrographs were
taken for EDAX analysis of areas of interest. Areas of interest
were those that contained
opaque and silicate mineral phases that may contain the proxy
elements or areas
indicative of the metamorphic processes on which this study is
focused.
Photomicrographs are used to interpret bit maps and backscatter
images generated by
EDAX to orient bitmaps to a specified area on the slide. Using
chemistry of the mineral
assemblage identified from petrographic analysis, bit maps
indicate how much of the
-
21
proxy element was substituted into the mineral structure. Data
for mineral assemblages
are included in Appendix F.
-
22
CHAPTER III
RESULTS
Analyzed samples can be compared based on their bulk elemental
composition,
petrography, statistical trends, and spatial relationships.
These factors allow assumptions
to be made on how well groups of samples fit the model for proxy
elements and their
ability to predict gold occurrence.
X-ray Fluorescence and ArcGIS
Because sampling in the Homestake Mining District may not
reflect the true
concentrations for the proxy elements around the Homestake Mine
due to restrictions
in sampling at that area, all interpretations in this area are
based on the data directly
surrounding the area to the south. Areas used as focal points
for this analysis are the
Homestake Mine near the cities of Lead and Deadwood, the
Rochford Mining District
near the city of Rochford, the Keystone District northeast of
the Harney Peak intrusion
near the city of Keystone (Koschman, 1968), and the Deerfield
Lake area which is about
10 km south of Rochford and 10 km northwest of Hill City.
Gold was mapped using concentrations ranging from 0 to 20 ppm
(Figure 8). Areas
containing high concentrations of gold are represented in dark
brown and areas of low
concentration appear in yellow. The two areas that serve as a
reference for gold are
the Homestake Mine and the Rochford Mining District. Data show
that gold is in high
abundance in the Homestake mining area with concentrations
centered around 15 ppm
-
23
Figure 8. Gold concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota using the kriging
method. Red dots indicate sample locations collected by L.M.
Friberg (1977) and M. Armstrong (1998). Black line indicates the
Precambrian boundary.Dashed lines indicate metamorphic isograds.
Blue line indicates outer boundary of Harney Peak Granite. Green
lines indicate Rochford and Homestake Mining Districts
-
24
to 20 ppm. In the Rochford Mining District the concentration
range for gold is between
10 ppm to 15 ppm. The western portion of the mineralized zone
indicates that there is a
high concentration of gold south of the Rochford Mining
District, in the Deerfield Lake
area, which has concentrations of gold varying from 10 ppm to 20
ppm. This extends
southeast to the Hill City Mining District, but concentrations
of gold tend to drop below
10 ppm farther south as one approaches the Harney Peak
intrusion.
Arsenic was mapped and has a range of concentrations between 8
and 36 ppm (Figure
9). To the north, in the Homestake Mining District, arsenic hass
concentrations between
12 ppm at its outermost extent to 25 ppm near the mine itself.
Concentrations around the
Rochford Mining District vary from 10 ppm near the city of
Rochford and increase to an
average of 20 ppm to the south toward Hill City.
Betweem the city of Rochford and Hill City the concentrations of
arsenic reache a
high of 30 ppm to 36 ppm and then decrease with distance from
the district. Near the
Harney Peak intrusion in the south there is an area of highly
concentrated arsenic on the
eastern side of Harney Peak that continues north into the
central Black Hills. This area
contains concentrations that range from 32 ppm to 36 ppm and can
be correlated to the
Keystone Mining District, which is located on the northeast side
of the Harney Peak
intrusion
Barium was mapped and varies in concentrations from 0 to 1,750
ppm (Figure 10).
To the north, in the Homestake Mining District, concentrations
range from 1,500 ppm
to 1,750 ppm centered around the cities of Lead and Deadwood and
decrease sharply as
distance increases from this central point toward the north. As
distance increases to the
south, there is a more gradual decrease in concentration to 600
ppm to 800 ppm.
In the Rochford Mining District, concentrations of barium hold
fairly consistent,
ranging between 800 ppm and 900 ppm. Southwest of the Rochford
Mining District
concentrations increase, ranging from 1,000 ppm to 1,200 ppm
toward the western border
-
25
Figure 9. Arsenic concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
26
Figure 10. Barium concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
27
of the mineralized zone. A trend exists near Keystone, extending
from the west to the
southeast and staying to the north of Harney Peak, that
increases to 1,450 ppm to 1,750
ppm as you approach the eastern border of the mineralized
zone.
In the northeastern portion of the Black Hills, concentrations
range from 1,750 ppm at
the edge of the mapped area, and decrease slowly to 900 ppm
toward the west and south.
This area is dominated by Tertiary thermal dikes known to have a
high occurrence of
barite (Dewitt et al., 1996).
Chlorine was mapped and has concentrations varying from 3,900
ppm to 14,900 ppm
(Figure 11). In the Homestake Mining district, chlorine
concentrations range between
8,000 and 11,000 ppm that increase sharply to 14,900 ppm 15 km
to the southeast. This
trend continues to the eastern border of the study area where
chlorine concentrations
fluctuate between 10,500 ppm and 14,900 ppm across a 20-km2
area
In the Rochford Mining District, concentrations of chlorine
range from 11,000 ppm
to 14,900 ppm. This trend continues through the western portion
of the study area to
the extent of sampling on the western margin of the Black Hills
and continues south for
20 km. As sample points approach the south toward Harney Peak,
the concentrations
decreased sharply to averages of 4,000 ppm to 6,000 ppm and
eventually declined to
3,900 ppm at the southernmost extent of the study area.
Manganese was mapped and was found to have a concentration
variation from 220
ppm to 4,300 ppm (Figure 12). In the Homestake Mining District,
concentrations of
manganese have a range of 500-800 ppm. Concentration changes
abruptly 10 km to the
southeast to an average of 2,200 ppm and then tapers off to 300
ppm to 400 ppm. This
trend continues to the eastern border of the study area.
In Rochford Mining District, the concentrations of manganese
range from 2,200
ppm to 2,400 ppm near the city of Rochford. The concentration of
manganese increases
consistently to the west to its upper limit of 4,300 ppm and
continues to the south for 35
-
28
Figure 11. Chlorine concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
29
Figure 12. Manganese concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts
-
30
km, having a small variation between 3,900 ppm and 4,000,
increasing to 4,300 ppm west
of the Harney Peak intrusion. South of Harney Peak,
concentrations fall to the lower
limit of 220 ppm with a slight increase to 350 ppm to the east
and west.
Lead was mapped and was found to have concentration variations
between 10 ppm
and 125 ppm (Figure 13). In the Homestake Mining District, lead
reaches upper limits of
125 ppm near the town of Galena and within the Lead-Deadwood
area. Concentrations
of lead drop as one moves to the southeast and drop to their
lower limit in the central
portion of the mapped area. A small area in the northeast
portion of the map shows
concentrations increase to 80 ppm and taper off slowly in all
directions until reaching the
lower limit in the central area.
The Rochford Mining District has average concentrations beteen
20 ppm and 40 ppm.
These concentrations continue to the western extent of the study
area and decline to 10
ppm in the central portion of the area. Five km north of the
city of Rochford there is a
slight increase in concentration to an average of 50 ppm, which
tapers off to the east and
west to 20 ppm.
In the southern portion of the map are two areas of higher
concentration of lead
converging near the Harney Peak intrusion. The area that enters
from the east extends
westward for 20 km and has a maximum concentration of 125 ppm at
the center of
the area and steps down between 90 ppm and 100 ppm near the
center of the mapped
area. The area that enters from the west extends eastward for 15
km and has a slightly
lower concentration density that increases to a maximum of
95-100 ppm at its center.
Concentrations tapers off between 70 ppm and 80 ppm before
converging with the
eastern area at Harney Peak. The southernmost portion of the
study area holds a near
constant concentration of 60 ppm.
Sulfur was mapped and was found to have a concentration varying
between 2,300
ppm and 7,400 ppm (Figure 14). In the Homestake Mining District
sulfur concentrations
-
31
Figure 13. Lead concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
32
Figure 14. Sulfur concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
33
ranged between 4,000 ppm to7,400 ppm, having a high
concentration of 7,400 ppm
centered near the city of Deadwood and dropping rapidly to 4,000
ppm to the west
toward Lead. This trend continues southeast for 30 km as the
concentration of sulfur at
7,400 ppm along the northern border decreases to 5,000 ppm
toward the west.
Within the Rochford Mining district, the concentration of sulfur
varies between 5,000
ppm and 7,400 ppm. High concentrations continue north to the
edge of the study area.
To the south of Rochford, concentrations fluctuate from 3,700
ppm to 5,000 ppm. In the
southern portion of the study area lower concentrations range
from 2,300 ppm to 4,000
ppm.
Zinc was mapped and has concentration variations between 20 ppm
and 175 ppm
(Figure 15). In the Homestake Mining District, concentrations
varied between 70 ppm
and 85 ppm. Moving 5 km southeast, concentrations increase to
values between 85 ppm
and 175 ppm and then sharply decrease to 20 ppm.
In the Rochford Mining District the range of concentrations are
between 100 ppm
and 120 ppm, centered near the city of Rochford. Moving south to
southwest, there is
an increase to 175 ppm until it tapers off northeast of Harney
Peak. South of there, it
decreases to a range between 20 and 40 ppm.
Statistical Analysis
Using Sigma Plot and Microsoft Excel software, statistical
analysis was conducted
using XRF data for gold, manganese, sulfur, arsenic, chlorine,
lead, zinc, and barium.
Full statistical analyses are in Appendix C. Gold is the
dependent variable and
manganese, sulfur, arsenic, chlorine, lead, zinc, and barium are
the independent variables.
Graphs indicate a positive or negative correlation as well as
the variance.
-
34
Figure 15. Zinc concentration (ppm) contour map of the
Precambrian area of the Black Hills, South Dakota, using the
kriging method. Red dots indicate sample locations collected by
L.M. Friberg (1977) and M. Armstrong (1998). Black line indicates
the Precambrian boundary. Dashed lines indicate metamorphic
isograds. Blue line indicates outer boundary of Harney Peak
Granite. Green lines indicate Rochford and Homestake Mining
Districts.
-
35
The significance (P 0.05) of the correlation coefficient was
tested using the
following null and alternative hypothesis:
Ho: r = 0; there is no correlation
H1: r 0; thre is a correlation
In the initial multiple regression analysis of the data,
arsenic, barium, and chlorine
exhibited significant positive correlations. Manganese, zinc,
and sulfur had positive
correlations that were not significant. This was believed to be
attributed to lead, which
had a probability above 0.05 and a poor positive correlation.
The multiple regression for
manganese, zinc, sulfur, and lead were recalculated to see if
their correlation coefficient
and probability improved without manganese, barium, and arsenic
as variables. After
recalculation, the probabilities of the manganese zinc and
sulfur improved significantly.
Lead still showed a poor correlation and poor probability after
samples were recalculated
(Table 1).
The data can be separated into three groups; a strong positive
correlation (> 0.400),
a moderate positive correlation (0.100-0.399), and a weak
positive correlation (<
0.099). The elements with strong positive correlations are
chlorine, zinc, and barium.
The element that exhibits the best correlation is chlorine with
an r-value of 0.514 and a
probability of < 0.001. Zinc was next with an r-value of
0.4034 and a probability of <
0.001. Barium had an r-value of 0.403 and a probability of <
0.001.
The elements that had a moderate positive correlation were
arsenic, manganese,
and sulfur. Arsenic had an r-value of 0.269 and a probability of
0.005. Manganese had
an r-value of 0.254 and a probability of 0.001. Sulfur had an
r-value of 0.224 and a
probability of 0.029.
-
36
Tabl
e 1.
Cor
rela
tion
coe
ffici
ents
bet
wee
n pa
ired
ele
men
ts: r
is s
igni
fica
nt w
hen
P
0.05
.
-
37
Lead was the only element that had a weak positive correlation.
It had an r-value of
0.0977 and a probability of 0.237. Because the probability
exceeded 0.05, there is no
significant relationship between gold and lead.
Petrographic Analysis
Of the 222 samples, 120 samples were analyzed using a polarizing
petrographic
microscope. General mineral assemblage and modal estimation for
all samples can be
seen in Appendix D. Of the 120 samples, 16 samples were chosen
based on XRF data for
their high gold concentrations and subjected to detailed
petrographic analysis. Detailed
analysis determines mineral phases present, source rock
(protolith), and metamorphic
grade. The locations of the 16 samples can be seen in Figure 16.
Order of crystallization
was determined using fabric, inclusion relationships, and
cross-cutting relationships.
Sample BH-2 is black to gray with visible hematite staining in
the sample. It is fine
to medium grained, holocrystalline, and has subhedral to
euhedral crystals. It is non-
foliated with porphyroblasts of garnet grains visible in the
sample and show that garnets
were being resorbed by quartz and micas. The sample has an
idioblastic texture with
a blastoporphyritic relict texture. The protolith for this
sample was a pelite and was
subjected to dynamo-thermo metamorphism. The dominant mineral
phases present in
this sample are muscovite, biotite, quartz, and chlorite. This
rock is a chlorite, biotite,
muscovite schist.
Hematite staining is present along quartz veins in the sample.
Muscovite occurs
as euhedral crystals with an average crystal size of
-
38
Figure 16. Precambrian area and Metamorphic Isograds of the
Black Hills, South Dakota. Red dots indicate sample locations
collected by L.M. Friberg (1977) and M. Armstrong(1998). Black line
indicates the Precambrian boundary. Dashed lines indicate
metamorphic isograds. Blue line indicates outer boundary of Harney
Peak Granite.
-
39
continued late stage crystallization. Quartz grains are
subhedral and were most likely
present in the protolith with some recrystallization as
metamorphism occurred. Biotite
crystallized along with muscovite. Late stage chlorite and
muscovite cross cut all other
minerals and early foliation and was last to crystallize.
Sample BH-4 is greenish black on a fresh surface, medium
grained, and is
hypocrystalline. It is non-foliated with visible hematite
staining along grain boundaries,
and some quartz veining is also present. It has a heteroblastic
texture lacking relict
texture. The protolith for this sample was basalt, and it was
subjected to dynamo-thermo
metamorphism. The dominant mineral phases present in this sample
are hornblende,
biotite, quartz, plagioclase, and magnetite. This rock is a
magnetite, biotite, hornblende
amphibolite.
The first minerals to crystallize were pyrite and biotite. Veins
of an opaque mineral
are present in the sample. Hornblende crystallized next and
exhibited euhedral to
subhedral crystals indicating that this was most likely the peak
of metamorphism because
muscovite, although present in the sample, was in low abundance
and cross cuts the
biotite. Late quartz veining is seen within the sample, cross
cutting the fabric of the rock
and may have aided in the oxidation of the magnetite and
biotite.
Sample BH-5b is black to gray on a fresh surface, with red bands
of garnet present. It
is fine to medium grained, hypocrystalline, and contains
subhedral to anhedral crystals.
There is evidence of relict bedding in the sample. It is
foliated with slight hematite
staining along grain boundaries. It has a lepidoblastic texture
with a blastopelitic relict
texture. The protolith for this sample was a pelite, was
subjected to dynamo-thermo
metamorphism, and is garnet grade. The dominant mineral phases
present in this sample
are muscovite, quartz, biotite, graphite, and ilmenite. This
rock is an ilmenite, graphite,
biotite, muscovite schist.
-
40
Ilmenite formed along with the micas and exhibited subhedral
crystals. Biotite
and muscovite crystallized and exhibited subhedral crystals that
are sub-parallel to the
foliation of the rock. Quartz was present from the protolith and
was also re-crystallized
during metamorphism. Late fluids were introduced causing the
oxidation of the iron-
bearing minerals along grain boundaries. Graphite in the sample
is present along veins,
indicating that the graphite was a result of the metamorphic
fluids intruding into the rock.
The reduction of the carbonates lead to the formation of
graphite within the sample.
Sample BH-15 is black to gray on a fresh surface with bands of
garnet present. It
is fine grained and occurs as subhedral to anhedral crystals. It
is foliated with sub-
parallel alignment of the micas with foliation. The sample has
been altered by late fluids,
converting pyrite and ankerite into hematite. It has a
mimetic/lepidoblastic texture
with relict bedding present. The protolith for this sample was a
marl, was subjected to
dynamo-thermo metamorphism, and is a biotite grade. The dominant
mineral phases
present in this sample are muscovite, chlorite, quartz, pyrite,
hematite, biotite, and
and iron carbonate. This rock is an iron carbonate, biotite,
hematite, pyrite, chlorite,
muscovite banded schist.
Iron carbonate grains are blastoporphyritic, and relict quartz
grains are also present
from the protolith. Pyrite crystallization began early and is
present as inclusions
within the micas and exhibits euhedral crystals. Biotite was the
next to crystallize
followed by muscovite and are both in sub-parallel alignment
with foliation. Late stage
fluids infiltrated the sample indicated by quartz veining
associated with retrograde
metamorphism. Resorbtion of biotite is shown by chlorite cross
cutting the biotite grains.
Sample BH-15b is black to gray on a fresh surface with bands of
almandine present.
It is fine to medium grained and has a subhedral to anhedral
texture. The sample is
foliated with sub-parallel alignment of the mica grains with
late stage cross cutting
porphyroblasts of biotite. It has a lepidoblastic texture with
relict bedding present. The
-
41
sample shows metasomatism occurring between biotite and
chlorite, with the biotite
grains showing resorbtion by chlorite. The protolith for this
sample was a marl subjected
to dynamo-thermo metamorphism and is biotite grade. The dominant
mineral phases
present in this sample are muscovite, quartz, chlorite, pyrite,
ankerite, and biotite. This
rock is a biotite, iron carbonate, pyrite, chlorite, muscovite
schist.
The iron carbonate grains present are blastoporphyritic and
relict quartz grains
are also present from the protolith. Pyrite began to crystallize
early, exhibits euhedral
crystals, and is included in the micas. Muscovite and chlorite
were the next minerals to
crystallize and are in sub-parallel alignment with foliation.
Biotite was last to crystallize
and exhibits subhedral porphyroblasts that cross cut foliation,
indicating a reactivation
of dynamo-thermo processes causing prograde metamorphism.
Chlorite formed last and
cross cuts all minerals and the foliation.
Sample BH-19a is blackish green with visible quartz crystals. It
is fine to medium
grained, holocrystalline, having subhedral to anhedral crystals,
and has a heteroblastic
texture. The sample is non-foliated and exhibits hematite
staining around the magnetite
and ferroactinolite grains. The protolith for this sample was
basalt that was subjected
to dynamo-thermo metamorphism and is biotite grade. The dominant
mineral phases
present in this sample were ferroactinolite, plagioclase, iron
carbonate, magnetite, and
quartz. This rock is a magnetite, iron carbonate,
ferroactinolite metabasalt.
Ferroactinolite and magnetite were the first minerals to
crystallize. Plagioclase grains
are bimodal, indicating that some of the plagioclase present are
most likely from the
protolith with the smaller grains being a result of
metamorphism. Quartz is present as
anhedral crystals
Sample BH-63 is black with visible quartz and almandine bands in
the matrix. Garnet
grains exhibit anhedral crystals. It is medium grained, and
contains euhedral to subhedral
crystals. The sample is foliated and exhibits hematite staining
along the garnet-grain
-
42
boundaries. It has a poikiloblastic/lepidoblastic/snowball
texture. Garnet crystals
have been rolled and inclusions within the crystal have
preserved the original fabric
orientation. The protolith for this sample was a pelite that was
subjected to dynamo-
thermo metamorphism and is staurolite grade. The dominant
mineral phases present in
this sample are muscovite, quartz, biotite, magnetite, and
garnet. This rock is a garnet,
biotite, muscovite schist.
Quartz grains present in the sample are relict grains from the
protolith because
they are included in the garnet crystals. Magnetite crystallized
along with the micas
and exhibits euhedral crystals. Biotite and muscovite
crystallized and are included
in the garnet crystals. Garnet was to next to crystallize and
exhibits euhedral crystals
having an average crystal size between 2 and 3mm. Inclusions in
the garnet crystal
indicate that following crystallization, prograde metamorphism
continued, during which
garnet crystals were rolled. Metamorphic fluids and plastic
deformation caused the
recrystallization of the micas causing them to be parallel to
the new foliation direction.
Biotite, and muscovite were the last to crystallize, forming the
current orientation of the
foliation within the rock.
Sample BH-66 is black to gray on a fresh surface with visible
hematite staining. It
is fine grained, holocrystalline, and has euhedral to anhedral
crystals. The sample is
foliated and contains many quartz veins. It has a
lepidoblastic/mimetic texture with a
blastopsammatic relict texture. Garnet crystals have been
rolled, and inclusions within
the crystal have preserved the original fabric orientation. The
protolith for this sample
was a pelite that was subjected to dynamo-thermo metamorphism
and is staurolite grade.
The dominant mineral phases present in this sample were quartz,
muscovite, biotite,
magnetite, and garnet. This rock is a garnet, biotite, muscovite
schistose quartzite.
Quartz grains were bimodal and the larger quartz grains present
in the sample are
relict from the protolith, and the smaller quartz grains are the
result of recrystallization
-
43
during metamorphism. This assumption was made because quartz
grains are included
within the garnet. Biotite and muscovite crystallized along with
quartz, and formed
parallel to the foliation direction. Porphyroblastic garnet was
the next to crystallize
with crystals ranging in size from 1 to 4 mm. Garnet crystals
had many inclusions
of quartz, muscovite, and biotite, indicating that they formed
during and prior to
garnet crystallization. The garnet crystals were slightly
rolled, indicating some slight
deformation due to shearing forces.
Sample BHMA-97-27a is black to gray on a fresh surface with
visible quartz present.
The rock is fine to coarse grained and contains euhedral to
subhedral crystals. The
sample is foliated and has a decussate/heteroblastic texture
with a blastopelitic relict
texture. Hematite staining is present along grain boundaries.
The protolith for this
sample was a pelite that was subjected to dynamo-thermo
metamorphism and is garnet
grade. The dominant mineral phases present in this sample are
biotite, muscovite, and
quartz. This rock is muscovite, biotite schist.
Quartz and biotite grains have a bimodal distribution,
indicating two growth periods.
Biotite grains not along veins appear to be altered to chlorite
but occurs as euhedral
crystals. Muscovite formed next and is parallel to foliation.
Late stage fluid infiltrated
the sample causing veining that is predominantly quartz with
re-crystallized biotite grains
at the margins of the relict beds that crosscut foliation.
Chlorite, garnet and graphite are
also present in the sample in minor abundances. Chlorite was
also observed cross cutting
porphyroblasts of biotite.
Sample BHMA-97-31 is black to gray on a fresh surface and is
fine to medium
grained. The sample is non-foliated and has an idioblastic
texture with a blastopelitic
relict texture. The protolith for this sample was a pelite that
was subjected to dynamo-
thermo metamorphism and is biotite grade. The dominant mineral
phases present in
-
44
this sample are muscovite, quartz, biotite, opaques, and garnet.
This rock is a biotite,
muscovite schist.
Whole rock crystallization began with the crystallization of
pyrite which was included
in the micas and exhibits euhedral crystals with an average
crystal size between 1 and
2 mm. Tourmaline was the next to crystallize as euhedral
crystals having an average
crystal size between 1 and 2 mm. Muscovite and biotite were the
next to crystallize and
are in random orientation within the sample. Quartz crystallized
next followed by garnet.
Garnet grains contain many quartz, muscovite, and biotite
inclusions. Pyrite crystals
were observed breaking down into hematite which occurs as halos
around the pyrite
grains.
Sample BHMA-97-47 is greenish black on a fresh surface and is
fine to medium
grained. The sample is non-foliated and has a
porphyroclastic/mortar texture with a
blastoporphyritic relict texture. The protolith for this sample
was a greywacke that was
subjected to dynamo-thermo metamorphism and is biotite grade.
The dominant mineral
phases present in this sample are quartz, muscovite, biotite,
and plagioclase. This rock is
a biotite, muscovite metawacke.
Plagioclase and quartz grains present are bimodal, and the
larger grains are believed
to be relict grains from the protolith and lack preferred
orientation in the rock. Quartz
grains also exhibit a mosaic structure indicating quartz
recrystallization as metamorphism
progressed. Muscovite and biotite have average grain sizes <
1 mm. Biotite occurs
within or along quartz boundaries, and muscovite/biotite
intergrowths occur within relict
bedding.
Sample BHMA-97-54 is gray to black on a fresh surface and is
fine grained,
holocrystalline, and has crystals range from euhedral to
subhedral. The sample is
non-foliated and has a lepidoblastic/snowball texture and a
blastopelitic relict texture.
The protolith for this sample was a pelite that was subjected to
dynamo-thermo
-
45
metamorphism and is garnet grade. The dominant mineral phases
present in this sample
were biotite, quartz, garnet, and chlorite. This rock is a
chlorite, biotite schist.
Magnetite crystallized first and exhibits euhedral crystals with
an average grain
size of 1 mm. Biotite was next to crystallize followed closely
by the crystallization of
chlorite. There is a sub-parallel alignment of biotite grains
within the sample indicating
pressure and temperature were beginning to align the
minerals.
Sample BHMA-97-57b is greenish-black with relict bedding and
probable quartz
veins present in areas rich in garnet. It is medium grained and
has euhedral to subhedral
crystals. The sample is foliated and has a porphyroblasts of
rotated garnets. The protolith
for this sample was a pelite that was subjected to dynamo-thermo
metamorphism and
is garnet grade. The dominant mineral phases present in this
sample were muscovite,
garnet, biotite, quartz, and chlorite. This rock is chlorite,
biotite, garnet, muscovite schist.
Quartz is bimodal indicating the larger grains are relict quartz
grains from the
protolith. Plagioclase grains are also believed to be relict
grains from the protolith.
Biotite was the first to crystallize and exhibits euhedral
crystals that are in subparallel
alignment with foliation. Chlorite was next to form and shows
subhedral crystals.
Muscovite formed late, is observed cross cutting chlorite, and
is in subparallel alignment
with foliation. Garnet was last to crystallize and contains many
inclusions of quartz and
biotite. The garnet has been rolled, indicating late deformation
followed by muscovite
growth.
Sample BHMA-97-63 is black on a fresh surface with garnet
porphyroblasts present.
It is fine to medium grained and contains anhedral crystals. The
sample is foliated and
has a heteroblastic texture with a blastopsammatic relict
texture. The protolith for this
sample was a pelite that was subjected to dynamo-thermo
metamorphism. The dominant
mineral phases present in this sample are biotite, plagioclase,
garnet, quartz, and
muscovite. This rock is muscovite, garnet, biotite schist.
-
46
Crystallization began with the formation of biotite and
muscovite, which formed
parallel to the foliation. Garnet was next to form, and the
crystals present in the sample
exhibit euhedral crystals, contain many quartz, muscovite, and
biotite inclusions, and
have been slightly rolled preserving the original foliation.
Quartz is present in the sample
as bedding. Hematite staining is present along and within the
quartz veins. Biotite is
observed breaking down into chlorite when the grains are in
close proximity to the garnet
indicating that as the garnet formed, biotite and garnet were
being consumed and chlorite
formed.
Sample BHMA-97-98 is greenish-black with visible quartz and
plagioclase. It
is fine to medium grained with crystals that range from
subhedral to anhedral. The
sample is foliated and has a heteroblastic/mortar texture with a
blastoporphyritic
texture. The protolith for this sample was a greywacke that was
subjected to dynamo-
thermo metamorphism. The dominant mineral phases present in this
sample are quartz,
biotite, muscovite, plagioclase, and chlorite. This rock is a
chlorite, muscovite, biotite
metawacke.
Plagioclase grains present in the sample may be relict grains
from the protolith.
Quartz was bimodal, indicating that the larger grains may also
be blastoporphyritic before
metamorphism, and the smaller quartz grains recrystallized
during metamorphism. The
larger quartz grains appear to be strained as a result of
metamorphism. Crystallization
began with the formation of biotite and muscovite which exhibit
anhedral crystals that
are parallel to foliation. Chlorite formed next and is found to
be either sub-parallel to
foliation or as rims around the biotite grains. Trace amounts of
garnet are present in the
sample. Hematite staining is apparent at grain boundaries within
the sample.
Sample BHMA-97-101 is gray to black on a fresh surface and is
fine to medium
grained and has euhedral to subhedral crystals. The sample is
foliated and shows kink
banding near the garnet crystals. It has a lepidoblastic texture
with a blastopsammatic/
-
47
blastopelitic texture. Relict bedding is observed within the
sample. The protolith for this
sample was a pelite that was subjected to dynamo-thermo
metamorphism and is garnet
grade. The dominant mineral phases present in this sample are
biotite, muscovite, quartz,
garnet, and hematite. This rock is a hematite, garnet,
muscovite, biotite banded schist.
Biotite and muscovite crystallized early and occur as subhedral
crystals that are
parallel to foliation. Quartz was the next to crystallize within
the sample as veins.
Quartz veins have hematite staining along the vein boundaries.
Hematite is also present
in the sample as free standing crystals. Garnet was last to
form, and the grains contain
many inclusions of quartz and biotite.
Energy Dispersive X-ray Analysis
The sixteen polished thin sections used in petrographic analysis
were chosen based
on XRF data for their high gold concentrations and were analyzed
by energy dispersive
X-ray analysis (EDAX) to obtain the element distribution in the
sample, their abundance,
and dot maps which indicated their occurrence across the section
(Figure 7). These data
were collected to determine if the elemental concentrations of
proxy elements are higher
on the outside rims of the grains than in the central core of
the crystal.
The backscatter images are used to match location of the dot
maps with
photomicrographs. Petrographically identified minerals can then
be matched with
elemental distribution on the dot maps. This will also indicate
whether the mineralizing
fluids deposited the proxy elements that were deposited at the
rim of the grain. Silicate
minerals capable of including several proxy elements into their
structure are chlorite,
biotite, garnet, while sulfur and arsenic partition in the iron
bearing opaques.
The occurrence of an element can be represented as bitmaps. As
the sample is
scanned, X-rays interact with the elements present in the
sample. When an element is
-
48
identified, dots are added to the bitmap where that element
occurred. The higher the
concentration of these dots, the higher the abundance of that
element. Therefore, a high
density of dots in an area indicates high abundance, and a low
density of dots in an area
indicates low concentration for the elements being scanned.
Using photomicrographs taken of the slide and the bit maps
generated from EDAX,
occurrence of proxy elements in high concentration can be
correlated to the mineral
phases present in the sample. The elements that were chosen for
this comparison are
those that correlated best with gold: barium, sulfur, and
chlorine. One slide that best
illustrates this relationship was chosen for each element.
EDAX analysis does define elemental occurrence, but does not
indicate chemical
zoning between the rim of the grain and its center. Where fluids
would have interacted
with the rim of the grain, a higher concentration should be
observed. This shows that
dot map scans cannot define chemical zoning and further research
using line scans may
delineate chemical zoning.
-
49
CHAPTER IV
DISCUSSION
The ability of the proxy elements to predict gold is based on
their increased
abundance with proximity to known gold deposits. Using ArcGIS,
statistical analysis,
EDAX, and petrography, the proxy elements can be rank ordered to
predict the location
of gold deposits.
Chlorine as a Predictor for Gold
Chlorine is the proxy element that best correlates with the
occurrence of gold. The
correlation coefficient of chlorine with gold was 0.514 (Figure
17). This strong positive
correlation indicates that as concentration of gold increases,
concentration of chlorine
increases with it. The r2 value suggests that 26% of the
variation in gold can be attributed
to chlorine. The mineral that would likely accommodate chlorine
into its structure is
chlorite. Petrographic and XRF analysis indicates that where
there is a higher abundance
of gold, there is a corresponding high abundance of chlorine.
This occurrence of high
concentrations of chlorine may predict the occurrence of
gold.
-
50
Fig
ure
17.
Cro
ss p
lot g
raph
of
gold
ver
sus
chlo
rine
con
cent
rati
on.
Chl
orin
e is
the
inde
pend
ent v
aria
ble
and
gold
is th
e de
pend
ent v
aria
ble.
y=11
69.8x5
507.9
r=0.264
2
1000
0
1500
0
2000
0
2500
0
3000
0
3500
0
Cl(p
pm)
y=11
69.8x5
507.9
r=0.264
2
0
5000
1000
0
1500
0
2000
0
2500
0
3000
0
3500
0
05
1015
2025
3035
Cl(p
pm)
Au(ppm
)
-
51
Barium as a Predictor of Gold
Barium is also correlated with the occurrence of gold. This
strong positive correlation
indicates that as concentration of gold increases, concentration
of barium increases with
it (Figure 18). Thus, variation in barium content can also be
responsible for variation in
gold concentration. The mineral that would likely accommodate
barium into its structure
is the feldspars. In pagioclase, barium would substitute into
the calcium or sodium site
in the structure. Petrographic and XRF analysis indicates that
where there is a higher
abundance of muscovite and biotite, there is corresponding high
abundance of barium.
Based on statistical data, occurrence of high concentrations of
barium may predict
occurrence of gold.
Arsenic as a Predictor of Gold
Arsenic correlates somewhat with gold (Figure 19). This moderate
positive
correlation indicates that as concentration of gold increases,
concentration of arsenic
increases with it. The minerals that would likely accommodate
arsenic into its structure
are pyrite and arsenopyrite. Petrographic and XRF analysis
indicates that where pyrite is
present in thin section, there is corresponding high abundance
of arsenic. Arsenopyrite is
not present in the slides described in detailed petrographic
analysis, so a correlation with
this mineral cannot be made. Based on statistical data,
occurrence of high concentrations
of arsenic can possibly predict occurrence of gold.
-
52
Fig
ure
18.
Cro
ss p
lot g
raph
of
gold
ver
sus
bari
um c
once
ntra
tion
. B
ariu
m is
the
inde
pend
ent v
aria
ble
and
gold
is th
e de
pend
ent v
aria
ble.
y=23
.345
x+37
.074
r=0.162
7
400
600
800
1000
1200
Ba(p
pm)
y=23
.345
x+37
.074
r=0.162
7
0
200
400
600
800
1000
1200
05
1015
2025
3035
Ba(p
pm)
Au(ppm
)
-
53
Fig
ure
19.
Cro
ss p
lot g
raph
of
gold
ver
sus
arse
nic
conc
entr
atio
n. A
rsen
ic is
the
inde
pend
ent v
aria
ble
and
gold
is th
e de
pend
ent v
aria
ble.
y=1.43
38x+2.32
7r=0.072
3
406080100
120
140
As(pp
m)
y=1.43
38x+2.32
7r=0.072
3
020406080100
120
140
05
1015
2025
3035
As(pp
m)
Au(ppm
)
-
54
Manganese as a Predictor of Gold
Manganese also correlates with occurrence of gold (Figure 20).
This moderate
positive correlation indicates that as concentration of gold
increases, concentration of
manganese increases with it. The minerals that would likely
accommodate manganese
into its structure are garnet, chlorite, and biotite. Manganese
will preferentially substitute
into the garnet structure first, then into chlorite, and if
there is any remaining manganese,
into biotite. Petrographic and XRF analysis indicates that where
garnet is present in thin
section, there is corresponding high abundance of manganese.
Also, when garnet and
chlorite are both found in thin section there is an increase in
manganese concentration
indicating that substitution was occurring within both minerals.
Based on statistical data,
occurrence of high concentrations of manganese may possibly
predict occurrence of gold.
Sulfur as a Predictor of Gold
Sulfur correlates with occurrence of gold (Figure 21). This
moderate positive
correlation indicates that as concentration of gold increases,
concentration of sulfur
increases with it. The minerals that would likely accommodate
sulfur into its structure
are pyrite and arsenopyrite. Petrographic and XRF analysis
indicates that where
pyrite was present in thin section, there is corresponding high
abundance of sulfur.
Arsenopyrite is not present in the slides. A correlation with
this mineral cannot be made,
but arsenic is most likely substituted into the pyrite
structure. Based on statistical data,
the occurrence of high concentrations of sulfur may also predict
occurrence of gold.
-
55
Fig
ure
20.
Cro
ss p
lot g
raph
of
gold
ver
sus
man
gane
se c
once
ntra
tion
. M
anga
nese
is th
e in
depe
nden
t var
iabl
e an
d go
ld is
the
depe
nden
t var
iabl
e.
y=17
6.3x10
64.7
r=0.064
4
6000
8000
1000
0
1200
0
1400
0
1600
0
1800
0
Mn(ppm
)
y=17
6.3x10
64.7
r=0.064
4
0
2000
4000
6000
8000
1000
0
1200
0
1400
0
1600
0
1800
0
05
1015
2025
3035
Mn(ppm
)
Au(ppm
)
-
56
Fig
ure
21.
Cro
ss p
lot g
raph
of
gold
ver
sus
sulf
ur c
once
ntra
tion
. S
ulfu
r is
the
inde
pend
ent v
aria
ble
and
gold
is th
e de
pend
ent
vari
able
.
y=37
7.28
x+10
29.4
r=0.050
2
3000
0
4000
0
5000
0
6000
0
7000
0
8000
0
S(ppm
)
y=37
7.28
x+10
29.4
r=0.050
2
0
1000
0
2000
0
3000
0
4000
0
5000
0
6000
0
7000
0
8000
0
05
1015
2025
3035
S(ppm
)
Au(ppm
)
-
57
Zinc as a Predictor of Gold
Zinc correlates with occurrence of gold (Figure 22). This
moderate positive
correlation indicates that as concentration of gold increases,
concentration of zinc
increases with it. The minerals that would likely accommodate
zinc into its structure are
the sulfides. Petrographic analysis defined opaques, which most
likely are hosts for zinc.
Although correlation between zinc and gold is not as strong an
indicator as the other
proxy elements, based on statistical data, occurrence of high
concentrations of zinc may
predict occurrence of gold. Extremely high concentrations of
zinc may be attributed to
zinc introduction during sample preparation.
Lead as a Predictor of Gold
Lead correlates weakly with occurrence of gold (Figure 23). This
weak positive
correlation indicates that as concentration of gold increases,
concentration of lead
increases with it. However, the r2 of the correlation suggests
that less than 1% of the
variation in gold can be attributed to lead.
Spatial Analysis of the Proxy Elements
Gold has previously been mined from the four mining districts in
the Black Hills
Precambrian core. These areas are the Homestake, Rochford,
Keystone, and Hill City
Mining Districts in the Black Hills, SD. These districts were
used as reference sites
for comparison with concentration levels of the proxy elements
within the Black Hills.
-
58
Fig
ure
22.
Cro
ss p
lot g
raph
of
gold
ver
sus
zinc
con
cent
rati
on. Z
inc
is th
e in
depe
nden
t var
iabl
e an
d go
ld is
the
depe
nden
t var
iabl
e.
y=9.69
64x5
1.30
7r=0.209
200
300
400
500
600