DIFFUSE, LOW-TEMPERATURE HYDROTHERMAL DEPOSITS
ON THE JUAN DE FUCA RIDGE AND PLATE
Catherine Erma Channing B.Sc., Carleton University, 2001
A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of
MASTER OF SCIENCE
In the Department of Earth and Ocean Sciences
O Catherine E. Channing, 2004
University of Victoria
All rights reserved. This thesis may not be reproduced in whole or in part, by photocopy or other means, without permission of the author.
Supervisor: Dr. Kathryn M. Gillis
ABSTRACT
Hydrothermal circulation in ocean crust results in significant geochemical exchanges between hydrosphere and lithosphere. This process begins at the mid-ocean ridge and continues as basaltic crust ages and is subducted, significantly altering the chemical composition of both fluid and rock. In the on-axis environment, heated crustal fluids with a composition altered from that of seawater vent as either high temperature
(> 100 C ) , focused flow or low temperature (< 100 oC) diffuse flow. Reaction between warm fluids and basalt results in the alteration of the rock, manifested as the breakdown of glass and primary minerals and the deposition of secondary minerals. In the off-axis
environment (crust > 1 Ma), crustal fluids discharge locally at seamounts, where
extensive manganese oxides can precipitate. Both types of mineral deposits record the time-integrated history of diffuse fluid-rock interaction, and in addition, Mn-oxide deposits are useful for estimating the longevity of hydrothermal activity.
The effects of low-temperature diffuse fluids on the basaltic crust was examined at both young (Axial Volcano) and mature (Main Endeavour field) on-axis
hydrothermal sites. In general, alteration was very minor (< 2%), with the M E F basalts
showing slightly more abundant and diverse mineral assemblages that that at Axial, due to the presumed longer period of low-temperature basalt-water reaction. Interaction of basalt with diffuse, low-temperature fluids resulted in only minor chemical changes in basalt. Chemical fluxes for basalt alteration at Axial Volcano is insignificant when compared to the cumulative on-axis fluid flux and fluxes associated with older, off-axis basalt alteration. Diffuse fluids from Axial were examined using geochemical modeling to determine how parameters such as fluid temperature, pH and degree of mixing and
water-rock reaction influence alteration mineral precipitation. Results show that precipitation of the observed alteration assemblages requires a combination of minerals
precipitating directly from diffuse fluids, from mixing of fluids and seawater (at < 20 %
seawater) and fi-om diffuse fluid - basalt reaction at water-rock ratios > 200: 1.
Manganese oxide crusts were examined at Baby Bare seamount to investigate the history of hydrothermal venting. Baby Bare acts to focus crustal fluids, which precipitate extensive Mn-oxide deposits. Textural, mineralogical and chemical evidence indicates a hydrothermal origin for these deposits, with a diagenetic signature as well.
Minimum ages of Mn-oxide crusts, using calculated growth rates (324 to
-
1800mm/Ma) and manganese outcrop thicknesses, indicate that Baby Bare has been
hydrothermally active for at least 0.5 Myr, and possibly since its formation at
-
1.7TABLE OF CONTENTS . . ... ABSTRACT 11 TABLE OF CONTENTS ... v
...
LIST OF TABLES ... VIII LIST OF FIGURES ... ixACKNOWLEDGEMENTS ... x
1
.
INTRODUCTION ... 11.1. Oceanic Hydrothermal Systems ... 1
1.2. Study Areas ... 5
1.3. Diffuse Hydrothermal Fluids ... 8
1.4. Thesis Objectives ... 9
1.5. Contributions and Publication Plan ... 10
Z
.
ALTERATION AND MASS TRANSFER AT AXIAL, LOW-TEMPERATURE DIFFUSE HYDROTHERMAL SITES ... 112.1. Introduction ... 11
a
1.1. Geologic Setting and Sample Suites ... 12... 2.1.2. Dfluse Hydrothermal Fluids 16 ... 2.2. Methods 19 ... 2.2.1. Analytical Methods 19 ... 2.2.2. Geochemical Modeling Methods 2 1 ... 2.3. Results 26 ... 2.3.1. Alteration Mineralogy 26 ... a3.2. Bulk Rock Chemistry 29 2.3.3. Chemical Change Calculations ... 33
2.3.4. Geochemical Models ... 34
2.3.4.1. Model I - Speciation of Dzfluse Fluids ... 34
2.3.4.2. Model 2 - Dguse Fluid - Seawater Mizing Model ... 37
2.3.4.3. Model 3- Dguse Fluid - Basalt Reaction Model ... 42
2.4. Discussion ... 45
2.4.1. Alteration ofBasaltic Crust ... 45
2.4.9. Chemical Fluxes ... 48
2.5. Conclusions ... 53
3
.
HYDROTHERMAL MANGANESE OXIDE DEPOSITS FROM BABY BARE SEAMOUNT IN THE NORTHEAST PACIFIC OCEAN ... 543.1. Introduction ... 54
3.2. Regional Setting ... 55
3.2. 1
.
Baby Bare Geologic Setting ... 55... 3.3. Sample Suite 59 3.4. Analytical Methods ... 64 ... 3.5. Results - 6 5 ... 3.5.1. Mineralogy 65 ... 3.5.2. Bulk Chemistry 66 ... 3.6. Discussion 70 3.6. I
.
Classzfication as Diagenetically Inzuenced Hydrothermal ... Crusts 70 3.6.2. Elemental Sources and Growth-Conditions ... 75 ...3.6.3. Calculation of Crust Growth Rates 80
...
vii ... 4
.
CONCLUSIONS 84 ... FUTURE WORK 86 ... REFERENCES CITED 87APPENDIX A: SAMPLE LOCATION SUMMARY FOR AXIAL VOLCANO ...
AND MAIN ENDEAVOUR FIELD 106
...
APPENDIX B: SAMPLE DESCRIPTIONS 108
...
APPENDIX C: PRECISION AND ACCURACY 115
...
APPENDIX D: ELECTRON MICROPROBE DATA 119
...
APPENDIX E: LASER ABLATION ICP-MS RESULTS 127
...
APPENDIX F: GEOCHEMICAL MODELING METHODS 129
... F
.
1.
Introduction 129 ... F.2. Initial Steps 129 ... F.3. Sensitivity Calculations 133 ... F.4. Mineral Suppression 134 ...F.5. Geochemist's Workbench Suite 135 ...
F.6. Calculation Check 139
APPENDIX G: ALTERATION MINERALOGY SUMMARY FOR AXIAL ...
VOLCANO AND MAIN ENDEAVOUR FIELD 141
... APPENDIX H: RESULTS O F CHEMICAL CHANGE CALCULATIONS 143
...
viii
LIST OF TABLES
Number
...
Table 2.1. Fluid chemistry used for geochemical models 18
...
Table 2.2. Basalt chemistry used for geochemical models 25
Table 2.3. List of minerals suppressed in geochemical models ... 25
Table 2.4. Alteration minerals and fluid temperatures during basalt collection ... 28
Table 2.5. Bulk rock chemistry of basalt from Axial Volcano ... 30
Table 2.6. Bulk Chemistry of basalt from the Main Endeavour Field ... 31
Table 2.7. Conditions at which fluid-seawater mix became seawater dominated ... 41
... Table 2.8. Comparison of flux calculations in the Axial region 51 Table 3.1. Sampled Mn-oxide crust locations at Baby Bare seamount ... 61
Table 3.2. Major and trace element chemistry for Mn-oxide crusts ... 67
Table 3.3. Comparison of Baby Bare Mn crusts with those from other locations ... 74
...
LIST OF FIGURES
Figure 1.1. Hydrothermal circulation in the axial region ... 3
Figure 1.2. Hydrothermal circulation in ocean crust with age ... 4
Figure 1.3. Map of sampled sites ... 7
Figure 2.1. Maps of Axial Volcano and the Main Endeavour Field ... 15
. ... Figure 2.2 Solubility plot for SiOq,, ) for all hydrothermal fluids at 15 OC vs pH 36 ... Figure 2.3 Results of diffuse fluid-seawater mixing model for Cloud (2998) 38 .... Figure 2.4 Results of water-rock reaction model for diffuse fluid from Cloud (1998) 43 Figure 3.1. Location of Baby Bare seamount ... 58
Figure 3.2. Photographs of Mn-oxide deposits at Baby Bare seamount ... 60
Figure 3.3. Photographs of Mn-oxide crust samples from Baby Bare seamount ... 63
Figure 3.4. Chondrite-normalized rare earth element plot for Mn- oxide crusts ... 69
Figure 3.5. Ternary diagrams classifying Mn-oxide crusts ... 73
Figure 3.6: Cu + Co + Ni concentrations (grade) versus calculated growth rates in Mn-oxide crusts ... 76
Figure 3.7: Chondrite normalized REE plot for average Mn crusts, spring fluids and Baby Bare bottom seawater ... 78
ACKNOWLEDGEMENTS
I wish to thank first and foremost Kathy Gillis for being a great mentor, and for her help and guidance while supervising this work. I really enjoyed my time here and I am thankful for the opportunities she gave me to experience life on the ocean. Russ McDuff is also thanked for agreeing on short notice to be my external examiner. Comments by Meg Tivey, Russ McDuff, Dave Butterfield, Dante C a d , Kevin Telmer and Kathy Gillis greatly improved the final draft. Technical assistance from Richard Cox, Steve Calvert, Brent Gowen, Mati Raudsepp and Sergei Mateev are greatly appreciated. This work also benefited from advice from and discussions with Meg Tivey, William Seyfried Jr., Susan Humphris, Dave Butterfield, Karen Von Damm, Jay Cullen and Laurence Coogan. I would also like to thank J.D. Chadwick and Michael Perfit for generously sharing data from their work at Axial Volcano.
I would also like to thank Bob Embley and Paul Johnson for inviting me on their oceanographic research cruises; Matt Heinz, Will Sellers and other members of the ROV Jason I1 group and the ROPOS team for assistance with sample collection at sea;
crews of the R/V Thomas G. Thompson and R/V Atlantis; Kevin Telmer for his
assistance with the geochemical modeling program; Dante Canil for use of his laboratory for sample preparation; Verena Tunnicliffe for sharing her cruise reports fi-om previous cruises to Axial Volcano; and Brian Cousens for inspiring me to do ocean science in the first place.
Thank you also goes to my friends: Laurie and Ian Gallagher, Colleen Delaney, Heather Paul, Bill Martin, Karen Patterson and Mitch Selly, whose friendships have made my time here absolutely fabulous. Finally, I could not have completed this work
without the love and support, both financial and emotional, from Kyle Fitzgerald and my parents.
This study was financially supported through the National Science and Engineering Research Council Discovery Grant to Kathy Gillis.
Chapter I
1. INTRODUCTION
1.1. Oceanic Hydrothermal Systems
Hydrothermal circulation in ocean crust is one of the largest geochemical cycles on earth and is the major method of geochemical exchange between the oceanic
lithosphere and hydrosphere (Delaney et al., 1998). This process removes heat and chemical species from ocean crust as it ages (Stein and Stein, 1994), acts to cool the lithosphere, and exerts a control on the chemistry of the oceans (Fisher and Becker, 2000). Heat that drives hydrothermal circulation is supplied at the mid-ocean ridge by a
magma source, and in the off-axis (crust > 1 Myr old), by the cooling of the lithosphere.
Uppermost ocean crust at mid-ocean ridge spreading centres is porous and permeable, which permits cold, oxygenated seawater to convect rapidly in and out of the crust (Mottl, 1983). Seawater penetrates basement through faults and fractures, is heated while descending through the crust and reacts with the surrounding rock, altering the chemical compositions of both the fluid and rock. These fluids are eventually emitted at the seafloor with a chemical composition that is altered from that of seawater
(Butterfield et al., 1990). Basalt alteration is manifested by the breakdown of primary minerals and glass and deposition of secondary (alteration) phases in the rock. With age, porosity and permeability in the crust decreases as alteration minerals precipitate in void spaces, reducing the vigor of the hydrothermal circulation process (Davis, 1992).
Crustal circulation may persist in the off-axis environment to an age of
-
65 Myr (SteinIn the axial region, hydrothermal vents are typically located in the neo-volcanic
zone and are either high or low temperature (Fig. 1.1). High temperature (> 100 oC)
vents have focused flow at high velocities where fluids are emitted from sulfate or
sulfide chimney structures. Low temperature (< 100 oC) vents are diffuse, where fluids
are emitted through holes and cracks in the seafloor. These fluids may be spatially associated with high temperature sites where fluids seep through sulfides or basalt, or more distal from high temperature sites where they vent only through basalt.
In the off-axis ridge-flank environment (crust > 1 Myr), hydrothermal
circulation is less vigorous, fluid temperatures are much lower (< 60 oC) and fluid
chemistry is much closer to that of seawater (Butterfield et al., 1997). Hydrothermal fluids are diffuse, venting either directly from the basaltic seafloor or diffuse through the overlying sediment cover (Fig. 1.2) (Butterfield et al., 1997; Delaney et al., 1992; Mottl et al., 1998; Wheat et al., 2002).
Figure 1 . 1 : Hydrothermal circulation at the mid-ocean ridge axis. Seawater that penetrated basement is heated in the subsurface, rises to the seafloor, and is vented from either high- temperature sulfide or sulfate structures, or low-temperature, diffuse sites on the seafloor (Kelley et al., 2002).
Distance from Ridge Axis (km)
Figure 1.2: Eastern flank of the Juan de Fuca Ridge. As the crust ages, it becomes progressively
covered with sediments, which act to restrict the free interaction between seawater and the ocean crust. Also with age, porosity and permeability are reduced in the crust due to secondary minerals precipitating in void spaces. These act to reduce the vigor of hydrothermal circulation in the off-axis (Davis, 1995).
High temperature mineral precipitation at vent sites has been extensively studied in the axial region (Tivey, 1995a, b; Tivey et al., 1995; Tivey and McDuff,
1990), whereas in the diffuse environment, it has been largely ignored. Diffuse fluids in the on-axis may contribute as much or up to three times more heat flux as high
temperature flow (McDuff et al., 2002), and is certainly the major heat contributor in
the off-axis region (Davis et al., 1997).
1.2. Study Areas
Three study areas were examined, which represent three different stages of evolution of a hydrothermal system. All areas are located in the Northeast Pacific ocean (Fig. 1.3) and are associated with the Juan de Fuca plate. Lava ages provide the
distinction between young and mature sites. Axial Volcano and the Main Endeavour Field (MEF) represent on-axis hydrothermal sites (situated on the Juan de Fuca ridge). Axial Volcano represents a young hydrothermal system (5 years), and is currently both seismically and magmatically active. The M E F represents a mature, on-axis
hydrothermal system (5000-8000 years old (Volpe and Goldstein, 1990), is also seismically active, but is currently in a period of tectonic, rather than magmatic
spreading (Delaney et al., 1992). Finally, Baby Bare seamount (-1.7 Myr) represents a
mature, off-axis site, located
-
100 km from the Endeavour segment of the Juan de Fucaridge.
Two types of diffuse, low-temperature fluid flow were examined at these three sites. T h e first was associated with the on-axis sites, where seawater-derived fluids
subsurface before venting at the seafloor as low-temperature diffuse flow (Butterfield et al., 1990). The second was associated with the off-axis site, where seawater-derived
fluids < 25 OC vent through basalt and sediment cover at the summit of the seamount.
In this region, seawater is recharged into basement locally (Fisher et al., 200.3) and fluid temperatures are much lower as they are associated with older and cooler crust.
Hydrothermal circulation in the off-axis results in two-thirds more heat loss than that at ridge crests, and nearly 20% of the total predicted global ocean heat flux (Elderfield et al., 1999; Fisher and Becker, 2000; Stein and Stein, 1994).
Figure 1.3: Map showing locations of Axial Volcano, the Main Endeavour
Field and Baby Bare seamount on the Juan de Fuca Ridge and plate (Wheat and Mottl, 2000).
1.3. Diffuse Hydrothermal Fluids
In the on-axis environment, diffuse fluids are primarily a result of mixing of high temperature fluids with fresh, oxygenated seawater in the subsurface (Butterfield et al., 1990, in press). Over-printing of mineral precipitation/dilution reactions and biologically mediated processes also add to the chemical signature of diffuse fluids (James and Elderfield, 1996; Sedwick et al., 1992; Tivey, 199513). Diffuse fluids may also result from conductive cooling of high temperature fluids or conductive heating of seawater (Cooper et al., 2000), however, at Axial Volcano, there is no evidence for this
in the diff~lse fluids (Butterfield et al., in press). Evidence for the dilution process comes
from the presence of chemical species and minerals in diffuse fluids that result only from high-temperature reaction of fluids with basalt (Butterfield et al., 1990, 2001, in press). Reaction with basalt causes most of the chemical changes that occur in the seawater- derived fluids, so basalt plays an important role in determining fluid chemistry (Von Damm, 1995). Many elements are exchanged between the fluid and basaltic rock during
the hydrothermal circulation process. For example, Mg, K and sulfate are removed from
solution, whereas elements such as Ca, Si, Fe and Mn are extracted from rock through dissolution and exchange reactions (Butterfield et al., 1997; Mottl, 1983; Seyfried Jr. et al., 1988; Seyfried Jr. and Bischoff, 1979). Any M g present in diffuse fluids is inferred to be the result of high temperature fluids mixing with fresh seawater, as high temperature
reactions at temperatures > 300 OC results in the complete removal of M g into chlorite
(Butterfield et al., in press).
In the off-axis environment, diffuse fluids are a result of the subseafloor heating
of seawater by the cooling of the lithosphere. Here, chemical exchange between these
et al., 2000; Wheat et al., 2000, 2002, 2003; Wheat and Mottl, 2000). These exchanges
typically occur at temperatures below 100 "C, with seawater losing elements such as
Mg, Na and K, and gaining Ca and Mn from oceanic crust. Elements such as Mn, Fe,
Na, Ca and K are also transferred from the sediment cover to the diffuse fluids (Wheat
and Mottl, 2000).
1.4. Thesis Objectives
The overall goal of this study was to examine low-temperature mineral precipitates from diffuse hydrothermal vent site. This involved two types of diffuse hydrothermal sites: on-axis sites where low-temperature fluids vent through basaltic seafloor and sulfide/basalt talus mounds; and an off-axis site, where diffuse fluids vent through sediments and precipitate manganese oxide deposits. At the on-axis young and mature sites (Axial Volcano and the MEF, respectively); goals were to characterize the alteration mineral assemblage, to determine the chemical changes associated with basalt alteration and to determine how these are influenced by differences in such parameters as tectonic setting and system age. In addition to this, the compositions of diffuse fluids from Axial Volcano were thermodynamically modeled in order to examine what fluid characteristics and types of reactions influence the precipitation of alteration minerals. At the mature, off-axis site (Baby Bare seamount), the goal was to determine the origin and evolution of the manganese oxide deposits near the summit area where low-
temperature hydrothermal fluids once vented freely from the sediment cover. At all hydrothermal vent sites, low-temperature alteration minerals provide information on the chemical reactions that occur between the hydrosphere and
important questions will be addressed, such as: how does age of the hydrothermal system influence mineral precipitation and chemical fluxes between fluid and rock? What role does geologic setting play in the extent and nature of fluid-rock reactions? What are the chemical fluxes between the fluids and rock and what parameters control these fluxes?
1.5. Contributions and Publication Plan
Chapter 2 - First author, submitted to Geochimica et Cosmochimica Acta, July 3oth, 2004. D.A. Butterfield supplied fluid data for this study.
Chapter 2
2. ALTERATION AND MASS TRANSFER AT AXIAL, LOW TEMPERATURE DIFFUSE HYDROTHERMAL SITES ON THE JUAN DE FUCA RIDGE
2.1.
Introduction
Movement of heated seawater through ocean crust is the major pathway of geochemical exchange between the oceanic lithosphere and hydrosphere. The upper basaltic ocean crust is porous and permeable, and permits seawater to penetrate to deep levels, where it is heated by a heat source at mid-ocean ridges. Hot, buoyant fluids rise to the seafloor, where they vent from hydrothermal vents at both diffuse, low
temperature (< 100 oC) and focused, high-temperature (> 150 oC) sites. In axial vent
fields, diffuse flow is typically distributed over a much larger cross-sectional area than focused, high temperature venting and may contribute as much as or up to three times more heat flux than high temperature venting (Veirs et al., 2001). T h e chemical composition of diffuse fluids is primarily determined by mixing of hot fluids with cold seawater-like fluids beneath the seafloor and low-temperature water-rock reaction
produce secondary effects (Edmond et al., 1979a ; Butterfield et al., 1997; in press).
Based on the prevalence and areal extent of diffuse venting, it is clear that the mass flux
associated with this type of venting is significant. A critical unresolved issue is whether
the rates of low-temperature water-rock reaction at the seafloor are fast enough to have a significant impact on chemical fluxes at mid-ocean ridges.
As diffuse fluids react with basalt, primary minerals breakdown and secondary minerals precipitate. The chemical and mineralogical effects of reactions between diffuse fluids and rock have not been explicitly examined, and it is not known, for example, how
parameters such as fluid chemistry and the age of the hydrothermal system influence mineral precipitation or the extent and nature of basalt alteration and chemical exchange. Understanding the chemical evolution of diffuse fluids and the impact on basaltic crust is important to resolve if we are to accurately predict the mass flux that occurs between the hydrosphere and lithosphere in these environments.
For this study, alteration and chemical changes associated with basalt interacting with warm, diffuse fluids are examined at young and mature axial hydrothermal sites. Alteration mineralogy and bulk geochemistry are examined for basalt samples recovered from Axial Volcano and the Main Endeavour Field, Juan de
Fuca Ridge. In addition, diffuse fluid compositions from three sites at Axial were used
for geochemical reaction path modeling to determine how temperature, fluid
composition, degree of mixing and other factors affect basalt alteration. Comparison of minerals predicted in different models to those observed provide insights about the relative importance of seawater mixing and water-rock reaction on the formation of the a1 teration assemblages.
2-1. I. Geologic Setting and Sample Suites
The Juan de Fuca is an intermediate spreading ridge (-29 mm/yr) located in the
Northeast Pacific Ocean (Fig. 2. la). Both young and mature, on-axis hydrothermal sites
are found along the ridge, represented by Axial Volcano and the Main Endeavour Field,
respectively. Axial Volcano is a young ( 5 years old) hydrothermal system that lies at the
ridge intersection with the Cobb-Eickelberg seamount chain. In January 1998, an eruption occurred along the South Rift Zone, in the southeast portion of the volcano's three-sided caldera (Dziak and Fox, 1999; Embley et a]., 1999), where the East
hydrothermal field is hosted. Erupted lavas extend over 9 km along the rift zone in a primarily diffusely venting area that was previously hydrothermally active (Embley et al., 1999). Basalt samples were collected from four diffuse vents in 2002 in the new lava
flow: Cloud, Village, Snail and Marker 3 3 (Fig. 2. l b and Appendix A). Fluid
temperature, composition, and flow velocity at Cloud, Marker 3 3 and Snail vents has been well documented since the eruption (Butterfield et al., in press) and basalts were collected in contact with diffuse fluids, where possible. Cloud vent is located in an area
of collapsed lava tubes and pillars, where
-
20 OC fluids emanate from a 1-meter widehole in the basalt carapace. Snail and Marker 33 both vent from cracks in sheet flow
lavas. Fluids are now < 25 OC at both Snail and Marker 3 3 vents, but Marker 3 3 was
one of the hottest and most vigorous vents after the eruption, with fluid temperatures
up to 78 "C in 1999 (Butterfield et al., in press). Village vent, discovered in 2002, is a
-
2meter high pile of basaltic talus, where
-
40 OC fluids vent through cracks betweenbasaltic rubble. Rocks collected from the 1998 eruption zone provide the unique opportunity to study alteration where basalt age is unequivocally known and fluid compositions are available.
The Main Endeavour Field (MEF) is the largest of 5 vent fields found on the
Endeavour segment and represents a mature end-member of a hydrothermal system, as
lavas are 5000-8000 years old (Volpe and Goldstein, 1990). In the MEF, 5 vents sites
were sampled: S&M, Southeast Hulk, Milli-Q, West Grotto and Easter Island (Fig. 2 . lc). S&M and Milli-Q are high-temperature sulfide structures with extensive talus mounds of sulfide, basalt talus and hydrothermal sediments. Basalt samples from these
vents were collected in or near diffuse fluid flow (< 20 oC) at the base of the structures
were also associated with organisms such as tubeworms, allvenilid worms and white filamentous bacteria. S.E. Hulk, W. Grotto and Easter Island are all diffuse sites,
venting from the basaltic seafloor. The basalt from S.E. Hulk was collected
-
10 meterssoutheast of the Hulk sulfide structure from a diffusely venting crack in sheet flow,
where fluids were < 40 oC. Biological organisms such as tubeworms and white
filamentous bacteria were associated with venting fluids here. Basalt from W. Grotto
was collected from a diffusely venting pile of basalt talus found between the Grotto
sulfide structure and the west axial valley wall, in contact with fluids < 40 C. Basalt at
this vent was completely covered in white filamentous bacteria. Finally, Easter Island is
a diffusely venting area -80 m2 in sheet flow and pillow lavas. Basalt was collected from
the southeast area of the vent, lying loosely in hydrothermal sediments approximately 1
meter away from the venting crack. Fluids near where this sample was collected were <
50 oC.
At all sampled diffuse vents from both young and mature sites, the presence of active diffuse venting, biological communities (bacterial mats and filaments, tubeworms and allvenilid worms) and/or surface mineral stainingdcoatings on basalt indicate that samples were at one point in contact with diffuse fluids. Basalt samples were collected in
2002 during R/V Thomas G. Thompson cruise TN-149 using the ROV ROPOS; in
2003 during the R/V Atlantis voyage 7, leg 20 using the ROV Jason 11; in 2003 during
the R/V Thomas G. Thompson cruise TN-158 using the ROV Jason-11; and finally, in 2003 during the R/V Thomas G. Thompson cruise TN-159 using the ROV ROPOS.
I'W
I
Figure 2.1: Maps of the Juan de Fuca Ridge, Axial Volcano and
the MEF. A: Map of the Juan de Fuca ridge showing locations
of Axial and M E F (Baker et al., 2004). B: Map of Axial Volcano
showing sampled diffuse vents in the East Field: 1: Marker 33
and Snail vents; 2: Cloud and Village vents. 1998 Lava flow
indicates the extent of eruptive material (Embley et al., 1999). C:
Map of the M E F showing sampled diffuse vent sites (Lilley et
9.1.9. Dguse Hydrothermal Fluids
In the axial region, diffuse fluids form primarily from the dilution of zero-Mg, sulfide-rich high-temperature fluids with seawater in the subseafloor (Butterfield et al., 1997, 2001, in press; Edmond et al., 1979a, b). Their chemistry can generally be
accounted for by the mixing of ambient bottom seawater with high-temperature fluids (Edmond et al., 1979a), with specific sites being affected by mineral precipitation, microbial activity (Butterfield et al., 1997; Butterfield and Massoth, 1994) and/or dissolution of existing hydrothermal minerals (James and Elderfield, 1996). Evidence for the high temperature component in axial diffuse fluids includes depleted M g and elevated concentrations of chemical species and minerals that result only from high- temperature reaction of fluids with basalt (Butterfield et al., 1997, 2001, in press; Edmond et al., 1979a, a; Lowell et al., 200s).
Diffuse fluid chemical data for Cloud, Snail and Marker 33 vents from Axial are
described in detail by Butterfield et al. (in press). Fluids were collected from Cloud and
Marker 3 3 vents one year before the eruption and for three years afterwards. Only data
from the year 2000 was available for the Snail vent. In general, major element chemistry
for these fluids is very similar between sites and years, with significant variations seen in concentrations of Has, COe, SiOq,,) and the trace elements Mn, Fe, A1 and Ba between vents (Table 2.1). All diffuse fluids sampled in the Axial S.E. caldera eruption area have chloride concentrations at or below ambient seawater, which indicates that they contain a vapour-rich component and phase separation has occurred at depth. Hydrothermal fluids commonly pass through two-phase conditions at depth in the seafloor and separate into a low-chlorinity vapour phase and a high-chlorinity liquid or
the presence of a high-temperature fluid component (Butterfield et al., in press). Fluid temperatures were measured at the time of basalt collection and are generally similar to those measured in the diffuse fluids over the four years of fluid sampling with the
exception of Marker 33 in 1999, where fluids were much hotter immediately after the
Table 2.1: Fluid Chemistry used for geochemical models." Vent Sample # Year Maximum T (OC) Average T (OC) pH (measured at 22 OC) Alkalinity (meqlkg)
cr
Na'so4'-
SiO~(aq) M~~+ ca2+ K+ H~SW ~e'+ (umoUkg) ~n'+ (umollkg) ~a'+ (umollkg) ~1~' (umoUkg) &(a,, (umollkg) Calculated logjOzcg) Axial Bottom Marker33 Marker33 Marker Marker Seawater 33 33 Cloud Cloud Cloud Cloud Snail *Data from Butterfield et al. (in press). Values are in mmoVkg unless otherwise stated. aValue of Mn taken from Snail 2000 sample R547b19 because no Mn data was available for this particular sample, which had the most complete chemistry.%slues
ofPz(,, were calculated assuming equilibrium with HZ(aq)2.2. Methods
9-2. I . Analytical Methods
For basalt samples from Axial Volcano, the light-grey, "fresh interiors were analyzed along with its paired altered outer rims, for major and trace element
compositions. The altered rims were several mm thick, and included the sample exterior where surficial alteration minerals were scraped off: The rims were separated from the cores based on colour differences, using a diamond saw. For the M E F basalts, altered
rock halos were < 2 cm thick. Fresh glass was used as an analogue for the fresh
interiors for these samples, which was picked off samples by hand.
Major elements and select trace elements (Cu, Cr, Ni, and Zn) were analyzed by
X-Ray Fluorescence (XRF) with a Philips PW244.0 4 kW automated XRF spectrometer
at McGill University in Montreal, Quebec. Fused beads were prepared from a 1:5 sample: Lithium tetraborate mixture. Precision is within 0.5% relative for each element
and accuracy is within 0.5 % for silica, 1 % for other major elements and 5 % for trace
elements (Cu, Cr, Ni, and Zn). Select trace elements (V, Ba, Sc, Rb, Sr, Y, Zr, Nb, Cs, Hf,
Ta, T h and U) and rare earth elements (REE's) were analyzed by Inductively-Coupled Plasma Mass Spectrometry using a Thermo Instruments PQII ICP-MS with a Gilsonm auto-sampler and peristaltic pump at the University of Victoria, in Victoria, British Columbia. Dissolution procedure follows the hotplate methods of Taylor et al. (2002), except that HF was added only once during dissolution. In some samples, a fine, white, insoluble fluoride precipitate appeared. These samples were centrifuged and only the clear solute that formed was used for analysis. Final solutions were mixed to a final
weight of 50
g
in polycarbonate Falcon tubes using 1% HN03 and 1 rnl of a complexingSr, Y, Zr, Nb, Ba, U and most REEs is < 10% (Eu, Tb, Ho, T m and Lu were 15-58%)
and 11-28% for Sc, V, Cr, Rb, Cs and Th. Accuracy is I 10 % for Sc, V, Rb, Sr, Y, Zr
and Ba; 11 to 40 % for Cr, Ni, Cs, Gd, Dy, Ho, Er, Eu, Y, Hf, T h and U; and > 40 % for
Tb, Tm, Lu, T a and Pb (Appendix C).
Major elements compositions for M E F glasses were analyzed by microprobe using a JXA-8900 Superprobe at the University of Alberta, Calgary, Alberta, at 15 kV
acceleration voltage with a 15 nA beam current and a beam diameter of 1 pm. Reported
results are an average of two analyses (Appendix D). Trace elements were analyzed
using Laser Ablation Inductively-Coupled Plasma Mass Spectrometry using a Thermo Instruments PQII ICP-MS at the University of Victoria. Operating conditions followed
those of Chen et al. (~ooo), except that 20 Hz laser output frequency and He carrier gas
were used (Appendix E). Precision is better than 10 % and accuracy is within 10 %
relative for all elements, except T i (30 %) and Sc (20 %) using MPI-DING glass
standards (Canil et al., 2003).
Alteration minerals coating exterior and fracture surfaces were identified using three methods. First, where enough material permitted, coatings were analyzed using a Siemens D.5000 powder X-Ray Diffractometer at the University of British Columbia, at
40 kV, 30
d
Cu-K alpha with a monochromatized scan from 3-50 2 00; samples wererun glycolated and unglycolated to identify clay phases. Second, for very small quantities, coatings were analyzed using Scanning Electron Microscope (SEM) with both a Philips X30, with a Princeton Gamma-Tech EDS system at the University of British Columbia, and a Hitachi S-3500N with an Oxford ISIS EDS at the University of Victoria. Elemental spectra were used to identify potential minerals. Finally, grain mounts of minerals coatings and fresh glass were analyzed using the JXA-8900
Superprobe at the University ofAlberta, at 15 kV acceleration voltage with a 15 nA
beam current and a beam diameter of 1-3 pm. Thin sections hosting secondary minerals,
and plagioclase crystals in all samples were also analyzed by electron microprobe (Appendix D).
aaz.
Geochemical Modeling MethodsSeawater basalt interaction at mid-ocean ridges has been explored by using chemical equilibria and mass transfer computer models, but most studies have focused
on high-temperature systems (Bowers and Taylor, 1985; Janecky and Seyfried Jr., 1984;
Janecky and Shanks 111, 1988; Tivey, 1995a; Tivey, 199.513; Tivey et al., 1995; Tivey and
McDuff, 1990). These models have also been used to examine the effects of host rock composition on secondary mineralogy and fluid chemistry in the subsurface (McCollom and Shock, 199'7; Wetzel and Shock, 2000). One purpose of these types of models is to understand the interaction between hydrothermal fluids, seawater and basalt by examining coexisting fluids and deposits at active sites where both can be directly sampled (Tivey, 1995a; Tivey et al., 1995). In this study, low-temperature water-rock reaction was examined using reaction path modeling in order to examine the conditions under which observed alteration minerals precipitated. Models were performed using diffuse fluid compositions from Cloud, Snail and Marker 33 vents at Axial Volcano only, as no fluid data was available for Village vent or the MEF. Reaction path modeling was carried out using the Geochemist's Workbench software package (Bethke, 2002), utilizing the thermo.dat database compiled by the Lawrence Livermore National Laboratory (Delany and Lundeen, 1990). The database used for modeling provides the thermodynamic data for aqueous species, minerals and gases, equilibrium constants for
reactions to form these species, and data required to calculate the activity coefficients of species in solution. The extended form of the Debye-Huckle equation was used to
calculate activity coefficients of aqueous species and the activity of water was set to 1.
More detailed explanation of modeling methods are discussed in Appendix F.
Default pressure in the program is 1 bar for temperatures less than 100 OC
(Bethke, 1996). Generally, a pressure difference of a few hundred bars does not significantly affect the thermodynamic properties of most mineral phases. Molar
compressibility's for most minerals can be assumed to be zero up to
-
1 kbar, with theexception of sulfate minerals, and the effect of pressure in reactions involving only solid
phases can be assumed to be negligible (Langmuir, 1997). The change in log K values is
considered to be small when compared to the error on these values. A 170 bar database,
supplied by M.K. Tivey, used for comparing log K values at 1 bar and 170 bar, showed
that the change in log K due to this pressure increase was negligible (Appendix F).
Finally, changes in log K values for seawater are negligible up to 1 kbar at 25 OC
(Millero, 1973). Pressure does have dramatic effects on gases, however, which is accounted for in the models by setting the partial pressure (fugacity) of oxygen, by
assuming equilibrium between Ha(aq) and He0 based on the reaction
He
+
'/Z 0 2 + He0and using measured He(,) in the diffuse fluids. Setting the oxygen fugacity determines
the redox state of the system.
The goal of all model types was to investigate what conditions are required in order for the observed minerals assemblage to precipitate in the low-temperature diffuse environment at Axial. Three modeling approaches were used to investigate mineral precipitation in the diffuse environment at the three Axial vent sites. Results of all
models are determined using a set of concentrations that satisfy equilibrium equations
for each possible reaction in the system (Bethke, 1996). The first model involved
speciating the diffuse fluids, which determined the distribution of species in solution, mineral saturation states, pH, Eh and gas fugacities. Results of this model illustrate an equilibrium attained within the fluid, but one that is meta-stable with respect to mineral precipitation (Bethke, 1996); not all minerals predicted to be supersaturated will
precipitate (Tivey, 1995a). The second model involved mixing diffuse fluids with seawater, simulating the natural environment where diffuse fluids exit the vent orifice. This model was run using the diffuse fluid as the initial system and seawater as a reactant, and again, in reverse, using seawater as the initial system and the diffuse fluid as the reactant to ensure different mineral phases would not precipitate. The final model
involved the titration of 100 g of basalt into various quantities of diffuse fluids in order
to simulate the reaction of fluid and rock at the vent orifice, which is a fluid dominated environment.
Diffuse fluids were constrained by entering the fluid mass, in-situ measured values of temperature, pH (measured at 22 oC), alkalinity, and concentrations of major components and the trace components Mn", Few, A13+ and Ba2+ (all in mol/kg). For
the water-rock reaction models, 100 grams of fresh basalt with a major element
chemistry based on a Ti02 free rock was entered as weight of oxides in grams (Table 2.2). Titanium was not measured in the diffuse fluids as samplers are made of Ti, so this element had to be excluded from the composition of the rock.
Suppression of certain mineral phases was required in the models in order to have those mineral phases known to precipitate in the low-temperature environment
mineral stabilities, environment of formation and reaction kinetics. A limitation of the program is that phases known to precipitate at higher temperatures may appear in models that were run at low-temperatures, due to the fact that the program converges
on the lowest AG assemblage. These high temperature phases may be stable at low
temperatures, slow kinetics prevent them from precipitating. hat form only at low temperatures.
Table 2.2: Basalt chemistry used for geochemical models.* Vent Sample Si02 A1203 FeO MnO MgO CaO NazO K20 Ba
co2
Total (g)Marker 33 Cloud Snail
52-08-09 R674-18 R549-12-1 50.33 50.25 50.26 14.86 14.85 14.87 11.30 11.34 11.29 0.20 0.20 0.2 1 7.65 7.68 7.69 12.51 12.52 12.5 1 2.86 2.87 2.90 0.22 0.2 1 0.22 0.003 0.003 0.003 0.07 0.08 0.06 100 100 100
*Bulk chemistry is based on a TiOz free basis and scaled to total 100 g. All values reported as weight percent.
Table 2.3: List of suppressed minerals in geochemical models.
Mineral Phases Reason for Suppression
Quartz, Tridymite, Cristobalite, Chalcedonya
Dolomite Pyrite Pyrrhotite Hematite, Magnetite
Kaolinite, Illite, Muscovite, ~ - f e l d s ~ a r ~ Dawsonite*'
T r ~ i l i t e * ~ ~ h e n ~ i t e * ~
Allows for the formation of amorphous silica Precipitation too slow at low temperatures Allows for the formation of Fe-oxyhydroxides Allows for the formation of Fe-oxyhydroxides Allows for the formation of Fe-oxyhydroxides Not typically found in seafloor environment Not typically found in seafloor environment allows for the formation of goethte and it is
only found in meteorites
Not twicallv found in seafloor environment *These phases were suppressed only in water-rock models.
aMottl and McConachy, 1990 b ~ u r r a y , 1988
9.3. Results
2.3.1. Alteration Mzneralogy
Basalt samples from both young and mature sites are very fresh with alteration
manifest primarily as surficial coatings that were generally < 3 mm thick. Some samples
have slight internal alteration, visible in hand sample as a dark-grey halo, when compared to the fresh, light-grey coloured interior of the sample. Also, secondary minerals partially line vesicles and fractures in some basalts in halos. Surficial coatings
on M E F basalts were generally thicker (2-3 mm) than those found on Axial samples (< 1
mm) and comprise a diverse mineral assemblage. Saponite and nontronite were the most abundant alteration phases at both locations (Table 2.4 and Appendix G).
All Axial basalt samples are very fresh (visual estimate of haloes is < $2 % altered,
whereas alteration for the entire sample is < 1 %) with minor vesicle linings of Fe-Si
oxyhydroxide and glass being only slightly devitrified. Clinopyroxene and olivine are very fresh. An extensive but discontinuous white coating on all samples at Axial consists primarily of saponite, anhydrite and a magnesium-silicate phase, very close in composition to talc or stevensite. The common assemblage (in order of decreasing
abundance) is saponite
+
nontronite+
Mg-silicate+
particulate sulfur+
anhydrite+
zeolites
+
Fe-oxyhydroxide (Table 2.4 and Appendix G).Basalt samples from the M E F are fairly fresh (visual estimate of alteration is <
10 % in haloes and < 4 % for entire sample) with glass that is partially devitrified and
rock that has minor vesicle linings of Fe-Si oxyhydroxide, pyrite and a saponite- celadonite-beidellite mix. Clinopyroxene and olivine are very fresh. The common
assemblage on surfaces (in order of decreasing abundance) is Mn-oxide
+
saponite+
Fe-pyrite (Table 2.4 and Appendix G). The Al-silicates, identified by SEM, have traces of
Ca, K, Na and Fe in various combinations and may possibly be zeolites.
Minerals common to both young and mature hydrothermal sites include zeolites, saponite and nontronite. Particulate sulfur and an Mg-silicate are observed only at Axial, while Mn-oxide is observed as a major mineral phase only at the MEF. Saponite, nontronite and Fe-oxyhydroxides are the most abundant phases at both locations.
Table
2..
Site Axial MEF
klteration minerals and fluid temperatures at the time of basalt collection. Vent
I
Surface Coatings Fluid T Chemical Changes-
Chemical Changes-
Vesicle Linings (OC) gained bv altered rock lost bv fresh rock Anh, Sap, Fe-oxy, S, Bar, Non, Eri, Cal, Mg-Si-0 phase Fe-Si oxy S, Fe-oxy, Anh, Sap, Non, Phil, Mg-Si-0 'loud phase Marker 33 Sap, Non S, Sap, Anh, possibly Phil, Non, Mg-Si- Snail 0 phase Mn-ox, Am.Si, Sap, Bar, Anh, Fe-Si Fe-Si oxy, Sap-cel, S&M oxy, Al-Non, Al-Fe-Si-0 phase Sap-cel-beid l~e-S-0 phase, Sap, Mn-Fe-ox, Bar, Am. Si., Al-Na-Si-0 phase, Mg-S-0 phase, Milli-Q IFe-oxy, Fe-Si oxy, Al-Fe-Si-0 phase, Sap-cel-beid, Py Al-Non Chl, Fauj, Anh Mn-ox, Non, Sap, Bar, Am. Si., Fe-Si S'E' Hulk oxy, Gor, Phil, HyTc Fe-Si oxy Mn-ox, Na-Al-Ca-Si-0 phase, Sap, Anh, W. Grotto Phil Bar, Anh, Sap, Fe-Si oxy, Mn-ox, Non, Py, Sap-cel, Fe-Si laster Island Al-Na-Si-0 phase, Fe-oxy, Py, Na-Mg- Si-0 phase, Gor, Chl, Eri OXY Ba SO2, Fe203, MgO, CaO, Na20, K20 SO2, TiO,, MgO, Na20 CaO, Na20 SO2, Fe203, MgO, CaO K20 Si02, Fe203, MnO, MgO, Ba MnO, Ba MnO, Ba MnO, MgO, K20, Ba MnO, MgO, K20, Ba MnO, MgO, K20, Ba SO2, Ti02, Fe203, CaO Si02, Ti02, Fe203, CaO, Na20 Si02, Ti02, Fe203, CaO, Na20 SiOz, Ti02, Fe203, CaO, Na20 I Minerals are in no particular order. Abbreviations: A1-Si phase = unidentified Al-silicate; Al-Non = Al-rich Nontronite, Am-Si = Amorphous Silica, Anh = anhydrite, Bar = Barite, Cal = Calcite, Chl= Chlorite, Eri = Erionite, Fauj = Faujasite, Fe-oxy = Fe oxyhydroxide or Ferrihydrite, Fe-Si oxy = either a mix of Fe-oxy and Am-Si., or a distinct Fe-Si oxyhydroxide phase, Gor = Gormanite, HyTc = Hydrotalcite, Mn-ox = Mn oxide, Non= nontronite, Phil = Phillipsite, Py = pyrite, S = elemental Sulfur, Sap = Saponite, Sap-cel-beid = saponite-celadonite-high A1 Beidellite mix, Sap-cel = Saponite-Celadonite mix.2,3,2. Bulk Rock Chemistry
Bulk rock compositions for Axial samples are presented in Table 2.5 and M E F
samples in Table 2.6. Axial basalt samples analyzed in this work are very similar chemically to basalt samples collected in the new lava flow, away from diffuse vents (Chadwick et al., 2000). Our M E F samples are chemically identical (within analytical uncertainty) to other basalts from the Endeavour area (Karsten et al., 1995). Axial
basalts are transitional between E and N-MORB (Rhodes et al., 1990), whereas MEF
basalts are typical E-MORB (Karsten et al., 1995). Groundmass plagioclase in samples from both locations have igneous values, with the An content of cores ranging from 0.79 to 0.90 and rims from 0.70 to 0.84.
At both Axial Volcano and the MEF, low-temperature alteration of the basalts resulted in only minor chemical changes in basalts. At Axial, major element
concentrations between fresh and altered pairs differed only slightly, whereas essentially no change was observed in the trace elements, with the exception of Ba (Table 2.5). At the MEF, slightly greater changes were observed in the major elements, and differences in most trace elements and REE's are negligible (Table 2.6). In general, the variation in elemental concentrations between fresh and altered rock in basalt samples from both young and mature hydrothermal sites are only slight.
Table 2.5: Bulk rock chemistry of basalt from Axial Volcano. Vent Site Type Sample - SiOz TiOz A1203 Fez03 MnO MgO CaO NazO K20 p2os co2 LO1 Total Cu Ni Zn
v
C r Ba Sc Rb Sr Y Zr Nb Cs La Ce Pr Nd Sm Eu Gd Tb DY Ho E r Tm Y b Lu Hf Ta Thu
*< dl1 isMarker 33 Marker 33 Snail Snail Village Village Cloud Cloud altered fresh altered fresh altered fresh altered fresh 08-09E 08-09i 12-1E 12-li 674-13E 674-131 674-183 674-181
0.12 0.22 0.14 0.37
is than detection limits. Trace elements for Cloud (altered) are an average of two analyses. Major
-
les
Table 2.6: Bulk Chemistry of basalt from the Main Endeavour Field.
MnO MgO CaO NazO K20 LO1 Total
Vent
Site
Sample SiOz Ti02 AlzO3 Fe203
0.19 0.18 0.26 0.14 0.24 0.19 0.22 0.19 0.18 0.15 0.19 0.17 0.20 0.17 0.20 0.18 7.30 7.04 7.22 6.94 6.36 6.25 6.33 6.28 7.21 6.57 7.04 6.99 5.89 6.33 7.35 6.99 11.73 11.81 11.78 11.64 10.71 10.84 10.76 10.65 11.71 11.37 11.79 11.63 10.79 10.91 11.69 11.87 2.83 2.83 2.79 2.83 3.12 3.17 3.21 3.21 2.88 3.15 2.94 3.01 3.27 3.20 2.88 2.89 0.41 0.38 0.50 0.48 0.60 0.58 0.66 0.58 0.44 0.48 0.43 0.38 0.61 0.57 0.43 0.40 < dl nla 0.66 nla < dl nla 0.05 nla 0.02 da 0.20 nla 0.57 da dl nla 100.28 98.83 99.95 99.18 99.78 98.69 99.82 99.45 99.91 99.12 99.92 99.65 99.78 99.15 100.16 99.45 S.E. Hulk S.E. Hulk W. Grotto W. Grotto Easter Easter Easter Easter M-Q M-Q M-Q M-Q S&M S&M S&M S&M 68-153 68-151 68-163 68-163 68-173 68-171 717-23 717-2i 717-43 717-4i 717-33 717-33 717-1E 717-li 730-1E 730-li 49.53 49.68 49.84 50.68 49.46 49.75 49.59 50.47 49.31 49.95 49.40 50.35 49.21 50.07 49.29 49.91 1.76 1.73 1.66 1.67 2.17 2.14 2.18 2.23 1.76 1.99 1.77 1.81 2.18 2.18 1.76 1.87 14.88 14.55 14.82 14.62 14.44 14.34 14.44 14.47 14.85 14.34 14.95 14.52 14.52 14.37 14.77 14.52 11.37 11.52 10.07 11.01 12.29 12.31 11.99 12.31 11.27 12.01 10.94 11.69 11.94 12.29 11.51 11.71 *n/a indicates that element was not analyzed. Major elements are reported in weight percent and trace elements in ppm. "E: analysis was on the altered exterior of the basalt b. I: analysis was on fresh interior of basalt
Table 2.6: Bulk Chemistrv of basalt from the Main Endeavour Field (continued). Vent Site Sample Cu Ni Zn Co
v
Ba Sc Rb Sr Y Zr Nb Cs La Ce Pr Nd Sm Eu Gd Tb DY Ho Er Tm Y b Lu H f Ta Thu
nla nla n/a nla nla 40.11 45.1 1 2.59 155.85 25.15 94.01 8.22 0.03 6.26 11.75 1.89 9.77 2.74 0.92 3.18 0.5 1 3.49 0.70 2.04 0.29 2.00 0.28 1.76 0.45 0.42 0.07 \ / i.E. Hulk S.E. Hulk W. Grotto W. Grotto Easter Easter Easter Easter M-Q M-Q M-Q M-Q S&M S&M S&M S&M 68-153 68-151 68-163 68-16i 68-173 68-173 717-2E 717-21 717-43 717-43 717-33 717-31 717-1E 717-li 730-1E 730-li 98.00 n/a 110.00 n/a 107.00 112.00 da 95.00 nla 102.00 nla 31.00 nla 30.00 43.00 nla 90.00 38.00 n/a 39.00 333.00 nla 332.00 114.00 67.52 150.00 27.53 40.44 37.01 5.84 4.32 5.82 182.65 172.15 237.57 26.58 31.60 27.83 162.74 136.70 219.99 17.54 13.60 22.89 0.05 0.04 0.00 14.13 10.22 13.09 33.40 17.94 32.26 4.44 2.78 4.54 21.22 13.96 19.41 5.67 3.74 5.26 1.46 1.17 1.26 7.41 4.17 6.44 1.13 0.66 1.13 7.58 4.49 6.04 1.45 0.90 1.21 4.23 2.56 3.73 0.61 0.36 0.53 3.87 2.54 3.41 0.60 0.36 0.47 5.06 2.41 5.02 1.64 0.79 1.97 1.43 0.76 1.31 0.48 0.13 0.92
2.3,3. Chemical Change Calculations
The chemical change that occurred during alteration, between fresh and altered rock, were calculated using Gresens' equation (Gresens, 196'7) modified by Grant
(1986). This method normalizes chemical data to determine relative elemental gains and losses from fresh rock. Grant (1986) modified this method to relate the concentration components in altered rock to that in the original, fresh rock, such that
c;
=MYM*
(c:+
ACJ
where CiA = concentration of component i in altered rock; Mo = mass of fresh rock; MA
= mass of altered rock; Cio = concentration of component i in fresh rock; and ACi
=
change in concentration of component i. Elements that are immobile during alteration defines the mass change term (MO/MA). Combining this with the chemical changes calculated in the basalt determined how much of each element was gained or lost by the fresh rock during alteration. At both Axial Volcano and the MEF, Ale03 was used to calculate the mass change term.
Results of chemical change calculations are discussed as the relative difference from fresh rock, in percent, in order to illustrate the magnitude of chemical change. For the Axial samples, altered rock haloes were compared to fresh, grey interiors. Altered
halos in most samples generally lost Fee03 (< 2 %), gained NaaO (< 3%) whereas SiOe
(k 1 %), MgO (k 1 %), K,O (+ 25 %), Ba and Sr (+ 50 % each) were variable. Based on
these chemical changes, basalts from Axial Volcano have experienced < 1 % alteration.
For M E F basalts, altered rock haloes were compared to fresh glass, which was used as an analogue for fresh interiors. The M E F samples show more element mobility than the
Axial suite, where altered halos show a loss of SiOa (< 5 %), Ti02 (< 15 %), Fee03 (3-10
trace elements, Ba, Rb and Sr showed the greatest mobility, with altered rock gaining
up to 200 % generally for each element. One exception is the S&M sample 'i 17-1, where
Ba changed by
-
2200 %. Changes in the REEs and some other trace elements werevariable in the M E F samples, but generally REEs were gained by altered rock as was
Hf, Y, Zr, Nb and Cs. Based on the chemical change calculations, basalt from the M E F
have experienced < 4 % chemical change due to low-temperature alteration.
A general trend between both young and mature sites is a loss of SiOe and Fee03 from fresh rock and a gain of Ba and Sr to altered rock. The largest chemical changes at both young and mature sites are observed in Ba and Sr. Essentially no chemical change occurred in most trace elements in Axial basalts, whereas altered rock showed an overall gain of most trace elements at the MEF.
2.3.4. Geochemical Models
Geochemical modeling was conducted to investigate the processes that lead to precipitation of low-temperature alteration minerals on the Axial basalt samples. The three model approaches used here consider minerals that precipitate directly from diffuse fluids, as a result of mixing of fluids and seawater, and by reaction of basalt with diffuse fluids. Emphasis is on the Axial suite only, as diffuse fluid data was available over a four-year period and the rocks have a precise age and known history. For all models, the predicted mineral assemblage was compared to the observed alteration assemblage.
2.3.4.1. Model I - Speciation of D @use Fluids
Diffuse fluids were speciated in this mode1 in order to determine what minerals phases would precipitate directly from them. Results show that there is essentially no
oxygen in solution, as the oxygen fugacity was very low (ranging from
-
lo-"
to 10-70) in the diffuse fluids. Oxygen fugacity for bottom seawater was calculated to be lo-*.". Alldiffuse fluid speciation results are listed in Appendix I.
When fluids were speciated with mineral precipitation disabled, diffuse fluids were supersaturated with many phases, such as clinoptililite, muscovite, beidellite, pyrophyllite, mordenite, illite and kaolinite. Supersaturation with many phases is commonly observed in low-temperature fluids (Bethke, 1996). When any minerals were allowed to precipitate, barite, pyrite, quartz and either muscovite or kaolinite
precipitated in all diffuse fluids. Finally, fluids were speciated with mineral precipitation enabled, but with some phases suppressed (Table 2.3). This resulted in barite and mordenite precipitation, and silica and pyrite supersaturation. Amorphous silica
precipitated in addition to these phases from the Marker 33 (2001) fluid. A SiOq,,)
stability plot reveals that all fluids are near saturation with respect to amorphous silica (Fig. 2.2).
Parameters such as temperature, pH and@qg) were varied from measured values in this model to investigate how changes in these control mineral precipitation.
In all diffuse fluids, temperatures greater than 50 OC and pH increases (up to 9) resulted
in the precipitation of clays such as Mg-rich saponite, beidellite and talc. In addition,
barite disappeared in all but the Marker 33 fluids and amorphous silica precipitated in
the Marker 33 (1999) fluid (only at 50 oC). When temperatures were increased to > 125
oC, anhydrite precipitated in all fluids, in some cases with saponite and talc. Temperature increases combined with a pH decrease resulted in only barite and mordenite precipitation. Measured temperatures accompanied with a pH increase to neutral values resulted in talc, barite and mordenite forming, while a pH decrease
resulted in barite only. Finally, increases inflqg) values resulted in the precipitation of barite, mordenite and nontronite in all diffuse fluids.
In summary, diffuse fluid speciation results show barite and mordenite
precipitate fi-om fluids at Marker 33, Cloud and Snail vents, at the conditions measured
at the time of fluid sample collection. In addition to these phases, amorphous silica
precipitated in the Marker 33 (200 1) fluid. Of all the predicted minerals, only zeolite is
observed at Cloud and Snail vents at Axial Volcano. None of the phases predicted were
observed at Marker 33 vent (Table 2.4), possibly because other processes not considered
in this particular model have influenced mineral precipitation at this vent.
Amorphous silica
-2%
2 4 6 8 I0 12
Figure 2.2: Solubility plot for SiOq,,) vs. pH at 25
C. All diffuse fluids plot on the boundary between
2.3.4.2. Model 2 - Dguse Fluid and Seawater Mixing Models
T o simulate mixing of bottom seawater with diffuse hydrothermal fluids as they exit the vent orifice, a model was used that illustrates possible mixing ratios between these two fluids. This permits us to determine the mixing ratios required to produce the observed mineral assemblages. Seawater was speciated first and added to the diffuse fluid (the initial system) as a reactant. Models were also run in reverse to ensure that no different phases were not precipitated. Results of mixing illustrate a sequence of mineral assemblages that precipitate from the mixed fluid as the proportion of seawater
increases (Fig. 2.3A). Minerals listed in Table 2.3 were suppressed and models were closed system, where precipitated minerals were permitted to back-react with the fluid and dissolve. A sliding temperature path was used, where the initial and final
temperatures were those measured in the diffuse fluids and bottom seawater, respectively.
Results show that, in general, similar minerals precipitated in each diffuse fluid- seawater mixture and quartz and dolomite are always supersaturated (Appendix I). In all mixtures, barite and mordenite precipitated, with barite remaining stable at all mixing ratios in all fluids. Mordenite, in contrast, precipitated at all mixing ratios in the Snail fluid and all Marker 33 fluids, with the exception of 1998. In all Cloud fluids and
the Marker 33 (1998) fluid, mordenite precipitated only when there was < 5-20 %
seawater, temperatures were > 10-1 8 oC, p H was > 5.7-6.8 and$Iqg was < (Fig.
2.3B). Na-rich nontronite, pyrolusite and goethite precipitated in these fluids when no mordenite was present. Nontronite precipitated at all mixing ratios in the Snail fluid
and the Marker 33 (1999) fluid. In Marker 33 (2000) and ZOO^), nontronite precipitated
A Temperature ('C)
0
Diffuse Fluid - Seawater Ratio
Mixing fraction
Figure 2.3: Results of mixing between seawater and Cloud (1998) diffuse fluid. A: mineralogical changes that occurred in the diffuse fluid as seawater reacted into it. B: changes in
Pe(gj, and pH as seawater is reacted into the diffuse
fluid. Goethite, nontronite and pyrolusite precipitate once mordenite dissolves, whenJ3qg) begins to increase to seawater values. Talc precipitated when goethite dissolved.
precipitated at > 20 % seawater, temperatures > 14 "C, p H < 6.1 and $la(g) < 10-lo; and
goethite did not precipitate. Talc precipitated in all Cloud fluids and in the Marker 33
(1999) and (2001) fluids when goethite dissolved, shown by the reaction
4 FeO(0H)
+
4.5 Mg2++
6 SiOq,q) + 1.5 MgsSi40lo(OH)a+
H++
4 Fez++
Oq,,)and when temperatures were < 6 oC, pH > 6 andflq,) at lo-+..". Finally, amorphous
silica precipitated in the Marker 33 (2001) fluid at seawater < 35 %, temperatures > 17
"C, pH < 6.1 andflqg) < In general, mordenite, nontronite and amorphous silica
precipitated at reducing conditions at < 35 % seawater, whereas pyrolusite, goethite and
talc precipitated under more oxidizing conditions at > 20 % seawater. In all models,
only very small amounts (-0.001 g) of the predicted minerals precipitated from the mixed fluids, requiring a time-integrated effect for hydrothermal minerals to be evident on rock surfaces.
As seawater concentration increased in the mixtures, concentrations of SiOq,,), Mgw, Few and Mnw decreased due to dilution with seawater, or their consumption by precipitating minerals; concentrations of other species were relatively stable. The amount of seawater required in the mixtures forflqg) to reach seawater values increased generally with increasing concentrations of He(,,) and HzS(,,) in the diffuse
fluids, and with decreasing pH (Table 2.7).
Mixing of diffuse fluid and seawater precipitated barite, mordenite, nontronite, pyrolusite, goethite, talc and amorphous silica in various combinations: barite,
nontronite and mordenite precipitated in all fluids; pyrolusite and goethite only in
Cloud fluids and the Marker 33 (1998) fluid; talc only in Marker 33 (1999) and (2001);
and the Mg-silicate are observed at all three vents and zeolite is observed at Cloud and Snail vents. No other predicted phases are observed.
Table 2.7: Conditions a t which diffuse fluid-seawater mix became seawater- dominated*.
Diffuse Fluid Sample
I
Fluid H2S(,,) Hz(aq)s e a z t e r p~ (mmol~kg) (mmol~kg) U•‹C) P2M
*Seawater dominated conditions in the mixtures are when pH and reached seawater values of 7.8 and 10-+5, respectively. The parameters listed here are those measured in the diffuse fluids.
Snail 2000 Cloud 200 1 Cloud 1999 Cloud 2000 Cloud 1998 Marker 33 2000 Marker 33 1998 Marker 33 200 1 Marker 33 1999 pp 7 6.1 0.09 0.07 14.7 -68.92 15 6.9 0.02 0.26 9.6 -71.32 16 5.8 0.49 4.2 1 19.9 -68.17 25 6.0 0.12 0.50 15.8 -68.94 28 5.3 0.75 10.16 20.6 -67.73 32 6.2 0.70 2.25 33.4 -67.13 3 3 4.9 1.94 6.08 19 -67.65 44 5.6 0.20 8.84 30 -66.88 47 4.83 2.32 11.88 68.3 -56.45
2.3.4.3. Model 3- Dzfuse Fluid - Basalt Reaction
Reaction of basalt with diffuse fluid is simulated in this model, where conditions are fluid dominated (large amounts of fluid flow by and interact with small amounts of rock). Diffuse fluids were first speciated and basalt was then added as a reactant. Water-
rock ratios ranging from 10: 1 to 1000: 1 were examined using 100 g of basalt (reacted
homogeneously) and 1 kg of diffuse fluid (however ratios as low as 1: 1 were examined;
at ratios less than
lo:
1, many mineral phases that were not observed precipitated).Mineral suppressions (Table 2.3) were made before reaction with basalt. The model assumes local equilibrium and is closed system, with the reaction path tracing chemical and mineralogical changes that occur in the diffuse fluid as basalt is reacted into it.
At all water-rock ratios and for all diffuse fluids, saponite nontronite, mordenite
and barite precipitated (Fig. 2.4A and Appendix I). Saponite, nontronite and mordenite
were the most abundant phases, producing up to 15 g of each mineral per 100 g of basalt
reacted. Barite (0.001
g
per 1 kg fluid) and gibbsite (0.001 g per 1 kg fluid) wereproduced in very small amounts. The predicted mineral assemblage best matched the
observed assemblages when water-rock ratios were greater than 200: 1 and pH < 7.5 in
all fluids, with the exception of Marker 33 (1999). Saponite, nontronite, mordenite, gibbsite and barite all precipitated at these high water-rock ratios. At ratios less than this, phases not observed (calcite, alabandite, rhodochrosite) precipitated. In the Marker 33 (1999) fluid, the predicted assemblages best matched the observed at water-rock
ratios > 100: 1, where saponite, mordenite, Mg-rich beidellite, nontronite, barite and
gibbsite precipitated. Talc precipitated in the Marker 33 (2000) fluid at water-rock
ratios > 500: 1 and amorphous silica precipitated in the Marker 33 (2001) at water-rock
Water-rock ratio
Mass reacted (grams)
Figure 2.4: Results of reaction between basalt and Cloud (1998) diffuse fluid. 100 g of fresh basalt was reacted into 1
kg of diffuse fluid. A: mineralogical changes that occurred in
the diffuse fluid as basalt was reacted into it. Phases
precipitated upon rock reaction. B: jQg) decreases steadily
