American Mineralogist, Volume 67, pages 137-141, 1982
Identification of fluid inclusion daughter minerals
using Gandolfi X'raY techniques
Mrcneel
E. ZoI-eNsxY AND R. J' BooNen
D ep artme nt of Geoscienc es
The Pennsylvania State University
(Jniversity Park, P ennsylvania 16802
Abstract
Unequivocal identifications of individual daughter minerals extracted from fluid inclusions have been obtainedusing microtechniquesin conjunctionwith the Gandolfi X-ray
camera. This developmentrepresentsa major advancein fluid inclusion researchbecause
solid phasesfrom preselectedfluid inclusions may now be routinely and unambiguously
identified. Application of this technique allows the worker to not only measure the
homogenizationand freezing temperaturesof a fluid inclusion of known origin but also to
identify any daughter minerals that might be present in the same inclusion, provided that
these phases are recognizable both before and after heating/freezingtests, and that the
heating/freezing tests have not affected the daughter phases. This procedure yields
considerably more information about the physical and chemical conditions attending
mineral deposition than is presently available from"conventional fluid inclusion analyses.
both heatinglfreezing data and daughter mineral
identification for the same inclusion. Furthermore,
elements lighter than sodium cannot generally be
Fluid inclusions are samples of the fluids from
which the host crystal grew or samplesof fluids that determined, and polytypes and polymorphs cannot
engulfed the crystal at some time after its forma- be distinguishedwith this method. Raman spectrotion. Becausetheseinclusions are actual samplesof metricanalyses(Rosascoet al.,1975 Rosascoand
the mineralizing fluids, valuable information on the Roedder, t979) have provided nondestructive
chemical conditions attending mineralization may daughtermineral identifications, but these analyses
be obtainedif the composition of the inclusions can are limited to minerals containing molecular units
be determined.This, of course, requires determina- previously fully charactetized by laser-excited Ration not only of the fluid composition but also of the man spectroscopy and enclosed within materials
compositionof all solid phases,i.e., daughtermin' which do not fluoresce in the laser beam. An
erals, that have precipitated from that solution.
additional limitation of this procedure is that ramaFluid inclusion daughter minerals are most often logs specially equippedfor daughter mineral identiidentifiedfrom their optical and physical properties. fication are not generally available. Single-crystal
Although rapid and nondestructive, this procedure X-ray photographs of materials containing fluid
permits only a few daughter minerals to be identi- inclusions may show a superimposedpowder patfied with certainty. Color change following X-ray tern from inchided daughter phasesprovided they
irradiation has been used to characterize daughter are present in sufficient quantity. Roedder and
crystals (Vorob'ev and Lozhkin , 1976)but is appli- Coombs (1967)used this observation to verify the
cable to only a few minerals (e.9., halite, sylvite, occurrenceof halite daughter crystals within incluand villaumite) and, even for these species, is not sionsin feldsparcrystals.Coveneyand Kelly (1971)
totally diagnostic. In recent yeaxs, the SEM and used this technique to identify dawsonite enclosed
electron microprobe have been used extensively for within small chips of quartz. These workers also
identification of fluid inclusion daughter phases successfully extracted large polycrystalline aggre(e.g.,Metzgeret al.,1977),butthesetechniquesdo gates (several hundred microns across?) of this
not allow for preselection of fluid inclusions to be daughter phase from fluid inclusions and verified
examined, eliminating the possibility of obtaining this identification using standardpowder diffraction
Introduction
0003-0M)vE2l0
10?-o
l 37$02.00
t37
l3E
ZOLENSKY AND BODNAR: FLT]ID INCLUSION DAUGHTER MINERALS
techniques.Various other techniques have been
suggestedfor the identification of fluid inclusion
daughter crystals and are summarizedby Roedder
(re72).
At present, X-ray powder diffraction is the single
most complete and generally applicable method of
mineral identification. In this paper a method for
routinely obtaining unequivocal identifications of
fluid inclusion daughter minerals extracted from
preselectedfluid inclusions using microtechniques
in conjunction with the Gandolfi X-ray camera
(Gandolfi,1967)is described.
Experimentalprocedure
Doubly-polishedthin-sectionsand cleavagesections were examined under a microscope and fluid
inclusions containing daughter crystals were located. Inclusionslying closeto the upper surfacewere
usedwheneverpossibleto minimize the amount of
host material that had to be removed to reach the
inclusion. Preliminary identifications were made,
when possible, from an examination of the optical
and physical properties of the daughter phases.
Heating/freezingtests were also conducted. Occasionally, daughterphasesfail to renucleatein inclusions or exhibit different morphologies following
heatingtests.This behaviorlimits a worker's ability
to continuously monitor a particular phase during
experiments.It was not found to be a problem in the
presentstudy owing to the types ofdaughterphases
examined. Following successful heating/freezing
tests, the sample was mounted onto a glass slide
and the inclusion was relocated under the microscope.A small, foot-switch controlled. hand-held
drill with a tapereddental diamond burr was used to
drill a "moat" around the preselected fluid inclusion,thusisolatingthe inclusionwithin a pedestalof
host material. The powder produced during this
drilling, a combination of drill grit and ground
sample, was periodically swept from the working
area so that the inclusion could be continuously
observedthrough the polished upper surface of the
pedestal. It was desirable that a pedestal of the
smallestpossible volume remain, but it was also
necessarythat the inclusion walls not be breached
as the contents of the inclusion would then become
lost among the drilling debris. When the moat had
beenextendedto a level below that of the inclusion"
a finely-sharpenedsteel probe was used to break
the pedestaloffat the base.The inclusion,enclosed
within this small fragment of the host crystal, was
then transferred to a clean glass slide, using the
steelprobe, and coveredwith a secondglassslide,
producing a simplified version of the crushing stage
described by Roedder (1970). While the inclusion
was being observed under the microscope, sufficient pressure was applied to the top slide to
fracture the host crystal grain through the inclusion,
liberating the daughter crystals. A certain amount
of restraintwas necessarywhen applying pressure
to the samplechip, so that when the grain fractured,
the daughter crystals would not be crushed. After
separationof the slides, a finely-drawn glass fiber
was used to isolate the daughter crystals on one of
the slides.
In preparationfor mounting the daughtermineral,
a Lindemann glass capillary tube was set into a
Gandolfi camera sample pin and was drawn to a
diameter comparable to that of the crystal to be
mounted. This capillary tube thinning was accomplished by touching the free end of the tube to a
heatedplatinum wire and drawing the melted glass
to a fine fiber. The crystal was then mounted onto
the glass fiber *ith clear fingernail polish or a
solution of collodion in amyl acetate, a procedure
developed by Smith and co-workers (Smith and
Wein, 1974).The crystal mounting procedure requiredpractice,with resultsbeingoptimizedby the
use of the minimum amount of mounting medium
and the smallestworkable glassfiber diameter.To
minimize scattering of the X-ray beam by the glass
fiber, it was desirable to mount the crystal onto the
end of the fiber. If the mounted crystal was several
times larger in at least two dimensionsthan the fiber
diameterat the point of attachment,however, the
amount of glass scattering was not found to be
significant, and placement of the crystal at the end
of the fiber was not essential.
The mounted sample was then mounted in a
Gandolfi camera. Careful centering of the sample
within the camerawas absolutely critical for obtaining a pattern of the maximum sharpnessin the least
time. Employing CuKa radiation and a 57.3 mm
cameracassettewith operating conditions of 20 mA
and 40 kV, exposure times varied from three to six
days, dependingupon crystal density and volume.
It was necessaryto evacuatethe cameraand use an
X-ray film with a low fog level to minimize film
darkening due to the scattering of X-rays by air,
which would otherwise have rendered the film
unreadableafter such long exposures.
ZOLENSKY AND BODNAR: FLUID INCLUSION DA(IGHTER MINERALS
139
Applications
During the course of the development of the
technique described above, several solid phases
were extracted from fluid inclusions and analyzed.
In the early part of this study, inclusionscontaining
opaquedaughterphaseswere chosenfor examination because,once these crystals were extracted
from the inclusions,they could be easilyrecognized
among the host mineral debris. As the handling
techniquesimproved, attempts were made to isolate less distinguishable daughter phases whose
identitieswere not previously known.
Fluid inclusionsfrom porphyry copper deposits
commonly contain an opaque phase that is identified as chalcopyrite basedupon its triangular crosssectionand optical properties(Bodnar and Beane,
1980),and semi-quantitativeSEM analyses(Reynolds and Beane, 1979).Figure la shows two inclusions in qvartz from Copper Creek, Arizona, containing several daughter phasesincluding a triangular opaquemineral. This crystal was removed from
the inclusionusing the proceduredescribedabove,
mountedon a glassfiber (Fig. 1b), and a Gandolfi
pattern was obtained. Inspection of this diffraction
pattern showed that the crystal was indeed chalcopyrite.
Another common solid phase observed in fluid
inclusionsfrom porphyry copperdepositsis a translucent red, hexagonalcrystal identified as hematite
(Figs. lc and d). Extraction and X-ray analysisof
severalsuch crystals verified this identification.It
has not yet been establishedthat these hematite
crystals are in all casestrue daughter minerals, but
rather may result from auto-oxidation of the inclusion due to hydrogen leakagethrough the inclusion
walls.
Coexisting with the hematite-bearing inclusions
observed in quartz from Copper Creek, Arizona,
were inclusionshaving phaserelatigns similar to the
hematite-bearingtype except that they contained an
octahedralopaquemineral instead of a red, translucent mineral (hematite). Even though this mineral
did not respond to a magnet (this particular sample
was not subjectedto heatingtests), the crystal was
tentatively identified as magnetite. X-ray analysis,
however,proved this phaseto be hematite(variety
martite-a pseudomorph of hematite after magnetite). An SEM examination of this sample would
have revealed only the presenceof iron and, combined with the observedcrystal morphology, this
Fig. 1. (a) Multiphaseinclusionsin quartz from Copper Creek,
Arizona, containinghalite (H), triangular opaque(chalcopyriteC), a vapor bubble (V), and several unidentified phases'
Photographtaken in transmitted plain light. Scalebar represents
20 pm. (b) Triangular opaque (C) from inclusion shown in Fig'
la, now mounted onto glass fiber in preparation for X-ray
analysis. Photograph taken in reflected plain light. Scale bar
represents l0 pm. (c) Multiphase inclusion in quartz from
CopperCreek, Arizona, containinghalite (H), a translucent,red,
hexagonal plate (hematite-Hm), and a vapor bubble (V).
Photographtaken in transmitted plain light. Scalebar represents
20 pm. (d) Red hexagonalplate (Hm), from inclusion shown in
Fig. lc, mounted onto glass fiber. Photograph taken in
transmitted plain light. Scale bar represents 15 pm.
mineral would probably have been incorrectly identified as magnetite.
The origin of the martite cannot be determined
with certainty from the available data. However,
the random distribution of red, translucenthematite
in some inclusions and opaque martite in others
suggeststhat magnetitedaughtermineralsformed in
some inclusions but failed to nucleate in others.
140
ZOLENSKY AND BODNAR: FLUID INCLUSION DAUGHTER MINERALS
Hematite becamethe stable iron oxide phase as the becauseof the presentsize limitation of this proceinclusioncompositioncrossedthe magnetite-hema- dure.
tite phaseboundary, due to cooling and/or diffusive
Summary and recommendations
hydrogenloss. This causedalterationof the magnetite to hematite in martite-bearing inclusions and
The Gandolfi X-ray procedure described above
deposition of hematite in some inclusions in which yieldsunambiguousidentificationsof fluid inclusion
magnetitedaughter crystals had originally failed to daughter minerals and is generally applicable to all
nucleate. It is possible that auto-oxidation from crystals larger than 2-5 p.m in at least two dimendiffusivehydrogen loss in laboratory heating tests sions. This technique also permits different polycould also cause magnetite to alter to hematite. types to be distinguishedand pseudomorphismto
However, as describedabove, the samplecontain_ be recognized.The practical limitation of the mething these inclusions was not subjectedto heating od describedhere is the excessiveamount of time
tests.
required for each daughtermineral characterization
Many fluid inclusions from porphyry copper sys_ as compared to most other identification techtemscontaintwo clear, cubic daughtercrystalsand, niques.From one hour to severaltens ofhours may
most often, one of these phases is considerably be required to extract a daughter crystal and mount
larger than the other. It is generallyassumedthat it onto a glassfiber, ready for X-ray analysis.The
the larger crystal is halite and the smaller one is actual length of time dependsupon crystal size as
sylvite. Theseassignmentsmay be verifiedby heat- well as the physical and optical properties of the
ing tests, due to the significantlydifferent dissolu- daughtermineral relative to that of the host materition rates of these two phases with temperature al, i.e., hardness,refractive index, luster, crystal
(Roedder, l97l), or freezing tests, becauseNaCl habit, cleavage,etc. Furthermore, this method reforms a highly birefringenthydrate, NaCl.2H2O,at quires considerable manual dexterity and patience
low temperaturewhile sylvite does not (Roedder, on the part of the operator. All of these factors
1963).The larger of two such crystals in an inclu_ affectthe successrate, which can be expectedto be
sion from the Copper Creek porphyry copper sys- initially quite low (approximately l}Vo) and which
tem was removed and analyzed. This mineral will probably never exceed about 50Vo.
proved to be sylvite, revealingthat the larger ofthe
Handling techniques,the limiting factor in this
two clear, cubic daughter phasesin porphyry cop_ study, could be greatly improved with the use of a
per inclusionsis not always halite and that the trVNa better drilling arrangementand the introduction of a
ratio of the mineralizing fluids in this casewas very micromanipulator. In addition, the extremely long
high as comparedto most porphyry copper deposits X-ray exposure times required could be at least
(Bodnarand Beane, 1980).If the K/Na ratio of this halved with the use of commonly available highinclusion had been determined by assuming the intensityX-ray generators.Suchgeneratorsmay, in
larger crystal was halite without verification from fact, be necessaryfor the successful identification
heating and then measuringthe volumes of the two of mineralscontainingpredominently low scattering
phases,an incorrect value would have been ob- elements,i.e., elementslighter in electron density
tained, although it is unlikely that any workers than sodium.
would use this procedure.
A principal advantage of the present method is
Topazfrom Volnya in the U.S.S.R. is particular- the possibility of identifying crystals from individly well known for the large daughter phase popula- ual fluid inclusions that have been previously subtions of its fluid inclusions (Zakharchenko, l97l). A jected to heating/freezingtests. However, it must
massof brownishplates,approximately20 microm_ be independentlyverified that the heating andfreezetersin diameter,from one multiphaseinclusionin ing procedureshave not affected the daughter crysthis topaz was extracted and found to be the 2M1 tals.
polytype of phlogopite. This mineral had not previ_
Acknowledgments
ously been described from this locality. Under
We would like to thank E. Roedder and R. E. Beane for
1000x slagnification several reddish-brown laths
approximately one micrometer in length could be providing samplesused in this study. In addition, E. Roedder
distinguishedwithin the phlogopite. An X-ray pat- provided many useful suggestionsfor the improvement of this
research.We also thank D. K. Smith, W. C. Kelly, J. B. Murtern could not be obtained from these crvstals. owchick, L. Anthony, T. J. Reynolds,
and R. C. Newton for
ZOLENSKY AND BODNAR: FLL]ID INCLUSION DAUGHTER MINERALS
Roedder, E. (1972) Composition of fluid inclusions' In M.
Fleischer,Ed., Data of Geochemistry,6th ed., U'S. Geological Survey ProfessionalPaper 440JJ.
Roedder,E. and Coombs,D. S. (1967)Immiscibilityin granitic
melts, indicated by fluid inclusions in ejected granitic blocks
References
from AscensionIsland. Journalof Petrology,8,417-451.
Bodnar, R. J. and Beane, R. E. (1980)Temporal and spatial Rosasco,G. J. and Roedder, E. (1979)Application of a new
microprobe spectrometerto nondestructiveanalysisof sulfate
variations in hydrothermal fluid characteristics during vein
and other ions in individual phases in fluid inclusions in
filling in preore cover overlying deeply buried porphyry copminerals. Geochimica et CosmochimicaActa, 43, 1907-1916.
per-type mineralization at Red Mountain, Arizona. Economic
Rosasco,G. J., Roedder,E. and Simmons,J. H. (1975)LaserGeology,75,876-893.
excited Raman spectroscopyfor nondestructivepartial analy(1971)
Dawsonite
as
a
daughC.
Coveney,R. M. and Kelly, W.
sis ofindividual phasesin fluid inclusionsin minerals. Science,
ter mineral in hydrothermal fluid inclusions. Contributions to
190,557-560.
334-342.
Mineralogy and Petrology, 32,
Smith,D. K. andWein, W. (1974)Experiencewith microparticle
Gandolfi,G. (1967)Discussionupon methods to obtain X-ray
identification-a coupled efort of X-ray diffraction with many
"powder patterns" from a single crystal. Mineralogica et
other methods. (abstr.) American Crystallographic AssociaPetrographicaActa, 13, 67-74.
tion Winter Meeting Program and Abstracts, 6, 32.
Essene,
E.
J.
B.
E.
and
W.
C.,
Nesbitt,
Metzger,F. W., Kelly,
Vorob'ev, E. L and Lozhkin, V. I. (1976)The identificationof
(1977)Scanningelectron microscopy of daughter minerals in
daughter minerals of fluid inclusions by means of X-ray
fluid inclusions.EconomicGeology,72, l4l-152.
coloration. Ezhegodnik Sibirskoe Institut Geokhimii, 281-283
Reynolds, T. J. and Beane, R. E. (1979) The evolution of
(in Russian).Translatedin Fluid Inclusion Research,Proceedtime
at
the
Santa
through
characteristics
hydrothermal fluid
ings of COFFI, 1976,9,253-254.
Rita, New Mexico, porphyry copperdeposit.(abstr.)GeologicondiA. I. (1971)On time and physico-chemical
Zakharchenko,
ll,
502.
with
Programs
Abstracts
Society
of
America
cal
,
tions of mobilization, transport and precipitation of tungsten
Roedder,E. (1963)Studiesoffluid inclusionsII: freezingdata
and tin in postmagmaticprocesses(exemplifredby intragranitand their interpretation.EconomicGeology, 58, l61Jll.
ic chamber pegmatites).In Mineralogy and Geochemistry of
Roedder, E. (1970)Application of an improved crushing microTungstenDeposits, p. 287-306.Leningrad University Publishgases
Schweifluid
inclusions.
the
in
to
studies
of
scopestage
ing House (in Russian). Translated in Fluid Inclusion Rezerische Mineralogische und PetrologischeMitteilungen, 50,
search,Proceedingsof COFFI, 1973,6, 191-194.
l,4l-48.
their critical reviews of this paper. Partial funding was provided
by The Experimental Station of The College of Earth and
Mineral Sciences,The PennsylvaniaState University.
Roedder,E. (1971)Fluid inclusionstudieson the porphyry-type
ore depositsat Bingham, Utah, Butte, Montana, and Climax,
Colorado. Economic Geology, 66, 98-120.
Manuscript received,April 16' 1981;
acceptedfor publication,September2I ' 1981.