I '31?! ‘ ' 1

‘ “V.w———v l

 

 

‘ ‘U" ..
1], ‘ 7‘9 0:

   
     

  

LIBRARY

M irrhigan State
University

This is to certify that the
thesis entitled

238U/230Th Isotope Systematics of
Rhyolites from Long Valley, California

presented by

James Baranowski

has been accepted towards fulfillment
of the requirements for

Masters degree in Geology

4M

Major professor y I

 

 

 

Date February 24, 1978

 

0-7639

230

238U/ Th ISOTOPE SYSTEMATICS OF

RHYOLITES FROM LONG VALLEY,

CALIFORNIA

BY

James Baranowski

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Geology

1977

ABSTRACT

 

238U/ 230Tb ISOTOPE SYSTEMATICS OF
RHYOLITES FROM LONG VALLEY,
CALIFORNIA
I“
By A
James Baranowski
‘L

Dates obtained by 2380 - 23oTh systematics agree

with K-Ar dates for two post collapse intercaldera
rhyolites from Long Valley, CA. Analytical precision
was not as great as that obtained by K-Ar methods, but
the possibility of systematic error present with the use

230T 232 238U/

of K-Ar methods was reduced. [ h/ Th], [
Th] values for the glass phase fell above the isochron,

232
possibly due to uranium loss during eruption. The use of
hornblende-glass pairs to indicate eruption ages
(Allegre, 1968) was not suitable for the rhyolites

234U/ 2380 fractionation was

investigated. Possible
observed. This fractionation could only occur before
emplacement, as no evidence for post-depositional uranium
migration was observed.

(230Th/ 232T

indicating that Th/U fractionation is not taking place at

h)o values were similar, at .94,

 

James Baranowski

a substantial rate among the rhyolites in the Long

Valley magma chamber.

 

 

To Hilary

ii

 

 

 

ACKNOWLEDGMENTS

I thank my advisor, Dr. Russel H. Harmon, and
Dr. Tom Vogel and Dr. John Wilband for their assiStance
during the course of this study. I would also like to
express my thanks to Dr. Rodd May and Dr. Brent
Dalrymple of the U.S.G.S. at Menlo Park, California, for
freely sharing geologic and geochronologic information
on Long Valley with me. Major element chemical analysis
was performed by Dr. Robert McNutt, McMaster University.
Uranium and Thorium analysis was performed by Dr. H. T.
Millard of the U.S.G.S. at Denver, Colorado.

I am indebted to Rich Lively for being my
surrogate advisor and I thank Sue Leo for her support
and friendship.

My family, especially my wife Cheryl, deserve
special thanks for the sacrifices they have made for me
and the help they have been to me over the course of

this study.

iii

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . .
LIST OF FIGURES . . . . . . .
Chapter

I. INTRODUCTION . . . . . .

II. GEOCHEMISTRY OF URANIUM AND THORIUM

Chemical Behavior of U and Th .
Isotopic Behavior of U and Th .

U/Th Disequilibrium Dating Methods

III. ISOTOPIC ANALYSIS OF URANIUM AND THORIUM

Sample Collection and Preparation.

Chemical Extraction of Uranium and Thorium

Separation of Radionuclides .

Separation of Major Elements .

Contamination . . . . . .
Reagents . . . . . . .
Glassware . . . .

Analytical Procedure, Silicic Samples

Alpha Counting Procedure . .
Preparation and Use of Spike .
U-concentration Measurement .
Accuracy of Method . . . .

IV. ISOTOPE RATIOS AND ABUNDANCE DATA FOR

URANIUM AND THORIUM . . . .
V. DISCUSSION AND CONCLUSIONS . .
Discussion of Sample Results .

Comparison with Previous Results
Suggested Improvements . . .

“FERENCES O O O O O O O O 0

iv

Page
vi

vii

meh 0" l-‘

22

27
29
31
31
32
32
45
51

58

63
78
78

83
83

86

Page

APPENDICES . . . . . . . . . . . . . . 94
I. GEOLOGY OF LONG VALLEY, CALIFORNIA . . . . 95
Sample Collection . . . . . . 106
Chemical and Petrographic Description . . . 106

II. CALCULATION OF ANALYTICAL ERROR . . . . . 112
Sample Calculation . . . . . . . . . 117

 

Table

IV

LIST OF TABLES

Activity ratios, abundance, and sample
yields obtained by a-spectrometry . . .

228Th/ 232Th activity ratios for unspiked
samples . . . . . . . . . . .

Internal isochron regression data for
samples 76A003 and 76A004 . . . . .

Uranium and Thorium abundance figures .
Major element abundance figures . . .

Sample location and description . . .

vi

Page

66

67

68
69
103

107

Figure

1.

12.

13.

14.

15.
16.

LIST OF FIGURES

[230Th/232Th], [238U/232Th] isochron . .
238U, 232Th, and 235U decay series . .
Graph of [(230Th/232TH)01, [ext] . . .

Flowchart--mineral separation procedure .
Flowchart—-chemical extraction procedure
Flowchart--a1ternate chemical procedure .
Uranium a-spectrum . . . . . . . .
Thorium a-spectrum . . . . . . . .
2320_238

Th spike d-spectrum . . . . .

Uranium concentration calibration curve .
238U 232T

Internal [230Th/232Th], [ / h]
isochron for sample 76A003 . . . . .
Internal [230Th/232Th], [238U/232Th]

isochron for sample 76A004 . . . . .

Internal [230Th/232Th], [238U/232Th]
isochron for sample 76A003 . . . . .

Internal [230Th/232Th], [238U/232Th]
isochron for sample 76A004 . . . . .

Regional map of Long Valley area . . .

Map of Long Valley including sample
locations . . . . . . . . . . .

vii

Page

19
24
33
35
46
49
54
59

70

72

74

76
97

99

CHAPTER I

INTRODUCTION

Accurate dating of Quaternary igneous rocks has
long posed a problem. K/Ar methods are subject to large
errors caused by small amounts of excess argon (Damon,
1969; Dalrymple and Lanphere, 1969). Analytical accuracy

40

is also critical due to the long half-life of K and

low 40Arrad abundances encountered. Many dates for,
samples of different geologic age appear contemporaneous

l4C dates are poor. 14

and correlation with C dating,
though more accurate, is of limited use since only wood,
soil profiles, ashfalls, or tuffs may be dated, and the
range of the method is limited to rocks younger than
40,000 years.

Uranium-series disequilibrium geochronologic
techniques may be applied to volcanic rocks younger than
300,000 years. Given an initial state of isotopic dis-
equilibrium and subsequent maintenance of closed system
conditions, nuclides within the 238U, 235U, and 232Th
decay series will show a progressive change toward a

230

state of secular equilibrium with time. Th is the

most useful nuclide in this respect, due to its half-life

1

of 75,200 years (Atree et a1., 1962). The systematics of

230Th growth back to equilibrium with its parent, 238U:

may be treated in a manner similar to Rb/Sr methodology

23

(Allegra, 1968). Thus, the individual [230Th/ 2Tb] -

2 2
[238U/ 3

rock will describe an isochron, with a sloPe which changes

Th] activity ratios for the mineral phases of a

from 0 to 1 as a function of time (Figure 1). Whole rock
and initial whole rock values may also lend insight into
the geochemical evolution of a magma chamber (Allegre and
Condomines, 1976).

The use of U/Th disequilibrium studies may offer
advantages over the use of standard K/Ar geochronology.
These are:

(1) Migration of U should not occur as readily
as migration of Ar.

(ii) The shorter half-lives of the nuclides
involved with U/Th disequilibrium systematics
theoretically should result in greater
accuracy in age determinations.

(iii) Many materials unsuitable for K/Ar dating
are suitable for U/Th dating.

(iv) The analysis of several phases is necessary
for the construction of an isochron and this
provides an internal check on the assumption
that the mineral phases have been closed to
isotope migration. K/Ar single phase
analyses do not possess such an internal
check.

(v) The use of U/Th systematics also provides
petrogenetic information in addition to the
age determination, whereas K/Ar analysis
will only provide age information.

Figure l.--U/Th isotopic behavior in the system
230 23 238U/232

( Th/ 2Th) - (
separate mineral phases of a volcanic

Th) for the
rock initially at isotopic equilibrium
with different U/Th ratios, but the

same (230Th/232

Th) ratios. Filled
circles represent the system at t = o,
the crossed circles at some elapsed
time t = t, and the Open circles at
secular equilibrium where t = w. The
slope of the line connecting the
individual phases is defined by the

chromatic function (l-e-xt) and gives

the age of the system at any time t.

 

 

 

 

u - — — u E

on 32... . .

a I.

Oufi « s x

9 ¢
... . 1. TIM Esmmmkhommv

f
7 . .u ago-m callmmmuv
I“, .. 5.03

2b -_ u one.»

\\

 

 

 

The post-collapse volcanics of Long Valley
caldera, California, have been chosen as a sample site for
this study because excellent age control was available for
a number of samples suitable for U/Th dating. In addition,
various geologic models have been proposed to explain the
origin and relationship of the chemically bi-modal suite
of volcanic rocks found in Long Valley. The application
of the U/Th systematics to the area should provide informa-
tion concerning the age of the associated magma chamber,
as well as the amount of fractionation or mixing that has
occurred within the magma chamber. In addition, the
source of the magma types may be ascertained.

The primary objective of this thesis was to
develop an analytical method for U/Th isotopic analysis
of igneous rock and to test the validity and accuracy of
the U/Th method as an alternative to K/Ar dating of
volcanic rocks between 10,000 and 200,000 years old. A
secondary objective was to use the U/Th systematics as a
petrogenetic indicator to test various models for the
evolution of the bi-modal volcanic suite of the Long Valley,
California, resurgent cauldron. The second goal was only
partially achieved as it required complete analysis of the
total range of rock types present. This was not possible
with the time constraints placed upon this study due to
problems encountered with the development of the analytical

techinque.

CHAPTER II

GEOCHEMISTRY OF U AND TH

Chemical Behavior of U and Th

 

The general geochemical behavior of U and Th is
similar in strongly reducing conditions, such as is
present in the magmatic environment. This is due to the

similarity of ionic radii in the 4+ valence state: U4+ =

4+ = .96A. U and Th are not readily par-

.935 and Th
titioned toward the mineral phase, but rather tend to be
concentrated in the liquid phase, obtaining their highest
concentrations in pegmatites and aplites.

There is usually an increase in U and Th content
during differentiation and an increase in Th/U ratio with
differentiation over a wide range of intrusive and extru-
sive rock suites. However, some oceanic island rock
suites, such as Hawaii, exhibit a decrease in Th/U ratio
(Cherdyntsev and Senina, 1970, 1973; Nishimura, 1970).
Cherdyntsev and Senia (1973) attribute this to the erup-
tion of a Th-enriched residual magma. Preliminary data
from Long Valley, California (Table 4) indicate that
Uranium and Thorium abundances follow a normal pattern.
Uranium abundances range'from 1 ppm for basalts to over

7 ppm for rhyolites. Thorium abundances range from 2.5

6

as

 

ppm for basalts to over 20 ppm for rhyolites. These
abundances were enough to permit measurement of U and
Th isotope ratios.

Previous studies (Rogers and Adams, 1957;
Taddeucci et a1., 1967; Allegre and Condomines, 1976) show
that U and Th abundance was distributed evenly throughout
the phases present in basic and silicic extrusive rocks.
Only small amounts of U and Th are present in uranium
minerals. Glass is usually enriched in U and Th relative
to the mineral phases, though accessory minerals such as
allanite, apatite, zircon, and sometimes magnetite and
biotite, act as hosts of U and Th. Quartz and feldspar
tend to have low abundances of U and Th. Th/U ratios in
phases vary, especially in accessory minerals. Thus, a

238U/ 230

any group of phases present in the Long Valley volcanics.

Th isochron may be constructed using virtually

U and Th behave much differently in the near-
surface environment. Th is always found in the 4+ oxida-
tion state, whereas U may be oxidized to a 6+ state.
Hexavalent U is usually found as the uranyl (003+) form
(Seaborg and Katz, 1954; Christ, et a1., 1955), which is
discriminated against in the formation of rock-forming
minerals, even as a trace element, and tends to remain in
solution. In near-surface, hydrothermal, and supergene

environments, the amount of U in the 6+ state may reach

100%, which will allow migration of U relative to Th.

U may be leached from mineral phases without
destruction of the crystal structure in metamorphic,
hydrothermal, and supergene environments (Szalay and
Samsoni, 1969; Cherdyntsev, 1971). This is due to the
uniform distribution of U and Th in rock-forming and
accessory minerals through isomorphous substitution,
especially in growth sites, strain defects or micro-
fractures, and microcapillaries (Cherdyntsev, 1971).

The behavior of U in glass phases is somewhat different.
A study by Kovalev and Malayasova (1971) on extrusive
rocks indicate that U exists as uranium oxide compounds
rather than silicate chains in vitric glasses. U does
not leach as readily from these glasses as from mineral
phases. Kovalev and Malayasova attributed this to a pro-
tective 'silica gel' coating on vitrified glasses. How-
ever, large amounts of U migration accompany the devitri-
fication of glass (Rosholt and Noble, 1969). Therefore,
samples showing any evidence of alteration or devitrifi-

cation must be regarded with caution.

Isotopic Behavior of U and Th

 

The concentration of the various isotopes of U
and Th is controlled primarily by their placement in one

of the natural radioactive decay series, shown in Figure

2. 238U, 234U, and 230Th belong to the 238U decay chain.

232Th and 228Th belong to the 232Th decay chain. In a

Figure 2.--238U, 232Th, and 235U radioactive

decay series.

l()

 

 

Ross.u..
haao~

1

.

.

_
new..~_.

«a
n-
L .
xv~.a_c
fia-
1 .
“as. .Nv
u<a-
a .

as o. x m~. no

“use. a n~.sc

amn~

 

magma awn"

Ao_enuav
as

“ea

 

 

nozgauuv
Ame—.ce haoo~
osczm w
7. _
. .
1..¢.m. 11 ~°._.
~M a -
a
Auum.nc Ax.c~ x ”m.kc
*- gthN
O
a
Assn”... .xmo. x h..~e
SN" s =Qn~
a
1 T
X o
A=~s. new“ 0. an M_.~S
~n~ - svn~
. A
a
A.
A.“ ow Au ~. «No
Wm” nevnN
» a
“same. nem._v Axoo_ x “n.4V
~n~ sanN
mmuamm es mu_¢um a

«nu 3N

 

 

11

234

system at secular equilibrium, the activities of U
and 230Th will be equal to 238D activity. 228Th activity
will be equal to 232Th activity. Therefore, 230Th

activity is related to uranium, not thorium, concentration

for a system at secular equilibrium. Since neither 238U

232 238 232

nor Th are produced by a-decay, the U/ Th ratio

in igneous rocks is solely effected by chemical processes

controlling the bulk distribution of U and Th in igneous

238 232

rocks. Thus, 0/ Th ratios are effectively the same

as the bulk U/Th ratios.
Secular equilibrium may be disturbed by a number

of processes. Chemical fractionation of the isotopes may

230

occur. For instance, Th may be thrown out of equilib-

2340 during crystallisation from a melt or any

rium with
other chemical process that fractionates U relative to Th.

Chemical fractionation of an intermediate element
may occur. This is the mechanism for disequilibrium

between 228Th and 232Th. 228Ra, intermediate in the decay

232 228

chain between Th and Th, is chemically active and

subject to migration, which will result in disequilibrium
of its daughter product, 228Th.

Isotopic fractionation of chemically identical
elements may also occur due to the Slizard-Chalmers, or
alpha-recoil, effect (Slizard and Chalmers, 1936;
Cherdyntsev, 1971; Kigoshi, 1971). Atoms produced by

a-decay occupy damaged sites in the mineral and will be

12

more subject to migration than an atom that has not under—

gone decay. This effect is pronounced in the case of

234U, as U is itself initially concentrated in structural

defects. As a result, 2340 will be preferentially in-

corporated into microfractures and microcapillaries where

it will be more subject to leaching than 238U. In addi-

tion, a higher percentage of 234D relative to 238U is

found in the 6+ oxidation state, either due to oxidation

234

during decay (Kigoshi, 1971) or to U location in micro-

fractures and microcapillaries (Cherdyntsev, 1971).

2340/ 238U fractionation cannot occur in recently

234

formed minerals, since the U present is incorporated

into the mineral at the time of formation and is in the

238 234

same state as U. Over time, decay—produced U will

accumulate so that after one half-life, 50% of 234U will
be decay-produced.

A similar effect will be encountered if an
element is transformed to another element by decay. The
daughter nuclide may occupy a position in the mineral
that will retain the parent element, but will be unstable
when occupied by the daughter element. This occurs in

decay produced 230Th.

The 2340/ 238

U ratio will change in samples under-
going weathering or hydrothermal alteration. Whole-rock
leaching experiments show a decrease in 2340/ 238

234

the sample and a marked enrichment of U in the extract

U in

l3

(Szalay and Samsoni, 1969). Basalts and andesites
(Cherdyntsev et a1., 1967; Cherdyntsev et a1., 1968) and

rhyolites (Rosholt and Noble, 1969) which have undergone
2340/ 238

waters which drain these areas exhibit

U ratio, whereas
234

alteration show a decrease in
U enrichment.

Fumarole gases and minerals which crystallize from them

are also enriched in 234U (Kupstov and Cherdyntsev, 1969).

A controversy exists concerning the operation of
the Slizard-Chalmers effect in the magmatic environment.

Studies on various volcanic rocks young enough to exhibit

23oTh disequilibrium show no fractionation between 2340

and 238U (Oversby and Gast, 1968; Nishimura, 1970; Allegre

and Condomines, 1976). However, studies by Cherdyntsev

234

et a1. (1967; 1968) indicate that U fractionation does

occur in recent volcanics, in conjunction with Th/U

230 232

fractionation and Th/ Th fractionation. Cherdyntsev

and Senina (1973) attribute this to fractionation of 234U

and 2380 between the liquid melt and solid phases in con-
tact with the magma either through contamination or at the
2340/ 238

tored in this study to ascertain if this type of frac-

site of partial fusion. The U ratio was moni-

tionation occurred in the Long Valley volcanics.

234 238

The U/ U ratio was also used as an indi-

cator of radio-element migration in this study. If 234U

238U, o-recoil 230Th was

probably also fractionated relative to 232Th and large

was fractionated relative to

14

scale fractionation of U relative to Th will also have
30Th/

Th)o ratio, or scatter in the isochron plot due to
230Th/ 232

Th ratios of the phases after

occurred. This would result in either a false (2
232

variance in the initial Th ratios of the phases,

2380/

formation. For this reason, samples that exhibited non-
234U/ 238

or changes in the 232

equilibrium U ratios were regarded with caution.

U/Th Disequilibrium Dating Methods

 

Uranium series disequilibrium techniques have
long been used to date sedimentary rocks. The basis of
most methods in disequilibrium of 23oTh and 231Pa due to
their insolubility in water. Deep sea sediments have
been dated using excess 23oTh and 231Pa created by decay
of soluble uranium in seawater and subsequently incor-
porated in bottom sediments (Piggot and Urry, 1942; Ku,
1965; Broecker, 1965; Broecker and Ku, 1969; Ku, 1976).
Likewise, many carbonates are deposited from water with

little or no initial Th. 230Th and 231Pa will grow into

equilibrium with their parents 234U and 2350, respec-
tively, after the carbonate has been deposited. This
method has been used to date corals and oolites (Broecker
and Thurber, 1965; Ku, 1968; Messolella et a1., 1969; Ku
et a1., 1974), speleothems (Ford, 1973; Schwarcz et a1.,
1975; Thompson et a1., 1975), marls (Kaufman, 1971),
caliche (Ku, 1975), and mollusc shells (Broecker, 1963;

Broecker et a1., 1968; Kaufman et a1., 1971).

15

The change in 230Th activity in relation to 238U
activity may be described by:*
(1) 230Th = ZBQThoe ’xot + 238U [l—e-Aot + (238U/ 2340‘1)
Ao -(>\ -A )t
X—IT—' Ore C) 4 )l
o 4
where:
230Th, 2380, 234U = activities of 23oTh, 238U,
and 234U.
230Tho = activity of 23oTh at time of formation
10,14 = decay constants of 230Th and 234D
t = age of sample
If little or no 234U fractionation relative to 238U has

2380/ 234

occurred, ( U - 1) becomes small and equation (1)

reduces to:

230 230 A t 2

(2) Th = Thoe‘ o + 380(1 - e—Aot).

This equation may be divided by 232Th to obtain:

 

*Activity units, rather than weight or absolute
abundance units, are used throughout when describing
isot0pe ratios or abundances. Elemental ratios or
abundances are given in terms of absolute abundances,
however.

16

t 238

(3) 230Th/ 232Th.= (230Th/ 232Tm0e-AO + U/ 23

2Th (l—e-Aot) .

Various means are used to find a value for the unknown

230

quantity Tho in the dating methods above. However,

238 232 230 232

complex variations in U/ Th and Th/ Th ratios

in the formation of igneous rocks make assumptions of
(230Th/ 232Th)o ratios impossible.

The comparison of two minerals crystallizing from
30Th/ 232

from equation (3). In order for an accurate date to be

the same melt may be used to eliminate (2 Th)

0

obtained, the following conditions must apply.

230 232

(i) The minerals must have identical Th/ Th

ratios at the time of formation.

(ii) The minerals must have formed at essentially
the same time in relation to the total time
span.

(iii) The minerals must behave as a closed system
with respect to uranium and thorium after
formation.

Previous U-series disequilibrium studies of
igneous rocks have yielded mixed results. Cerrari et a1.
(1965) found that an age obtained for a tuff by analysis
of resistate minerals in a beach sand deposited from the
tuff was consistant with the age assigned on the basis
of geological and geomophological evidence. Kigoshi
(1967), studying successive acid extracts of a whole-rock

sample, obtained an age for a rhyolitic tuff which agreed

with a K/Ar date for the same sample.

17

Other results have not been as successful.
Taddeucci et a1. (1967) obtained only general agreement
between U/Th and K/Ar dates for five rhyolites near Long
Valley, California. Kupstov et a1. (1969) analyzed
trachyliparite and basalt, and obtained discordant and
unintelligible results, which they attributed to a complex
232

Th) values of minerals

history and different (230Th/ 0

at the time of formation. As a result, Kupstov et a1.
and others (Cherdyntsev et a1., 1967, 1968) resorted to
calculation of a maximum age based upon the assumption
that no 230Th was present in the mineral at the time of
formation. Maximum ages so obtained will be considerably
older than the actual ages if a considerable amount of
230Th was present at formation.

Allegre (1968) noted that equation (3) described
38U/ 23 230

2Th] (figure 1). The slope of the isochron is defined

an isochron in coordinates of [2 2Th]. [ Th/

23
by the relationship:

(4) slope = l - e-Aot.

Minerals which do not satisfy the three basic conditions
listed previously will plot off the isochron. Fukoka
(1974) and Fukoka and Kigoshi (1974) found that zircon
plotted off the isochron in a study of dacite pumices.
This was interpreted as relic zircon grains included in

the dacite lava. Delitala et a1. (1975) was able to

18

construct a separate isochron of infinite age for

clinopyroxene and magnetite contained in a pyroclastic

230

tuff with a Th date of 41,500 y. using other mineral

phases. This was also attributed to the magnetite and
clinopyroxene being relics of an earlier event.

Allegre and Condomines (1976) have used initial

230T 32

h/ 2 Th ratios to derive information concerning the

development of a suite of andesites in Costa Rica. If a

magma is enriched or depleted in U relative to Th as a

result of formation, fractionation, or contamination, the
Th ratio of the melt will no longer be at

238U/ 230
238 230 230 232

equilibrium. The 0/ Th (and Th/ Th) ratio

to the melt will then progressively change back to a new

equilibrium value (Figure 3). This progression will be
-Aot 2380/ 230

the melt is disturbed once and the melt subsequently

behaves as a closed system. The (230Th/ 232Th)o value of

230 232

linear with respect to e if the Th ratio of

a rock will reflect the Th/ Th ratio of the magma

at the time of crystallization (Allegre and Condomines,

232

1976). Therefore, the (230Th/ Th)o ratios of a suite

of rocks eminating from a common magma may be used to

define the petrogenesis of a magma by analysis of its

230 232

change in Th/ Th through time.

19

232

Figure 3.--Initial (230Th/ Th)o ratios as a

function of ext. Five arbitrary stages
(to + t4) are shown for successive
eruptions of a magma derived from partial
melting in secular equilibrium and
rapidly transported to the surface

(case 1), successive eruptions of a
magma from a differentiating magma
chamber with U/Th constant and higher
than the source value (case 2) and for

a U/Th ratio lower than the source

value (case 3).

20

 

05:9: .222. Co 2.5.

3.295
oceaom Stone

‘5

 

 

 

 

 

 

 

Aeemmme
geomm

CHAPTER III

ISOTPOIC ANALYSIS OF U AND Th

One of the goals of this thesis was to develop an
analytical method for U/Th isotopic analysis of igneous
rock using the facilities available at Michigan State
University. Analytical procedures were largely adapted
from procedures used by Thompson (1973a, 1973b) and
Kaufman (1964) for use with carbonate materials. Modi-
fication of these procedures was required due to the
complex chemical nature of the volcanic rocks compared
to carbonates. A

The method developed was satisfactory for the
analysis of rhyolitic rock, and splits of biotite, quartz,
sanidine, and rhyolitic glass. Rocks with mafic and
intermediate composition were subject to low yields and
sample loss. Two analyses of whole rock splits from a
rim rhyodacite and three analyses of whole rock splits
from basalts were attempted. A satisfactory yield was
obtained only once, for the uranium split of a basalt.
Sample loss also resulted from the use of the method
followed by Allegre and Condomines (1976). Subsequently,
U/Th isotopic analysis of basic rocks was not attempted.

21

22

It is recommended that the analysis of mineral splits

only should be considered for future studies of this

type.

Sample Collection and Preparation

 

Four silicic samples were analyzed in this study.
All were collected from the post-collapse intercaldera
rhyolites of Long Valley, California. Samples of basic
and intermediate composition were from the basalt and
rhyodacite sequences within the caldera. Sample loca-
tions and descriptions are given in Appendix I along
with a description of the geology of Long Valley.

At the time the samples were collected, it was
thought that 5-10 kg samples would yield sufficient
material for dating purposes. Due to sample losses in
trimming and crushing, and the use of sample splits for
thin-section and whole-rock analysis, it became clear
that at least 10-20 kg samples should have been collected
depending on the relative proportions of mineral phases
present in the sample. After weathered surfaces were
trimmed, a split was set aside for thin-sectioning. A
second split of about 100 g was crushed to less than 100
mesh, first with a metal hammer, then with a ball mill.
This split was used for all whole-rock analysis. The
remainder of the sample was used for analysis of indi-

vidual mineral phases. A flow chart of the mineral

 

23

separation procedure is given in Figure 4. The sample
was crushed to between 40 and 200 mesh, with several
smaller size ranges for samples that were to be mag-
netically separated. Sieving the samples was necessary
in order to increase the efficiency of the separation
methods and also to keep the apparatus from clogging.
Several samples were run through a Frantz isodynamic
separator to separate mafic minerals. The magnetic
phase included some glass as well as sanidine-magnetite
and quartz-magnetite intergrowths. Although several
passes through the separator were required, the sample
still required density separation. For this reason,
later samples were processed entirely by density separa-
tion in heavy liquids. Bromoform mixed with dimethyl-
sulfoxide to obtain densities of 2.5, 2.6, and 2.75 g/ml
was used. Separations were performed in a separatory
funnel in approximately 2 g increments. Glass was
separated as the low-density phase at 2.50 g/ml. Sani-
dine was then separated from heavier minerals at 2.60
g/ml. Quartz plus any plagioclase present in the rock
was separated as the light phase from biotite plus any
magneitite and hornblende at 2.75 g/ml. An attempt was
made to separate enough material to have at least 5 g of
each mineral phase except quartz. It was desirable to
collect 10 g of quartz. Due to the small quantity of

mineral phases present in these samples, this was not

24

Figure 4.--Flowchart of sample preparation and

mineral separation procedures.

25

SAMPLE F LOW
PREPARATION

0 m

 

lunweathe redJ

 

 

thin—section]

 

 

    
   

40 mesh

 

magnetic

glass
{with quartz,

aanidine)
L __ __ __ (if Eadie) _ __ )

non- rmagnetic

 

 

 

 

 

 

 

 

 

 

 

 

biotite A quartz
hornblende plagioclaise

magnetite

 

26

always possible. No attempt was made to separate plagio-
clase from quartz or to separate mafic phases. This did

238 230 '

not affect the placement of the [fifilL] - $3522]
Th Th

isochron for the sample under study if the individual
phases plotted on the isochron. The isotope ratios of
the mixed phases were averaged and this value was also
plotted as a point on the isochron for the sample under
consideration. For this reason, very high purity was not
required of the mineral splits. Purity was checked by
optical methods, and with the exceptions above, was found
to be 90% or better.

Chemical Extraction of Uranium
and Thorium

 

 

Before accurate analyses of the d-spectra of the
uranium and thorium isotopes of interest could be per-
formed, chemical purification was required. This was
because radioisotopes with a-energies between 4.0 and 5.5
MeV would overlap the q-energies of the nuclides of
interest (Figures 7 and 8). The elements responsible for
interference are radium, protactinium, radon, and polon-
ium. Uranium will also mask the thorium spectrum and
vice versa, as the decay energies of the elements overlap.
The activity of interfering nuclides should be reduced
to less than .1% of the activity of the desired nuclide

in order to reduce systematic errors. Chemical

27

purification was also required since self-absorption
would result if the source contained more than 100 ug/cm2
of material. This would result in attentuation of the
sample activity and 'tailing' of the original particle
energies, causing loss of resolution. This is known as

a 'thick' source. Any element that can be plated on a
disc and is present in quantity will cause 'thick source'
problems, or problems with the extraction procedure. The
elements that must be reduced in quantity by the largest
degree, due to their concentrations in rhyolites, are:

Si, Al, Fe, Ca, Mg, K, Na, and Mn.

Separation of Radionuclides

 

Uranium and thorium were separated quantitatively
by the use of the strong-base anion exchange resin Dowex
l-X8 in 9M_HC1 (Kraus and Nelson, 1956; Kraus et a1.,
1956). U was adsorbed onto the resin while Th remained
in solution in the elutate. U was subsequently eluted
by .1M_HCl. Following this, Th was removed from the U
elutate by extraction in .ZM thenoyltrifluoroacetone (TTA)
(Poskanser and Foreman, 1961). At pH 1.2, 100% of the Th
was extracted, while less than 5% of the U was extracted.
U was also removed from the Th elutate of the 9M’HC1
column by feeding the Th solution through the same
resin in a 10% SMIHNO3 - 90% methanol form (Tera et a1.,

1961). U remained in solution while Th was adsorbed to

28

be eluted later in 1M_HNO Through these steps U and

3.
Th were quantitatively separated.

Since 226Ra contributed to 230Th and 234U

activity it was removed. This was accomplished during
the Fe(OH)3 precipitate step where Ra remained in
solution. Ra also remained in solution under anion

exchange, and was not extracted by TTA (Hagemon, 1950).

A fraction of 231Pa activity will overlap the

activities of 228Th, 230Th, 2320, and 234U. The presence

of 231Pa can be detected from observation of the 5.01-

5.05 MeV peaks, but correction was not necessary since
Pa was separated from Th by adsorption with U on the Cl-
column (Kraus et a1., 1956) and was separated from U by
extraction with Th into TTA at pH 1.2 (Poskanser and

Foreman, 1961).

210 232 228

Po will obscure the U and Th peaks if

present. Po was removed by volitalization during
evaporation of the sample onto a planchlet and subse-

quent heating of the planchlet to red-hot before counting

(Kaufman, 1964). However, 2108i will decay into 210Po

with a 138 day half-life. Bi was partially separated

from Th by incomplete adsorption onto anion exchange

resin in C1- media and almost no adsorption in HNO3 -

methanol media (Kraus and Nelson, 1956; Tera et a1.,

1965). Bi was partially separated from U by incomplete

elution in .1M.HC1. Since some 21oBi was present in the

 

29

the sample, especially in U samples, the samples were

plated and flamed immediately before counting. Thompson

21opa (t = 134.8

i

days) by counting 232U several months after plating, and

(1973a) monitored the presence of

no change in U activity was reported.
Since radon is an inert gas, it was not plated
on the planchlet and constituted no contamination prob-

lem.

Separation of Major Elements

 

The presence of Si during the chemical extraction
process may cause problems as gels will form at many
stages of the extraction procedure. Fortunately, Si was
volatilized during the dissolution process by production
of SiF4 during HF fuming. In addition, Si did not absorb
onto anion resins and tended to remain in solution during
Fe(OH)3 precipitation. When silica or alumina gel was
present in small amounts, it was assumed that negligible
U or Th was contained in the gel and the gel was dis-
carded. No difference in sample yields were noticed
between samples with small amounts of gel and gel-free
samples. However, samples with large amounts of gel were
frequently lost, either by U or Th loss or the impossi-
bility of continuing the extraction procedure as the
.presence of gels and precipitates also interferred with

solvent extraction steps.

30

Most gels and precipitates encountered were
probably due to the presence of Al, rather than Si.
A1(OH)3 was precipitated along with Fe(OH)3 in the pre-
cipitation step. A1 should elute while in 9M.HC1 form
in the anion exchange step, but small quantities were
noticed in U elutates from columns that were not rinsed
correctly. The presence of large quantities of A1 will
interfere with the N03- - methanol anion exchange or the
Cl— cation exchange methods. For this reason, Al was
removed as A1(OH)4- form by washing the sample while
in the hydroxide precipitate form with NaOH before
adsorption onto the Th adsorbate columns (Kaufman, 1964).
The possibility of large precipitates of insoluble resi-
dues was reduced by eliminating evaporation steps from
the procedure until relatively pure U or Th absorbates
were eluted from their respective columns.

Although iron was added to iron-poor samples as
a carrier in the Fe(OH)3 precipitate step, iron will
adsorb onto the anion resin along with U in 9M.HC1 media,
saturating available exchange sites rather quickly. For
this reason, Fe was extracted into isopropyl either
before the sample was placed on the Cl_ exchange column.
Because the initial ether extraction was usually incom-
plete, and Fe was extracted with U into TTA at pH 3.0
(Poskanser and Foreman, 1961), thick source problems were

avoided by a second ether extraction using lower volumes

 

31

of liquid before extraction into TTA. The presence of
Fe could easily be detected in solution, in precipitate,
and on exchange resin by its reddish color.

Any other element that was present in the sample
in quantity would also present a problem in analysis,
either by creating a thick source or by interfering with
the chemical procedure. However, no other steps were
taken for removal of these elements as most, with the
exception of Mn, Ni, and Mg, were removed in the hydroxide
precipitation step by remaining in solution. In addition,
the ion exchange and TTA extraction steps further dis-
criminated against other elements present in the rock

samples.

Contamination

 

Reagents

All reagents used in the chemical process were
reagent grade obtained from Baker, Fisher, or Mallincrodt.
The hydrofluoric acid was Target electronic grade
(similar to reagent grade). These reagents were not com-
pletely free of uranium or thorium, but any contribution
due to reagent or glassware contamination was negligible.
This was ascertained in reagent blank runs, spiked with

2320/ 228

distilled and deionized. Reagents were stored in either

Th, conducted at this lab. All water used was

polyethelene or factory supplied containers.

32

Glassware

 

All evaporations were carried out in vycor or
teflon glassware to minimize adsorption of thorium onto
the glassware. For other uses pyrex glassware was used
as well. No qualitative adsorption problems were
encountered provided the samples were in either strong
acid or precipitate form. It was noticed during the
preparation of thorium standards that thorium could be
adsorbed if the standard was not strongly acidic (greater
than 1N HCl). Glassware used in the final (TTA) steps of
the procedure were cleaned by acetone, followed by
vigorous soap scrubbing, 8N HNO3 wash, 8N HNO3 rinse,
distilled water rinse, and a final acetone rinse. Vycor
and teflon glassware were cleaned by storage in 8N HNO3,
followed by soap wash, acid wash, acid rinse, and dis-
tilled water rinse. All other glassware were cleaned by
soap, followed by acid and distilled water rinses.
Liquinox phosphate-free soap was used, as uranium will
complex with phosphorus. Reagent preparation is after

Thompson (1973b).

Analytical Procedure-~Silicic Samples

 

Rhyolites were analyzed from the following pro-
cedure, adapted in part from Thompson (1973a; 1973b),
Kaufman (1964), and Tera, et a1. (1961). A flowchart of

the procedure is given in Figures 5 and 6.

33

Figure 5.--Flowchart of chemical procedures
used in chemical extraction of

uranium and thorium.

34

SAMPLE FLOW - U-Th EXTRACTION

 

 

insolubles ‘
NaCO fusion
'*
m

 

If
[33d spike, Fe carriefl

  
  
    
  
 
  
    
     
   
      
 

hydroxides

organic

     
   

isopropyl
eths r
extraction

aqueous

absorbate elutate

 

‘JTh,AI K Metal ions

 

 

 

s ups rnatant

sum:
hydroxides

supernatant

m
hyd roxidss m

10']. 5M HNO3
90$ H's-ethanol

 

 

 

 

 

 

 

 
 
 

 

 
   

 

   

 

. a} uno
pH-l. 2 ‘5'9'53'. elutate
. 25M TTA organic 91"“ m
snfaction m m.
I evaporate
. m "no;
93's). 2 ‘
. 25M TTA “INN“ . 25 TTA aqueous
anti-action PW Bi. 9" 3‘ sxttr‘actb Po. Bl, Pa, Rs
evaporate evaporate

m

35

Figure 6.-—Flowchart of alternate chemical
procedures used in chemical

extraction of Thorium.

36

 

 

Manges 0‘
a l
N .Humm
mozm 2S .

r «anemone

- 0333

 

 

 

       

0:

0390925

 

Noam dam
. 08m dam
, .3 muomm 25

. A

Bow undue e43 .

 

 

 

 

 

 

 

 

 

 

E \/ 320
mcom 332 _.I wXJSom XoBOQ . Aah.

 

147 332» Ao‘ 8383s 1

Hummfl
pl) v l
N some, EOmZ Eoum «AQV

 

 

 

533be .2...

measures 3382+.

37

1. Weighed samples, preferably between 5-129
(samples in this study ranged from 2-36g) and crushed
beyond 100 mesh, were placed in teflon beakers. Approxi-
mately l80-200m1 concentrated HF was slowly added, with
agitation, to minimize splattering. The sample was
covered, heated, and fumed for a minimum of 24 hours,
with occasional stirring. Additional HF was added to
maintain the original volume.

2. After fuming, approximately lg boric acid was
added to neutralize the F- ion. The sample was then
evaporated to & volume. To prevent the formation of
insoluble fluorides, approximately 100ml concentrated

HNO was added and the solution was evaporated. While

3
stirring, 150ml concentrated HCl was added and evaporated.

The sample was then wetted with concentrated HNO3,

stirred and evaporated, and then taken up in approximately
150ml 5M|HNO3. All solids were stirred into suspension
and allowed to settle. A glass stirring rod was placed
in the sample to check for F-. After an hour the super-
natant was decanted into a large beaker and more SM'HNO3
was added to the residue. This process of adding addi-
tional SMHNO3 was repeated until the sample was quanti-
tatively in solution. Generally 4-5 washes of SMHNO3

are required to gain the desired solution. Insoluble

residues at this point were fused with NaCO at 850°C in

3
platinum crucibles. It is important that all the sample

38

goes into solution, since uranium will be preferentially
accumulated in the supernatant before the tracer is added.

3. Once the sample was in solution, it was
diluted to 4M HNO3 and a 2ml Fe3+ carrier was added to
all Fe-poor samples (all but biotite and mafic whole-rock
samples). 2ml dilute spike was added to all samples.
This amount of spike results in peaks that are approxi-
mately equivalent in size to a 5-lOg rhyolite sample.

The sample was then allowed to stand for 8 hours to allow
the spike to reach chemical equilibrium.

4. The sample was placed on a magnetic stirrer
and concentrated NH3OH was slowly added until the pH
reached 6.5. The solution turned brown and a tan to
light brown precipitate was formed. This precipitate
consisted of Fe, Al, U, Th, Mg, Mn, and other hydroxides.
When adding NH3OH, care must be taken to avoid splatter-
ing and also to prevent the formation of localized areas
of high pH.

5. The precipitate was allowed to settle. The
solution was filtered using a funnel and 19cm Wratten #1
filter paper. The supernatant was discarded.

6. The filtered precipitate was dissolved with
9M'HC1 into a clean 250ml pyrex or vycor beaker. Fre-
quently, insoluble gel substances were encountered at
this point. In that case, the funnels were rinsed and

left 4-8 hours to allow all of the supernatant to drain

39

through, then the gel was discarded. Since no difference
in yield was noticed between samples with and without
insoluble gels, it was assumed that the gels are composed
primarily of silica and alumina and do not contain appre-
ciable uranium or thorium.

A 7. The 9M HCl sample was placed in a 250ml
separatory funnel. An equal amount of isopropyl ether
was added, and the funnel was shaken vigorously, venting
frequently to relieve pressure. The phases were allowed
to separate and Fe was extracted into the organic phase.
This step was repeated until the ether showed no change
in color. The ether dissolved in the sample was evapo—
rated by warming.

8. An ion exchange column for separation of U
and Th was prepared by stopping the bottom of the column
with glass wool, slurring 7cm anion exchange resin
(BioRad AGl-X8) into it, and rinsing the column with
30ml 9M'HC1. The rinse was discarded. The sample, in
9M.HC1, was added to the column and the drip rate adjusted
to 50ml/hour. 30-50ml 9M_HC1 was added as rinse. U and
Fe were adsorbed on the column, while Th and other
elements were eluted. To elute U and Fe, 100ml of .1M
HCl was added. When the color band separating the acid
phases was 4cm from the bottom of the column, the col—

1ecting beaker containing Th was removed and saved. A

40

clean beaker was placed under the column and the U and Fe
was collected.

Note: The adsorbate may be dissolved into solu-
tion if too much sample and rinse is allowed to pass
through the column. For this reason, the volume of
sample and rinse was limited to 170ml for small (8mm x
7cm) columns and 250ml for 12mm x 8cm columns. If the
sample was of greater volume, it was split between two
columns.

9. The initial elutate containing Th was trans-
ferred to a larger beaker and concentrated NH4OH was
added until the pH reached 6.5. Hydroxides were
precipitated as in step 4. The solution was filtered
and the supernatant discarded.

10. The contents of the funnel were washed twice
with 75ml hot 3M NaOH. Al was removed into solution as
A1(OH)4- and the supernatant was discarded.

11a. If the remaining precipitate filled more
than i of the filter paper, it would not go into solution
in a small quantity (less than 30ml) of acid and an
alternate method (described in 11b) was used. For low
volume precipitates, the sample was dissolved in a
minimum amount of 5M HNO3 by adding 10ml 5M HNO3 to the
funnel, taking care to wash the sides of the filter paper.
With a spitzer (a disposable lml pipet), the precipitate

was rinsed down the sides of the paper using the remaining

41

solution in the funnel. This procedure was repeated,
recycling the same solution. 7m1 fresh 5M HNO3 was then
added to the funnel and the walls were washed as before.
If an insoluble gel remains, it may be presumed to be a
silica or alumina gel and discarded. An ion exchange
column was prepared as previously described, except the
column was set in 10% 5M__HNO3 - 90% methanol instead of
9M'HC1. The sample was diluted with methanol to obtain
a 10% SMHNO3 - 90% methanol mixture and then added to
the column. The column was rinsed with 30ml 10% SM'HNO3.
The elutate containing impurities was discarded. Th
had been adsorbed onto the column and was eluted with

100ml 1M.HNO So as not to cook the Th residue, it

3.
was evaporated to dryness over low heat or a steam bath
in a vycor beaker. The sample was then ready for step 12.

Notes:

(1) The 5_M_,HNO3 solution should be run through
the Th-column as soon as possible or practical. It may
be oversaturated and precipitate with time.

(2) If a slightly larger volume of acid is
needed to dissolve the sample, use 20ml SMHNO3 rinsed
with 10ml 5MHNO3 and feed the sample onto two exchange
columns.

(3) Care should be taken to insure that the sample

WilJl dissolve while in hydroxide form, as the hydroxides

42

are much more soluble than the nitrates or chlorides at
this stage.

In early trials, the precipitate was taken up into
HCl or HNO3, evaporated, and taken up in HCl or HNO3 for
placement on the exchange columsn. Insoluble residues,
gels, and precipitates were common, and Th was often lost.

11b). If the precipitate was too large to dis-
solve in a small quantity of acid, the following technique
was used. The precipitate, after NaOH wash, was dissolved
in 100-150ml 3M|HC1. The wash was repeated as necessary,
using fresh or recycled 3M|HC1. If an insoluble gel was
present, it was discarded. A column was prepared using
cation exchange resin (BioRad 50W-X8) and 3M_HC1. The
sample was fed onto the column and rinsed with 30ml 3!
HCl, followed by 30ml 6M_HC1. The original elutate was
discarded and the column rinsed. Th was adsorbed on the
column and next eluted with .75M oxalic acid. 10ml con-
centrated HNO3 and 500 ug La was added and the solution
evaporated in a vycor beaker. When approximately 10ml

of sample remained, 50ml BM’HNO was added and re-

3
evaporation continued. When 2-5m1 remained, the sample
was again diluted with 50ml 8M|HNO3. This step was

repeated an additional 3-4 times or until all the oxalic
acid had been evaporated. Finally, the sample was care-

fully evaporated to dryness over low heat. The presence

of oxalic acid in the evaporating solution was

 

 

43

characterized by a honey-yellow color and a large amount
of white vapor as the solution neared dryness. Small
white crystals in the residue also indicated oxalic acid.
The sample was now ready for step 12.

Note: Since oxalic acid forms precipitates in
benzene which is used in the following step, and thorium
oxalate is insoluble, sample loss will result if all
oxalic acid is not removed from the solution before
residues form during evaporation. For this reason, the
3M_HC1 columns should be used only when necessary.

12. The Th residue was carefully dissolved in

2m1 .lM'HNO (pH adjusted to 1.2) and transferred to a

3
centrifuge tube with a spitzer. An additional lml .1!
HNO3 was used to rinse the beaker and was transferred to
the centrifuge tube. If the sample did not dissolve,

it was wetted with 8M_HNO3, evaporated, and taken up in
.1M_HNO3; or the sample was dissolved in a small amount
of SM'HNO3 and diluted with a few drops concentrated

NH OH to adjust to pH 1.2. The pH was brought to 1.2

4

by a stepwise addition of .3m1 dilute NH OH. 2ml .25!

4
TTA (thenoyltrifluoroacetone) in benzene was added and
the solution was thoroughly mixed using a spitzer and a
vortex stirrer for 3 minutes. The time at which mixing
was started was recorded to the nearest minute. The

contents were centrifuged and Th extracted into the

organic phase which was transferred to a 10ml beaker

 

 

44

with a spitzer, care being taken not to transfer any of
the aqueous phase. lml of the TTA solution was again
added to the aqueous phase, the sample was mixed for 5
minutes and centrifuged, and the organic phase was trans-
ferred to the 10ml beaker. The aqueous phase was dis-
carded and the organic phase was evaporated to approxi-
mately .lml over low heat. This was taken up with a
spitzer and evaporated drop by drOp onto a stainless
steel planchlet warmed from the edges. Care was taken

to insure that evaporation did not go to the edge of the
planchlet and was uniformly thin. Lastly, the organics,
radon, and polonium were volatilized by flaming the
planchet until red-hot over a bunsen burner. The planchet
was now ready for the a-counter.

13. The U elutate in .1M|HC1 from step 8 was
carefully evaporated in a vycor beaker to dryness over
low heat so as not to overdry the residue. The residue
was taken up in 5m1 9M|HC1 and Fe was again extracted by
isopropyl ether using a test tube and spitzer. The
aqueous phase was returned to the beaker, carefully
evaporated, wetted with several drops concentrated HNO3,
and carefully re-evaporated.

14. The U-containing residue was dissolved in

.1M'HNO (pH 1.2) and transferred to a test tube the same

3
way as the Th residue. lml of the .25M_TTA was added,

the contents mixed, and centrifuged. The aqueous phase

45

was then transferred to a second test tube, and the
organic phase was discarded. The pH of the aqueous

phase was adjusted to between 3.0 and 3.5 by the step-
wise addition of .3M_NH4OH. Uranium was extracted into
the TTA solution and plated onto a stainless steel
planchlet in the same manner as the Th split. The plated
U disc was then flamed in the same manner and was then

ready for a-counting.

Alpha Counting Procedure

 

A typical U spectrum is shown schematically in

232

Figure 7. U activity plus background activity was

measured by summing all counts from the base of the 232U

peak (D) to the base of the 234U peak (C). In this way,

activity due to tailing of 232U was counted. 2340 plus

96.7% 235U activity was counted from C (base of 234U

peak) to B (base of 238U peak). 2380 plus 3.3% 235U

238U and 235U

are found in fixed proportion to each other, the 2350

contribution to 2340 was eliminated by dividing the

238

activity was counted from A to B. Since

U activity by 21.7 and subtraction of this figure

234 235 2380

from the U activity. U contribution to

activity is negligible. In order to minimize errors

caused by tailing of 238U and 234U, the number of

channels included in the 2380 and 2340 activities were

the same. Some error resulted from 234U activity in

46

Figure 7.--Graphic representation of the
uranium alpha-spectra of a

typical rhyolite.

 

47

 

$9.622 ozEmummpz.) w w new w
_ r. } 1 _
oh >22 2.555 oh

_ _

—0
_q 18922
022

h V. _"Lasz

 

 

ml..._...o>xm Ado-arr
22235..

«JP—bum m (In—44

 

322

 

bSZ

993 m

 

 

n89?

WBNNVHO
SlNflOO

48

the 2380 region, and 232U activity in the 238U plus

2340 regions, but this was considered negligible for a

thin source.

Thorium spectra (Figure 8) were counted in the

same manner: 94.5% 224Ra was counted between D and E,

5.5% 224Ra and 228Th was counted between C and D, 230T

was counted between B and C, and 232Th was counted

between A and B (same number of channels as 230Th).

224 228

h

Ra will grow into equilibrium with Th with a 3.64

day half-life after TTA extraction. Since a correction

for 224Ra activity in the 228Th energy level must be made,

and some tailing of 224Ra occurs under 228

Th decay, Th
samples were counted immediately after TTA extraction.
All activities also include a background activity,
mostly due to contamination of the detector over time.
To compensate for this a blank disc was counted for 12
hours every week. It is important that background
counts cover the same channels that the readings cover
or substantial errors may result. In this manner a
separate background activity for each isotope counted
was obtained. These activities were averaged over three

consecutive weeks. Sample calculations and error

analysis are given in Appendix II.

49

Figure 8.--Graphic representation of the
thorium alpha-spectra of a

typical rhyolite.

50

 

 

 

 

 

 

 

 

 

 

Z Z 7.. 2 Z Z Z
”whn m n ma
mmosoaz ozEuummpz.‘ uo_.. n_. o n w w__. n_u n_u
o a. om >2. 2. 552m oh 0 1.
.l . _ x i1 _ L
< < w o Ni]
_ _
a
Z
d9 W0 O
o w
_ m
u 1...
a. __ _
339:1”. 43...: o z z a
.22on m. n
«581% 51.1.3 t w. w.
. _
2.
7w
8

Q1

 

51

Preparation and Use of Spike

 

Uranium and thorium yields were monitored by the

232 228

use of a 0/ Th spike. This isotope pair was used

232U, 232

because 228Th is the daughter product of U does

not occur in nature, and the a-energies of 228Th and 232D

are near to those of 232Th, 230Th, 2380, and 234U.

232U/

However, there are disadvantages to the use of a

228Th spike, notably that common 228Th occurs in the

samples as a member of the 232Th decay chain. However,

if it is assumed that common 228Th and 232Th are in equi-

librium, 232Th activity may be subtracted from the total

observed 228Th activity to obtain only the 228Th

activity due to the spike. 228Th and 232Th are separated

by 228Ra (ti = 6.7 years) and 228Ac (t = 6.13 hours).

i
228Th will grow into equilibrium

with 232Th within 40 years, therefore, extensive 228Ra

In a closed system,

migration may occur during the history of the rock, as
long as no fractionation has occurred in the last 40
years. Thompson (1973a) and Oversby and Cast (1968)
found no evidence of disequilibrium in pleistocene

samples. However, Cherdyntsev et a1. (1967; 1968) found

228 232

wide variations in Th/ Th ratios.

This problem was resolved by analysis of two

228 232T

unspiked samples and measurement of their Th/ h

ratios. The data (Table 2) suggest that 228Th is in

equilibrium for the rhyolites of Long Valley, California.

52

When using a 228Th enriched spike, a correction

228T

factor must be added to correct for unsupported h

decay. Since unsupported 228Th decays with a half-life

of 1.90 years, the following correction equation was

used:
228Th = 228Th (1 _ e-lt)
corr meas
_ 228 At
- Thmeas ( )
where:
228 _ . .
Thcorr — act1v1ty corrected for unsupported
decay
228Thmeas = measured activity
_ 228 _

A - decay constant ( Th — 9.927 x
10‘4d'1)

t = elapsed time in days between start
of sample counting and uranium
separation (9M HCl column for
silicic procedures; HF wash for
basic procedures)

228

Th spike is unsupported after it is separated from

232 228

U in the 9§'HC1 column, but

common 228Th is discriminated against in the Fe(0H)3

Ra supporting the

precipitation step. Since less than 1% of the original

228Ra will accumulate in one month, and intermediate 228Ac

4!!

53

will decay rapidly, the equation describing the decay of
common 228Th simplifies to the equation above. This
correction is small for samples counted less than a

month after U and Ra separation from Th. Since U was
separated from Th only a day or two after Ra was separated
from Th and a small amount of Ra remained with the Th
after the Fe(OH)3 precipitation step, the time of U
extraction was arbitrarily chosen for use.

Another problem in using the spike was that Th
tends to adsorb onto glassware, drawing the spike out of
equilibrium. For this reason, the spike was stored in
borosilicate glass and acidified with GH’HCl. In addi-
tion, the spike isotope ratio and activity was calibrated
periodically by direct measurement, as described below,
after Thompson (1973b).

lml of prepared spike was evaporated to near dry-
ness, 4 drops distilled H20 was added, the solution was
transferred to a spitzer, and the solution was slowly
evaporated onto a stainless steel planchlet. The
evaporation must be thin and uniform, as minimal tailing
is necessary to resolve the peaks. The planchlet was not
flamed. The sample was counted until at least 10,000

counts were recorded on the 224Ra peak (see Figure 9).

228 232 224

Although Th and U cannot be resolved, Ra can be

considered in equilibrium with 228Th. Therefore:

54

Figure 9.--Graphic representation of the alpha

spectra of the 232U - 228Th spike.

>02 2. >m¢wzw

 

 

 

55

 

 

 

 

 

9m
—
o
_ .
a u u
I O .7
9 U
u u w
-3
Sim fimuw «nu

(chumam <Im4<

 

 

56

228 224

 

Th = Ra

232U = (232v + 228Th) - 224Ra
and the spike ratio, 232U/ 228Th is:

232U = (232U + 228Th) _ 224Ra

2”Th 224Ra

224

The spectrum was counted as follows: Ra was
counted from the base of the 224Ra peak to the base of
the 232U + 228Th peak (C to D). 2320 + 228Th was counted

from C to channel zero. In addition, the tail due to

2320 + 228Th was counted from channel zero to B, which

has the same number of counts as C. Since 5.5% of 224Ra

decays occur in the 2320 + 228Th energy range, 5.8% of

the 224Ra peak was added to the 224Ra peak and sub-

tracted from the 232U + 228Th peak. An emperical

approximation to correct for 224Ra tailing under the

2320 + 228Th peak was performed as follows: 1/3 of the

counts in the 232U + 228Th tail (0 to B) was assumed to

be the number of counts in the 224Ra tail. This was

subtracted from the 232U + 228Th peak and added to the

22433 peak. The [corrected 224Ra peak] was taken to be

the 228Th spike activity and the [(corrected 228Th +

232U) - the corrected 224Ra peak] was assumed to be 2320

activity.

57

U-concentration Measurement

 

U concentration in the samples was determined
internally by fitting sample activity to an activity/
concentration curve.

This calibration curve was established by direct
measurement of U standards for each detector and detector
geometry used. Concentration estimates obtained using
this method are limited to approximately 5-10% accuracy,
the limiting factors being variations in the thickness
and geometry of the sample on the planchlet.

For this reason, U and Th abundances of some
samples were measured by neutron activation analysis, as
described elsewhere. An attempt at estimating Th
abundance by a-activity was also made by using the same
method. However, problems were encountered in setting
up a smooth working curve, possibly due to adsorption of
Th onto the glassware. Subsequently, Th abundances were

238U/ 232Th

estimated using the U abundance value and
ratio.

Preparation of uranium standards follows the
method of Thompson (1973b).

.52739 of 002(N03)2 ° 6 H20 was dissolved in
H20. 50ml concentrated HNO3 was added and the solution
was diluted to 1 liter giving a 25 ug U/ml stock solution.
To obtain a 25 ug U/ml working solution, 10ml of stock

solution was diluted to 100ml. 1, 2, 5, 7, and 10 ml of

58

the working solution was pipetted into separate 10ml
beakers and slowly evaporated to dryness. The residue

was dissolved in 2ml .lg'HNO and the pH of all the

3
solutions was adjusted to 3.0. U was carefully extracted
into .25! TTA in benzene three times. The solutions were
plated and flamed as before (Step 12, procedure), care
being taken not to lose any sample and to keep deposition
on the disc as thin and uniform as possible. 2380
activity of the samples were counted to an accuracy of

i 1%. Activity corrected for background ug U was plotted
for each sample. The points should describe a straight
line (see Figure 10). Scatter in the diagram is most
likely a reflection of differences in the geometry of

the sample on the disc, indicating the inaccuracy
reflected in this determination. Activities must be
measured and a new calibration curve must be made for
each new detector and source/detector geometry. However,
only one set of standards were made, and these were re-
used, since no daughter nuclides have energies in the

238U range.

Accuracy of Method

 

As with most analysis for dating work, the stated
error associated with an analysis is based upon the
estimated instrumental error, based upon counting sta-

tistics or similar variations and is stated at one

59

Figure lO.--Uranium concentration calibration

curve .

60

9:42.! cum «#2300

 

cm NN ON 0. w. v. N. o. o w v N
_ — 4| q A q 4 _ d u _ a
..
z
.6/
o9 ’o
.9
~L.\O\o

_.oz mahoukmo
zo:.<¢._.zuozoo 23.251:

u>¢ao zo_._.<mm.._<o

 

On

00.

on.

as:

OON

onN

con

61

standard deviation. This method will be described in
Appendix II. Other sources of error are not accounted
for by this estimate.

Systematic errors may result from contamination
of the sample, or tailing of other isotope activities
under the activity of interest. These errors are kept
to a minimum by following the chemical and counting pro-
cedures outlined previously. For replicate samples,

67% had activity ratios within one standard deviation of
each other. This, plus reagent blank results, indicate
that these errors are small compared to the calculated
error.

There are two exceptions to this. One is that
excessive tailing of 224Ra under 232U + 228Th is
encountered when a spike calibration is analyzed.
Although a correction for 224Ra tailing was incorporated
into the spike ratio, systematic error may result. This

234U 232

error should increase the error in the / Th and

2380/ 232

1969; Oversby and Gast, 1968).

Th ratios by no more than 2% (Rosholt and Noble,

In addition, several sample splits of 76A021 and
76A017 were processed at the same time as several uranium
ore samples, and it is suspected that cross-contamination
resulted, either through vapors or improperly washed
glassware, since these samples yielded anomalous isotope

ratios, approximately the isotope ratios of the ores.

62

Subsequent re-analysis yielded results in the same range
as other samples. For this reason, analysis of 76A017
and 76A021 are considered suspect.

The estimated errors associated with the slope of
the isochron line, or t, and the (230Th/ 232Th)O ratio
are not directly associated with the isotOpe ratio
errors. The errors are calculated in the scatter of the
various mineral phases from the regression line. There-
fore, the error associated with the age and (230Th/
232Th)o values of the sample was indirectly based on the
true total error associated with the activity ratios and
not the estimated error associated with the activity
ratios.

The error for t and (230Th/ 232

Th)o was stated

at the 90% confidence level. The coefficient of variation
of t varies from 12% to 47% depending upon the sample and
mineral phases used. This is in the same range of error
associated with Allegre and Condomines (1976), using U/Th
methods and in the same range or somewhat less accurate
as Bailey et a1. (1976) using K/Ar methods at the 90%
confidence level. Although no replicate isochrons were
run, the ages for both samples agree with ages obtained
by K/Ar methods (Dalrymple, personal communication) at
the 90% confidence level (Figure 13). Sample 76AOO3 was
younger than the K/Ar age, sample 76A004 was older than

the K/Ar age.

CHAPTER IV

ISOTOPE RATIOS AND ABUNDANCE

DATA FOR U AND TH

Activity ratios of U and Th isotopes obtained by
a-spectrometry are given in Tables 1 and 2.

Preliminary data indicate that the samples

analyzed are in equilibrium with respect to 228Th.

2340/ 238D ratios are in equilibrium for most

2340/ 238

234U

U values were

/ 2380 values more

samples, though disequilibrium
observed in a few sample splits.
than 2 standard deviations greater than 1.00 were noted
for the glass and sanidine phases of sample 76A003.

2340/ 238

noted in two mineral splits and 76A017 whole-rock. It

U values significantly less than 1.00 were

is possible that variation of this ratio may be due to

experimental error.
[2380/ 232T

for samples 76A003 and 76A004 are given in Figures 11 and

230 232

h] — [ Th/ Th] internal isochrons
12. Regression data is given in Table 3. The internal
isochron for sample 76A004 defines an age of 127,000

: 16,000 y.b.p. Sample 76A003 has an age of 89,000 :
42,000 y.b.p. when sanidine values were used and

63

 

64

86,500 : 20,000 when sanidine values were neglected.

(230Th/ 232Th)o values are equivalent for both samples

at .94. Glass falls above the line in both cases and

was not used in calculation of the isochoron. Sanidine

also fell above the isochron in sample 76A003. Since a

considerable amount (about 25%) of glass was noted in the

sanidine split, the error is likely due to glass con-

tamination. The whole-rock split does not occur at the

intersection of the isochron with the equiline and thus

is not an average of the phases present in the rock.

Instead, the whole-rock value is slightly enriched in

U and 230Th.
[234U/

data is presented in Table 3 and Figures 13 and 14. An

23 230 232

2Th] - [ Th/ Th] internal isochron

age of 107,000 :_38,000 y.b.p. was defined for sample
76A004, while an age of 85,000 : 27'000 y.b.p. was

230 232

defined for sample 76A003. Initial Th/ Th values

of .91 were calculated for both samples. Other aspects

of the data based upon 234U are similar to data based

upon 238U. IsotOpe data for samples 76A017 and 76A021
was incomplete and did not define an isochron. Problems
were encountered in the analysis of these samples as well.
For these reasons, internal isochrons were not calculated
for these samples.

Uranium and thorium abundance data for whole-rock

(Table 4) indicate no major variation in U and Th abundance

 

65

among Long Valley rhyolites. Abundance data for mineral
and glass splits (Table 1) indicate U and Th partition

among the various phases present. Feldspar was slightly
depleted in U and Th. Quartz exhibited strong depletion
of U and Th. Glass was enriched in U and Th. The mafic
phase was strongly enriched in Th, but slightly depleted
in U. Other phases did not exhibit strong fractionation

between U and Th.

 

66

 

 

omv. mmo.s ohm. mnm.s moo. smm.s Hmo. smm. mm smm smm m.m smm H.m smm manuscmm
moo. moo. coo. coo. mmo. mom. mmo. NHO.H so wvm smm m.m smm macs:
Hmosos
oom. mam. so wms smm o.~ mos wusuosm
ohm. omH.H mom. 6mm.H mom. mmo.s ass. mmm. smm so smm s.m smm ~.v has waswscmm
won smm smm o.o smm 0.0 was Nusmso
ems. mmm. How. mmm.s nmo. mmm. «NH. mvs. smm so smm m.HH smm H.m cos macs:
kmosoa
mmo. Nqs. ovo. ems. mvo. mmv. smm. Hon. smm smm smm o.H mms wasuosm
mos. mum. mos. mum. mos. 0mm. Hmo. moo.H mom woo smm o.ms smm ~.o mos mcssssmm
smm. HRN.H vsm. ons.s mom. NNH.H smm. who.s mam smm smm H.~ cos nusmso
mmo. mom. mmo. mom. «so. onH.H mmo. mom. sum smm smm m.om smm m.m mam mmmso
smm. oo~.H mom. mmm.a mno. mmH.H ooo. mmm. was sm smm N.¢H smm m.m voH macs:
soosok
moo. who. moo. “no. mmo. mmv. omo. 6mm. smm was smm m.m mam massesm
omo. Ham. mvo. has. one. «mm. «mo. mmH.H wvm wqm smm «.ma smm n.v «Hm assessmm
Hoo. mam. omo. mvs. who. mmm. mso. mmo.~ smm smm smm ~.m smm m.s Nam Nassau
moo. mum. «mo. mam. mmo. moo.H Hmo. ems.s ssm wow smm h.nm smm H.o mam mmmsu
Ava. vms.s svs. mmH.H mas. mvo.a mmo. mmm. was mum smm H.ms smm m.o Hos macs:
moosok

. mmmusa . mmwuss u mmmusa . mmmus asmH» osmH» .ozoo .ozoo .oz .02
+ emwuo + mm~-= + cmmnss + ¢m~-s as a as a .msq msmzsm

 

 

.muuofiomuommmumnmam >3 nocwmuno mpaofl> mamsmm can .mCOwumuucmocoo .mowumu >uw>wuo¢nn.a mqmda

 

TABLE 2.--238Th/232Th activity ratios - unspiked mineral
splits.

76A004 Glass 1.055 : .068

76A004 Quartz .990 + .070

 

 

6E3

 

 

 

as as m m m 1mm .>.o
ooo.os ooo.oa ooo.m ooo.m ooo.m .mc.H
ooo.vm ooo.¢m ooo.m0H ooo.mos ooo.m0H Ame mom sans
«so. mmo. H
oAnBNMN.
Nam. mam. ovmsm. mum. smm. .1111:
0mm
c s s s s ‘H
000 on ooo MHH ooo so ooo mm 000 we Ass u
. :
ooo.~mH ooo.vwa ooo.oss ooo.mms ooo mos Ass 0
mm NH mm sq mm Ass .>.o
ooo.mm ooo.oH ooo.- ooo.mv ooo.om Assn“
ooo.moH ooo.n~H ooo.mm ooo.mw cow.om ”sou
mmm. mama. mom. momm. room. as
mommovmmo. monHosmo. mmosmvvvm. omosooHom. Nommmnmvm. Hg
0
mmm. mam. mac. mmq. mac. a
x00» wHona xoou oaonz ocapwcmm msflpwcdm
ocwowcmm ocwpwcmm wuumsv uunmsv nuumsw
NauQSU Nuumsv xoou oaoz3 xoou maon3 xoou maon3 muwamm
musuosa wusuosn musuosn wusuomn musuosn magmas:
A29 D :9 D :9 D :9 D a D con 00m
«mmx «mus A mmmx smm. 1 mmmx ammo 1 ~m~\ mmmc 1; ~m~\ mmmc s H
voosmn voosoh mooson moosmn moose” uwnssz msmsmm
. VOO<QF mug MOO¢®F mOHnmEmmllmCOHSOOmH HMGNOUCH NON MUM” COHmmOHOOMII . m 3&9

€59

 

 

 

 

 

.moaumflumum
ocflucsoo co comma .coflumuucoo mo ovmucooumm mm Ummmoumxm cofiumfi>m0 pumocmum mac n sawumwum> mo ucmwufimwooo n .>.o
onmscsoo m3.68 ..m.o.m.s .oumaawz .a.=
at
AcoHumowcsano assemuwmv was m

C
nu nn nn nn nu nn nn nu m.m m H.H Hm ~.v mao¢o>
nu nn nn nu nn nu nn nu o.m 0H o.H mm o.~ hector
nu un nu nn nn nu nn nu H.m m N.H mm m.m ooomwh
uHmmmm
o.m o N.N ms N.HH nn nu nu o.m e m.~ ms m.m «NOUthoHoson
oufiompoxsm
nn nn uu nn un nn nn nn nn nu un nn nn anodes
m.~ m m.o m o.oa nn nn nu uu nu nu nn nu nooownxomoson
H.¢ m m.m m m.ms nu nn nu nn un nn nu nn ssoomm\mao<mh
nn nu nu nn nn nu nu nn m.m m m.m oH o.m~ haotoh
v.m N m.m h H.H~ m.m m.m N.vH H.N m m.h 0H m.oa mOOUNh\voo¢mh
o.m m m.o m 9.0m h.m m.o H.ma nn nn nu nn nu m000m5\moo¢wh
muflaoxm

D\ss .>.o smm a .>.o smm as axns smm a smm as o\ns .>.o smm a .>.o smm as

n. onEMm

a COAum>wuo¢ couusmz

>uuweouuowmmuo

ac cowum>wuo¢ couusmz

 

.moocmvnsnd £9 and Dun.v Wanda

70

232 238U

Figure ll.--[230Th/ Th], [ 232

/ Th] internal

isochrons for sample 76A003.

71

0.0

 

39540210:
uthoi

  

20¢:00n. 4(28Uh2.

uh30>rc >u.__.._<> 0204
n00<0h

 

'n -|"""'I-""

on...osh
mu<so «nu

Tgwl qua

 

     

.xo.

5.

09d.

1 0..

 

72

23 238 232

Figure 12.-~[230Th/ 2Th], [ U/ Th] internal

isochrons for sample 76A004.

73

0d 0..

zocruow. adzcmhz.
95405.3. >u44<> 020..

Q00<0b

g... «MN

 

0..

 

 

 

NOZHJOZRO...
uh.h0.0

 

 

0.0

74

23 232

234U/ Th] internal

Figure 13.--[230Th/ 2Th]. I

isochrons for sample 76A003.

75

 

 

 

film's...
shunu
. . a
o u n _ v3 o.. ad ad
— 4 u u 4 d 4 J u d u q u 1 4 J d 1 0.0
3232.5: ..
22.303. .3255. 3.33 .
330:... >33... 26.. n n o
noo<ms n
L
3.1.3.3.... .. .—
33.... n «nu
lpllg
2.653 n on"
ma. . a: n n. I l
39.... «nu n 0..
. w 3.3.. n
offio
race «.03 n
600 n
900
)l l
4;
o
n a.

76

23 234U 232

2Th]. n /

isochrons for sample 76A004.

Figure 14.--[230Th/ Th] internal

77

 

      

«nu
. . 3 .
.uu a. c»« c. an.
1| q q u 4 1 u q q d + s q .
uozuanzco:
92.0.0
zocroom.4<2¢uhz. n
uh...0>z¢ >u44<> 020.. n
t00<os :
1
«.0. n a...
«nu n
um<4003<41 0mm

N...¢¢:0

  

 

V

 

 

 

.LN Arm*ln
user: mm<.e

 

o.0

an“
on“

 

0..

CHAPTER V

DISCUSSION AND CONCLUSIONS

Discussion of Sample Results

 

It is unlikely that recent uranium loss occurred

228 232

in samples 76A003 and 76A004. Equilibrium Th/ Th

ratios indicate little or no alteration within the past
60 years. 234U depletion, indicative of U loss, was
observed only once, for the biotite split of 76A004.
Textural evidence indicates little or no deuteric
exchange with either hydrothermal or meteoric waters,
and the sample locations are located on a topographical
and structural high away from the locus of current hydro-
thermal activity.

There is a slight possibility that uranium migra-
tion occurred soon after eruption. Bailey et a1. (1976)
and Lachenbruch et a1. (1976) suggest that the possi-
bility of hydrothermal alteration was greatest at that

234 238

time. U/ U fractionation should not accompany

alteration at this time, since only a small amount of
234U will have been created by radioactive decay and will
be occupying a-recoil damaged sites. Uniform uranium

abundances, linear sample isochrons, and textural and

78

79

field evidence tend to argue against uranium migration
after formation.

2340 observed in several samples

Disequilibrium
may occur before eruption of emplacement. Such fraction-
ation could occur in several ways. Excess uranium may
be introduced into the melt during initial partial melt-
ing, as residual minerals are leached in U, and the melt

234

is preferentially enriched in U by the Szliard—Chalmers

effect. Excess 234U may also result from leaching of the
wall rock. The above mechanisms would probably give

similar 2340/ 238

U ratios for all mineral splits of the
same sample. Early formed minerals could be leached of

U by the remaining melt. The melt composition could

also change with time, either through addition of new
material, fractional crystallization, or loss of volatiles.

2340/ 238U

These processes could lead to variation of
ratios in mineral phases from the same sample.

Mineral phases used in this study were well suited
for U/Th disequilibrium geochronology. Sufficient spread
was present in the Th/U ratio to allow reasonably accurate
internal isochrons. The linearity of all phases, except
glass, indicated that formation of phases took place at

230 232Th

approximately the same time, that initial Th/
fractionation did not occur, and that subsequent U

migration either did not occur or was minimal.

80

The anomalous appearance of glass above the iso-
chron was probably a result of chemical fractionation of
the liquid after crystal formation close to the time of
eruption or emplacement. It is possible that U was
preferentially incorporated into a gas phase and subse-
quently lost through fumarole activity, consequently the
remaining melt was depleted in U. Upon cooling of the
melt, the system with respect to glass closed. However,
glass remained open to chemical exchange longer than the
minerals, which crystallized at depth, causing glass to
fall off the isochron line.

The glass in the samples analyzed was probably
not leached of U by modern rain water, as the glass was
vitric and showed no evidence of alteration in thin-

2340/ 238

indicated that large scale U leaching of the glass had

section. U ratios were not less than unity which
not occurred.

There was a possibility that U was lost as UF4
gas during dissolution of the glass phase. No isotopic
fractionation was present due to complete dissolution
of the sample. Such a uranium loss would result in the
displacement of the glass phase to the right, above the
isochron. However, the dissolution of silicate glass by
HF for U analysis has been used successfully in the past

(Tadeucci et a1., 1967; Allegre and Condomines, 1976).

In addition, similar behavior would result from

81

dissolution of whole-rock samples or mineral phases.
This was not the case. U loss during dissolution is
therefore improbable.

With one exception, ages obtained in this study
were equivalent to corresponding K/Ar ages at the 90%
confidence level (Table 3). Sample 76A004 yielded a
U/Th age older than the corresponding K/Ar age, however,
the reverse was true of sample 76A003. Better fit of
mineral phases to the regression line were obtained

238U/ 232

in lower experimental error. However, no significant

difference was found between [2380/ 232T 230 232T

2340/ 232 230Th/ 232
232

Similar initial Th/ Th values indicated a

through the use of the Th ratio. This resulted

h] [ Th/ h]

ages and [ Th] [ Th] ages.

230

common origin for the two rhyolites studied. Three

models may be proposed to explain the lack of change in
230Th/ 232

10,000 years apart.

initial Th value for samples erupted at least

(i) The samples were erupted from an old magma
chamber that had reached isotopic equilibrium
with respect to U and Th.

(ii) The samples were erupted from a magma chamber
where relative fractionation between U and
Th had not taken place.

(iii) The samples were produced from separate
partial melting events from a homogeneous
source, subsequently with identical
petrogenesis.

82

The possibility of the third model is remote due to the
coincidences involved. The second model may possibly
have occurred, but the first model is in agreement with
existing evidence that magma has existed continuously in
the Long Valley chamber for over 106 years (Bailey et a1.,
1976; Lachenbruch et a1., 1976).

Models (i) and (ii) require that isotopic
equilibrium was maintained for some time preceeding and
during the eruption of samples 76A003 and 76A004. Conse-
quently, the source magma did not undergo change in U or
Th content through fractionation or contamination.

Differentiation of a silicic magma to produce
intermediate rock types, as proposed by Bailey et al.
(1976) is not in agreement with the data presented here,
unless differentiation proceeded so slowly that the

230Th/ 232

samples analyzed. The production of intermediate magma

change in the Th ratio was undetected for the
types through contamination as proposed by Eichelberger
and Gooley (1977) is in agreement with the data presented
in this paper, if the contamination did not affect the
parent magma of the rocks used in this study.

If 23oTh was at equilibrium during mineral forma-'

tion, then an initial Th/U value of 3.3 may be calculated

for the parent magma at the time of mineral formation.

83

Comparison with Previous Results

 

The ages calculated in this study are in far
greater agreement with K/Ar dates than previous U/Th
ages presented by Taddeucci et al. (1967) and Allegre
(1968). This is partially due to the inaccuracy of the
method of calculation used by Taddeucci et al. (1967)
and Allegre (1968), but may also be a reflection of the
unreliability of K/Ar dates of such young rocks. Glass
in this study falls above the isochron, while glass fell
below the isochron in the data presented by Taddeucci
et al. (1967). This casts doubt on the possibility of
using glass-hornblende pairs to estimate the eruption
age, as suggested by Allegre (1968). Analytical error
associated with the ages calculated in this study was
higher than analytical error for corresponding K/Ar ages.
The correlation of U-Th ages with previous K-Ar studies

lend support to the accuracy of both U-Th and K-Ar ages.

Suggested Improvements

 

The accuracy of the dates obtained in this study
could be improved in several ways. Samples could be
counted with greater accuracy through slightly higher
sample sizes, longer count times, and the use of detectors
with lower background count rates and greater sensitivity.
A greater amount of data points may be obtained through

separation and analysis of more mineral phases, such as

84

magnetite. Artificial "phases" may be separated,
possibly by separating the glass by various physical
methods, including color, magnetic susceptibility, or
sample size after crushing. Slight differences in U and
Th content may result, even if the glass is essentially
homogeneous. Several splits of the same sample or
several samples of the same cooling unit may be analyzed
and placed on the isochron, increasing the power of con-
fidence and possibly increasing the accuracy of the age.

Isochron lepe and intercept has been calculated
by linear regression using the least-squares technique.
This is a completely unbiased estimator of slope and
intercept only if the standard deviation of the errors
in the data points are equal. Since this error varies
somewhat from sample split to sample split, another
method would be to weight individual samples on the basis
of analytical error and perform a linear regression on
the isochron points using the method of maximum likeli-
hood. This has been used to treat Rb/Sr data (York,
1969; Brooks et al., 1972).

Errors associated with the determination of the

228T 232

h/

of lml of spike to insure minimum tailing, and counting

U spike may be reduced by plating .lml instead

for long periods of time, up to one week.
The partial success of one basalt sample and

preliminary success by another worker at Michigan State

85

University (8. Leo, in preparation) with isotopic
analysis of Hawaiian basalts indicates that further
studies of the caldera rocks are possible. Analysis of
contemporaneous rhyodacites and basalts in addition to
the rhyolites would yield information on the relationship
of the rhyodacites to the basalts and rhyolites.

In summary, this thesis has shown that excellent
correlation between U/Th and K/Ar dates may be obtained
for silicic volcanic rocks. Analytical accuracy will
not be significantly better than K/Ar accuracy with the
present procedures and instrumentation, however. U/Th
dating may reduce the possibility of systematic error
inherent with K/Ar dating of young rocks due to argon
loss or experimental error by the use and correlation of
many different sample splits to obtain an age.

In addition, it has been shown that two rhyolites
in Long Valley caldera were erupted approximately 100,000
years ago. The major mineral phases were formed at -
roughly the same time, after which U was lost from the
magma, probably at emplacement. Migration of U or Th
after emplacement probably did not take place. The
original magma had a Th/U ratio around 3.3 and subse-

quently lost Th relative to U prior to emplacement.

REFERENCES

86

REFERENCES

Allegre, C.J., 1968. 230Th date of volcanic rocks: a

comment. Earth Plant. Sci. Letters, v. 5, p. 209.

Allegre, C. J. and M. Condomines, 1976. Fine chronology
of volcanic processes using 233U- Th systematics,
Earth Plant. Sci. Letters, v. 28, p. 395-406.

Attree, R. W., M. J. Cabell, R. L. Cushing, and J. J.
Pieron, 1962. A calorimetric determination of
the half-life of Th-230 and consequent revision
of its neutron capture cross-section. Can. J.
Phys., v. 40, p. 194.

Bailey, R.A., 1974. Preliminary geologic map and cross
sections of the Casa Diablo geothermal area,
Long Valley Caldera, Mono County, CA, Open file
map, 2an U.S. Geol. Surv. Reston VA.

Bailey, R.A., G.B. Dalrymple, M.A. Lanphere, 1976.
Volcanism, structure, and geochronology of Long
Valley Caldera, Mono County, CA. J. Geophys.
Res., v. 81, p. 725-744.

Bateman, P.C., L.D. Clark, N.K. Huber, J.G. Moore, and
C.D. Rinehart, 1963. The Sierra Nevada Batholith:
A synthesis of recent work across the central
part. U.S. Geol. Surv. Prof. Pap. 414-D
Dl-D46.

Broecker, W.S., 1963. A preliminary evaluation of
uranium series inequilibrium as a tool for
absolute age measurement on marine carbonates.
J. Geophys. Res., v. 68, p. 2817-34.

Boecker, W.S., 1965. Isotope geochemistry and the
Pleistocene climatic record, in H.E. Wright, Jr.,
D.G. Frey (Editors). The Quaternary of the
United States. Princeton Univ. Press,C§rinceton,
NJ, pp. 737—53.

 

87

88

Broecker, W.S. and T.L. Ku, 1969. Caribbean cores
P 6304-9: New analysis of absolute chronology.
Science, v. 166, p. 404-6.

Broecker, W.S. and D.L. Thurber, 1965. Uranium series
dating of corals-oolites from Bahaman and Florida
Key limestones. Science, v. 149, p. 58-60.

Broecker, W.S., D.L. Thurber, J. Goddard, T.L. Ku, R.K.
Matthews, K.J. Mesolella, 1968. Milankovich
hypothesis supported by precise dating of coral
reefs and deep-sea sediments. Science, v. 159,
p. 297-300.

Brooks, C., I Wendt, W. Harre, 1968. A two-error
regression treatment and its application to
Rb-Sr and initial 87Sr/855r ratios of younger
variscan granitic rocks from the Schwarzwald
Massif, Southwest Germany. J. Geophys. Res.,
v. 73, p. 6071-6084.

Cerrai, E., R.D. Lonati, F. Garrazzini, E. Tongiori,
1965. Il metodo iono-uranio per la determinazione
dell'eta dei minerali vulcanici recenti. Rend.
Soc. Min. Ital., v. 21, p. 47.

Cherdyntsev, V.V., 1971. Uranium-234 [Translated from
ORAN-234 (Russian)] Israel Program for Scientific

Translations. Ltd-Keter Press, Monson, Jerusalem,
308pp.

Cherdyntsev, V.V., G.I. Kishtsina, V.M. Kuptsov, E.A.
Kuzmina, V.L. Zverev, 1967. Radioactivity and
absolute age of young volcanic rocks. Geochem.
Int. v. 4, p. 639-46.

Cherdyntsev, V.V., G.I. Kishtsina, V.L. Zverev, 1967.
IsotOpic composition of uranium and thorium in
rocks and products of active volcanism. Akademia
Nauk SSSR Doklady, v. 172, p. 177.

Cherdyntsev, V.V., V.M. Kuptsov, E.A. Kuzmina, V.L.
Zverev, 1968. Radioisotopes and protactinium
age of neovolcanic rocks of the Caucasus.
Geochem. Int., v. 5, p. 56-64.

Cherdyntsev, V.V. and N.A. Senina, 1970. Radioactivity
of young volcanic rocks of the Pacific Islands.
Geochem. Int., v. 8, p. 652-63.

89

Cherdyntsev, V.V. and N.A. Senina, 1973. On the
behavior of uranium-234 in volcanic processes.
Geochem. Int., v. 10, p. 18-23.

Christ, C.L., J.R. Clark, H.T. Evans, Jr., 1955.
Crystal structure of rutherfordine. Science,
v. 121, p. 472.

Christensen, M.N. and C.M. Gilbert, 1964. Basaltic cone
suggests constructional origin of some quyots.
Science, v. 143, p. 240-42.

Christiansen, R.L. and P.W. Lipman, 1972. Cenozoic
volcanism and plate-tectonic evolution of the
Western United States. II. Late Cenozoic.
Phil. Trans. R. Soc. Lond. A., v. 271, p. 249-84.

Dalrymple, G.B., 1967. Potassium-argon ages of recent
rhyolites of Mono and Inyo craters, CA. Earth
Plant. Sci. Letters, v. 3, p. 289-298.

Dalrymple, G.B., A. Cox, and R.R. Doell, 1965. Potassium-
argon age and paleomagnetism of the BishOp Tuff,
CA. Geol. Soc. Amer. Bull., v. 76, p. 665-73.

Dalrymple, G.G. and M.A. Lanphere, 1969. K-Ar Dating.
W.H. Freeman, New York.

 

Damon, P.E., A.W. Laughlin, J.K. Percious, 1967. Problem
of excess argon 40 in volcanic rocks, in'
Proceedings of a symposium on radioactive dating
and methods of low level counting. IAEA, March,
1969, p. 463-82.

Delitala, M.C., C. Federici, A. Taddeucci, 1975.
Uranium and thorium isotopes in minerals from the
Cetara-Serrara Fontana Tuff (Ischia Island,
Italy)-a tentative radiometric dating. Bull.
Volcan., v. 38, p. 1015-1022.

Eichelberger, J.C. and R. Gooley, in press. Evolution of
silicic magma chambers and their relationship to
basaltic volcanism. Am. Geophys. Union Monogram
Series.

Ford, D.C., 1973. Development of the canyon of the South
Nahanni River, N.W.T. Can. J. Earth Sci.,
v. 10, p. 366-78.

Friedman, I., 1968. Hydration rind dates of rhyolite
flows. Science, v. 159, p. 878-80.

90

Fukoka, T., 1974. Ionium dating of acidic volcanic rocks.
Geochem. J., v. 8, p. 109.

Fukoka, T. and K. Kogoshi, 1974. ‘Discordant Io-ages and
the uranium and thorium distribution between
zircon and host rocks. Geochem. J., v. 8, p. 117.

Hageman, F.T., 1950. The isolation of actinium. J.
Amer. Chem. Soc., v. 72, p. 768.

Hildreth, W. and F. Spera, 1974. Magma chamber of the
Bishop Tuff: Gradients in T,P , and P ,
Geol. Soc. Am., v. 6, p. 795. total H20

Huber, N.K. and C.D. Rinehart, 1967. Cenozoic volcanic
rocks of the Devil's Post Pile Quadrangle,
Eastern Sierra Nevada, CA. U.S. Geol. Surv. Prof.
Pap. 554-D, Dl-D21.

Kaufman, A., 1964. 234U/230Th dating of carbonates from
Lakes Lahontan and Bonneville. Ph.D. Dissertation,
Columbia University, New York.

Kaufman, A., 1971. U-series dating of Dead Sea Basin
carbonates. Geochim. Cosmochim. Acta, v. 35,
p. 1269-81.

Kaufman, A., W.S. Broecker, T.L. Ku and D.L. Thurber,
1971. The status of U-series methods of mollusk
dating. Geochim. Cosmochim. Acta., v. 35,
p. 1155-83.

Kigoshi, K., 1967. Ionium dating of igneous rocks.
Science, v. 156, p. 932.

Kigoshi, K., 1971. Alpha-recoil thorium-234: Dissolu-
tion into water and the uranium-234 uranium-238
disequilibrium in nature. Science, v. 173, p. 47.

Kovalev, V.P. and Z.V. Malyasova, 1971. The content of
mobile Uranium in extrusive and intrusive rocks
of the East margin of the South Minusinsk Basin.
Geochim. Int., v. 8, p. 541.

Kraus, K.A. and F. Nelson, 1956. Anion exchange studies
of the fission products, in Proc. Int. Conf.,
Peaceful uses of Atom. Energy. Geneva, 1955,

v. 7, P/ 837, p. 113-25.

91

Ku, T.L., 1965. An evaluation of the 234U/238U method as
a tool for dating pelagic sediments. J. Geophys.
Res., v. 70, p. 3457-74.

Ku, T.L., 1968. Protactinium-231 method for dating
coral from Barbados Island. J. Geophys. Res.
v. 73, p. 2271-76.

Ku, T.L., 1975. 230Th/234U dating of desert caliche.
Presented at Geol. Soc. Am. Penrose Conf.,
June 8-13, Mammoth Lakes, CA.

Ku, T.L., 1976. The U-series methods of age determina-
tion. Ann. Rev. of Earth and Plant Sci., v. 4,
p. 347-80.

Ku, T.L., K.G. Knauss, and G.G. Mathiew, 1974. Uranium
in Open ocean: concentration and isotopic com-
position. EOS Tran. Am. Geophys. Union, v. 55,
p. 314.

Kupstov, V.M. and V.V. Cherdyntsev, 1969. The decay
products of uranium and thorium in active
volcanism in the USSR. Geochim. Int., v. 6,
p. 532.

Kupstov, V.M., V.V. Cherdyntsev, E.A. Kuzimina, L.D.
Sulerzhitskiy, 1969. Ionium ages and the process
of formation of volcanic rocks. Geochem. Int.,
v. 6, p. 671.

Lachenbruch, A.H., J.H. Sass, R.J. Monroe, and T.H.
Moses, Jr., 1976. Geothermal setting and simple
magmatic models for the Long Valley Caldera.

J. Geophys. Res., v. 81, p.

Leeman, W.P. and J.J.W. Rogers, 1969. Late Cenozoic
alkali-olivine basalts of the Basin-Range
province U.S.A. Contr. Miner. Petr., v. 25, p. 1.

Messolella, K.S., R.K. Matthews, W.S. Broecher, D.L.
Thurber, 1969. The astronomical theory of
climatic change: Barbados data. J. Geol.,

v. 77, p. 250-74.

Nishimura, S., 1970. Disequilibrium of the 238U series
in recent volcanic rocks, Earth Plant. Sci.
Letters, v. 8, p. 293-300.

92

Noble, D.C., M.K. Korringa, C.E. Hedge, G.O. Riddle,
1972. Highly differentiated subalkaline
rhyolite from Glass Mountain, Mono County, CA.
Geol. Soc. Am. Bull., v. 83, p. 1179-84.

Oversby, V.M. and P.W. Gast, 1968. Lead isotope compo-
sition and uranium decay series disequilibrium
in recent volcanic rocks. Earth Plant. Sci.
Letters, v. 5, p. 199-206.

Piggot, C.S. and W.D. Urry, 1942. Time relations in
ocean sediments. Geol. Soc. Am. Bull., v. 53,
p. 1187-1210.

Poskanser, A.M. and B.M. Foreman, Jr, 1961. A summary
of TTA extraction coefficients. J. Inorg. Nuc.
Chem., v. 16, p. 323-35.

Reinhart, C.D. and D.C. Ross, 1964. Geology and mineral
deposits of the Mt. Morrison quadrangle, Sierra
Nevada, CA. U.S. Geol. Surv. Prof. Pap.,
v. 385, p. 1-106.

Rogers, J.J.W. and J.A.S. Adams, 1957. Autoradiography
of volcanic rocks of Mt. Lassen. Science,
v. 125, p. 1150.

Rosholt, J.N. and D.C. Noble, 1969. Loss of uranium
from crystallized silicic volcanic rocks. Earth
Plant. Sci. Letters, v. 6, p. 268-70.

Schwarcz, H.P., R.S. Harmon, D.C; Ford, 1975. North
American climate during the last 200,000 years
inferred from stable isotope studies of
speleothem. Geol. Soc. Am. Abstr., v. 7,

p. 851.

Seaborg, G.T. and J.J. Katz, 1954. The Actinide Elements.
National Nuclear Energy Series, McGraw-Hill,
New York.

 

Sheridan, M.F., 1968. Double cooling-unit nature of the
Bishop Tuff in Owens Gorge, CA. Geol. Soc.
Am. Spec. Pap., v. 115, p. 351.

Slizard, L. and T.A. Chalmers, 1934. Chemical separation
of the radioactive element from its bombarded
isotope in the Fermi effect. Nature, v. 134,
p. 462.

93

Smith, R.L. and R.A. Bailey, 1968. Resurgent cauldrons
in_Studies in Volcanology. Geol. Soc. Am.
Mem., v. 116, p. 613.

Szalay, S. and Z. Samsoni, 1966. Investigation of the
leaching of uranium from crushed magmatic rock.
Geochim. Int., v. 3, p. 613-23.

Taddeucci, A., W.S. Broecker, D.L. Thurber, 1967. 230Th
dating of volcanic rocks. Earth Plant. Sci.
Letters, v. 3, p. 338-42.

Tera, F., J. Korkisch, F. Hecht, 1961. The distribution
of thorium between alcohol-nitric acid solutions
and the strongly basic anion exchanger Dowex-1.
Separation of thorium from uranium. J. Inorg.
Nuc. Chem., v. 16, p. 345-49.

Thompson, P., 1973a. Speleochronology and late pleisto-
cene climates inferred from O,C,H,U, and Th
isotopic abundances in speleothems. Ph.D.
dissertation, McMaster University, Hamilton,
Ontario, 340pp.

Thompson, P., 1973b. Extraction and isotopic analysis
of trace amounts of uranium and thorium from
speleothems. McMaster Univ. Tech. Memo 73-9.

Thompson, P., D.C. Ford, H.P. Schwarcz, 1975. 234U/ZBBU
ratios in limestone cave seepage waters and
speleothem from West Virginia. Geochim.
Cosmochim. Acta., v. 39, p. 661-69.

WOod, S.H., 1977. 14C dating of the Mono-Inyo Craters.
Geol. Soc. Am., v. 88, p. 89.

APPENDICES

94

APPENDIX I

GEOLOGY OF LONG VALLEY, CALIFORNIA

95

APPENDIX I

GEOLOGY OF LONG VALLEY, CALIFORNIA

Long Valley caldera is a 450 km2 elliptical
depression on the east front of the Sierra Nevada which
straddles the Sierra Nevada and the Owens Valley
(Figure 15). The caldera is approximately 30 km south
of the Mono craters area, the postulated location of a
submerged magma chamber in an earlier stage of develop-
ment than Long Valley (Bailey et al., 1976).

The dimensions of the caldera are 17 km north-
south and 32 km east-west. The walls of the caldera are
well-defined, ranging from 500 to 1200 m with lower
elevations to the east and southeast where the area is
drained.

Pre-caldera host rocks in the vicinity of the
caldera (Figures 15 and 16) include the Jurassic and
Cretateous granites and granodiorites of the Sierra
Nevada batholith and Paleozoic and Mesozoic metamorphic
rocks of the Mount Morrison and Ritter range roof pen-

. dants (Bateman et a1., 1963; Rinehart and Ross, 1964).
Basalt and andesite flows ranging in age from 2.6 to 3.2
m.y.b.p. by K/Ar methods which reach thicknesses of 250 m

96

97

Figure 15.--Geologic map of the region around

Long Valley, from Bailey et al. (1976).

m

Bishop Tull

m

Rhyolite of Glass Mountain

lIIIHHIII]

Tertiary volcanic rocks

Geophysical boundary of Mono Basin
and Long Valley caldera alter Pakiser
(1961)

EXPLANATION

Mesozoic plutonic and metamorphic rocks

l::|

Alluvium and glacial deposits

Mono Craters ring lracture zone
ol Kistler (1966b)

[:3

Holocene rhyolite and rhyodacite

 

Postcaldera basaltic rocks

Postcaldera rhyolite and rhyodacite

+
Normal fault
Ball and bar on downthrown side

98

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

we
5 1 l in ‘ 1 l ‘ ‘ l 1 1 1
s. 7,352,.v,,..«,€..,?,m..n......n.«...«.«u>, w
r /I/\ x .z \. II \ zlzzxx n.
m. /\ \ “/V\II’—A’—\ /\ \ W \~\~ ~
m r...
K
m]
0 s
n. 'I w
I r
I
l I.
l I.
.w
l‘
O .H
II
I '
I I
I ‘ a
l L
l u I.
l A
I" m V
l
I A
"
w.
P
I.

 

            

nnl'l""‘l||||[l" A

“W

In»

'5! "ill
lllliin

 

h

 

""l

       

.\t
I '
l\.n I

  

119‘15'

99

Figure l6.--Geologic map of Long, Valley,
California, showing sample locations.
Numbers on map correspond to the last

two digits of the sample number.

EXPLANATION

 

[:3 Alluvium, glacial deposits. and caldera run 0 rhyolite
Volcanic vents I rhyodacite
' - it
I. ‘i Holocene rhyolnte rhyodac e ‘ ba salt-an d esite
S .
\\\\\\\ Late basaltuc rocks 3. sample locality
' Rum rhyodacntes 0 Drill hole
7 -" Moat rhyolutes _L. Direction of dip oi strata
Early rhyolites tutts: line dotted
flows: coarse dotted / General direction oi
Bishop Tull "”3” °' '3"
. . . . dome "0V6: ""0 lined Normal fault - ball and bar
Rhyolnte 0' Glass Mtn tufls: coarse lined on downthrown side
‘ .
m 7W3" V°'°a"‘° '°°"‘ I- Outline on Long Valley

I
Jurassic-Cretaceous granitic rocks I caldera floor
Paleozoic-Mesozoic metamorphic rocks

     
 

119'00'
\\\

100

 

 

////////””’ j I
///////////// ////////
//////////”’/”/’/
/////////////;;/I
/

   

 

 

  
  
    

\\\\\\~

\\

   
      
    
   
  
 
 

usuu§§§§§§§§;-

  

I O
Ill/ll ’ u _
xz/n/ /” 1 an
z/////////”/ n “.9
/////////”” . ()C)
////////”” I U
///////,///l -’ ,,
Ill/l/ /”” .1 I.
’7’, //// m
x/ , , /// o
// /’ I

.....

 

  
   

  
 

{,6 %I"‘ ~/ 7) /

/

g b
m

   

10 KM

8

101

occur in the caldera area, including the caldera itself.
The distribution of the mafic flows does not suggest a
direct correlation with the associated magma chamber and
caldera itself (Bailey et al., 1976).

A thick sequence of rhyolites on the northeast
rim of the caldera can be related to the Long Valley
chamber (Bailey et al., 1976). These rhyolites, which
underlie Glass Mountain, range in age from .9 to 1.9
m.y. old. The rhyolites are aphyric to sparsely por-
phyritic, high in silica, peraluminus, and highly dif-
ferentiated (Noble et al., 1972).

The Bishop tuff is a rhyolitic ash flow sheet
which erupted from vents in the caldera. The tuff is
crystal-rich, rhyolitic, and homogeneous in composition
(Huber and Rinehart, 1967; Sheridan, 1968). Based on
geochemical and mineralogical evidence (Hildreth and
[Spera, 1974), Bailey et al. (1976) calculated that 600
km3 of magma was ejected. Although great volumes of
material were ejected in more than one event (Sheridan,
1968), K/Ar dates and geochemical evidence indicate that
the Bishop tuff was emplaced within a time span of a
few centuries, or less, about 700,000 years ago
(Dalrymple et a1., 1965). Formation of the caldera took
place immediately after or during eruption of the Bishop

tuff, as the roof of the magma chamber collapsed.

102

Crystal-poor rhyolite tuffs, flows, and domes were
erupted within the caldera almost immediately after
collapse. K/Ar dates indicate that these early rhyolites
had been erupted over a span of 100,000 years and less
than 100,000 years after eruption of the Bishop tuff and
collapse of the chamber. Chemical analyses (Table 5)
show that the rhyolite is homogeneous in chemical compo-
sition and high in silica, although Bailey (1974) was
able to map three mineralogical facies. Glass in the
early rhyolite domes and flows is partially devitrified.

Uplift into a resurgent dome in the west-central
part of the caldera accompanied emplacement of the early
rhyolites (Bailey et al., 1976; Smith and Bailey, 1968).

After resurgence, hornblende-biotite rhyolite,
richer in phenocrysts and slightly poorer in silica than
the early rhyolites, was erupted. These rhyolites are
highly pumacious with a vitrophyric texture. Three
groups of rhyolites were erupted, in the south-east
around 300,000 years ago, in the north-east about 500,000
years ago, and in the west around 100,000 years ago by
K/Ar dating.

Basaltic magma was erupted in the west moat
associated with a chain of basaltic magmatism extending
from Devil's Postpile to the south-west of Long Valley

(Huber and Rinehart, 1967), to the Mono basin, 45 km

1113

.moosoas on now oooooH o conunsou so soon n .H.o.n

..oooa. woos one someones.

 

 

nn nu m..o on. nu nn oo.o Ho. nu oo.~ oo.m nu ma. H.o.n
nn un mm. on. oo. ms. nu nn o.. nn uu om. nn ow:
on. o.m mo. mo. oH. mm. mm. mm. so. ed. on. mo. ea. memm
NH. on. os. ma. oo. oo. vo. mo. mo. so. so. No. no. on:
v.. o.~ on.. so.. so. on. me. mo. ma. ow. vs. oH. oH. «One
o.H ~.~ ~o.s oo.H o.e o.m oh.m va.v m.v oo.v m>.o ~.m oH.m ems
m.m o.o ov.m oH.o m.o o.m so.~ -.o o.m oo.n oo.m m.m em.m cmoz
m.os H.o oo.o m~.o e.~ >.~ no. Hm.~ on. oH.H mo. on. he. one
m.e o.m oo.o No. H.H o.H mo. eo.a om. em. em. on. on. om:
o.o H.o oo.o mo.o o.v e.m om.a oo.m s~.H ~o.~ NH.H ov.~ no.~ momom
o.n~ o.oH me.nn o~.eH o.>H o.oH Nm.ms .m.oH m.ma mm.oH oo.MH m.oH on.ma memss
.m.ov .e.om oa.~m ~o.om .n.oo so.mo om.~e ma.>o .v.oe -.mm om.o> .~.vp mm.oe menu
moan: moon: oaosoe oooson o~onz oomnz oHosos Noosoe wean: naosoh ooosoe Hanna oaosou

 

 

 

 

monsoon opus

mouwoooomnm 5.2m .

mounaomnm one:

mounaomsm mason

 

.ouoo Houseman xaomnn.m mamas

104

north (Christensen and Gilbert, 1964). Eruption of the
basalt was contemporaneous with the eruption of the last
group of moat rhyolites. These rocks are similar to
other Cenozoic mafic rocks occurring throughout the basin
and range provence and are thought to have a common upper
mantle origin (Leeman and Rogers, 1969; Christiansen and
Lipman, 1972).

Also contemporaneous with eruption of rhyolite
and basalt is the eruption of crystal-rich rhyodacite
along the caldera rim and moat. Textural and field
relationships as well as discordant K/Ar dates on sani-
dine and biotite for some samples suggest that these
rocks are at least partially a product of mixing of a
rhyolitic or rhyodacitic magma with the basalt magma
(Bailey et al., 1976; Eichelberger and Gooley, 1977).
Several stages of differentiation may also exist, as
suggested by Bailey et al. (1976). The effect of these
processes on the petrogenesis of the intermediate magma
is unclear, however.

The most recent volcanic products of Long Valley
are the Inyo domes, a series of rhyolite-rhyodacite domes,
related in time (12,000-720 y.b.p.; Dalrymple, 1967;
Friedman, 1968; Wood, 1977) to the rhyolites of the Mono
craters. The domes are chemically and petrogenetically
heteroqeneous, consisting of rhyolite obsidian similar

to the rhyolites of Mono craters and rhyodacite similar

105

to the earlier rim rhyodacites of Long Valley. These
textures suggest an incomplete mixing of the two magmas.

In summary, the Long Valley caldera was formed
.7 m.y.b.p. during the eruption of the Bishop tuff.
Volcanism continued within the caldera in the form of
crystal-poor rhyolite flows and tuffs in conjunction
with doming in the west-central area of the caldera.

This was followed by eruption of three groups of rhyolites
near the caldera most with higher amounts of H20 and
phenocryst content. Contemporaneous with the last group
of rhyolites, rhyodacites exhibiting disequilibrium
crystallization textures were erupted along the caldera
rim, and basalts were erupted in the caldera area in con-
junction with basalt eruption in other areas. Recent
eruption of rhyolite-rhyodacite in the caldera is similar
to the volcanism at Mono craters to the north.

The basaltic, rhyodacitic, and last of the three
moat rhyolite sequences are suitable for age dating by
the U/Th disequilibrium method, because they were erupted
within the age range of the method (10,000-200,000 y.b.p.).
They possess a sufficient number of phases, and they have
undergone little alteration. In addition, U/Th dis-
equilibrium dating of the rhyodacite phase of the Inyo
dome holocene volcanics may reveal the presence of relic

minerals of an older age than that given by K/Ar dating.

106

Sample Collection

 

At the time sample collection was undertaken, it
was assumed that the U/Th isotope analysis of at least
5 or 6 samples of felsic and mafic composition would be
possible. Consequently, a sampling program was set up to
define the compositional and isotopic changes exhibited
within the caldera with composition and with time. Nine-
teen samples averaging 10 kg each were collected from
within the caldera (Figure 15 and Table 6).

Six basalts were collected with as wide a spread
in location and time as possible. Four rim rhyodacites
were collected, two from the flanks of Mammoth Mountain
and two from the summit of Mammoth Mountain. Six moat
rhyolites were collected, one from the north-east, one
from the south-east, and four from a wide variation of
locations in the west, including one from a rhyolite
dome (76A020). In addition, two early rhyolites were
collected. Samples were taken from identical locations
of samples used for K/Ar dates. Control was established
using field maps and notes of B. Dalrymple and R. May.
Samples were taken from fresh, unweathered surfaces when

possible.

Chemical and Petrographic Description
Major element chemical data for the moat rhyolites

is given in Table 5. The moat rhyolites are high in

107

.osooouoooo oasmuoom o.an>oo ..me~ .m

 

..mv .9 .a .m .x 32 .x mm .x 32 .xum: oqu no nuuo: odes H usuooou a aouu .Hooaoh no mean ouuoo uoc wwoomh moo<mh

.oaucnuogo cemwuuox .uz Tm...“ .m Tmm .H. .mm .m .x mm so. 3m

.x 3m .oouo unausou sodomy oxosqsuuoos nuance souu oouooaaou .ocoxouhm one convened: _I

nocafi .uuwu0nn .ommHUOAmnHm uo mumauooconm nun) ouaooooazu nausea 30am menu .>neoao .m.n.>.a one. + mva. macOMh aoocoh
meadoovoxnm Ed...

.osmomuooso somsuuoz .ux...mo~ .x ..mn .9 .Nn .m .x 3m .x x: .m m no

season .mocnumm uo: uuosuwnz uo 322 am. .oocusqom u>wuaauo :uoumoonuooa ecu cw 30au e Bonk I.

.mummuooconm on non“ ouoa xauzunau umooxo .noocoh cameo» o» uoaaanu ouMHo>nu oauacmouua> .m.n.>.a oao. + own. mooomn H~o¢oo

.uaocouousa camwuuox .uz

..mn~ .m ..mm .9 .NN .m .x 32 .x :2 .m m and mo noucou .noxuq nuanux mo nuke: vuuoOOH

.oososqom o>num=uo opened; on» on deep ouaaoxnu m scum .uounaouoaa osmoacda one uneducauoam
mo mumfimcoo xwuuofi one .omoHUmeuam mo num>uooconm sud: ouwao>£u oUMDOHQnovcoanuox .m.n.m.a moo. + ooa. hooomh owoaon

.oamcauoulo oawmueom

m.an>oo ..u- .s ..mm .9 .o .m .x mm .x mm .x 32 .cnoucsox noon «0 on.» 3m ago so .oosoovoo

o>nudauo cuuunoa ecu scum .woouo nozuo a. can» ouwn ca nomad. one nuoxuooconm .oocoancuo:
one .oosuosn .oooooo .oommosou no moosuoososm osoosono no.2 oussosso onussmouuno .ouns: .m.n.s.a ooo. + mas. ssoowh ososoe

6.9.9330 canauuo: .uz Tmm~ .m ..mm .9 .Nm .m 3. 3m

.x mz .x mm .soscx manuoaot no 3mm .am.a .oocuavoa o>wumauo snounounuuoc on» a. soda e Ion»
.sosoouooso uo moosm osoo ocssosm umooxo .moosoh 0» soasssa ounHossu onsssmouuns .ousss .m.o.s.s oao. + moo. asoome esoaoe
.ososouoosa soonuuoz .u: ..mh~ .m ..mm .9 .ea .m .x mm .x 32 .x u: soon umooxo .noosos on snow .m.n.s.a ooo..H woo. oooows oooson

.oaosouooso someone: .ux ..mh~ .a ..mn .e .5“ .m .x mm .x m: .x no

.modi nuoaadz no nuuoc Haosx o Bouu ocuooaaou .eocosvon o>wumauo :uounea on» on 30am u louh
.oocososoon oco .oonuoso .nuuooe .osnossom mo nonsuoosonm no.3 sunflosso onussmouuns .ouss: .m.n.>.s «oo. + nos. oooo~u noosoe
nouaamxnm one:
coauaooq one newumauouea cannon send u¢\u . c.02uwmmnum .oz damadm

 

.ucOAumauouoo ounsamnu.o mamas

108

.co«n«ooum

Heonuxaece com: oomen .co«ue«>oo oueoceum oco on cue u¢\x on» c« nouue Hecau>~e=4 ..co«ueo«csaaoo «encouem. eHQIhuden ..05mu. .He no 5euaen noun:
C

 

assuage: xoouusoo ..us~ .m ..m~ .a .mN .m .x m2 .x as no season .oooaun uo

.eaoceuoeso
.8 enql a.

ocsoummseo annuam men we xoouo uo ouose .m aouu oouooaaou .eouwaou04l eaeauoaoeam undocsne
neweucoo xauue: .ocoxouhm ocwueamou >Houoaqsoo uufiauuflbn .ocoxou50 co umcau sawuoeou
mauou cocoancuou .nunmuoococm ocoxoumm one oueaoo~ueam vowed new: oneuen uensoanoa .anueuo

.oamseuoeso connuuo: .u: ..u5~ .z ..ma .9 .Na .m .x :2 .x Hz .J a «0 uouceu .mon .u.a

no .3 aaonx Haeau e Souk .nunhuoocosm uneducameam one ocoxouxm ooaenlu nun) uaeeen accede

.oaoceuoeso sonnuuo: .ux ..un~ .m ..mn .e .H .m .x 32 .x 3m .x 3m .Hmob :- aouu ”Hoax Adela
e co .nun5uooconm uneducaoeam one ocdxon>n Haeau nun: uaenen ueH:u«u9> ode aanenu 5H£o«=

.393qu oceans. .ux rumw .u
..mn .9 .Hm .m .x mm .x :2 .x 3m .oossumm no: canons coco no em: use: a. neon». no new. .2 none

.munxuoocozm uneducaoeam one .ocuuouam .u:u>wao eased sud) uueeen ue«:u«no>

.oaoceuoeso sonauuox .u: ..Mm~ .m ..mm .9 .HM .m .x 3m .3 Hz .1 3m .naaa Ounewo

.cueuonecuu

eneo eeone

sasoosnmoumnoouou .ooonumm no: canons once so am: as“: H.~ season .0 one. .2 noun oooooudoo
.ooo<oh ou ue~«e«n xuuue: .auauuooconm uneducaoeam one ocuxouhm can: agenda occaeuonecah

.eaoceuoisa ceawuuo: .9:
..uo~ .m ..mn .a .Hn .m .x 3m .x as .x 3m .Hasa canons oooo soaon snaooscmouosoouu-

.umCAumm Do: caneuo eneo uo zmz nudes m. aeouua no down .2 Bonn oouoeaaoo

.ocoxouhn

one .oocoancuo: .omeHUOHOeHm .uuuuocuea uo mcwununcoo unuuea sun: .udeuen oceaeuonocuh

.ouoceuoesa eawmunoa a.d«>oo
.53:st fies-sex uo pun-5.. Bonn .usuaoun ace 3 soaueuouae aesu umouxe .0835 ue ofiem

educeuoeso oauaueom u.u«>ua .caeucao: zuolaa.
no unison Bosh .nunhuoococm ouwuo«n oououae one uneducaoeam so»: euqoeooanu 5nneau .u:u«q

.m.n.>.a

.m.n.5.a

.m.n.h.l

amenehea

eman0%0.

eman0%e.

onene>o.

omufle>ea

.Ho..n «on.

33 oo:

mso..u no”.

ooueo no:

Mac. + N00.

mno. + ONa.

oosomma so
«00. + 0Nd.
ecaoaceu no
«no. + 000.

confines so
0N0. + moo.
ocaoucea so
0H0. +.0m0.

0000n5

0000n5

n~00n5

vH00n5

NH00n5

NNOUn5

n~00n5

m~0¢w5

N~0¢O5

~H0405

000¢05

500105

000105

nuances «use

0a0¢05

000105

109

silica, approximately the same amount as the early
rhyolites. The moat rhyolites contained lower amounts
of K20 and higher amounts of volatiles when compared to
the early rhyolites.

Thin-section petrographic analysis indicated that
the early rhyolites were suitable for dating, though K/Ar
data indicated that the north-east rhyolite group was too
old to date by U/Th disequilibrium. None of the samples
taken showed evidence of major alteration, and the glass
in the flows was vitrophyric. Phenocrysts of sanidine
(about 15% by volume), plagioclase (5-15%), lesser
amounts of quartz, and biotite or hornblende (l-2%) were
present. Phenocrysts averaged up to 5 mm in length.
Minor fine-grained magnetite was also present. Glomero-
porphyritic textures were uncommon. Sample 76A020,
collected on a rhyolite dome, was slightly different,
with a matrix of plagioclase and sanidine microlites,
with no void space. Plagioclase phenocrysts were zoned,
and plagioclase rather than sanidine was the primary
phenocryst phase.

The last basalts analyzed were similar in compo-
sition to other basin and range basalts. Two types were
present, quartz normative tholeiite, represented by
sample 76A006, and olivine normative basalt, represented

by sample 76A018.

110

Thin-section petrographic analysis indicated that
the basalts were suitable for U/Th dating, though mineral
separation could be difficult due to the fine grain size.
The basalts were fine grained, usually with abundant
glass. Pyroxene was the predominant mineral phase,
occupying 5-20% of the whole rock. Minor amounts of
olivine were observed in some samples. Hornblende was
found in all samples, in places as reaction rings on
pyroxene. Mafic phenocrysts were generally less than
1 mm in length, though larger in some samples (76A008,
76A018). The matrix consisted of roughly equal amounts
of andesine plagioclase microlites less than .5 mm and
glass, with 1-5% fine-grained magnetite. Only minor
amounts of alteration were present. Sample 76A018, from
the north moat, was slightly different. It contained 10%
plagioclase (An 35) phenocrysts, in addition to pyroxene,
and the matrix was almost totally crystallized. The
crystallized matrix consisted of 70% andesine plagioclase
microlites, 10% pyroxene, and minor amounts of hematite,
magnetite, and glass.

Chemical analysis of the rim rhyodacites indi-
cated that the rhyodacites were of intermediate composi-
tion and show some chemical variation. Sample 76A010,
from the summit of Mammoth Mountain, was richer in silica,

K O, and volatiles while poorer in Fe O

2 2 3' 2
CaO, and MgO. This may have been due to hydrothermal

TiOz, Na 0,

alteration.

111

The mineralogical composition of the rim rhyoda-
cites varied, but they generally contained 10-20% zoned
plagioclase, 15% or less hornblende, and varied amounts
of augite, hypersthene, biotite, magnetite, sanidine,
quartz, and olivine in some places. Reaction rings and
embayed reaction surfaces were prevalent. The matrix
consisted of high Fe glass with varied amounts of
plagioclase microlites. In some samples, the groundmass
exhibited a heterogeneous flow banded texture. Alteration

was present in some areas (76A009).

APPENDIX II

CALCULATION OF ANALYTICAL ERROR

112

APPENDIX II

CALCULATION OF ANALYTICAL ERROR

The number of alpha-decays produced or counted in

a given time always follows the Poisson distribution,

N
(No)

W(N) = N! exp (-No)

 

which approximates a Gaussian distribution in N is large
(Jenkins and DeVries, 1969). The standard deviation

(0) of the distribution is equal to No or approximately
equal to N. This means that, neglecting background
effects, the standard deviation of: 100 counts is i 10
counts, or i 10%: 1,000 counts is i 31.7 counts, or

i 3.17%; 10,000 counts is i 100 counts, or i 1%. With
this relationship as a rule of thumb, an effort was made
to count until 1000 decays had accumulated under each
peak to be counted. Samples used for calibration were
counted until 10,000 decays had accumulated under each
analyzed peak. Since N = RT where R is the activity and

T is the counting time,

0 (R) _ o (N)

R N

 

113

114
if 0 (T) is considered negligible. Therefore,
-5
o (R) — T'

There is a similar error associated with the counts due
to background. When the background is subtracted from

the total count rate, the errors are propagated accord-

 

ing to:
o 2 o 2
s Rb
where: CT = error in corrected activity
as = error in sample activity
ab = error in background activity'
RS = sample activity
= background activity
since a = g , this may be reduced to:

 

o.=/

sample count time

6' an”
silo?”
filo?"

m

where: T
s
Th = background count time

T

= error in the sample activity

115

The first term expresses the uncertainty in the total
activity. The second term is an expression of the un-
certainty in the background activity calculated from
the background runs. The third term expresses the uncer-
tainty in the number of background counts contribution
to the sample activity.

Error in activity ratios are propagated according

to:

 

2 2
A 0A 0B
0 = / //<-) + (-—0
A/B B RA RE

where: p73 is the activity ratio

R

A' RB are the act1V1t1es

OA' OB are the errors
It should be noted that all errors of isotope activities
or ratios discussed are 1 standard deviation, or at the
68.3% confidence level. All errors are approximately
twice as large at the 95% confidence level.

230Th/ 232T

the method suggested by Allegre (1968) and Kigoshi (1967).

Ages and ( h)o ratios are calculated by

The slope (b1) of the regression line is calculated by
a least-squares linear regression in two dimensions,

238 232 234 232 230Th/ 232T

using U/ Th or U/ Th as X and h

as Y:

116

b = 2(Xi - X) (Yi - Y)
— 2
Z(Xi - X)

 

The Y-intercept (be) is calculated as:

b =-l—(2Y.-b
n l

EX.)
0 1

1

The (230Th/ 232

Th)o value is the point where the isochron
meets the equiline, or X = Y for the isochron.

Error in the slope is calculated as:

 

2 _.
3(b1)’ _2

s(b1) is distributed as t with n-2 degrees of freedom.
This means that s(b1) cannot be treated as one standard
deviation of a Gaussian distribution, which is the way
most errors associated with isotope ratios and K-Ar dates
are expressed. Direct comparison may be made by comparing
similar confidence intervals, however.

230Th/ 232T

estimating a confidence interval at h, the

h) ratio was found by

2380/ 232Th

Error in the ( 0

value at the intersection of the isochron and equiline:

 

117

These positive and negative values were then extrapolated

parallel to the isochron until they intersected the

230T 232

equiline. The difference between the h/ Th ratios

at these points was then divided by two and used as the
230Th/ 232T

The age of the sample is related to the slope of

error associated with the ( h)o ratio.

isochron by:

where 10 = decay constant of 230Th. The sample age does

not vary linearly with the isochron slope, with more
variation in age for a given change in slope as the
sample is older. Since the error in the slope is
symmetric around b1, the calculated lower age limit is
closer to the calculated age than the upper limit. These
ages have been averaged by calculating ages for the upper
and lower slope limits, taking the difference, and

dividing by two.

Sample Calculation

 

(1) Raw counts were calculated from instrument read-outs

as follows:

238U

1350

235U _ 238

U / 21.7

62

118

234 234 235

u = u - u = 1382 - 62 = 1320
raw
232U = 1272
232Th 2117
230Th = 2327
_ 224 _ 224
224Ra ‘ Raraw ( Rabackground activity)
(Th )

count time
= 667 - (.1597) (1479) = 431

224
hraw - ( Ra) (.06)

228 228T

Th =

= 4587 - 26 = 4561.

(2) Raw counts were converted to counts per minute:

238U = 238 / 238

cpm counts U

count time (minutes)

1350 / 953.07 = 1.4165 cpm

238Th = 238Th / 233

Th
cpm counts

count time (minutes)
= 2117 / 1479 = 1.4314 cpm.

all other activities were treated in a similar

manner .

(3) Background activity was subtracted:

238 _ 238 _ 238
U — Usample Ubackground

1.4165 - .0240 = 1.225cpm

119

all other activities were treated in a similar

manner .

(4) 228Th activity was corrected for decay of unsupported
228Th as follows:
228Th = (228Th ) e‘zzat
uncorrected
228 (9 927x10’4 d-1)(18d)
Th = (2.7690 cpm) e '
228Th = 2.8217

(55) Error in sample activity was calculated as follows:

233 238 238
U Ub d Ub d
0238 = t + —E_—g_. + _—-t—_L
U bgd U

 

1.225 .0240 .0240

953.07 + 2633733 I 933267“: '0352 C9“

other activities are calculated similarly.

((5) 228Th activity due to spike addition was calculated

as follows:

228 _ 228 23

232
Thspike —

Th- n 2Th)(228Th/ Th)

unspiked

228T /232

h Th ratio in an unspiked sample is

usually 1.

‘71

(7)

120

2.8217 - (1.4197)(1) = 1.4020 cpm

error:

 

spike = //( 228)2 + ( 232)2

 

 

 

 

O228
_ / 2 2 _
- (.0450) + (.0314) — .0549 cpm
230Th/ 232Th activity ratio was calculated as follows:
23°Th = 1.5088 = 1 063
232Th 1.4197
errOr:
0 = 2”Th //1 °230)2 + ( 0232)2
230 232 230 232
232 Th Th Th
= 1 186 //, .0337)7 + ( .0314_§ _ 034
° 1. 5088 1. 4197’ ‘ °

2340/ 238 228 232

U, Th/ Th ratios are calculated

in the same manner.

121

 

 

 

 

 

 

 

(8) 238U/ 232Th activity ratio was calculated as follows:
238U = 238U x 228Th spike x S ike
232 232 232 9
Th Th U
238
U _ 1.1225 1.4020 =
232 ‘ (1.4197) (1.3176) ('998) '82?
Th
spike = measured 232U/ 228Th ratio in spike
calibration
associated error:
2 2 2'
238 o o o .- _
0238/232 = 2322”'//<2§%§’ + (232229 + (22:28 spike)
Th U Th Th spike
o 2 a 2
+ ( 232) + ( spike)
232U spike

 

.0352 2 .0314 2 .0549 2 .0378 2

= '825 //FfiITZZ? + fiiiififi' * 9:352? + fiifiififi

 

2
.02 _
+ (.998) "' 0054.
2340/ 232Th ratio was calculated in the same

manner .

122

(9) Sample yield was calculated as follows:

232

 

. _ U = .11 1.3176 cpm
% yield — —-—————(SU) (X) x 100 (12.76 CPR/m1)(.2 ml) x100
= 26%
SU = measured activity of 1 ml concentrated spike

X
II

amount of concentrated spike added to the
sample.

Th yield was calculated in the same manner.

(10) Uranium concentration was calculated as follows:

(2320) (c)

U concentration (ppm) (% yield) (sample weight)

(1.3176 cpm)(12.01 g/cpm)
(26%)(5.3 g)

10 ppm

 

WWIllllfllHlH

6241

1

1|

vmo
mm3
u“0
Em