lNculDlNG

ISOTOPE GEOSCIENCE
Chemical Geology 126 (1995) 119-135

The boron isotopic composition of altered oceanic crust

H. Jesse Smith a**, Arthur J. Spivack b, Hubert Staudigel ‘,‘, Stanley R. Hart d
aScripps Institution of Oceanography, University of California at San Diego, La Jolla, CA 92093-0220, USA
b Department of Earth Sciences, 601 South College Road, University of North Carolina at Wilmington, Wilmington, NC 28403-3297, USA
’ Centerfor Isotope Gemology,Netherlands Research School of Sedimentology, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
d Woods Hole Oceanographic Institution, Woods Hole, MA 02543 USA

Received6 January 1995;accepted31 July 1995

Abstract

Samples of drilled oceanic crust, from DSDP Holes 417A, 417D and 418A and ODP Hole 735B, and oceanic crust from the
Oman and Cyprus ophiolites, were analyzed for B contents and S”B. Composite samples from DSDP Holes 417A, 417D and
418A were used to represent the upper 550 m of altered oceanic crustal Layer 2A. Whole-rock samples from the Troodos
ophiolite, Cyprus, and the Oman ophiolite were selected to represent crustal Layer 2B dikes. Composite samples from ODP
Hole 735B were used ~torepresent crustal Layer 3.
The B content of the DSDP composites ranges from 7.2 ppm to 104 ppm and correlates with both 6”O and K, showing that
it is a good indicator of the extent of low temperature alteration. The S”B of the DSDP composites varies between - 2.5%0 and
5.4%0.The B content of the samples from the Troodos ophiolite ranges from 2.4 ppm to 8.1 ppm; S”B varies from - 0.9%0 to
7.8%0.The B content of the Oman ophiolite samples ranges from 5.0 ppm to 11.1 ppm; S”B varies from - 1.6%0to 16.9%0.The
B content of the samples from ODP Hole 735B ranges from 1.1 ppm to 7.1 ppm; 6”B varies from -4.3%0 to 24.9%0. The
general pattern displayed by these samples is one of greatest (and most variable) B enrichment at the top of the crust and least
enrichment at the bottlom of the section. All of these samples are enriched compared to unaltered MORB, which is believed to
have a B content of approximately 0.5 ppm. The S”B values of deeper samples, from Layers 2B and 3, are more variable and
generally higher than those from Layer 2A. Boron contents and G”B are not correlated.
The data from the DSDP Site 417/418 composites indicate that the S1lB of fluid circulating in the upper crust changes only
slightly during alteration, increasing by an average of 5.1%0 with an accompanying decrease in B concentration of 7%. Low
temperature alteration’ appears to be a water-dominated process resulting in minor modification of circulating seawater. A
minimum water-rock ratio of 400 is calculated for these samples, implying a minimum low-temperature seawater flux through
the upper oceanic crust of 3.4X lOI l/y.
The average B content of altered oceanic crust, as represented by these samples, is 5.2 f 1.7 ppm and the average S”B is
3.4 f 1.1%0.This average isotopic composition is measurably different from the apparent average of oceanic sediments, supporting
the idea that S”B cou.ld be useful for identifying the source(s) of B in island arcs.

1. Introduction Boron is an ‘excess volatile’; the B inventory of the

earth’s surface is too great to have been derived solely
from the weathering of igneous rocks and must prin-
* Correspondingauthor.
’Also at: Institute for Geophysicsand PlanetaryPhysics,Scripps cipally result from outgassing of the earth’s interior
Institution of Oceanography, University of California, La Jolla, CA (Goldschmidt and Peters, 1932; Rubey, 195 1) . Chem-
92093-0225, USA ical differentiation has resulted in the concentration of

0009-2541/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved

SSDIOOO9-2541(95)00113-l
 120 H.J. Smith et al. /Chemical
B in the ocean and crust; the B content of the upper
mantle is argued to be 0.1-0.2 ppm, while concentra-
tions in the principal surface reservoirs are two to three
orders of magnitude higher (Harder, 1978; Broecker
and Peng, 1982; Spivack, 1986). The distributions of
B and S”B are controlled by a dynamic cycle of weath-
ering, uptake by sediments and oceanic crust, subduc-
tion and degassing (Agyei, 1968; Spivack, 1986;
Spivack and Edmond, 1987; Spivack et al., 1987;
Chaussidon and Albarede, 1992; You et al., 1993). One
consequence of this cycling is that there are large iso-
topic fractionations between the major terrestrial B res-
ervoirs. Boron concentrations and B isotope ratios are
useful in the study of the low temperature alteration in
oceanic crust, including the evolution of circulating
fluid, the secular evolution of seawater, subduction-
related magmagenesis and the flux of volatiles from
subducting lithosphere into the mantle (Thompson and
Melson, 1970; Spivack, 1986; Spivack and Edmond,
1987; Palmer, 1991; Edwards et al., 1993; Ishikawa
and Nakamura, 1994). All of these applications require
an understanding of the distributions of B and S”B in
altered oceanic crust.
Fresh MORB has a low B abundance and becomes
progressively more enriched by low-temperature alter-
ation as it moves off-axis. In contrast, B is efficiently
extracted from oceanic rocks by seawater at tempera-
tures higher than 375°C and perhaps higher than only
150°C even at water-rock ratios as low as one (Sey-
fried et al., 1984; Berndt et al., 1988; Berndt and Sey-
fried, 1990). Initial’ studies of oceanic basalt suggest
that the B concentrated in secondary phases formed
during low-temperature submarine weathering is iso-
topically fractionated by approximately - 40% from
seawater (Spivack and Edmond, 1987; Ishikawa and
Nakamura, 1992). Variations in S”B may depend on
differences in mineralogy, fluid evolution and altera-
tion temperature, though the relative importance of
those controls has not been assessed. Furthermore, the
distributions of B and B isotopes in the oceanic crust
have not been well established.
The utility of S”B as a tracer of subduction-related
magmagenesis can only be exploited if the B isotopic
compositions of the sources of B in island arc volcanic
rocks are well known. Boron enrichments in island arc
lavas are consistent with a model in which aqueous
fluids, mobilized during the subduction of oceanic crust
and sediments, chemically enrich, and promote melting
Geology 126 (1995) 119-135
of, the overlying mantle wedge (Tatsumi et al., 1986;
Tatsumi, 1989; Morris et al., 1990). Coupled B-“Be
data have been used to demonstrate that subducted
young sediments are involved in the process of island-
arc magmagenesis, probably as sources of material
transported by fluids (Morris et al., 1990). This sug-
gests that older sediments are involved, too. It has been
postulated that altered oceanic basalt might also be an
important source of water and other volatile species
(particularly B ) mobilized during subduction (Spi-
vack, 1986; Tatsumi, 1989). This potentially important
source of B may be identifiable on the basis of its S”B.
Knowledge of the B content and 6”B of the oceanic
crust is crucial for understanding the subduction-driven
(re)cycling of B and, possibly, quantifying the sources
of B in arc volcanic rocks.
The goal of this work is to determine the large-scale
distributions of B and S”B in oceanic crustal Layers
2A, 2B and 3. Samples from DSDP Sites 4 17 and 418,
ODP Hole 735B and the Troodos and Oman ophiolites
were analyzed for this purpose, as well as to infer
changes in fluid composition during alteration and to
estimate the flux of seawater through the oceanic crust.
A B budget for the upper oceanic crust is calculated
and the 6r’B of the oceanic crust is considered in terms
of its role as a tracer of subducted volatiles.
2. Materials and methods
2.1. Materials
Two distinct approaches are available for determin-
ing geochemical variability on scales of meters to hun-
dreds of meters: either a large number of bulk samples
can be analyzed and arithmetically combined in the
proportions in which they occur in the oceanic crust or
a smaller set of samples (‘composites’) can be pre-
pared by combining material from a broader range of
locations before analysis. While information on inter-
nal structure is lost with composites, their use reduces
the number of analyses required to generate a first-order
description of the area sampled. Both composite and
bulk samples were used in this study. Additionally, two
types of composites were used. The DSDP Hole 417A,
417D, 418A composites were prepared by combining
individual samples from various depth ranges in pro-
portions in which those rock types are observed to
 H.J. Smith et al. /Chemical Geology 126 (1995) 119435 121

occur. The ODP Hole 735B composites, referred to as Table 1

B in DSDP Site 417/418 composites
‘strip composites’ for clarity, were prepared simply by
powdering approximately meter-long strips of core Sample Depth B 6”B Method Weight
material. (m) (ppm) (X0)
Composite samples from DSDP Holes 417A, 417D
and 418A were analyzed for B content and a1 ‘B. DSDP DSDP:
417A-24 10 104*5.3 1.6f0.4 HF 0.021
Holes 417A, 417D ar?d 418A were drilled in 1 lO-m.y.-
417A-32 86 52k2.7 0.6kO.4 PY 0.020
old crust at the southern end of the Bermuda Rise, 417A-44146 187 69f3.6 2.4kO.4 PY 0.076
immediately north of the Vema Gap. Mineralogic and 417D-22127 I 26f 1.3 - 1.7f0.4 PY 0.077
isotopic studies have: shown that these cores are mid- 417D-39 121 40-12.1 3.1 f0.4 PY 0.100
ocean ridge basalt (MORB ) which has undergone per- 417D-59/60 281 25f1.3 -2.5kO.4 PY 0.149
418A-15 1 30f1.7 2.3f0.4 HF 0.080
vasive low-temperature alteration ( <50”(Z) by
418A-40 174 23fl.2 3.t3+0.7 PY 0.074
seawater (Muehlenbachs, 1980; Friedrichsen and 418A-73175 11.8kO.4
458 1.2kO.4 HF 0.200
Hoernes, 1980; Lawrence, 1980; Humphris and Mar- 418A-86 536 7.220.2 0.4*0.5 PY 0.200
riner, 1980; Alt and Honnorez, 1984). In general, the O-100 MF/P &lo0 4Ofl.O 2.0*0.4 HF
major weathering products are smectites and carbon- loo-300 MF/P 100-300 34+ 1.0 1.6kO.4 HF
300-550 MF/P 3w550 10.2 f 0.5 0.5kO.6 HF
ates with variable amounts of alkali feldspars, iron
O-100 HC/PB O-100 59ztl.O 0.3kO.5 HF
hydroxides and zeolites. Iron-containing phases indi- 10&300 HC/PB 100-300 64kl.O 2.6f0.4 HF
cate that both reducing and oxidizing conditions 30&550 HC/
occurred during various stages of alteration (Humphris PB 30&550 16f0.5 5.4f0.4 HF
and Marriner, 1980; Alt and Honnorez, 1984; Gillis Integrated 26.2f0.5 0.8 f0.4 0.997

and Robinson, 1988).

Depths are relative to the top of the crust. Methods: PY = pyrohydrolysis
These drilled sites are well-suited for characterizing followed by methyl borate distillation and mixed bed ion exchange;
the B distribution in the altered upper oceanic crust for HF = hydrofluoric acid dissolution with mannitol followed by ion exchange.
several reasons. First, they are located in old crust that Weights arc based on the observed abundances of material of the types con-
has had ample time to undergo submarine alteration tained in those samples and ate used for the calculation of integrated values.

(Staudigel et al., 1981; Alt and Honnorez, 1984). Sec-

ond, based largely on mineralogy, al80 and elemental
distributions in secondary phases, Holes 417D and
4 18A seem to represent typical altered oceanic crust,
while Hole 417A was more highly altered under oxi-
dizing conditions. The extensive alteration at 417A has
been attributed both to unusually high fluid flux and to
a longer period of fluid circulation (Muehlenbachs,
1980; Donnelly et al., 1980). Third, these holes all
sampled the top of oceanic crustal Layer 2, the layer
expected to contain most of the low-temperature alter-
ation. All holes had high recovery rates, ranging from
61% (417A) to 72% (417D and 418A) (Donnelly et
al., 1980).
Two types of composites were used to study the
samples from DSDP Sites 417 and 418, ‘depth com-
posites’ and ‘lithological composites’. Depth compos-
ites were developed to examine the composition of the
crust over small intervals, of approximately l-7 m, at
specific depths. They are composed of 6% volcaniclas-
tics, 30% massive flows and 64% pillows, proportions
chosen to reflect the ‘composition of the crust recovered
from these cores. Volcaniclastics contain pillow frag-
ment breccias, pillow breccias and hyaloclastites. They
are more porous than and, consequently, more suscep-
tible to submarine alteration than massive flows or pil-
lows. They also contained abundant (now altered)
volcanic glass, which is unstable in the presence of
seawater. These depth composites are used to estimate
the composition of average oceanic crust by weighting
them according to their inferred abundances in these
holes and arithmetically combining them to give ‘inte-
grated’ values of B content and a1 ‘B (Table 1) . Lith-
ologic composites were developed to examine
variations between volcaniclastics ( ‘VCL’) and sub-
marine flows ( ‘FLO’). VCL and FL0 material were
mixed in proportions typically found in these core
materials: interpillow material, margin material and
more crystalline basalt samples were included in their
respective abundances. They were prepared to reflect
the composition of distinct lithologies over broad
ranges of depth. A total of sixteen different composites
were analyzed. These results are supplemented by anal-
122 H.J. Smith et al. /Chemical Geology I26 (1995) 119-135

Table 2 (Schiffman et al., 1987). The mineralogy of these sam-

The K and S’*O compositions of DSDP Site 417/418 composites
ples is discussed in detail by Schiffman et al. (1987, ,
S’s0
1990) and Schiffman and Smith (1988). Their al80
Sample K (so)
values are listed in Table 3. The second suite of crustal
DSDP: Layer 2B samples analyzed is from the Oman ophiolite.
417A-24 3.64 17.99 The dike swarms of the Oman ophiolite consist of 100%
417A-32 1.35 13.12
dikes with no intervening country rock (Coleman,
417A-44146 0.90 13.2
1977). Their primary mineralogy has been nearly com-
417D-22127 0.63 10.67
417D-39 0.69 11.45 pletely replaced as a result of extreme hydrothermal
417D-59/60 0.25 7.87 activity. The petrography and mineralogy of the Oman
418A-15 0.5 1 10.77 ophiolite dikes is discussed by Lippard et al. ( 1986).
418A-40 0.70 10.53
The ophiolite samples must be used as representa-
418A-73/75 0.14 8.8
8.27
tives of typical oceanic crust with caution. Ophiolites
418A-86 0.11
O-100 MF/P 0.92 11.64 may be formed from crust made at a mid-ocean ridge,
100-300 MF/P 0.65 10.26 in a back-arc basin, an arc or a fore-arc (Reinhardt,
300-550 MF/P 0.09 7.69 1969; Miyashiro, 1973, Miyashiro, 1975; Cameron et
O-100 HC/PB 1.51 18.7
al., 1980; Coleman, 1981; Pearceet al., 1981; Alabaster
100-300 HC/PB 1.71 15.18
300-550 HC/PB 0.41 11.6
et al., 1982; Hawkins and Evans, 1983; Moores et al.,
Integrated 0.47 9.87 1984). Many ophiolites, including the Troodos and
Oman ophiolites, are now thought to originate in supra-
Oxygen isotope analyses by B. Smith. subduction environments (Pearce et al., 1984). Oce-
anic crust formed at a supra-subduction spreading
yses of K contents and oxygen isotopic composition center could be significantly enriched in B, in the same
(Table 2). Trace element abundances, 87Sr/86Sr and way that other subduction-related lavas are (Palmer,
‘43Nd/144Nd are presented in Staudigel et al. ( 1995). 1991; Ryan and Langmuir, 1993; Ishikawa and Naka-
Suitable material for the study of B in the lower mura, 1994; Smith, 1994). If that is true, their primary
oceanic crust is rare. ODP Hole 504B, analyzed for B B and 6”B inventories of ophiolites may be different
content and S”B by Ishikawa and Nakamura ( 1992), than that of average altered MORB. Although the Tro-
is the only location where drilled oceanic crustal Layer odos and Oman ophiolite samples are enriched com-
2B dikes have been successfully recovered. ODP Hole pared to MORB, their B concentrations are not as high
735B is the only source of drilled samples taken directly as most island-arc lavas. This suggests that they did not
from intact oceanic crustal Layer 3. The other major
source of oceanic crust is from ophiolites. The Troodos Table 3
and Oman ophiolites, in particular, have good expo- Troodos ophiolite B compositions
sures of the lower crust and have been thoroughly
Sample B (ppm) S”B (%Q) 6’so (%0)
investigated geologically, geochemically and tectoni-
cally. 7-15-84-6 (15-8) 5.13io.05 0.3 f 0.4 9.9
Two suites of samples, from the Solea Graben, Tro- 7-15-84-9 (15-11) 8.06*0.08 7.2 f 0.4 11.5
odos ophiolite, and from the Oman ophiolite, were ana- 9-19-85-01 6.65 f0.07 3.4*0.4 7.7
9-21-85-02 5.03 io.05 4.6 f 0.4 8.0
lyzed in order to determine the B composition of dikes
9-23-85-03 3.35 f 0.03 -0.1*0.4 4.2
from crustal Layer 2B. The samples from the Troodos 9-23-85-04 4.56 f 0.05 0.0 f 0.4 8.8
ophiolite are all highly altered diabasic rocks from 9-24-85-02 3.1s*o.o3 1.6f0.4 5.0
within the Troodos Sheeted Intrusive Complex (SIC) 80103 3.13*0.03 3.5
(Wilson, 1959), a region thought to be a remnant 80104 2.41 f 0.02 7.8 f 0.4 3.5
76-84-5 (6- 11) 3.11*0.03 -0.9*0.4 7.5
spreading center from a supra-subduction environment
Average: 4.46kO.15 3.7* 1.2
(Moores et al., 1984). The pervasive alteration
observed within the SIC is evidence of vigorous, high- Oxygen isotope analyses from Schiffman et al. ( 1990) and Schiff-
temperature circulation of seawater through the crust man (pers. commun.).
 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135 123

Table 4
possess the unusually high primary B inventories one
ODP Hale 73SB B data
might expect if they lhad been enriched by inputs from
subducted lithosphere. This implies that it is acceptable SUllpk Minimum Maximum B 6”B K
to use ophiolite samples as representatives of typical depth (m) depti (m) (ppm) (W.) (PPm)

oceanic crust, at least with respect to B.

B 18.59 21.29 4.95 fO.05 -2.210.4 1064
Composite samples from ODP Hole 735B were cho- 24.01 25.30
sen to examine the B and S”B of Layer 3 gabbros. This C 29.66 34.17 7.05 +0.07 5.01to.4 1342
1 192.20 194.29 1.85f0.02 5.9f0.5 388
hole, drilled into 12-m.y.-old crust at the Atlantis II
I 215.01 216.26 1.71 io.02 2.7 f0.4 658
Fracture Zone on the Southwest Indian Ridge, contains K 229.21 230.58 I .79 f 0.02 8.4 f 0.4 583
six distinct lithological units, all composed of either L 256.80 257.93 2.39 f 0.02 -1.7f0.4 2219
M 262.27 263.49 1.92 f 0.02 -4.3Ito.4 1323
foliated metagabbros or cumulate gabbros (Robinson
0 319.97 320.58 2.63 + 0.03 7.6f0.4 374
et al., 1991). Gabbros from ODP Hole 735B have P 362.21 363.41 l.ixfO.O1 24.9f0.4 281
undergone early, near-solidus deformation and altera- Q 394.83 396.7 I 2.11*0.02 19.2 f0.4 366
T 444.58 445.96 2.73f0.03 11.4io.4 911
tion, low-temperature smectite formation and late, low-
V 481.11 482.43 2.19 i 0.02 11.3f0.4 649
temperature oxidative alteration (Dick et al., 1991). Average: 2.70 * 0. I 5.7 * 1.4
They have 6180 values (mostly between 3 and 7%0;
Kempton et al., 1991) which indicate alteration at high
temperature. These data, combined with data on min- metamorphic variations in Hole 735B in the propor-
eral chemistry and flluid inclusions, indicate that most tions in which they occur.
of the alteration occurred at temperatures between Twelve strip composite samples of ODP Hole 735B
300°C and 600°C at water-rock ratios close to unity were analyzed for B and S”B. Sr, Nd and Pb isotope
(Stakes et al., 1991; Vanko and Stakes, 1991). This compositions of 735B gabbros (Kempton et al., 1991;
early alteration was followed by later episodes of low- S.R. Hart, J. Blusztajn and H.J.B. Dick, in preparation)
temperature smectite formation and low-temperature are similar to those of N-MORB. Major and minor
oxidative alteration; the presence of pure green smec- element data for these samples (S.R. Hart et al., in
tites in some veins implies alteration at temperatures preparation) also resemble those of typical MOR-gen-
less than 5O”C,and later clay minerals, Fe oxides and erated rocks. The B and S”B data are complemented
calcite were formed at estimated temperatures of less by analyses of K content (Table 4). These factors
than 25°C (Robinson et al., 1991; Hart et al., 1994). If make the ODP 735B strip composites good candidates
the altered basalts from DSDP Sites 417 and 418 are for characterizing the distributions of B and S”B in the
good analogs, most of the secondary B in the ODP Hole lower oceanic crust.
735B gabbros is contained in these low-temperature ODP Hole 735B did not necessarily sample typical
minerals. The primary alteration phase is sub-solidus oceanic Layer 3 gabbro, however. The extremely flat
amphibole. Complete descriptions of the material topography of the 36 km2 platform on which the hole
recovered in this holle are in Shipboard Scientific Party was located suggests that this area was unroofed,
(1989). uplifted, eroded by waves and subsided to its present
The strip composites from ODP Hole 735B, like the depth of 700 m (Shipboard Scientific Party, 1989).
DSDP samples, provide a sample base at the meter scale The hole was drilled in oceanic crust that formed at a
instead of the centimeter scale, for which local varia- very slowly spreading ridge with a full spreading rate
bility resulting from mineralogical differences may of only 1 cm/y over the past 20 m.y. (Shipboard Sci-
obscure important general trends. These composites entific Party, 1989). However, since the B content of
were prepared by homogenizing l-cm strips of bulk- oceanic crust is largely a result of secondary, low-tem-
rock material, cut from core sections 0.5 to 4.5 m in perature alteration, and not a primary feature inherited
length Unlike for [the Site 417 and 418 samples, no from its parent magma, its slow-spreading origin may
attempt was made 1.0combine specific proportions of not be important in terms of the B systematics of these
differently altered material or different lithological samples. Still, the crust at this site may have been more
types, although the locations of the strip samples were highly fractured during its early cooling history than
carefully chosen to include all of the lithologic and crust formed at a fast-spreading ridge. Furthermore,
124 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135
Layer 3 is exposed directly to seawater at this site.
These processes could have produced a higher degree
of alteration in these samples than that occurring in
most of oceanic crustal Layer 3. Unfortunately, ODP
Hole 735B is the only hole from which in-situ Layer 3
gabbros have been recovered relatively completely, so
it cannot be compared to other samples in order to
establish whether or not it is typical. Although ODP
Hole 735B sampled crust which is atypical in some
ways, its B content is not obviously anomalous, so these
samples will be used with caution for characterizing
average oceanic crustal Layer 3.
Intact oceanic crust is difficult to sample and there
are few drilled locations at which high recovery rates
of oceanic crust have been achieved. The samples used
in this work are from some of the best, and only, struc-
tures appropriate for modeling the B and B isotopic
compositions of altered oceanic crust. Therefore, these
results will be used with the understanding that further
studies of intact oceanic crust are still needed to know
how well they represent average altered crust.
2.2. Methods
Two different methods were used to extract and
purify B from these ocean crust samples. The first,
referred to as the pyrohydrolysis method, involves
extraction of B from a sample of powdered rock by
melting in an induction furnace, collection of B with
water vapor, methyl borate distillation to isolate B and
B purification by ion exchange. The second, referred
to as the HF method, involves the isolation of B by
dissolution of powdered rock with HF in the presence
of mannitol, followed by multiple ion exchange steps
to purify the B. These procedures have been described,
in detail, by Spivack ( 1986), Nakamura et al. (1992)
and Smith ( 1994). The pyrohydrolysis method was
used for most of the DSDP Site 417/418 depth com-
posites, but was discontinued after procedural blanks
abruptly became too large (apparently due to contam-
ination from the furnace and support tubes) to allow
the accurate determination of S”B. The HF method
was used for all of the other samples. The methods used
for the analyses of these samples are listed in Table 1.
The B concentrations of samples extracted by induc-
tive melting were determined by high performance ion
chromatography. A split of approximately 10% of the
extracted sample was used for this. The split was first
cleaned by mixed bed ion exchange, with 0.1 ml of
resin, and eluted with 3 ml of analytical water. The
clean sample was then evaporated, under an infrared
lamp and filtered air, to a volume of roughly 0.5 ml. 50
~1 of the sample, typically containing between 10 and
100 ng B, was injected into a Dionex 4OOOi,all Teflon
ion chromatograph equipped with a Dionex ASl-ICE
column and ion micromembrane conductivity suppres-
sor with 3% tetrabutylammonium hydroxide regener-
ant. The sample was eluted with a 0.1 M mannitol/
0.037 mMHCl0, solution at a flow rate of 0.8 ml/min.
The precision of the measured B concentrations, based
on repeated analysis of boric acid standards, is _+5%.
The B concentrations of samples dissolved in HF were
determined by negative ion isotope dilution thermal ion
mass spectrometry (negative ion IDTIMS) using the
BO; ion, with precisions of approximately 1%. All
S”B determinations were made by positive ion TIMS
of Cs,BO:. The standard deviation of this measure-
ment, based on repeated analyses of SRM 951 boric
acid, is +0.2%0. We report two times the standard
deviation, f0.4%0, as a measure of precision, except
as noted in Table 1.
3. Results
The B content (and its variability) of the DSDP
depth composites generally decreases with increasing
depth (Fig. 1 and Table 1). In Hole 417A, B varies
from 104 to 52 ppm, in 417D it varies between 40 and
25 ppm and in Hole 418A it decreases monotonically
from 30 ppm at the top of the core to 7.2 ppm in the
deepest composite. The shallowest samples display the
greatest variability in B concentrations. The B concen-
trations of all the depth composites are at least an order
of magnitude larger than that of unaltered MORB,
approximately 0.5 ppm (Spivack and Edmond, 1987;
Ryan and Langmuir, 1993; Chaussidon and Jambon,
1994), the highest concentrations being found at the
most altered of the three sites, 417A. Hole 417A com-
posites are also the most enriched in ‘a0 and K
(Table 2). The lithological composites show a similar
depth dependence (Fig. 1) : the B content of FL0 sam-
ples decreases with depth, from 40 ppm to 10 ppm,
while the VCL vary from 64 ppm to 16 ppm. The
interval which displays the maximum B concentration
 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135 125

positively correlated in both the depth and lithological

0 rnA ccc composites with an average B/K ratio of 0.004. The
0 V B/K ratio of altered upper crust is approximately a
tm factor of two lower than that of seawater as noted by
100
ml
Donnelly et al. (1980) but well above the ratio in
.
E 200 0 v’
140
E
! I I I I

fj 300 .
.
n 120

400
ov i
A 100

500
.
.
E 80
I / I I 2
0 20 40 60 80 100 120 ti 60
___~ ~~
6, rwm 0 417A
40 ??417D
Fig. 1. B ( ppm) vs. depth into basement (m) for DSDP composites. . 41SA
0 FL0
The depths of the FL0 and VCL samples and the integrated value v VCL
? &X%&TED
are the average depths of ltheintervals from which they were sampled
0 TROODOS
(i.e. the O-100 m interval is plotted at 50 m, the 100-300 m interval
at 200 m, the 300-550 m interval at 425 m and the integrated value
at 275 m). Error bars are 20 values and are shown only when they 30
are larger than the symbols. PO

120 Fig. 3. 6”O vs. B (ppm) for DSDP Hole 417A, 417D and 418A
T depth composites, FL0 lithological composites, VCL lithological
f composites and the integrated value, and Troodos ophiolite samples.
100
Data compiled by Spivack and Redmond( 1987) for a suite of DSDP
Hole 417A bulk samples are also included (Donnelly et al., 1980;
80 Mueblenbachs, 1980).
30
E I V
ES0 V
25 A
ui

20 A
.
15 .

Am
=m 10
00

0 1 2 3 4 5

K, % 0
0
Fig. 2. K (% by weight) vs. B (ppm) for DSDP Hole 417A, 417D
and 418A depth composites, FL0 lithological composites, VCL lith- -5
ological composites and the integrated value, and ODP Hole 7358
composites. -10
0 20 40 60 80 100 120
among the VCL composites, 100-300 m, also contains B, ppm
the maximum K content (not shown). Fig. 4. B (ppm) vs. 6”B for DSDP Hole 417A, 417D and 418A
Boron abundanoes appear to be principally related depth composites, ODP Hole 735B composites, Troodos ophiolite
to the extent of alkration. Boron and K (Fig. 2) are and Oman ophiolite samples.
126 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135

Table 5 6’‘B are not correlated in the depth composites or litho-

Omen ophiolite B compositions
Sample 6”B (%o)
B (ppm)
OM 09-I 7.25 f 0.07 14.8 f 0.4
OM 09-2 8.34 f 0.08 11.6rtO.4
OM 09-3 7.31 f 0.07 16.9 f 0.4
OM 11-l 6.07 f 0.06 0.9 * 0.4
OM 12-4 7.69 f 0.08 - 1.6kO.4
OM 12-6 8.63 f 0.09 6.0 f 0.4
OM 13-1 4.95 f 0.05 8.5 f 0.4
OM 13-3 11.08rtO.11 6.8 f 0.4
Avenge: 7.67 + 0.22 8.0f 1.1
unaltered MORB ( 0.00 1, Ryan and Langmuir, 1993).
This suggests that K is more effectively scavenged from
seawater than B.
Boron correlates well with S180 in both composite
types (Fig. 3). The lowest S’*O measured in the depth
composites is 7.87%0, at the bottom of Hole 417D. The
highest S180 measured in the depth composites is
18.05%0at the top of Hole 417A. S180 varies between
7.72%0and 19.17%0in the lithological composites. The
VCL have higher iY80 values at all depth intervals than
the PLO composites. All samples have S’*O values
greater than fresh MORB, 5.8 f 0.1%0 (Taylor, 1968;
Muehlenbachs and Clayton, 1972). Data from DSDP
Hole 4 17A whole-rock samples (Donnelly et al., 1980;
Muehlenbachs, 1980) are also plotted in Fig. 3. The
composite data coincide with them and confirm the
relationship between B content and S”?Ofound in those
altered oceanic basalts.
There is no apparent relationship between 6”B and
sub-basement depth for depth composites. The S”B
values for depth composites vary between - 2.5%0and
3.8%0and show no trends with respect to depth, either
in individual holes or in aggregate. S”B is, however,
more variable at the top of the crust than below 400 m,
where the depth composites have values close to the
average. The data for the two types of lithological com-
posites, however, are isotopically distinct and vary in
opposite directions; the #‘B of volcaniclastics
increases (from 0.3 to 5.4%0) while the S”B of massive
flows decreases (from 2.0 to 0.5%0) with depth. The
reasons for this difference are not obvious, and
undoubtedly complex, but probably include variable
water-rock ratios and differences in the timing of alter-
ation, porosity and glass content between these two
varieties of crustal material. Boron concentrations and
logical composites.
The B contents of the Troodos samples vary from
2.4 ppm to 8.1 ppm. The four epidosites have the lowest
B concentrations, between 2.4 and 3.4 ppm, except for
the potassium feldspar-rich sample 76-84-5 (6- 11) ,
which has a B content of 3.11 ppm. 6”B varies between
-0.9%0 and 7.8%0 (Table 3, Fig. 4). The B content
of these samples is positively correlated with 6’*0,
though with a different slope than that of the DSDP
samples (Fig. 3). The average B content of the Troodos
samples is 4.5 ppm and the average S”B, weighted by
B concentration, is 3.7%0.
The B contents of the Oman ophiolite samples vary
between 5.0 ppm and 11.1 ppm. The S”B varies from
- 1.6 to 16.9%0(Table 5, Fig. 4). Boron and 6”B are
not correlated. The average B content is 7.7 ppm and
the average S”B, weighted for B concentration, is
8.0%0.There are no other data for these samples.
The B contents of the samples from ODP Hole 735B
vary from 1.1 ppm to 7.1 ppm while the 6”B values
range from - 4.3%0to 24.9%0(Table 4, Fig. 4). Boron
and 6”B are not correlated, and neither B nor S”B are
correlated with depth. Boron is not correlated with K
content, though a negative correlation between #‘B
and K does appear to exist. The oxygen isotopic com-
positions of these samples have not been measured.
4. Discussion
4.1. Boron distributions andjuxes
Two fundamental questions about the B geochemical
cycle are: how large is the B inventory of the oceanic
crust, and what is the flux of B from seawater into the
crust? The flux can be estimated if the B inventory of
average oceanic crust and the rate of crustal generation
are known. A complete estimate of the B inventory of
the crust must include the entire crust, including extru-
sive volcanics, dikes, gabbros and ultramafics.
Layer 2A. The depth dependence of B, a’*0 and K
apparent in the DSDP composite samples presents a
clearer picture of downhole trends than did earlier stud-
ies from DSDP Site 4 17,418 and 504 basalts (Donnelly
et al., 1980; Muehlenbachs, 1980, Barrett, 1983; Bar-
rett and Friedrichsen, 1983; Alt and Honnorez, 1984;
Alt et al., 1986). The combination of infrequent sam-
 H. J. Smith et al. / Chemical Geology 126 (1995) I I %135
pling and small-scale length compositional variability
prevented the identifcation of large-scale trends in
those investigations. The use of composite samples
overcomes this problem.
As noted previously, the B enrichment in the DSDP
composite samples d’ecreases broadly with increasing
depth, similar to 6%~ and K. This correlation between
B and S’*O (and K) confirms that B concentrations are
principally determined by the fraction of secondary
silicates formed at rel.atively low temperatures, as sug-
gested by Spivack and Edmond ( 1987). The decrease
in B content is (at least in part) a result of the higher
permeability of the top of the crust. Boron does not
decrease monotonically with depth, however (Fig. 1).
This is not surprising since the extent of alteration is
likely to be dependent on fracturing and fracturing is
highly irregular.
The average B content of the upper 550 m of altered
crust, determined by integrating the depth composites
from DSDP Holes 4 17A, 4 17D and 418A, is estimated
to be 26 ppm. In the integration, individual composites
are weighted according to their inferred abundances
(Table 1). The wet bulk density of these altered
basalts is 2.8 g/cm3, based on an average porosity of
10.1% (Shipboard Scientific Party, 1979). Assuming
a density of 2.8 g/cm3 for average altered upper crust,
the computed inventory is 0.37 mol/cm’ for a typical
500-m-thick column of altered Layer 2A. This value is
consistent with an e,arlier estimate derived indirectly,
based on the covariance of B and S’*O, by Spivack and
Edmond ( 1987). While these samples only represent
material from the upper half of the extrusives in a typ-
ical crustal section, tS’80, 6’lB and s’Sr/‘%r distribu-
tions indicate that seawater-derived components are
concentrated in this :section.
The B content of t!he volcaniclastic lithological com-
posites is systematically higher than that of the flow
composites. This is probably due to the generally
greater porosities of the VCL samples, as well as their
more highly fractured natures (resulting in greater sur-
face-volume ratios) and the reactivity of the abundant
glass they once contained.
The flux of B from seawater into the upper crust can
be estimated from the difference in the inventories of
altered and fresh crust. Assuming a present-day crustal
generation rate of 3.0 km2/y (Gaffin, 1987; S. Solo-
mon, pers. commu.n., 1994), that low-temperature
alteration of oceanic crust is complete before subduc-
127
tion (Staudigel and Hart, 1985, and Staudigel et al.,
1986, have estimated that most is complete within
approximately 15 m.y. of crustal formation) and that
B uptake is proportional to seafloor production, a global
flux of 1.0 X 10” mol/y is estimated for the upper 500
m of oceanic crust. This is equal to the flux associated
with sediment adsorption (Spivack et al., 1987). A
more complete estimate of the B flux from seawater to
the oceanic crust must also include the B flux into
oceanic crustal Layer 2B dikes and Layer 3 gabbros
and ultramafics.
Layer 2B. The B inventory of altered oceanic crustal
Layer 2B is estimated from the data on the Troodos and
Oman ophiolites. Samples of altered dikes from the
Solea Graben, Troodos ophiolite, Cyprus contain
between 2.4 and 8.1 ppm B, with an unweighted aver-
age of 4.5 ppm. The lowest B concentrations are found
in the epidosites (which are inferred to have been pro-
duced at alteration temperatures between 292°C and
325°C) while the greenschist samples (inferred to have
been produced at temperatures above approximately
230°C) and sub-greenschist samples (inferred to have
been produced at temperatures below approximately
23O“C) have higher B contents (Schiffman and Smith,
1988). Altered dikes from the Oman ophiolite have B
contents between 5.0 and 11.1 ppm with an unweighted
average of 7.7 ppm. The average B content of the Layer
2B, calculated as the average of all the Troodos and
Oman samples, is 6.1 ppm. This corresponds to an
inventory of 0.17 mol/cm2 less than half of the value
of 0.37 mol/cm2 calculated for crustal Layer 2A.
Layer 3. An important result of the analyses of ODP
735B strip composites is that these Layer 3 gabbros are
significantly enriched in B, with an average B content
of 2.7 ppm. If the degree and style of alteration seen in
the samples from ODP Hole 735B are typical for the
oceanic crust as a whole, a 4.5~km-thick section of
Layer 3 gabbros would have a B inventory of 0.34 mol/
cm2, nearly as much as Layer 2A.
The primary B content of ODP Hole 735B gabbros
is also of interest. It can be estimated by comparing the
concentrations of B and alteration-insensitiveelements,
such as Nd, Th, Y, Hf, Zr and Nb, in the core samples
and average N-MORB. The average depletion factor
(i.e. the average ratio of the concentration in 735B
compared to N-MORB) for these elements is 0.78. The
primary B content, since B is similarly incompatible,
should then by analogy be 0.39 ppm. This is similar to
128 H.J. Smith et al. /Chemical Geology I.26 (1995) 119-135
the values of 0.34 to 0.74 ppm found in MORB by
Chaussidon and Jambon (1994) and consistent with
what one might expect for Layer 3 gabbros.
The lower limit of the B content of oceanic gabbro
is probably that of unaltered gabbro, though there are
no data to confirm this. The bulk chemical composition
of isotropic oceanic gabbro is presumably the same as
MORB, so its B should be 0.5 ppm. Cumulate gabbro,
produced by crystal fractionation and settling from the
same type of parental magma, should have a lower B
content since B is an incompatible element. Ryan and
Langmuir ( 1993) have estimated that the solid/liquid
distribution coefficient for B during fractional crystal-
lization is less than 0.07. Bloomer et al. (1991) report
that ODP Hole 735B gabbros have 5-10% trapped liq-
uid. These values can be used to estimate the B con-
centration expected for typical cumulate gabbros. If it
is assumed that the B content of the parental magma is
initially 0.5 ppm, and it is also assumed that 25% of
this magma crystallizes as cumulus gabbro, then the
cumulus crystals would have between 0.04 and 0.05
ppm B, while the residual liquid would have a B con-
centration of 0.65 ppm. In this case, an assemblage of
cumulus crystals and 8% trapped residual liquid would
have a B content of 0.09 ppm. A similar value, 0.1 ppm,
was inferred by Spivack ( 1986), on the basis of a
depletion factor of 0.25 for lower crustal cumulates
with respect to N-MORB (Hart and Staudigel, 1982).
Therefore, the estimated primary B content of ODP
Hole 735B gabbros is consistent with our understand-
ing of B geochemistry and the nature of crustal Layer
3.
Combining the results for Layer 2B dikes and Layer
3 gabbros, the B inventory of this lower part of the
oceanic crust is estimated to be 0.5 1 mol/cm*. In abso-
lute terms, this is a B concentration of 3.3 ppm and a
B inventory of 1.8 X lOi mol, assuming an area of
3.6 X lo’* cm* for the oceanic crust. This implies a B
flux of 1.4 X 10” mol/y from the ocean into the lower
oceanic crust (assuming a crustal generation rate of 3.0
km*/y) ,indicating that the lower crust may be as large
a sink for B as the sediments or upper crust. Alteration
of the lower crust could, therefore, be an extremely
important process in the B cycle if temperature differ-
ences in the descending crust result in differential mobi-
lization of B from various crustal depths (e.g. Staudigel
and King, 1992). Combining all of the results for oce-
anic crustal Layers 2A, 2B and 3, the average B content
of the oceanic crust is estimated to be 5.2 ppm.
4.2. Boron isotopic composition
The equilibrium isotopic fractionation of B between
seawater and secondary minerals is thought to primarily
depend on the relative abundances of three- and four-
coordinated B in each phase. The isotopic compositions
of the aqueous three- and four-coordinated species dif-
fer due to differences in interatomic vibrational modes
and energies (Kotaka, 1973; Kakihana et al., 1977;
Spivack and Edmond, 1987; Oi et al., 1989). Different
minerals display different magnitudes of B isotopic
fractionation. Modern marine biogenic carbonates frac-
tionate B by roughly - 7%0 to - 25%0 (Vengosh et al.,
1991; Hemming and Hanson, 1992; Spivack et al.,
1993; Hemming et al., 1995) while estimates for authi-
genie marine silicates indicate B isotopic fractionations
of - 38 f 3%0 (Spivack et al., 1987). The B adsorbed
onto marine sediments is fractionated by approximately
- 25%0 (Spivack et al., 1987). These differences
between carbonates, marine sediments and altered oce-
anic crust have important implications for the B geo-
chemical cycle of the oceans.
There are two factors which could, in principle, cause
systematic variations of S”B in secondary phases:
changes in the fluid-solid isotopic fractionation factor
(due to temperature, pH or mineralogy) or variable
fluid composition (i.e. variable integrated WIT) (Spi-
vack, 1986). These will be considered below, after a
discussion of the B isotopic data.
Layer 2~. Relative to modern seawater, the B in the
DSDP composites is isotopically fractionated by
- 34.7%0 to - 42.0%0. The weighted average 6”B of
the upper crust sampled by the composites is
0.8 f 0.4%0, similar to the range of values measured in
dikes from the Troodos ophiolite and basalts from
DSDP/ODP Hole 504B (Ishikawa and Nakamura,
1992) but clearly distinguishable from the estimated
upper mantle value of - 3 f 1%0 (Spivack, 1986). This
average is also different than the S”B of four serpen-
tinites analyzed by Spivack and Edmond ( 1987),
which are between 8.3%0 and 12.6%0, and the B
adsorbed onto marine sediments, which is 14.8 + 1%0
(Spivack et al., 1987). Furthermore, the S”B of most
marine sediments has been found to lie between
- 5.6%0 and - 17.0%0 (Ishikawa and Nakamura, 1993;
 H.J. Smith et al. /Chemical Geology 126 (1995) 11W35
You, 1994; E. Zuleger, pers. commun., 1994). These
results suggest that 6;“B is clearly different in altered
oceanic crust and sediments.
Assuming that the S”B of seawater was the same at
the time of alteration as at present, the S”B observed
in the DSDP depth composites falls within a relatively
narrow range, *5%0 from a 40%0 fractionation from
seawater. The variations of #‘B observed in these sam-
ples is probably the result of (one or) two factors: the
isotopic composition of the circulating fluid may have
evolved and fractionation between the fluid and solid
may not have been constant. Crustal fluids may be
expected to evolve towards heavier S”B values as a
result of the net uptake of isotopically depleted B into
secondary phases (considered in more detail below),
while bulk solid/fluid fractionation is, in principle,
dependent on the mineralogy of secondary assem-
blages, fluid pH (pH determines the relative abun-
dances of three- and four-coordinated B) and
temperature. These complex controls may explain the
range of S”B seen in the composites.
The 611B values of the composites can be used to
constrain the pH of the circulating fluid if the experi-
mentally derived frectionations are appropriate for the
uptake of B into silicates. Utilizing pK, values for
B (OH) 3 corrected to a pressure of 300 atm and 25°C
(Millero, 1979), the pH values of altering fluids are
calculated to be lower than 7.6.
Based on the oxygen isotopic composition of calcite
veins, the alteration temperatures of these samples are
thought to have been between 14°C and 41°C
(Lawrence, 1980; Muehlenbachs, 1980). At this time,
data do not exist which would allow a quantitative
consideration of the role of temperature on B isotopic
fractionation, although the data from Troodos epidos-
ites and diabasic dike rocks, altered at 240-320°C
(Schiffman et al., 1990), show few differences in a1‘B
from the DSDP composite samples.
Layers 2B and 3. The most notable feature of the
samples from crustal Layers 2B and 3, the ODP Hole
735B gabbros and Oman ophiolite altered dikes in par-
ticular, is their large range of B isotopic compositions,
in striking contrast to the relatively limited range of
6”B in the DSDP Site4171418 composites, Part of the
reason that the DSDP 417/418 samples display a rel-
atively narrow range of 6’lB is that they are composi-
tional averages from a range of depths and lithological
compositions. This averaging should smooth out vari-
129
ation in isotopic composition. On the other hand, the
ODP Hole 735B gabbro samples are strip composites,
covering limited depth ranges, yet they display the wid-
est range of S”B of any of the sample suites. The range
of a1‘B seen in these samples can be better understood
by examining the different controls on the B isotopic
composition of secondary phases.
Though the temperature of alteration may affect the
B isotopic composition of secondary minerals, recent
work suggests that it does not have an important effect.
Two samples of sheeted dikes from DSDP Hole 504B,
altered at temperatures estimated to be between 200°C
and 4OO”C,have 6”B values of 0.2%oand 0.8%0,indis-
tinguishable from the DSDP Site 417/418 basalts
altered at lower temperatures (Ishikawa and Naka-
mura, 1992). Additionally, the B isotopic compositions
of three epidosites from the hydrothermally altered
dikes of the Troodos ophiolite have S”B values
between -0.8%0 and 7.2%0, similar in their range to
the greenschist and sub-greenschist samples from the
same area of the Troodos ophiolite.
Another possible control on the B isotopic compo-
sition of altered oceanic crust is variable pH. If only
the four-coordinated form of B is incorporated into
secondary minerals during the low-temperature alter-
ation of oceanic basalt by seawater, and the isotopic
fractionation factor is 0.960 (the value determined
from the isotopic composition of DSDP Site 417 and
418 basalts) , the pH of the altering fluid (seawater)
would have to be approximately 9 to produce a phase
with a S”B of 24.3%0,the highest measured one in our
samples. This seems unlikely, since the low-tempera-
ture submarine alteration of oceanic crust is essentially
a process in which cations and hydroxyl groups are
removed from seawater into secondary phases, thereby
making the remaining seawater less alkaline. The pH
of seawater circulating through the crust should, there-
fore, decrease, not increase, with increasing degree of
alteration. Thus, pH variation is not a likely mechanism
for producing the range of #‘B values observed in
these rocks.
The distribution of 6’lB values in these samples does
not, on the basis of the data discussed above, seem to
depend on their mineralogy. The scatter in 6’lB within
suites of similar mineralogical composition is just as
large as, or larger than, the variations seen between
suites of distinct mineralogy. Unfortunately, no data
exist on the B isotopic compositions of these individual
130 H.J. Smith et al. /Chemical
secondary phases, so the possibility of mineralogical
controls on S”B cannot be assessed quantitatively at
this time.
Using the model of Spivack and Edmond ( 1987),
variable composition of the circulating seawater could
account for the 6”B values seen in the rocks from
Layers 2B and 3. For a basalt-seawater system, as more
and more rock is altered by a given parcel of fluid, its
611B will increase and it will produce secondary min-
erals with ever heavier 6”B. Reasonable estimates of
CYand D give 6”BR of 25%0 when wlr equals 43.
Additionally, an evolved fluid with a S”B greater than
39.5%0would allow an even lower water-rock ratio to
produce a secondary phase with a S”B of 25%0. The
high 6”B values in the ODP 735B strip composites are
likely the result of this kind of fluid evolution. A simple
explanation for the wide range of 6”B observed in ODP
Hole 735B is that different parts of the crust experience
different water-rock ratios. Since the flux of seawater
through the crust is not likely to be uniform, either
laterally or with depth, the crust should experience dif-
ferent water-rock ratios in different locations, leading
to isotopically diverse B in the low-temperature sec-
ondary phases.
High 6”B values could, theoretically, be produced
by the complete removal of B from seawater trapped
in pore spaces in oceanic crust. This possibility can be
investigated best by examining the strip composites of
ODP Hole 735B, which display the widest range of B
isotopic composition. Locally closed-system reactions
would result in S”B values between 20%0and 36%0for
rocks with B concentrations between 1 ppm and 7 ppm,
respectively, the range of B contents in these samples.
The 6i’B values of most of the strip composites vary
from approximately -4%0 to 8%0, however, not like
what would be expected if this type of closed-system
process were responsible for the alteration they exhibit.
Even the composite with the highest S”B (sample P,
with a B concentration of 1.1 ppm and a S”B of
24.9%0) cannot have acquired its B isotopic composi-
tion in a closed system. If it had, 64% of its B (i.e. 0.68
pg B per gram of gabbro) would have had to come
from seawater (with a S”B of 39.5%0), requiring 0.15
g of seawater to be present for each gram of gabbro and
the B to be completely removed from the seawater into
secondary phases. Since this would necessitate poros-
ities much higher than any observed, closed-system
alteration is not considered a realistic mechanism for
Geology I26 (I995) 119-135
alteration in ODP Hole 735B. Furthermore, the corre-
lation between B and S’‘B which closed-system alter-
ation would produce is absent. It is not likely, then, that
the B in these samples was acquired during closed-
system alteration. The unlikelihood of this type of alter-
ation can also be seen by calculating the amount of
seawater needed to produce the B compositions deter-
mined in these samples. Consider, for example, sample
C, the most enriched of the ODP suite with 7.1 ppm B.
For this B enrichment to have occurred by closed-sys-
tem alteration, a minimum of 1.4 g of seawater would
have had to be present for every gram of rock, an
unrealistically large proportion of water for this crust,
so closed-system alteration is clearly impossible for this
sample on the basis of B content alone. It is possible,
though, that minor amounts of locally closed-system
alteration could have occurred in these rocks. Similar
arguments can be made for the samples from Troodos
and Oman.
A puzzling observation from the ODP Hole 735B
results is that the samples which are the least evolved
magmatically have the highest 6i1B values. In other
words, it seems that the B isotopic composition,
acquired during secondary processes, is somehow
related to the major element chemistry of the samples,
presumably a primary characteristic. Specifically, high
S”B correlates with high CaO and low Na,O, Fe,O3,
Nd, Y and Eu. Two possible reasons for these relation-
ships are either that the least evolved gabbros formed
secondary phases which did not isotopically fractionate
B as much as the more evolved gabbros or that less-
evolved gabbros (the samples with the highest S”B
values, from the bottom of the core) were altered by
more-evolved circulating fluids than the more-evolved
gabbros. Beyond noting that these relationships occur,
we have not yet developed satisfactory explanations of
their causes.
The weighted averages of the Oman and Troodos
samples are 8.0%~and 3.7%0,respectively. If the com-
bined Troodos and Oman ophiolite samples are consid-
ered as representative of crustal Layer 2B, an average
S”B of 5.%0 (weighted by B content) is computed.
The corresponding value for crustal Layer 3, repre-
sented by the ODP Hole 735B samples, is 5.7%0.These
values are different from that of altered oceanic crustal
Layer 2A, 0.8%0. Combining all the data for DSDP
Holes 417A, 417D, 418A, ODP Hole 735B, Troodos
ophiolite and the Oman ophiolite, the oceanic crust is
 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135 131

Table 6 by a model consistent with the data from Layer 2A.

Boron inventories and fluxes
The monotonic increase of S”B seen in the most highly
Reservoir B [B] S’iB Source altered DSDP 417/418 lithological composites may be
(10” mol) (ppm) (%o) evidence of the evolution of the B isotopic composition
of the circulating fluid. These composites, taken over
Mantle 0.45-0.8 0.25 -3fl 2 intervals between 100 m and 250 m in length, represent
Continental crust 0.19 10 -7 1.5
Sediments: 0.21 -1.3*4
the integrated compositions of altered crust at those
non-desorbable 0.2 -1.8*4 2 depth ranges and should reflect any large-scale
desorbable 0.00157 15fl 2 changes. The 6t’B of the volcaniclastics increases by
Altered oceanic 5.1%0between the top depth interval and the bottom, a
crust: 0.02’6 5.2 3.7
trend expected for alteration by a fluid which becomes
Layer 2A 0.013 26 0.8 7
Layer 2B 0.006 6.1 5.9 7 more evolved (heavier) at greater depths.
Layer 3 0.007 2.7 5.7 7 A summary of the variations observed in B content
Seawater 0.0057 4.5 39.52 f 0.36 2 and S”B, with respect to relative stratigraphic position
Atmosphere 2.5 >( lo-” 2 in the oceanic crust, is shown in Fig. 5. The B content
C of non-mantle 0.664 -2.2-14
of altered oceanic crust is highest, and most variable,
Flux B 6”B Source in the shallowest crust, decreasing in both abundance
(IO” mol/y) (%o) and range with increasing depth. The 6”B, on the other
hand, is most variable at the bottom of the crust, in
OCEAN INPUTS 0.74 f 0.1 12*10
Layer 3, where the B content is lowest. The average
Hydrothermal: 0.4 It 0. I 6.5*8
Non-sedimented S”B of the oceanic crust increases slightly with
ridges 0.09 *0.1 -4 3 increasing depth. The S”B values of these samples
Sedimented 0.3 10*10 6
ridges
Weathering: 0.34. 18f5
Igneous 0.04 1.7*6 2
Carbonate 0.3 20f5 6
OCEAN OUTPUTS 4.0 8.9
Oceanic crust
alteration: 2.5 3.7 2
Layer 2A 1.0 to.2 0.8 E
Y
Layer 2B 0.5 5.9 _
Layer 3 1.0 5.7 F3
Sediment: 1.5 17f4
Adsorption 0.9 15*1 % 1.1-7.1
(2.7)
4
Carbonates 0.6 20f5
MORB production 0.086 -3*1

Sources are: I, Harder ( 1978); 2, Spivack ( 1986); 3, Spivack et al.

(1987); 4, Vengosh et al. (1991); 5, Chaussidon and Albar&le
(1992); 6, You et al. (1993); 7, this work.
estimated to have a S”B of 3.7%~~. This value is listed
in Table 6, along with a summary of the estimated
major B inventories and fluxes.
The average S”B is higher in Layers 2B and 3 than
in Layer 2A, an observation consistent with a system
in which the fluid circulating through the crust evolves
toward higher 6”B with increasing depth. The rela-
tively high average tS”B values measured in these sam-
ples of crustal Layers 2B and 3 can thus be explained
0 20 40 60 80 100 0 10 20 30
B, ppm 6”B
Fig. 5. B ( ppm) and 6’iB vs. relative stratigraphic position for DSDP
Hole 417A, 417D and 418A depth composites, ODP Hole 735B
composites, Troodos ophiolite and Oman ophiolite samples. Depth
refers to au idealized depth within a section of average oceanic crust,
not the actual depth of the samples. The ranges observed are shown
by shaded boxes, and the average values by white lines, for each
crustal Layer. The numerical values are also shown (averages in
parentheses). The two averages shown for Layer 2B are for the
Troodos ophiolite and Oman ophiolite samples.
 132 H.J. Smith et al. /Chemical Geology I26 (1995) 119-13.5
converge, as the B content increases, to a value of
approximately 2%0, close to that of average oceanic
crust. This is interpreted as being due to the fact that
the DSDP composites are essentially averages of large
numbers of potentially dissimilar subdomains altered
at different integrated water-rock ratios by relatively
unevolved fluids. In other words, as deeper sections of
the crust are sampled, low-temperature alteration
becomes less pervasive and more irregular, and alter-
ation proceeds with smaller fluxes of more-evolved
fluids, on average. Therefore, the ODP Hole 735B sam-
ples from crustal Layer 3 have lower B contents but
higher and more variable S”B than the DSDP samples
from Layer 2A. Fig. 5 also illustrates the independence
of B content and 6”B, an important result of this work.
These two quantities are decoupled because the B con-
centration of a parcel of rock which has been altered
by seawater at low-temperature depends primarily on
the fraction of the parcel which is altered, while the
6l’B depends primarily on the B isotopic fractionation
factor of the secondary phase and the B isotopic com-
position of the fluid by which it is altered. The absence
of a clear systematic relationship between B and S”B
is thus expected for oceanic crust altered at low tem-
perature.
4.3. Water-rock ratios
The water-rock ratio, the extent of B removal from
circulating fluids and the low-temperature flux of sea-
water through the upper oceanic crust can be estimated
from the data on the DSDP Site 417/418 composites.
The water-rock ratio, w/r, can be calculated from the
6”B of the secondary phase by:
W D(l-a)
-=
1
r ln ( l/a)(6i1B,+ 1000)
6”B,, + 1000
as shown by Spivack and Edmond ( 1987). The distri-
bution coefficient (for B, between mineral and water),
D, depends on the temperature at which alteration
occurs and can be calculated from the relationship
between 6”O and B. At 1°C the S”O of smectite in
equilibrium with seawater is 30%~ (Muehlenbachs and
Clayton, 1972; Yeh and Savin, 1976) and the distri-
bution coefficient is 33 (Spivack and Edmond, 1987).
If the temperature of alteration is 4O”C, the S”O of
smectite is 20%0 and the distribution coefficient is cal-
culated to be 21. A reasonable estimate of the average
alteration temperature, 2S’C, gives a D of 29. A frac-
tionation factor of 0.960 (the value that produces a
solid phase 6iiB of -2.5%~ the lowest observed in
any of the shallowest DSDP Site 417/418 composites,
i.e. those presumably altered by unevolved seawater
with a 6i’B of 39.5%0) is assumed. These values can
be used to calculate a minimum water-rock ratio for
the integrated average depth composite ( S’iB of 0.8%0)
of between 300 (for D=21) and 460 (for D=33),
with a best estimate of 400 (for D = 29).
The extent of B removal from fluids circulating
through oceanic crust has been previously addressed
by Seyfried et al. ( 1984), who suggested that B is
quantitatively removed from seawater during low-tem-
perature alteration of oceanic basalt. If this were true,
the S”B of altered oceanic crust would approach that
of seawater, 39.5%0, as the crust is more completely
altered. This is not observed. The composite data con-
strain estimates of B removal into the upper crust to
less than approximately 7%, as given by the equation:
[Blcez [ 1
[Blw
where [B] is the B concentration of the altered fluid
and [B] sw is the B concentration of seawater.
The average B content of the depth composites can
be used in conjunction with the degree of B removal
from the circulating fluid to estimate a minimum low-
temperature flux of seawater through the upper crust.
This flux is given by:
M
F=-
where F is the volume flux, A4 is molar B inventory of
the altered oceanic crust and f is the fraction of B
removed from seawater during alteration. Using an
average B inventory of 0.34 mol/cm2 and an f of 0.07,
a minimum low-temperature seawater flux of 3.4 X 1014
l/y is calculated for a present-day crustal production
rate of 3.0 km2/y.
5. Conclusions
Samples from DSDP Holes 4 17A, 417D and 4 18A,
ODP Hole 735B, the Troodos ophiolite and the Oman
 H.J. Smith et al. /Chemical Geology 126 (1995) 119435
ophiolite have been analyzed to estimate the B content
and 6”B of average altered oceanic crust. The crust has
the greatest B enrichments in the upper 500 m of crustal
Layer 2A, where the a.verage is 26 ppm. The B content
of Layers 2B and 3 are 6.1 ppm and 2.7 ppm, respec-
tively. The B content of an idealized section of oceanic
crust, composed of 0.5 km of Layer 2A extrusive
basalts, 1 km of Layer 2B dikes and 4.5 km of Layer 3
gabbros is 5.2 ppm. 6”B is not correlated with B con-
tent. The #‘B of Layer 2A varies between - 2.5%0and
5.4%0, with an average of 0.8%0, while the S”B of
Layers 2B and 3 varies over a much wider range, from
- 0.9%0 to 24.9%0, with averages of 5.9%0 and 5.7%0,
respectively. The combined average S”B of these lay-
ers is 3.4%0.This average isotopic composition is mea-
surably different from the apparent average of oceanic
sediments, supporting the idea that 6’‘B could be useful
for identifying the source(s) of B in island arcs.
The data from the DSDP Site 4 17/418 composites
indicate that the 6”B of fluid circulating in the upper
crust changes only slightly during alteration, increasing
by an average of 5.1%0with an accompanying decrease
in B concentration of 7%. Low-temperature alteration
appears to be a water-dominated process resulting in
minor modification of circulating seawater. A mini-
mum water-rock ratio of 400 is calculated for these
samples, implying a minimum low-temperature sea-
water flux through the upper oceanic crust of 3.4 X 1014
l/y.
The results from this study suggest that 6”B might
be useful for studying subduction-related magmage-
nesis and recycling. The average 6”B calculated for
altered oceanic crust., 3.4%0, is considerably different
than the range observed in sediments, from - 17%0to
4.5%0 (for all but carbonates, which typically have
values between 10% and 32%0). If subducted sedi-
ments are more like the older sediments analyzed by
Ishikawa and Nakamura ( 1993)) which had 6”B val-
ues between - 5.6%~~and - 17%0,then the difference
between subducted altered basalt and sediment is even
more pronounced. Based on these measurements, and
on measurements of sediments from the Japan Sea
(Zuleger et al., 1994)1,the average 6’‘B of sedimentary
B appears to be approximately - 10%0.It may be pos-
sible, therefore, to identify and quantify the sources of
B in island-arc volcanic rocks on the basis of their B
isotopic compositions. Measurements of 6”B in
MORB could also help place limits on the amount of
133
B recycled back into the mantle by subduction, if the
average S”B values of the subducted oceanic litho-
sphere and the mantle are different. This information
is essential in the construction of an accurate B mass
balance for the whole earth.
Acknowledgements
We would like to thank Professor Peter Schiffman
for supplying the Troodos ophiolite samples, as well as
for the helpful discussions about them. We would also
like to thank Joris Gieskes, Chen-Feng You and Evelyn
Zuleger for their comments and suggestions. This is
NSG contribution number 950504. We thank Dr. B.
Smith for the oxygen-isotope data on the DSDP sam-
ples. We also thank M. Chaussidon and an anonymous
reviewer for their helpful comments.
References
Agyei, E.K., 1968. Isotopic and elemental composition of boron in
meteorites, tektites and terrestrial materials. Ph.D. Thesis,
McMaster University, Hamilton, Ont.
Alabaster, T., Pearce, J.A. and Malpas, J., 1982. The volcanic stra-
tigraphy and petrogenesis of the Oman ophiolite complex. Con-
trib. Mineral. Petrol., 81: 168-183.
Ah, J. and Honnorez, J., 1984. Alteration of the upper oceanic crust,
DSDP site 417: mineralogy and chemistry. Contrih. Mineral.
Petrol., 87: 149-169.
Alt, J.C., Muehlenbachs, K. and Honnorez, J., 1986. An oxygen
isotopic profile through the upper kilometer of the oceanic crust,
DSDP Hole 504B. Earth Planet. Sci. Lett., 80: 217-229.
Barrett, T.J., 1983. Strontium- and lead-isotope composition of some
basalts from Deep Sea Drilling Project Hole 504B, Costa Rica
Rift, Legs 69 and 70. Init. Rep. Deep-Sea Drill. Proj., 69: 643-
650.
Barrett, T.J. and Friedrichsen, H., 1983. Oxygen- and hydrogen-
isotope composition of some basalts from Deep Sea Drilling
Project Hole 504B, Costa Rica Rift, Legs 69 and 70. Init. Rep.
Deep-Sea Drill. Proj., 69: 637-642.
Bemdt, M.E. and Seyfried, W.E., 1990. Boron, bromine and other
trace elements as clues to the fate of chlorine in mid-ocean ridge
vent fluids. Geochim. Cosmochim. Acta, 54: 2235-2245.
Bemdt, M.E., Seyfried, W.E. and Beck, W.J., 1988. Hydrothermal
alteration processes at midocean ridges: experimental and theo-
retical constraints from Ca and Sr exchange reactions and Sr
isotopic ratios. J. Geophys. Res., 93: 4573-4583.
Bloomer, S.H., Meyer, P.S., Dick, H.J.B., Ozawa, K. and Natland,
J.H., 1991. Textural and mineralogical variations in gabbroic
rocks from Hole 735B. Proc. ODP (Ocean Drill. Prog.), Sci.
Results, 118: 21-36.
 134 H.J. Smith et al. /Chemical Geology I26 (1995) 119-135
Broecker, W.S. and Peng, T.-H., 1982. Tracers in the Sea. Eldigio
Press, Palisades, N.Y., 690 pp.
Cameron, W.E., Nisbet, V.G. and Dietrich, V.J., 1980. Boninites,
komatiites and ophiolites. Nature (London), 280: 550-553.
Chaussidon, M. and Albarede, F., 1992. Secular boron isotope var-
iations in the continental crust: an ion microprobe study. Earth
Planet. Sci. Len., 108: 229-241.
Chaussidon, M. and Jambon, A., 1994. Boron content and isotopic
composition of oceanic basalts: geochemical and cosmochemical
implications. Earth Planet. Sci. Lett., 121: 277-291.
Coleman, R.G., 1977. Ophiolites. Springer, Berlin, 229 pp.
Coleman, R.G., 1981. Tectonic setting for ophiolite abduction in
Oman. J. Geophys. Res., 86: 2497-2508.
Dick, H.J.B., Meyer, P.S., Bloomer, S., Kirby, S., Stakes, D. and
Mawer, C., 1991. Lithostratigraphic evolution of an in-situ sec-
tion of oceanic layer 3. Proc. ODP (Ocean Drill. Prop.), Sci.
Results, 118: 439-538.
Donnelly, T.W., Prichard, R.A. and Emmorman, R., 1980.The aging
of the oceanic crust: synthesis of the mineralogic and chemical
results of Deep Sea Drilling Project, Legs 51 through 53. Init.
Rep. Deep-Sea Drill. Proj., Vols. 51, 52, 53, Part 2, pp. 1563-
1577.
Edwards, C.M.H.,Morris, J.D. and ThirlwaR,M.F., 1993. Separating
mantle from slab signatures in arc lavas using B/Be and radio-
genie isotope systematics. Nature (London), 362: 530-533.
Friedrichsen, H. and Hoemes, S., 1980. Oxygen and hydrogen iso-
tope exchange reactions between sea water and oceanic basalts
from Legs 5 1 through 53. Init. Rep. Deep-Sea Drill. Proj., Vols.
51,52,53, pp. 1177-1181.
Gaffin, S., 1987. Ridge volume dependence on seafloor generation
rate and inversion using long term sealevel change. Am. J. Sci.,
287: 596-611.
Gillis, K.M. and Robinson, P.T., 1988. Distribution of alteration
zones in the upper oceanic crust. J. Geol., 16: 262-266.
Goldschmidt, V.M. and Peters, C., 1932. Zur Geochimie des Bars:
I, II. Nachr. Ges. Wiss. Glittingen, Math. Phys. Kl., pp. 402-407,
528-545.
Harder, H., 1978. Boron: granitic rocks and related effusives. In:
Handbook of Geochemistry, II-l. Springer, Berlin, pp. 5E5-
5ElO.
Hart, S.R. and Staudigel, H., 1982. The control ofalkalis and uranium
in sea water by ocean crust alteration. Earth Planet. Sci. Lett.,
58: 202-212.
Hart,S.R., Blusztajn,J., Dick,H.J.B. andLawrence,J.R., 1994.Fluid
circulation in the oceanic crust: contrast between volcanic and
plutonic regimes. I. Geophys. Res., 99(B2): 3163-3173.
Hawkins, J.W. and Evans, C.A., 1983. Geology of the Zambales
range, Luzon, Phillipine Islands: ophiolite derived from an island
arc-back arc basin pair. Am. Geophys. Union, Geophys. Mon-
ogr., 27: 95-123.
Hemming, N.G. and Hanson, G.N., 1992. Boron isotopic composi-
tion and concentration in marine carbonates. Geochim. Cosmo-
chim. Acta, 56: 537-543.
Hemming, N.G., Reeder, R.J. and Hanson, G.N., 1995. Mineral-
fluid partitioning and isotopic fractionation of boron in synthetic
calcium. Geochim. Cosmochim. Acta, 59: 371-379.
Humphris, S.E. and Marriner, G.F., 1980. The mineralogy and geo-
chemistry of basalt weathering, Holes 417A and 418A. Init. Rep.
Deep-SeaDrill. Proj., Vols. 51,52,53, pp. 1201-1213.
Ishikawa, T. and Nakamura, E., 1992. Boron isotope geochemistry
of the oceanic crust from DSDP/ODP Hole 504B. Geochim.
Cosmochim. Acta, 56: 1633-1639.
Ishikawa, T. and Nakamura, E., 1993. Boron isotope systematics of
marine sediments. Earth Planet. Sci. Lett., 117: 567-580.
Ishikawa, T. and Nakamura, E., 1994. Origin of the slab component
in arc lavas from across-arc variation of B and Pb isotopes. Nature
(London), 370: 205-208.
Kakihana, H., Kotaka, M., Shohei, S., Nomura, M. and Okamoto,
M., 1977. Fundamental studies on the ion-exchange separation
of boron isotopes. Bull. Chem. Sot. Jpn., 50: 158-163.
Kempton, P.D., Hawkesw0tth.C.J. andFowler,M., 1991. Geochem-
istry and isotopic composition of gabbros from layer 3 of the
Indian Ocean crust, Leg 118, Hole 735B. Proc. ODP (Ocean
Drill. Prog.), Sci. Results, 118: 127-143.
Kotaka, M., 1973.Chromatographic separationofboron and nitrogen
isotopes using pure water as an eluent. Ph.D. Dissertation, Tokyo
Institute of Technology, Tokyo.
Lawrence, J.R., 1980. Temperatures of formation of calcite veins in
the basahs fromDeep SeaDrillingProjectHoles417Aand417D.
Init. Rep. Deep-Sea Drill. Proj., Vols. 51.52.53, pp. 1183-l 184.
Lippard, S.J., Shelton, A.W. and Gass, I.G., 1986. The Ophiolite of
Northern Oman. Blackwell, Oxford.
Millero, F.J., 1979. The thermodynamics of the carbonate system in
seawater. Geochim. Cosmochim. Acta, 43: 1651-1661.
Miyashiro, A., 1973. The Troodos Complex was probably formed in
an island arc. Earth Planet. Sci. I_&., 19: 218-224.
Miyashiro, A., 1975. Origin of Troodos and other ophiolites: a reply
to Hynes. Earth Planet. Sci. IA%.,25: 217-222.
Moores, E.M., Robinson, P.T., Malpas, J. and Xenophontos, C.,
1984. Model for the origin of the Troodos massif, Cyprus and
other mideast ophiolites. Geology, 12: 500- 503.
Morris, J.D., Leeman, W.P. and Tera, F., 1990. The subducted com-
ponent in island arc lavas: constraints from Be isotopes and B-
Be systematics. Nature (London), 344: 3 l-36.
Muehlenbachs, K., 1980. The alteration and aging of the basaltic
layer of the sea floor: oxygen isotope evidence from DSDPI
IPOD Legs 51.52, and 53. Init. Rep. Deep-Sea Drill. Proj., Vols.
51,52,53, pp. 1159-l 166.
Muehlenbachs, K. and Clayton, R.N., 1972. Oxygen isotope studies
of fresh and weathered submarine basalts. Can. J. Earth Sci., 9:
172-184.
Nakamura, E., Ishikawa, T., Birck, J.L. and All&e, C.J.. 1992.
Precise boron isotopic analysis of natural rock samples using a
boron-mannitol complex. Chem. Geol. (Isot. Geosci. Sect.), 94:
193-204.
Oi, T., Nomura, M., Musashi, M., Ossaka, T., Okamoto, M. and
Kakihana, H., 1989. Boron isotopic compositions of some boron
minerals. Geochim. Cosmochim. Acta, 53: 3189-3195.
Palmer, M.R., 1991. Boron-isotope systematics of Hahnahera arc
(Indonesia) lavas: evidence for involvement of the subducted
slab. Geology, 19: 215-217.
Pearce, J.A., Alabaster, T., Shelton, A.W. and Searle, M.P., 1981.
The Oman ophiolite as a Cretaceous arc-basin complex: evi-
 H.J. Smith et al. /Chemical Geology 126 (1995) 119-135
dence and implications. Philos. Trans. R. Sot. London, Ser. B,
300: 299-3 17.
Pearce, J.A., Lippard, S.J. and Roberts, S., 1984. Characteristics and
tectonic significance of supra-subduction zone ophiolites. In:
Marginal Basin Geology. Geol. Sot., London, pp. 77-94.
Reinhardt, B.M., 1969. On the genesis and emplacement of ophiolites
in the Oman Mountains geosyncline. Schweiz. Mineral. Petrogr.
Mitt., 49: l-30.
Robinson, P.T., Dick, H.J.B. and Von Herzen, R.P., 1991. Meta-
morphism and alteration in oceanic layer 3: Hole 735B. Proc.
ODP (Ocean Drill. Prog.), Sci. Results, 118: 541-552.
Rubey, W.W., 1951. Geologic history of seawater: an attempt to state
the problem. Geol. Sot. Am., Bull., 62: 111l-l 148.
Ryan, J.G. and Langmuir, C.H., 1993. The systematics of boron
abundances in young volcanic rocks. Geochim. Cosmochim.
Acta, 57: 1489-1498.
Schiffman, P. and Smith, B.M., 1988. Petrology and oxygen isotope
geochemistry of a fossil seawater hydrothermal system within
the Solea Graben, northern Troodos Ophiolite, Cyprus. J. Geo-
phys. Res., 93(B5): 4.612-4624.
Schiffman, P., Smith, B.M., Varga, R.J. and Moores, E.M., 1987.
Geometry, conditions and timing of off-axis hydrothermal meta-
morphism and ore-delposition in the Solea graben. Nature (Lon-
don), 325: 423-425.
Schiffman, P., Benison, L.A. and Smith, B.M., 1990. Mineralogy
and geochemistry of epidosites from the Solea graben, Troodos
ophiolite, Cyprus. In: Ophiolites: Oceanic Crustal Analogues.
Geol. Surv. Dep., Min. Agric. Nat. Resour., Nicosia, pp. 673-
683.
Seyfried, W.E., Janecky, D.R. and Mottl, M.J., 1984. Alteration of
the oceanic crust: implications for geochemical cycles of lithium
and boron. Geochim. Cosmochim. Acta, 48: 557-569.
Shipboard Scientific Party, 1979. Proc. DSDP Init. Rep., Vols. 5 1,
52,53, Part 1. U.S. Gov. Print. Off., Washington, D.C.
Shipboard Scientific Party, 1989. Proc. ODP (Ocean Drill. Prop.),
College Station, Texas, Init. Rep., 118.
Smith, H.J., 1994. The E: composition of oceanic crust. Ph.D. Dis-
sertation, Scripps Institution of Oceanography, La Jolla, Calif.
Spivack, A.J., 1986. Boron isotope geochemistry. Ph.D. Dissertation,
Massachusetts Institute of Technology, Woods Hole Oceano-
graphic Institution, Woods Hole, Mass.
Spivack, A.J. and Edmond, J.M., 1987. Boron isotopic exchange
between seawater and the oceanic crust. Geochim. Cosmochim.
Acta, 51: 1033-1043.
Spivack, A.J., Palmer, M.R. and Edmond, J.M., 1987. The sedimen-
tary cycle of the boron isotopes. Geochim. Cosmochim. Acta,
51: 1939-1949.
Spivack, A.J., You, C.F. and Smith, H.J., 1993. Foraminiferal boron
isotope ratios as a proxy for surface ocean pH over the past 21
Myr. Nature (London), 363: 149-151.
Stakes, D., Mevel, C., Cannat, M. and Chaput, T., 1991. Metamor-
phic stratigraphy of Hole 735B. Proc. ODP (Ocean Drill. Prog.),
Sci. Results, 118: 15.3-180.
135
Staudigel, H. and Hart, S.R., 1985. Dating of ocean crust alteration:
strontium isotope ratios from 504B carbonates and a re-interpre-
tation of SrisotopedatafromDSDPSites 105,332,417 and 418.
Init. Rep. Deep-Sea Drill. Proj., 83: 297-303.
Staudigel, H. and King, S.D., 1992. Ultrafast subduction: the key to
slab recycling efficiency and mantle differentiation?Earth Planet.
Sci. Lett., 109: 517-530.
Staudigel, H., Muehlenbachs, K., Richardson, S.H. and Hart, S.R.,
1981. Agents of low temperature ocean crust alteration. Contrib.
Mineral, Petrol., 77: 150-157.
Staudigel, H., Gillis, K. and Duncan, R., 1986. K/Ar and Rb/Sr ages
of celadonites from the Troodos ophiolite, Cyprus. Geology, 14:
72-75.
Staudigel, H., Davies, G.R., Hart, S.R., Marchant, K.M. and Smith,
B.M., 1995. Large scale isotopic Sr, Nd and 0 isotopic anatomy
of altered oceanic crust: DSDP/ODP Sites 417/418. Earth
Planet. Sci. Lett., 130: 169-185.
Tatsumi, Y., 1989. Migration of fluid phases and genesis of basalt
magmas in subduction zones. J. Geophys. Res., 94(B4): 4697-
4707.
Tatsumi, Y., Hamilton, D.L. and Nesbit, R.W., 1986. Chemical char-
acteristics of fluid phase released from a subducted lithosphere
and origin of arc magmas: evidence from high-pressure experi-
ments and natural rocks. J. Volcano!. Geothenn. Res., 29: 293-
309.
Taylor, H.P., 1968. The oxygen isotope geochemistry of igneous
rocks. Contrib. Mineral. Petrol., 19: l-71.
Thompson, G. and Melson, W.G., 1970. Boron contents of serpen-
tinites and metabasalts in the oceanic crust: implications for the
boron cycle in the oceans. Earth Planet. Sci. Lett., 8: 61-65.
Vanko, D.A. and Stakes, D.S., 1991. Fluids in oceanic layer 3: evi-
dence from veined rocks, Hole 735B. Proc. ODP (Ocean Drill.
Prog.), Sci. Results, 118: 181-218.
Vengosh, A., Kolodny, Y., Starinsky, A., Chivas, A. and McCuIloch,
M., 1991. Coprecipitation and isotopic fractionation of boron in
modem biogenic carbonate. Geochim. Cosmochim. Acta, 55:
2901-2910.
Wilson, R.A.M., 1959, The geology of the Xeros-Troodos area,
Cyprus. Cyprus Geol. Surv. Mem., 1,135 pp.
Yeh, H.W. and Savin, S.M., 1976. The extent of oxygen isotope
exchange between clay minerals and seawater. Geochim. Cos-
mochim. Acta, 40: 743-748.
You, C.F., 1994. Lithium, beryllium and boron isotope geochemis-
try: Implications for fluid processes in convergent margins. Ph.D.
Dissertation, Scripps Institution of Oceanography, University of
California, San Diego, Calif.
You, C.F., Spivack, A.J., Smith, H.J. andGieskes, J.M., 1993. Mobi-
lization of boron in convergent margins: implications for the
boron geochemical cycle. J. Geol., 21: 207- 210.
2uleger.E.. You, C.F.,Gieskes, J.M. and Spivack, A.J., 1994. Boron
and boron isotopes in sediments and pore waters from ODP Leg
127 (Japan Sea). Eos (Trans. Am. Geophys. Union), 75(44):
740 (abstract)