Usgeorosicar',)Ljii;'?'Ix",Ihi'f; I.'""':';:I%:1L|;'Ffi].]:St""'Mser''
Total Page:16
File Type:pdf, Size:1020Kb
American Mineralogist, Volume 78, pages204-209, 1993 Holotype buddingtonite:An ammoniumfeldspar without zeolitic HrO J.H.L. VoNCKEN,* H.L.M. vAN RSERMUNDT** A.M.J. vAN DER EnnonN, J.B.H. Jl.NsnN Facultyof EarthSciences, Department of ChemicalGeology, Budapestlaan 4, 3508TA Utrecht,The Netherlands R. C. Eno uSGeorosicar',)lJii;'?'ix",iHi'f; i.'""':';:i%:1l|;'ffi].]:sT""'Mser'' Ansrucr Holotype buddingtonite from the Sulphur Bank Quicksilver deposit, Lake County, Cal- ifornia, was studied by XRD, IR, STEM, SEM, TGA, DTA and wet chemical analyses. Synthetic anhydrous ammonium feldspar was studied also for comparison. The holotype buddingtonite specimen(NMNH no. I1697400) contains admixtures of FeSr,anatase and, most importantly, montmorillonite. Montmorillonite is capableof reversibledehydration, and its admixture explains the zeolitic behavior previously ascribed to buddingtonite. It is concluded that buddingtonite is an anhydrous ammonium feldspar: i.e., free of zeolitic HrO, with no zeolitic properties. INtnooucrroN the HrO is not present in an ordered way (Kimball and Megaw, 1974). Moreover, structure and properties of Buddingtonite was first describedby Erd et al. (1964) synthetic ammonium feldspar are consistentwith the an- as an ammonium feldspar with zeolitic HrO, with the hydrous formula NHoAlSi3Os(Voncken et al., 1988; following formula: NHoAlSi3Os't/zH2O.The mineral was Voncken, 1990). In addition, the lattice parametersand studied by chemical analysis,XRD, thermal analysis,and X-ray diffraction pattern of synthetic ammonium feld- IR. Buddingtonite was found initially as a hydrothermal spar are essentiallyidentical to those of natural ammo- replacementof plagioclasoin certain parts of the Sulphur nium feldspar, suggestingthat differencesin properties Bank Quicksilver deposit,Lake County, California (White between natural and synthetic ammonium feldspar are and Robertson,1962), and later in oil shalesand hydro- due to impurities in the natural material. Pyrite, marca- thermal deposits (Gulbrandsen, 1974; Kimbara and site. anatase. and montmorillonite were found in the Nishimura, 1982;Loughnan et al., 1983).Buddingtonite sample initially characterizedby Erd et al. (1964). has gained importance recently becauseof spatial rela- This study presentsnew data on the natural holotype tionships with Au deposits (Kydd and kvinson, 1986). buddingtonite specimen studied originally in order to shed All authors have acceptedthe hydrous formula that was light on the nature of the zeolite-like behavior of the min- proposedfor buddingtonite.Pavlishin and Bagmut (1988) eral. Pure synthetic ammonium feldspar was used for studied the substitution of K by NHf, in feldsparsof gran- comparison. ites and pegmatitesbut did not mention the presenceof any structural HrO in their ammonium-bearing feldspars. Solomonand Rossman(1988), however, did find minor S,l,lupr,B DESCRIPTToN amounts of structurally bound HrO (0.02 wto/oHrO) in Two samples, taken from the holotype specimen of their ammonium-bearing feldspars, although there was buddingtonite(NMNH no. I1697400),were available for no evidence for HrO* substitution in the alkali-cation comparison with the synthetic material. The first sample srtes. was a 70-mg remnant of DW-1, initially describedby Erd Experimental studies of ammonium feldspars(Barker, et al. (1964). This is the purest fraction ofholotype bud- 1964; Hallam and Eugster, 1976) were inconclusive with dingtonite. The secondsample, designatedhere as 2821, respectto the zeolitic characterofbuddingtonite. A study has not been previously described.It consists of a 100- of the crystal structure of holotype buddingtonite con- mg remnant of material, split from the original holotype cluded that buddingtonite has a feldspar structure in which specimen, that was crushed, treated with cold 1:l HCI and l:l HNO3, and sedimentedin H'O for over an hour * Present address: Faculty of Mining and Petroleum Engi- to determine the associatedclay minerals. Buddingtonite neering,Department of Raw Materials Technology, Delft Uni- was the predominant mineral in this fraction. Fyrite and versity of Technology,Mijnbouwstraat 120,2628 RX Delft, The marcasite were eliminated and very minor anatasewas Netherlands. present.The smectite content was significantly geat- ** Present address: Laboratoire de P6trologie M6tamor- still phique,C.N.R.S. U.R.A., 736, Universit6Paris 7, 4 placeJus- er than in DW-1. sieu,75252Paris Cedex05, France. Synthetic ammonium feldsparwas preparedfrom NH. 0003-004x/93l0 l 02-0204$02.00 204 VONCKEN ET AL.: HOLOTYPE BUDDINGTONITE 205 Fig. l. SEMphotograph of a syntheticammonium feldspar Fig.2.rR spectra,]1l:H.*:":1,]* oooo-oo0 cm, ror crystalgrown at 600"C and2 kbar.Flaky material on both sides naturalbuddingtonite and syntheticbuddingtonite. of the crystalis carbonsubstrate material. solution and an Al-Si gel at 600 "C and 2 kbar using the In the IR-spectra of the natural ammonium feldspars method of Voncken et al. (1988).A rapid-quenchpres- (Fig. 2), thereare two distinct bandsat 3750-3600cm ', sure vesselwas used, allowing cooling to room tempera- which are characteristicof montmorillonite. Thesebands ture within 2 s. If qugnching is insufficiently rapid, sec- are most clearly visible in the spectrum of buddingtonite ondary tobelite develops (Voncken et al., 1988). The sample 2821. To highlight absorption features of the synthetic sample consistedof pure buddingtonite. montmorillonite impurity, a KBr pellet with synthetic buddingtonite was put in the referencebeam when nat- ANALYTICAL TECHNIQUES ural buddingtonite was analyzed.Many of the resulting X-ray powder diffraction photographs were obtained bands can be ascribed to montmorillonite, which may with an Enraf Nonius FR552 powder diffraction camera, have NHo*; in the interlayer (Table l). usingCuKa radiation. For SEM analysis,the sampleswere SEM analysesof DW-l reveal typical buddingtonite Au-coated and studied using a Cambridge Stereoscan, crystalsof - 5 pm in length, as well as blocky grains (Figs. operatedat 10 KeV. For STEM-analysis,the sampleswere 3a, 3b). Sometimes thick plates were visible. Figure 3b ground for 5 min in alcohol in an agate mortar, and a illustrates small thin flakes, visible as overgrowths on drop ofthe suspensionwas transferredto a Cu grid bear- feldspar faces.The grain size of the thin flakes is = I pm. ing a C film on one side. STEM analyseswere performed The amount of the flaky material in 2821 is estimatedto with a JEOL 200C TEM, equipped with an energy-dis- be four times as largeas in DW-I. persive X-ray analyzer(Link SystemsLtd.), and operated A STEM study of the samples DW-l and 2821 con- at 120 or 200 KeV. The sample preparation for IR-spec- firms the presenceof platesand flakesin both. The small troscopy was as follows: 2.5 mgof sampiewas mixed with thin flakesare extremelyunstable under the electronbeam. 250 mg of KBr. Pelletswere pressedin vacuum and sub- No analysescould be obtained. The larger and thicker sequentlydried for one night at 120.C and analyzedus- plates,already observedby SEM, were identified as bud- ing a Perkin Elmer 580 IR spectrometer.Thermal anal- dingtonite by means of electron diffraction patterns. Iron yses(TGA and DTA) were obtained in air with a Dupont sulfide and what was probably TiO, were identified a few thermal analyzer at a heating rate of l0 "C/min. Wet times using qualitative energydispersive X-ray analyses. chemical analysesof DW-l were carried out for NHi, If the amounts of NH; in wt0/0,determined by wet which was determined by colorimetry after acid decom- chemical methods, are recalculatedto (NHo)rO, values of position and Kjeldahl distillation. 8.5 and 8.3 wto/oare found for DW-l. This agreesvery well with the results of Erd et al. (1964, Table 3), who Rrsur,rs reported 8.34 wto/o(NH.),O. They reported that the ho- A typical euhedral synthetic ammonium feldspar crys- lotype buddingtonite contains small amounts of K, Na, tal is shown in Figure l. The XRD pattern of the holotype Ca, and Ba. sample DW-l only showed buddingtonite lines, as did Erd et al. (1964) recordeda total weight loss of l2olo those of previous studies (Erd et al., 1964: Kimball and for sample DW-I, whereas our TGA data for sample Megaw, 1974; Voncken et a1., 1988).In the pattern of DW-l shows a total weight loss of 10.8 wto/o(Figs. 4a, sample 2821, vaguebroad lines were visible with d values 4b, Table 2). Thermogravimetric analyses(Figs. 4a, 4b) of 4.45,3.04,2.59,and 1.50A, which areinterpreted as show a loss of bulk NH, and HrO from the material, belonging to montmorillonite. starting at about 500 "C. There is a difference between 206 VONCKEN ET AL.: HOLOTYPE BUDDINGTONITE TABLE1, lR vibrationsof DW-1. Inten- Refer- cm 1 sity Interpretation ence 3710 sh 3703 OH-stretchmontmorillonite '|1 I 36s5 OH-stretchmontmorillonite 3630 s 1 3640 OH-stretchmontmorillonite [ 3630 OH-stretch,tobelite 3 3400 SN 3410 H,O (montmorillonite) 3290 3290-3300 NHo* 3070 s 3070 NHi 4-1 0 2850 m 2850 NHo* 1640 m 1630 H,O (montmorillonite) 1430 s 1430- 1 435 NHi 4-10 1130 VS 1120 montmorillonite 1050 vsl ) 1035 montmorillonite(broad band) 1000 vsl 920 Sh 915 montmorillonite 2,11 695 880 montmorillonite 2 785 m 795 montmorillonite 2 735 sn 731 buddingtoniteor sanidine 2,4 705 m 711,715 buddingtonite 4, 5 692 montmorillonite 2 630 sn ozE montmorillonite 2 580 St 575 montmorillonite 2 540 520 montmorillonite 2 470 SI 464 montmorillonite 2 430 st 425 montmorillonite 2 Note;Abbreviations: s : strong,m : medium,w: weak,sh : shoulder Referencenumbers: 1 : Mortlandet al. (1963);2: Van der Marel and Beutelspacher(1976); 3 : Vonckenet al (1987);4 : Erd et al. (1964);5 : Vonckenet al. (1988);6 : Shigorovaet al. (1981);7 : Vedder(1965); 8 : Solomonand Rossman(1988); 9 : Chourabiand Fripiat(1981); 10 : Krohnand Altaner(1987); 11 : Serratosa(.1962). ' A sampleo{ syntheticbuddingtonite was in the referencebeam during the analvsis. the temperaturesrecorded at the start of the loss of NH3 and H,O from buddingtonite given by Erd et al. (1964) (608 "C) and the presentresults (500 "C). However, Erd et Fig.