MINERALOGICAL NOTES 28r

BIr,ruenot, H. W. (1969) Synthesis of lead pyrosilicate and other barysiliteJike

compounds. Amer. Mineral. 54, 510-b?l.
Bonnoeux,D., aNo J. Le.rznnowrcz(1969) Synthbsede la barysilite pb"SLOr.Bull.
Soc. F ranc. M iner aI. C ristallb gr. 92, gg3-3gb.
Gr,essnn,F. p. (1964) New data on barysilite. Amer. Mitwra;J.49, 14gfF14gg.
ho, J., exn C. FnoNnnr,(1967) Synthesesof lead silicates: Larsenite, barysilite
and related phases.Amer. Mineral. 52, l0ZZ-IOg4.
Lttznnowrcz, J. (1966) Etude par diffraction des rayons X et absorption infra-
rouge de la ba.rysilite, MnPb".3Si,Oz, et de compos6s isomorphes. .Acla
C ry stal\ogr. 20, 357-363.
Pnrrnn, W., eNn A. B. Eenxrr (1g71)Der Strukturtyp desBarysilit, Xyr(pb,Si,Or)*.
F or tschr. M ineral. 49, Bg-40.
Pnmnn, W., A. B. ElenNrr., axo U. Knppr,pn (1971) Die Kristallstruktur von Blei-
Barysilit, Pb"Si,O?.Z . Kris tallo gr. 133,44E4EB.
SrraxNow, R. D., er.rl C. T. Pnnwrrr (lg6g) Effective ionic radii in oxides and
fluorides.Acta Crgstallogr. BZ5, 925-946.

Amcrican Mineralogi,st
Vol 57, pp 281-283 (1972)

SAMPLEPREPARATION FORX-RAY FLUORESCENCE

ANALYSIS:LI-BORATEGLASSDISKS
Wrr,r,reuLooorNoANDDoucLAsW. Rrrsrr, Rutgers Oniuersity,
DepartmentoJGeology

Inrnonitcrron
X-ray fluorescence analysisis a widely acceptedtechniquefor ob-
taining rapid chemicalanalysesof geologicsamples.problems have
arisen,however,in samplepreparationthat is fast, reproducibleand
inexpensive,and which producesa durable, long-lived standard or
sample.
claisse (1956) pioneerdthe flux-fusion technique for preparation
of X-ray fluorescencesamples.Subsequentmodifications by Rose,
Adler, and Flanagan (1g62) have been widely adopted. However,
fluorescence samplespreparedby briquetting crushedgrassboulesare
susceptibleto contamination,are subjectto deterioration,and require
finely-machineddies and a laboratory press.A fusedglassdisk sample
preparationhas been describedby Norrish and Hutton (lg6g), but
the writers experienceddifficulty preparing disks that did not crack
upon remoyal from the graphite mold. The writers' proposedtech-
nique is similar to that developedby stephenson(1969), but elimi-
nates the oven annealing processdesignedto relieve stress within
282 MINERALOGICAL NOTES

the glassdisk sample.we also reducedthe LiBoz-to-sampleratio to

approachthe eutectic compositionin the system LiOz-BzOg-SiOz'
PnPPenertoN
Se,rupr,n
For major elementanalysisin silicate rocks a lithium- metaborate
flux is preparedin the foliowing proportions,and thoroughly mixed:
LiBO, 87.00
LarOa 12.20
LiNOs 0.80

100.00 wt Percent
Rock samplesare finely ground and mixed with the lithium metabo-
rate flux as follows:
FIux 3'000 g
Powderedrock samPle 0'450 g'
For Na and Mg analysis LiBOz rlisks are fused without heavy
absorber.Here we use a sample-to-LiBOz ratio of 1:1'

the glassdiskas it cools.SeeFigure 1'

a flat
An aluminum quenching plunger should be prepared with
approxinlateweight of 200 grams'
bottom, an OD of 27 mm and'
Aflux-samplemixtureisplacedinagraphitecrucible'andfusedi

/l\ r
I /t\
ON
@@

I
,1, -ln_

p--6Omm-->p-7Omm
50mm----->{

plunger'
Fig. 1. Dimensional view of graphite crucible and
 MINERALOGICALNOTES 2{]3

an oven at 1,000"Cfor 4 minutes.'Whenremovedfrom the oven,the

crucibleand molten sampleshould be swirled onceor twice to ensure
homogeneityand set on a horizontal surface to cool. While bright
cherry-red,the molten sampleis lightly quenchedwith the aluminum
plunger until the boule retains the form of a flat-topped glass disk.
Quenchingtime shouldbe minimizedto enhanceannealingof the glass
disk. When thoroughlycool,the crucibleis inverted and tappedlightly
to freethe disk.
The glassdisk may be embeddedin plastic by placing it facedown
in a metal ring mold and filling the mold with castingresin (compar-
able to "Cadco Casting Resin," Cadillac Plastic & Chemical Co.,
Detroit, Michigan). When cured, the disk is ground until the full
diameterof the glassdisk sampleis exposed. The exposedglassis then
readily polishedto give an optimum surfacefor irradiation.
Glass disks preparedin this manner are tough and homogeneous.
Electronmicroprobescansacrossglasssurfacesshowedno appreciable
variations in composition.Samplesembeddedin plastic are nearly
indestmctible.Surfacesfor irradiation may be quickly and repro-
ducibly preparedon all samples.Unlike boric acid pelleLs,glassdisks
apparently do not deteriorate with time; after six months of constant
use, X-ray fluorescencestandards made with USGS Standards II
consistently generate reproducible calibration curves of exceptional
fit. If contaminated,glass disk samplesmay readily be restoredby
polishingaway any contaminatedsurface.
With 8 or 10 crucibles,any number of sampledisks may be rapidly
preparedin assembly-linemanner.

RnrpnoNcns
Cr,erssn,F. (1956) Accurate X-ray fluorescenceanalysis without internal btandard.
QuebecDept. Mines Preli,m.nep.927.
Nonusrr, K. exo J. T. Hurrox (1963) An accurate X-ray spectographic method
for the analysis of a wide range of geological samples. Geoch.hn.Cosmoch'im.
4cta,33,431453.
Rosn, I[. J., f. Anr,en eNl F. J. Fr,er'recew (1g62) Use of LazOe as a heavy ab-
sorber in X-ray fluorescenceanalysis of silisate rocks. U.8. Geol. 9uru., Prof .
Pap. 45v8,80-82.
Srorrrnrsow, D. A. (1969) An improved flux-fusion technique for X-ray emis-
sion analysis.Arnl. Chem.41, 966-967.