Canadian MineralogistYol. ?.4, pp. 523-528 (1986)

Ansrnecr

Besides gold, platinum-group-element nuggets andmicrospherules have been encountered in heavy-mineralconcentratg panned from alluvium in tlethree rivers Mies-sijoki, Puskuoja and Sotajoki, and in tle Hlirkiiselkli areain northern Finland. We have found abundant compositePt-PtAs2 spherules or roundish grains. Compositionally,these form a series from pure sperrylite ptAs2 to inter-grown'As-bearing Pt and PtAsz, Morphologically, withthe decrease in As, sperrylite ch4nges from idiomorphiccryslals to perfect spherules showing a eutectoid intergrowthof Ft and FtAs2, These spherules and roundish grainsform during the drying of the last (heaviest) fraction in theminers'pans.

Keywords: Pt-PtAs2 alloy spherules,,,spherulization,,,sperrylite, placer, Lapland, Finland.

Sotrwernr

On trouve, en plus de I'or, des pepites et des microsph6-rules enrichis en 6l6ments du groupe du platine dans les con-centres de minfu4rrK lourds alluvionnaires des trois rivie-res Messijoki, Puskuoja et Sotajoki et dans la region deHiirkiiselkii, dans le Nord de la Finlande (Laponie). Il ya une abondance de sph6rules et de grains arrondis. Ils for-ment une sdrie allant de sperrylite pure (ptAs, i une inter-croissaoce de platine arsenifdre et de sperrylite. A mesureque Ia teneur €n As diminue, la sperrylite change de sa formeidiomorphe originelle i celle d,un sphdrule parfait i tex-ture eutedoiae. Ces sph6rules et grains arrondis se seraientform€s au cours de la dessiccation de la dernibre fraction(a plus dense) dans la cuvette des prospecteurs.

(Iraduit par la Rddaction)

Mots-clds: sphdrules, aliage Pt-PtAs2, .,sph6rulisation",sperrylite, placer, Laponie, Finlande.

INrnoougrroN

Placer gold has been recovered since the last cen-tury from river beds in the granulite complex of Fin-nish Lapland. Heavy-mineral concentrates suppliedto us by several minsls working deposits of placergold along the Miessijoki, Puskuoja and Sotajokirivers and in the Hiirkiiselkii area Qocalities shown

MAN-MADE Ft-PtAs, SPHERULES AFTER SpERRyLtTEFROM ALLUVIAL DEPOSITS IN FINNISH LAPLAND

YRJO WORELAINENMotinkatu 28, SF-02230 Espoo, Finland

RAGNAR TORNROOSGeological Sumey of Finland, Stenkarlsviigen 1, SF-02150 Esbo, Fintand

on the map in Fig. 1) represent several thous4ndtonnes of river gravel.

The granulite complex of northern FinnishLapland forms an arcthat extends north-$outh fromthe Norwegian border and intersects the Russianborder in an east-west direction (Fig. 1). This arcis bordered in the southwest by Archean greenstones,quartzo feldspathic schists and gneisses.

About 140,0fi) individuat platinum-group€lement(PGE) grains were separated from the materials byone of us (Y.V.) Approximately 9590 of them aresperrylite gf,ains or crystals, whereas the remaindercomprise many different platinum-group mineralsand alloys. Together with some very densemicrospherules, abundant Pt-PtAs, spherules androunded Pt-PtAs, grains have also been encoun-tered; these spherules and grains are the subject of

Ftc. l. Map of northern Finnish Lapland showing the allu-vial deposits: I Miessijoki, 2 Puskuoja, 3 Sotajoki,4Hiirkiiselkii area.

NORWAYNORWAY

F I N L A N D

., 1'f:) i"};

USSR

F I N L A N D

s23

5U THE CANADIAN MINERALOCIST

q{FrtE E 6 + 4

$f$r-s

F#**$ijffj

FIG.2. A series of secondary-electron images showing different stages of spherulization of sperrylite beginning with(A) an unaltered crystal (ength of bar 100 pm) and ending with (D) a complete Pt-PtAs2 spherule 0ength of barl0 pm).

Pt-PtAg SPIIERULES FROM ALLUVIAL DBPOSITS IN FINLAND 525

the present paper. The concentrates also includenumerous other artificial spherules (scrap products,welding sparks), which have to be "screened" out.Sperrylite crystals had earlier been described fromthese areas by Koivisto et al. (1980), but no men-tion was made of As-depleted gains.

Spherules, and especially PGE alloy spherules,have long interested scientists, and their origin hasbeen a subject of debate (e.9., Bird & Bassett 1980,Blanchardet al. 1980, Brownleeet al. 1984, Feather1976). This paper contributes to the discussion byshowing that the Pt-PtAs2 spherules and roundedgrains found in Finnish Lapland are man-made inorigin.

MareRrAm aNo Mlcnopnoss INvEsrtcATroNS

The artificial spherules (welding sparks, scrapproducts) axe - 100 4n to I mm in diameter, whereasthe PGE alloy spherules are 85 to 200 pm, and theperfectly spherical Pt-PtAq spherules are slightlysmaller (50 - 100 pm). The rounded Pt-PtAs,grains, on the other hand, are somewhat larger (150- 3@ pm).

Chemical analyses were conducted and scanning-electron photographs (SEM) were taken using aJEOL JCXA=733 microprobe equipped with aFrinceton Gamma-Tech energydispersion sp€ctrom-eter. The following standards were used: syntheticcompounds Pdr r SbrAsr, I{hSb, Ptl.0Fe0.6aCu6.36 forPd, Rh, As, Sb, Pt, Fe and Cu, the natural minerals.lsellingite for As, and laurite for Ru and S. The X-ray hqes measured were Kcu for Fe, Cu and S, andLo for the others. The accelerating voltage was setat 30 kV, and the probe current at 1 nA.

SPHERULIZATIoN oF SpnnnyrlTs

The roughly 133,000 sperrylite grains includeseveral more-or-less rounded grains with a markedvariation in proportion of Pt and As. Idiomorphiccrystals have the sperrylite composition PtAs, or arenear to it, but crystals that appear porous on the sur-face show a deficiency in As content. These crystalsare roundish, becoming more so with decreasing con-tent of As. Figure 2 shows such a series from an idio-morphic crystal to a complete As-deficient spherule.

Under the microscope, the As-bearing platinumcan be seen to form a eutectoid (vermicular) texturewith sperrylite (Fig. 3), beginning at the surface ofthe crystal. With the continued decrease in As, thePt and PtAs2 form a eutectoid texture, and theprocess ends with a spherule. The process is illus-trated in Figures 2 and3. Figure 2 shows a series ofSEM photographs depicting different stages in the"spherulization" process, beginning with an idio-morphic crystal of sperrylite and ending with a per-fect eutectoid Pt-PtAs2 spherule. The

IAELE I. C}IEMICAL COIPOSITION OF Pt.As PHASES OF "SPHERULIZED"SPERRYLITE FROM ATLWIAL DEPOSITS IN FINNISH LAPTAND

P t u t . t 5 5 . 0 8 0 . 0F e 0 . 4 1 0 . 4 6c u 0 . 1 1 6 0 . 4 8R b 0 . 4 4 1 . 6 0A B \ 2 . 9 I 4 . 3lo ta1 99 .21 96 .8q

Atoelg per ootrt

Pr 32 .28 64 .86F e 0 . 8 1 1 1 . 3 0c u 0 . 8 2 1 . 2 0n h 0 . u 9 2 . 4 5As 65 .5 '1 30 .19

96.9 53 .90 . 2 5 n d0 . 3 0 ! d1 . 2 6 0 . 9 10.97 43 .2

9 9 . 6 8 9 8 . 0 1

s : . 8 54 -6Dd dd!d !d

0 . 0 7 0 , 9 34 3 . 3 4 3 . 0 .

9 7 . 9 7 9 8 . 5 3

3 1 . 9 9 3 2 . 4 6

olgg r]o,r6 ? . 0 3 6 6 . 4 9

9 3 . 5 4 3 2 . 0 60 . 8 50 . 8 92 . 3 O 1 . O 22 . \ 3 6 6 . 9 1

!d

1 )

4 )

, )6 )

: oot deteoted

sperryl l te crysial , uraffeoted, aa 1n FiB. 2A.

Inleraedlaie phase, I Corrgapond io polntgAhost puro Pi pbase, I narked lD Flg. 3.Sperryl i te reEnant. I

Sperryl i le oryatal f roD saod dr led by 8€aiLo beat iDg.speryyl lbe relnani af ior aperryl l te oryatal f ronsand drtsd by ht€lae beai lng. Corneapouda bo polntl n F l g . 5 .

backscattered-electron images of polished sectionsin Figure 3 show various vermicular textures in thealtered sperrylite crains and spherules. Results ofanalyses of the various phases are given in Table IQ - 4 ) .

One process that could conceivably release arsenicand produce alterations such as those shown inFigure 2 and the textures in Figure 3 is heating inair. The variable degrees of rounding imply that thegains have been exposed to a range of either hightemperatures or heating times, or both. The ques-tions are: could heating possibly produce spherulesfrom sperrylite grains by the release of arsenic, andwhat has caused the heating?

One answer might be a forest fire, but because thematerial is recovered by panning several thousandtonnes of gravel, this can hardly be the main expla-nation. In fact, the panning and procedures used bythe niners suggest a more plausible one: the last, veryheavy fraction in the pan has to be dried, and tohasten this, miners commonly use a blow torch.Could this heating release arsenic and also be respon-sible for the textures?

We checked the possibility with some experiments.First, DTA experiments (AT = 5',/min) conductedon sperrylite crystals give a very strong exothermicreaction aI 4A8-572"C. In the temperature range350-900oC, simultaneous thermogravimetric deter-mination gave a total weight loss of 42.25t/0, whichtakes place in three steps. In the temperature inter-val 350-408'C, the weight loss was 0.8590, in theinterval 572-9W"C, 0.40V0, and during the strongexothermic reaction it was 41.0090. The total lossin weight is equivalent to the As content in sperry-lite. Thus a blow torch certainly can produce enoughheat to release arsenic.

Electron-microprobe sa4minatisn of the grainsafter the DTA-TG experiments revealed that As wasno Ionger present; only Pt was detected. Thus in this

526 TI{E CANADIAN MINERALOGIST

FIG. 3. A) and B) show backscattered-electron images of various eutectoid (vermicular) textrures in altered sperrylitegrains. C) backscattered+lectron image of a section though a Fl-PtAs2 spherule. A small domain of unaltered sper-rylite still remains. Numbers correspond to compositions in Table l. 2) intermediate phase, 3) almost pure Pt phaseafter sperrylite, 4) relict sperrylite.

temperature interval all the arsenic is released fromsperrylite. A backscattered-electron image of apolished section of such a grain is seen in Figure 4.The grains reveal their original morphology, withonly a sligbt rounding of the edges. The interior ofthe grains, however, is full of cavities as a result ofthe loss of As. No spherules were foundt

The next experiment was performed to reproducethe procedure used by the miners. Some sperrylitecrystals were mixed with wet sand, and the mixturewas their dried with a blow torch. In the first attempt,when the sand was heated until it was just dry, onlya slight reaction was visible on the surface of the sper-rylite crystals. The second drying was prolonged untilthe pan was glowing red (-5(X)'C). Now the sper-rylite crystals all became more or less round. Figure5A shows a secondary-electron image of one of thesespherules, and Figure 58 a backscattered-electronimage of a section through one of them. The ver-micular texture seen is similar to that in Figure 3C.We conclude, therefore, that the Pt-PtAs2 spher-ules are man-made and were formed during the dry-ing of the last (heaviest) fractions.

DISCUSSIoI+

An interesting finding in our experiments was that

527

Ftc.4. Backscattered-electron image of polished sectiontlrough sperrylite crystals after DTA experiment, show-ing a central part full of cavities surrounded by an outerunbroken rim. In no part of the remainder is any Asdetected. Length of bar 100 pm.

sperrylite crystals form spherules when heatedtogether with sand using a blow torch, but remainmorphologically more or less intact when heated(without sand) more slowly in a DTA-TG appara-

ft-PtAv SPI{ERULES FROM ALLT.MAL DEPOSITS IN FINLAND

FIc.5. Scanning-electron photographs of Pt-PtAsz spherules formed by heating with a blow torch in the laboratory.A) Secondary+lecron image of spherule, B) backscattered-electron image of a section througb a spherule. The eutectoidtexture corresponds to that illustrated in Figure 3. The composition of the relict sperrylite (6) is grven in Table l.Lengti of bar l0 pm.

528 THE CANADTAN MINERALOGIST

tus, developing, however, cavities iaside Xhe singlg- (Geological Survey of Finland) read the manuscriptcrystal grains. This may be due to the heating rate, and gave valuable comments. Miss Taina Koivistowhich is low (AT = Jo,/min) in the DTA-TG experi- .(Geol. Surv. of Finland) drew the figures, and Mrs.ment but high and very sudden with a blow torch. Gillian Hikli corrected the Engilish.

The Pt-PtAs2 spherules are smaller than therounded grains probably because the heating did notlast long enough to totally "spherulize" the larger RsFERENcsscrystals; only the small ones were transformed. In

*:]*::"9y":P:lT:ott' "large" (> 2@ pm in Brnp, J.M. & BessHr, W.A. (1e80): Evidence of a

diameter) crystals of sperrylite were also used, and "'?"i.funelfrtry'interresiria 6smium_iridium_

the heating had to be prolonged to totally "spheru- ilfi;il.1"t;. j. Geophys..l(es. E5, S46t-5470.lize" all of them.

Many scientists have pondered the origin of PGE-alloy microspherules. Some have proposed anextraterrestrial origin (ablation droplets: e.g., Brown-lee et al. 1984), whereas others favored desulfuriza-tion (reduction) of PGE sulfides (Stockman & Hlava1984). Feather (1970 described the texture of PGEmicrospherules from the Witwatersrand; he showedthat Os-Ir-Ru alloys inay alter into sulfarsenides,which would begin at the grain surfaces, and thusmay acquire a vermicular texture.

The outcome ofthe process described by Featheris the reverse of that seen in the Ft-FtAs2 spherulesfrom Finnish Lapland. Ablation droplets should notshow any eutectoid (vermicular) texture, and hencethis mode of origin is not applicable to our spherules.

AcKNowLEDGEMENTS

We are indebted to several minslg, who kindlyplaced their material at our disposal. Dr. Louis J.Cabri (CANMET, Ottawa) and Dr. KariKinnunen

Br,aNcnanp, M.8., Bnowru,er, D.8., Bmtcs, T.E.,Honcr, P.W. & Kvrs, F.T. (1980): Meteoroid abla-tion sphoes from deep-sea sediments. Earth Planet.sci. Lett. 6, 178-190.

Bnowr.rI,re, ,D.E., Beres, B.A. & Wnpet.ocr, M.M.(1984): Extraterestrial platinum group nuggets indeep-sea sediments. Nature 3W, 693-695.

Fnernpn, C.E. 097A: Mineralogy of platinum-groupminerals in the Witwatersrand, Soutl Africa. .Ecoz.Geol, 71, 1399-1428.

Kovrsro, H., Vuoner,anm, Y. & Sanaue, T.G. (1980):Alluvial sperrylite from Finnish Lapland. Mineral.Record U, 303-305.

SrocrueN, H.W. & Hr.eva, P.F. (1984): Platinum-group minerals in alpine chromitites from south-western Oregon. Econ. Geol. 79,491-508.

Received April 8, 1985, revised manuscript acceptedJanuary 20, 1986,