An electron microscopic study of anorthoclase spherulites

An electron microscopic study of anorthoclase spherulites

An electron microscopic study of anorthoclase S phen es I~flZUHIKO AKIZUKI LD HOS ,,,,1015:0024.4937.An ,9,,3,, c'lectron microscopic study of anor...

3MB Sizes 9 Downloads 53 Views

An electron microscopic study of anorthoclase S phen



LD HOS ,,,,1015:0024.4937.An ,9,,3,, c'lectron microscopic study of anorth,~lasc :~phemlites. Lithos, Vol. 16, pp. I

Sph~rulites consist~mg of fibr:ous alkali feldspar and silica minerals a~re produced by devitdf~cation of rlttyolite glass und~r hydrothermal conditions. The alkali feldspars (At~.~Orz~.oAr~.s, Abai.70~'14.oAn4.3) ~a spherufites from two Iocal~ities in Japan consist of triciini¢ anarthoclase showing fine cro~;s-hatched twinning and raoncclinic sanidine showing fine cress-hatching not attributable to twinni3g. The crosshatching, which corresponds to albite and periciine twinning, is produced in tne process of transition from a rnonoclinic to a triclinic phase. The spherul~te may develop at a temperature lower than about 200°C because the co-exi:ifing silica mineral is not quartz, but meta~aable tridymite. Acco~rding to the phase diagram of the alkali feldspars by MacKenzie (~952), the alk~i feldspars should have been triclinic during growth. However, the textures show that the alkali feldspar grew as a disordered monod_inic phase. Because of the high growth rate, the AI/Si dis3rdered stnmture was ~roduced during growth and afterwards tran~sfovmed into a triclinic s,~ructur¢ ~ith cross-hatched twinning.

Mizuhiko Akizuki, Institute of Mineralogy, Petrolo,~ and Economic Geology, Faculty of Science, Tohoku University, Sendai 980, Japan; 3rd February, 1983.

Spheralites consisting: of alkali feldspar and silica are formed by devitl~ication of acidic volcanic glasses. Lofgren (1971)produced ,~,,,, "~'~'~,'°':~....... on textures consisting of spherulites in natural rhyorite glass by hydrothermal treatment at 340700°C and 0 . 5 - 4 Kbat. Na, ural spher~ites from various local'ties have been studied by both optical microcopy and X-ray powder methods. Ta n nida (1961) described spherulite textures in weld-, ed tuff and rhyo|ite obsidian from Japan, and found the spherulites to be composed of sanidine, anorthoclase and silica minerals. Because the ,crystals are very fine an0 X-~my diffraction pe;.~:s are rather broad, detailed identification of the. minerals and observation of the texture are difficult. MacKenzie (1952) studied the transition tempezature of synthe~;,c, sodium-rich feldspars and natural anorthocla,~e and showed that the crosshatched '~vinning of natural anorthodase results from initial crystallization with monocrinic symmetry and subsequent transition to triclinic symmetry during cooling. Gray & Anderson (1982) studied tihe twin texture in a centimetre-sized anorthoclase megacryst from a basalt cinder and a weathered lava between crossed polars. A'.*though some d~fieulties are encountered in naming sodium-rich alkali feldspar, the names ano orthoclase and smfidine are used in this paper for the tricli'uMcand monc~:linic polymorphs, respectively, The objective of the present study was to ob 16 - Lithos4/83

serve the texture of alkali feldspar in spherurites by t~ransmission electron microscopy and to discuss the orion of the texture.

Specimens Sphcrulites were observed in obsidian, perlite or welded tuff from Hosak~, Shirataki, Wadatoge, Haneyama and other localities, Japan. Sphe~mlites originate at fine cracks in welded tufts from Wadatoge and Haneyama, while the cracks are not observable ifi the specimens from Hosaka and SMrataki. Optical wicroscopic rings, socalled 'liesegvng' rings, of amorphous silica (see Henis,-h 1970) are distinctly developed on the sphemlite from Haneyama (Fig. 1), are poorly observed in the speciimens from Shirataki and Wadatog~ and are not found in the Hosaka specimen. Transmission electron microscopy (TEM) of alkali feldspar in the four specimens :;hows similar textt, res and structures. The two specimens ~om Hosak~ ~md Shirataki were studied in deto21. H o s T k a s p e c i m e n . - Many sphev~,.~te,~:, several centimetres in tfiameter, are found in perlite from ~'he Hosaka opal mine, Fukushima Prefectufa, .Iaran. q'he spherdites include common op~ds and a few precious opals at the core (Fig. 2 A). TEM show~.d that the common opal con-


Mizuhiko Akizuki


16 (]19:83)

Observations ~ Ii~ "i~


i ~,~:~ ",~

• '

19] ! m,,'~, " " ~ _

big. I. Optica~ micrograph of th~in section olf spheruiite with licscgang rings. Deep coloured stripes are rich ~n amorphous or low crystalline silica. Plagioclase (pl) is inclad,ed in the spheru-

litc. Plane polarized light, tlaneyama,

sists of acicular, high t.ridymite (Akizuki & Shimada 1979). The chemical composition of the minerals in the spherulite analysed by electron probe mic~oanalyzer is feldspar: 75.5 weight % (AbT.~ 5Or2~0Anas)and silica: 24.5 weight %. Shirataki specimen. - Several spherulites, with a diameter smaller than 3 cm, are found, in black obsidian from Shirataki, Hokkaido,. Japan (Fi~;. 2B). The spherulites used for the study were those in the same hand specimen that Tanida (1961) studied by chemical and X-ray analyses, According to Tanida (1961), the chemical cornpo.sitior, of the minerals in the spherulite is feld.,,par: 66.1 weight % ~[Ab~t 7Orla0An4 3) and silica: 33.9 weight %.

Fibrous crystals radiate from the core to the rim of each spherulite. The thin se,ctions were prepared in a direction roughly paralk:l to the fibres a,ad were thinned b'y ion bombardment to facilit~tte observation under TEM. Two kinds of alkali feldspar were identified by electron diffraction. One exhibited a monoclinic diiffract~on pattern which is accompanied by triclinic, diffuse diiffraction spots of albite t~Anning (Fig. 3 A). The monoclinic structure corresponds to AI/Si disordered sanidine. The texture consisted of fine crosshatching the directions of which correspond to those of albite and pericline twins, and were similal to the cross-hatchings of orthoclase. In Fig. 3 B, pericii~ae twinned lamelllae occur in tlae upper and lower areas while albite twinned lamellae are dominant in the centre and the two types of lamellae interpenetrate at the boundaries. Although cross-hatched twinning of maximum microcline usually produces a Z-shaped reflection in the diffraction pattern (Laves 1950; Ak[zuki i[972), such a reflection iis not observed in the present specimens. Both albite and pericline twins show superlattices in some crystals (Fig. 3 A~,;~. The otlaer type of alkali feldspar sho~,s a tri.. clinic diffraction pattern consisting of albite and periclinte 'twinning and corresponds to an(~.rtho.. clase with an Al/Si highly ordered structulre (Figs. 4~A and 4B). The texture consists of fine albite and pericline twins which are intergrown with each other (Figs. 5 and 6). The cross-hatched twinning of microcline can be observed in a thin section parallel to (001).. w!hi!e the cross-hatched twinnin~.~ of anorth(3,cla,,;e

Fig. 2. Optic~ rrficrographs of thin sections o,f spheruli'tes seen between crossed potars. [A] The Hosaka mine., [B] SMralaki.

LITHOS 16 (1983)

Fig. 3. [A] Electron diffraction pattern of monoclinic sanidine with triclinic, diffuse spots and [B] micrograph (bright field) showing fine lamellae parallel or normal to the b-axis. The Hosaka mine.

is observable in a (100) thin section (MacKenzie 1956). Direction of b* and c* axes can be determined directly in the b*--c* diffraction pattern in the case of monoclinic sanidine, because of the difference between d020 and d0o2 lattice dimensions. However, both dimensions are very similar in triclinic anorthoclase and therefore the crystal has to be rotated around the b* or c* axis in order to determine the direction. Fig. 4 B shows the diffraction pattern rotated around the b* axis

Anorthocla~e spherulites


~[~ ~ ! :i:i~!:,~ i~ ~....... ~ 7 ..........

of Fig. 4A, from which the crystal orientation of Figs. 4 A and 5 was determined. It seems that one kind of lamella in the cross-hatching in monoclinic sanidine changes continuousl,¢ du~ng cooling to polysynthetic albite or pericline twinning in anorthoclase. The twin components shown in Fig. 6 have pericline twin-like fine inmellae crossing the albite tv.,hls, though the diffraction pattern does not shc, w twinned ~;pots but streaks: this texture is similar ~!o that of maxi-

Fig. 4. Electron diffraction p,attems of anorthoclase showing both albite and pericline twinning° The Hosaka mine (see details in text).

252 Mizuhiko Aki2ruki

Lrntos 16 (1983)



~i~~ ........ .:~ Fig. 5. Elecltro,i micrograph (bright field) showing anorthoelas¢ from the Hosaka mine width albite and perictirtc twinlaing, Corresponding diffraction l~,attern, which is eorrec.~iy oriented with respect to this micrograph, is shown in Fig. 4A. i

mum microcline (Akizuki 1972, Fig. 3). Crosshatched t~4~aning is common along the boundary between albite and pericline twinned areas in the anorth~:l~se, while some maximum microclines do not show such a cross-hatched twinning under TEM (Akizuki 1972).. In the diffraction patterns of the area with the cross-hatched twinning, lines joining two twinned spots in OkO and 001 reflecfinns are s!ightly tilted with respect tG the normal arrangement of the twinned spots, because of lattice distortion caused by tb-. cross..hatched twinning. These tiltings have been observed in some M-twinned microcline by precessio~i X-ray photographs. (Laves & Soldatos 19(:3). The dit-





~ Fig. 6. Electron micrograph (,bright field) shooting ~orlhoela~ from the Hosa~ka mine with albite and pericline twinning, q'he oriemation is the same as that ot Fig. 5 (see details in text)

fraction contrast due to the strain is strong along the boundary between albite- and periclinetwinned areas. The weakly strained areas near the boundary exhibit a weak zig-zag pattern repeatiing each twin lamella as shown by the black lines in Fig. 5. The interface between the twins of the anorthoclase is straight on the micrographs, but the interface between the microeline twins is not straight, but in general, finely irregular (Akizuki 19'72). Some twin lamellae are periodnc awd show a superlattice in the diffraction pattern (.Fig. 3 A). The components repeat roughly equaliy in I~.,,th albite ;.and pericline twinnings of some ax'eas, while c,ne is wider than the other in some other a r e a s . Fig. 7 shows the diffraction pattern from a paralleii growth of two crystals, in whicll both b*a x e s a n d (101) & (201) are parallel w i t h e a c h

other~ respectively. In ether crystals, both b*axes and (101) & (001) are parallel with each other, respectively. Altlaough the fibrc,,us aggregate showing parallel growth picked up from the thin section was examin~ed by a precession camera and diffractometer, the direction of the fibre axis of anorthoclase was not clearly defined. The fibre ax~is,however, may be paral|el to the b-axis, which is parallel or nearl~y parallel to the b*-axis~ because the diffraction patterns of crystals sbo:, : ing parallel growth share a common b*-.axis. Dttfraction patterns showing e?dstence of exsolution were not observed in the present specimens. An undevitrified obsidian consists of isolated particles which are abou.t ~30 A in diameter and rows.of' several particles. :Silica minerals co-existing with alkali feldspar from bot!h localities consist of very small particles, with diameters about 200/~, and a few laths showing weak contrast. Diffraction patterns of the small particles of silica ,adnerais correspond to those of so-called Opal(.T (Jones & Segnit 1971), th,cmgh some exhibit a weak tridymite-like pattern. The common opal within the spherulites from the Hosaka mine showed tlhe distinct diffraction pattern of high tridymite (Akizuki & Shimada 1979).

D i s c u s s i o n

Metastable tridymi~te and cristobatite can crystallize at a temoerature as low as 100--260°C, and the crystals e~rentwMiy transform into quartz at a higher temperature (see, Betterman & Liebau 1975). I n addition, Lofgren (t971) s t u d ~ . d t h e devitrification of rhyolitic obsidian in the tern°

Anorthoclase spherulites

LITHOS 16 (1983)

F/g. 7. Electron diff,'action l~,attern showing parallel growth of sanidine and twinned anorthoclase. The indices above the spots refer to monoclinic sanidinc; those below refer to triclinic, albite-twinned anorthoclase. The b*-axes of both diffraction patterns are parallel. The Hosaka mine.

perature range 240° to 700°C, and showed that quartz was unive]rsally present in the spherulites, Therefore, it is ~thought t:hat the spherulites of alkali feldspar wiitl.out quartz may have formed under hydrothermal conditions at a temperature below about 200°C~ According to the phase diagram determined by MacKenzie (1952), the anorthoclase should be ordered triclinic at the g~rowth conditions. However, the spherulites cov~sist of both monoclinic sanidine and triclinic anorthoclase with crosshatched twinning, which was produced during the phase llranshiion from monoclinic to triclinic structure. Thus, the texture shows that when it nucleated the al~:ali feldspar was disordered monoclinic. Thi:s discrepancy is interpreted as follows: since the growth temperature is low, atom mobility o.n the growth surface is slow, though the grow~th rate is rather high because the crystals are fibrous. Therefore the AI and Si atoms cannot occupy the ordered sites and a disordered structure results. Metastable, disordered, monoclinic sanidine transgorms iinto stable, ordered tficlinic, anorthoclase over long periods of time. The te,xture and struclure of ano,rthoclase in the spheru~ite results from the same growth mechanism a~ those of adularia (Akizuki & Sunagawa 1978) and the phase transiition occurred after the growth. Adulm'ia cry~stals con:fist of a higlfly diisordered structure inspire of a low growth temperature and,

with f¢w exceptions, have not been transformed into ar~ ordered structure. Sodium-rich alkali


feldspar can transform more quickly than potash feldspar, resulting in cross-hatched transition twinning (Akizuki 198!). The textures of the pres,mt hydrothermal anorthoclase observed by TEM we,re compared with those of volcanic anortht~clase studied between crossed polars by Gray & Anderson (1982), though the millii~ietre scale texture c,5served in the volcanic anorthoclase cannot be compared with that of the hydrothernlal one ~bserved under TEM. The hydrothermai arLortt.~oclase consists of only fine twinning whose wi~ith is less than about O.1 ~un. It therefore differs from the volcanic anorthoclase in which the twin width ranges from submicroscopic to thicknesses as large as three twins per miilimetre. In the hydrothermai anorthoczlase observed under the cross-hatched twinning is only com.. men along the boundary of albite and pericliaetype twinned areas (Fig. 5), though fine [amen!ae producing the streaks, as opposed to twinm~i spots in the diffraction pattern, cross the one ~f the two kinds of twinning (Fig. 6). In the volca~nic anorthoclase the cross-hatched twinning is di,:tinct in the region whose albite and pericline twinning are roughly equally dew~.}oped and also the twin individuals cross even the thick lame|la of the other twin over a wide range. However, some areas consist of only on~:." component of albite and pericline twinning; i.e. one twin set is absent. Differences in the twin texture between both-specimens may be ~ttributed to the crystal habit, scale and cooling rate which affect strain during coolling. Further detailed comparison of the specimens needs to be carried out un.d~r TEM. The '!iesegang' phenomenor~ means that the feldspar grew in a silica gel which was produced by alteration of natural glass under hydrothermal conditions, :hat is, the spherulites of sodium-rich feldspar were produced by gel growth.


Acl:nowledgements. - I thank Professor K. Aoki, Tohoku Uni-

versity, Japan for supplying the samples and Lr. G. E. Lob gren, NASA, U.,'LA. and an anonymom reviewer for critically reading and correcting the manuscript.

References Akizuld, M. 1972:Electronmicroscopic i~wesfigafior~of microcline twinning. Am. Minera~t.57, 79% :308. Ak~uki, M. & Sur~agawa, 1. 1978: Stt~dyof ,the sector structure in adulada by means of optical microscopy, infra-red ab.~,orpdon, and electron micros¢apy. Mineral.. Mag. 42, 453-~62.

\ 254

Mizuhiko Akizuki


Akizuki, M. & Shim~t~a, L 1979: Texture and mi~e~'alsin\~p~al from Hosaka, Fuku~;hima preSeclture, Japan. J. Jap~l.\4isscc. MiL~eral. Pe,¢ro,L Econ. Geol. 74, 274--279 (in J a p a n ~ with English abstra,c~). \ AkL~mki, M. 1981: Cl~Tstal growOt and phase transition ~ intermedi~te microcltine. N~ues J~hrb. Mineral. Monat. 181-\ 1[]9. \ BeUetmar~n, P. & Liebau, F. 1975: The transfonnatiq~n of ' amorphous silica to crystallin,: silica under hydrotherntai , c¢,mditi~)ns. Contrib. Mine~M. Petrol. 53, 25-36. Gray, N. H. & Ander.,on, J. B. 1982: Polysynthetic twin width d~stribt~tior~s in an~,:~oclase. Lithos 13, 27-37. Her~isch, H. K. 1970: Crystal Growth in Gels. Pennsylvania State L!niv. Press, Univ. lXa~'kand London. Jones, J. B. & Segnit, E. R. 1971: The nature of opal. I. l~iome~Iclatureand constitu~,nt phases. J. Geol. Aust. 18, 5763.

LrrHos 16 (r~3) Laves, F. 1953: %he lattice and twinning of Inicrocfine a~Jd other fe~ldsp~rs.1. GeoL 58, 548-571. Laves, F. & ~l¢~ttos, K. 1962: Ober '~,e~zen~e' MikroklinVerzwillingung and fiber unsymmetfiscl~eAlbitausscheidung in Kryptope'~d~t. Zeit. Krbt. 117, 209-7.'17. Lofgren, G. 1971: Experimentally produced devitrific~ttiontextures in natural rhyolitic glass. Bull. Geol. Soc. Amo 82, H1123. MacKenzie, W. $. 1952: The effect of temperature on the symmetry of ~gh-temperatu:te soda-ri~:h feldspars. Am. ]. \ Sci. Bowen VoluJne 319-342. MacKenzie, W. So 1956: The orientatioa of the peridine twin lpJnellae in uicl~nic ~dkafifeldspars. Mine:ral. Mag. M, 41-46. Tanida, K. 1961: A study on salic effusive roc]~s. Sci Report, Tohoku Unit,. Third series 7, ~7-100.