FLORE Repository istituzionale dell'Università degli Studi di Firenze How many alacranites do exist? A structural study of non-stoichiometric As8S9-x crystals. Questa è la Versione finale referata (Post print/Accepted manuscript) della seguente pubblicazione:

Original Citation: How many alacranites do exist? A structural study of non-stoichiometric As8S9-x crystals / P. BONAZZI; L. BINDI; F. OLMI; S. MENCHETTI. - In: EUROPEAN JOURNAL OF MINERALOGY. - ISSN 0935-1221. - STAMPA. - 15(2003), pp. 283-288.

Availability: This version is available at: 2158/312415 since:

Terms of use: Open Access La pubblicazione è resa disponibile sotto le norme e i termini della licenza di deposito, secondo quanto stabilito dalla Policy per l'accesso aperto dell'Università degli Studi di Firenze (https://www.sba.unifi.it/upload/policy-oa-2016-1.pdf)

Publisher copyright claim:

(Article begins on next page)

30 September 2021 Eur.J. Mineral. Dedicatedto thememory of 2003, 15, 283–288 LucianoUngaretti

Howmany alacranitesdo exist? Astructuralstudy ofnon-stoichiome tricAs 8S9-x crystals

PAOLA BONAZZI1, LUCA BINDI1, FILIPPO OLMI2 and SILVIO MENCHETTI1

1Dipartimento di Scienze della Terra,Universit` adegli Studi di Firenze, via LaPira4, I-50121 Firenze, Italy e-mail:[email protected] 2CNR –Istituto di Geoscienze eGeorisorse –sezione di Firenze, via LaPira 4, I-50121 Firenze, Italy

Abstract: Crystals ofarsenic sulphide (reportedas alacranite),coming fromthe burningdump of Kate ÏrinaMine (Czech Republic), have been investigated by single crystal X-raydiffraction and chemical microanalysis. Both analytical data and unit-cellparameters stronglysuggest the existence ofa continuous series between the high-temperature polymorph ( -As4S4)and the mineralalacranite (As 8S9).As the Scontent increases inthe series, the unit-cellvolume increases accordingly.Thestructural model has been obtained fortwo crystals, exhibitingdifferent unit-cell volumes (ALA15: a = 9.940(2), b = 9.398(2), c = 9.033(2), =102.12(2), V = 825.0(3), Robs =6.12 %;ALA2: a = 9.936(2), b = 9.458(2), c = 9.106(2), =101.90(2), V = 837.3(3), Robs =6.41 %).Wefoundthat the non-stoichiometriccompounds crystallize as adisordered mixtureof two kinds of cage-like molecules, packed togetheras inthe -As4S4 phase. Thefirst one is identical tothe As 4S4 molecule foundin the structures ofboth realgar ( -As4S4) and -As4S4.Thesecond molecule is chemically and structurallyidentical tothat foundin the As 4S5 compound. Thesimultaneous presence ofAs 4S4 (C2/c) and As8S9 (P2/c)microdomains could be areason forthe observed gradual change ofthe translation symmetry fromthe -phase toalacranite s.s.. Key-words: alacranite, crystal structure, arsenic sulphides, chemical composition, Kate Ïrina Mine.

Introduction Table1. Unitcell parameters for alacranites from KateÏrinaMine to- getherwith data from literature.

Threephases among the known crystalline modifications of 3 tetra-arsenictetrasulphide ( i.e. realgar,pararealgar, and - a (Å) b (Å) c (Å) (°) Vol. (Å ) phase1)occur as minerals. Realgar is the low-temperature ALA119.968(4) 9.317(2) 8.906(4) 102.45(4) 807.7(5) ALA129.963(9) 9.351(2) 8.984(5) 102.43(2) 817.4(9) form, -As4S4 (Hall,1966; Clark, 1970; Roland, 1972; Blachnik et al.,1980; Bryndzya &Kleppa, 1988), which ALA159.940(2) 9.398(2) 9.033(2) 102.12(2) 825.0(3) commonly occurs as asublimation product in active volca- ALA1 9.95(1)9.44(1) 9.07(1) 102.1(2) 833(2) nic areas,in low-temperaturehydrothermal deposits, and, to ALA2 9.936(2)9.458(2) 9.106(2) 101.90(2) 837.3(3) alesser extent, as aminor constituent of lead, silver, and ALA7 9.941(6)9.450(5) 9.110(2) 101.81(3) 837.7(7) ALA6 9.918(5)9.478(7) 9.145(6) 102.10(5) 840.6(9) gold ore veins. Pararealgaroccurs as ayellow filmcovering realgarcrystals which have been exposed to light (Roberts CL* 9.97(1)9.29(1) 8.88(1) 102.6(1) 803(2) PS 9.957(3)9.335(4) 8.889(5) 102.48(4) 806.7(6) et al.,1980; Bonazzi et al., 1995). The -As4S4 phase is the high-temperature form,stable inthe system As-S attemper- PPV 9.89(2)9.73(2) 9.13(1) 101.84(5) 860(3) atures higher than 256 5°C(Hall,1966). The occurrence BP9.943(1)9.366(1) 8.908(1) 102.007(2) 811.4(1) of anatural phase exhibiting adiffraction pattern quite simi- ZO 9.87(1)9.73(3) 9.16(2) 101.52(4) 858(4) larto that of the synthetic -As4S4 was firstreported by Note:CL =natural -As4S4 from Alacr`anMine,Chile (Clark, 1970); Clark (1970), who found both low- and high-temperature PS=synthetic -As4S4 (Porter& Sheldrick,1972); PPV =alacranite formsin the Ag-As-Sb vein deposit atAlacr an` (Chile). Ac- fromKamchatka (Popova et al.,1986);BP=natural -As4S4 (Burns &Percival,2001); ZO = alacranite from KateÏrina Mine (Z´Ï acekÏ & cording to Clark (1970), the natural -As4S4 phase formsir- regular masses and exhibits optical properties similarto Ondrus,Ï 1997).*indexed by thepresent authors. those of realgar,but the colour is slightly paler and moreyel- lowish than realgar.At the Alacr`an Mine the mineraloccurs closely associated to smithite, orpiment and arsenolamprite. 1 Muchconfusion exists in literature for the use of - and -descriptors.Fol- lowingDouglass et al.(1992)we willrefer tothe low-temperature form as Although the main properties of the mineral(XRD data, andto the high-temperature form as . chemicalcomposition, optical properties and Vickers’

0935-1221/03/0015-0283 $2.70 DOI:10.1127/ 0935-1221/2003/0015-0283 2003 E.Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart 284 P.Bonazzi,L. Bindi,F. Olmi,S. Menchetti hardness) weredefined by Clark (1970), the proposal of a new mineralspecies corresponding to the natural -As4S4 polymorph was not approved by the NMMN– IMACom- mission (Popova et al.,1986). Later,a new arsenic sulphide was found atthe Uzon caldera(Kamchatka, Russian Feder- ation) by Popova et al. (1986). These authors assumed this mineralto be identical to the species previously described by Clark (1970) due to the sim- ilarity of their XRD powder patterns. For this reason, the mineralwas named alacranite(Popova et al., 1986). As shown in Table 1, the latticeparameters of alacraniteresem- ble fairlythose of the synthetic -As4S4 (Porter& Sheldrick, 1972) and those of the natural -phase fromAlacr an` (Clark, 1970). In keeping with the greaterunit-cell volume, alacra- nite exhibits adifferentchemical composition (As 8S9). Ac- cording to Popova et al.(1986), alacraniteis monoclinic, P2/c,while the synthetic -As4S4 crystallizes in the C2/c space group. During arecentsampling of the seafloor around Lihir Island (Papua NewGuinea), aspecimen main- Fig.1. SEM micrographof a crystalof alacranite togetherwith ly consisting of pyrite, sphalerite, and galena, together with amorphousAs-S alloyfrom Kate ÏrinaMine, Czech Republic. red and orange arsenic sulphides, was recovered atthe top of Conical Seamount (Percival et al.,1999). According to KateÏrina Mine is located atRadvanice, near Trutnov and these authors, the XRD analysis of both the deep-red and or- belongs to the Lower-Silesian coal basin in the north-east- ange crystals revealed amixture of realgarand alacranite. ern part of Bohemia (Czech Republic). Disseminated sul- The subsequent detailed investigation (Burns &Percival, phide (pyrite, marcasite,pyrrothine, chalcopyrite, chalco- 2001) showed the mineralfrom Papua NewGuinea to be cite, bornite, galena and sphalerite) and uranium (earthy structurally and chemically identical to the synthetic - uranium oxides and rarecoffinite) mineralizations accom- As4S4.However, one maywonder whether this mineralis pany the coal measures of the Radvanice Group of Coals actually the samemineral species defined asalacraniteby (ZÏ a´cekÏ & Ondrus,Ï 1997). Asecondary mineralization origi- Popova et al. (1986). Indeed, the alacranite described by nated as aconsequence of morethan adecade lasting sub- Burns &Percival (2001) is quite similarto that fromAlac- surfacefire of alarge dump atthe Kate Ïrina colliery.Accord- ran` (Clark, 1970), but, as also noted by Jambor &Roberts ing to ZÏ a´cekÏ & OndrusÏ (1997), several “minerals”were (2002), itappears to differwith respect to chemicalformula, formedfrom escaping gasses and vapours, including native unit-cell volume and space group fromthe mineralap- elements (sulphur, selenium, Bi-antimony,bismuth and proved with the namealacranite by the NMMN– IMA lead), sulphides (galena, greenockite, antimonite, arsenic Commission (Hawthorne et al., 1988). sulfides, and monoclinic GeSnS 3 ),oxides (molybdite, arse- The complexity of the problem is furtherincreased due to nolite, hexagonal GeO 2),halides (salammoniac, cryptoha- the alteration induced by light on the tetra-arsenictetrasul- lite, bararite),sulphates (anglesite, mascagnite, letovicite), phides. Itis long known that both polychromatic and mono- and organic compounds (kratochvilite, kladnoite). Inaddi- chromatic light alterrealgar and the -phase up to parareal- tion, avariety of sulphates formedby alteration. Among the gar through an intermediate product ( -phase) (Douglass et arsenic sulphides, As-S alloy,realgar,alacranite s.s., orpi- al.,1992; Bonazzi et al.,1996; Muniz-Miranda et al., 1996). ment and a“monoclinic As 4S4”close to alacranitehave The light-induced formation of the -phase obtained from been reported ( ZÏ a´cekÏ & Ondrus,Ï 1997). the -phase occurs with astrong anisotropic increase of the The sample examined consists of asiltstone partially unit-cell volume. Therefore,crystals of -phase partially al- covered by acicularcrystals of anhydrite. Scattered euhed- tered by exposure to light exhibit unit-cell volumes greater raldeep-red transparent crystals of realgarwith avariable than the unaltered crystals. filmof pararealgarare coexisting with orange-red to orange For these reasons, wewere intrigued by anew recovery of crystals, with pinacoidal prismatichabit (Fig. 1). Crystals alacranite coexisting with realgar,and other As-Sphases ( Z´Ï a- arerarely transparent, with greasy luster. The orange miner- cekÏ & Ondrus,Ï 1997) fromthe burning dump of the Kate Ïrina alis extremely brittle, with orange-yellow streak; maximum colliery,Radvanice (Czech Republic). On the basis of its lat- size is approximately 100 µm.Associated mineralsare ticeparameters (T able 1), this mineralappears to be very simi- mainly orpiment and native sulphur, together with botryoi- lar to the true alacranitedescribed by Popova et al. (1986). dal or dendritic aggregates of amorphous As-S alloy similar to those already described by ZÏ a´cekÏ & OndrusÏ (1997).

Occurrence andsample description Experimental Asample fromKate Ïrina Mine (Czech Republic) was sub- mitted to our attention by Dr.Giovanni Pratesi(Museo di Several well-formedcrystals wereselected forX-ray single- Storia Naturale– Universit`adi Firenze, Italy). crystal investigations. Onthe whole, the diffraction quality How manyalacranites do exist? 285

Table2. Chemical compositions (wt. % el.)and atomic ratios for alacranites from KateÏrina Mine.

* As4S4 realgar ALA ALA ALA ALA ALA ALA 11 12 15# 7 2# 6 As70.03 69.98 69.59 69.47 - 68.27 - 67.80 S29.9730.02 30.41 30.53 - 31.73 - 32.20 As8.00 8.00 8.00 8.00 8.00 8.00 8.00 8.00 S8.008.02 8.17 8.22 8.42 8.69 8.70 8.88 Note:values are averagedon threespots; * calculatedvalues for stoi- # chiometricAs 4S4; atomicratios obtained from structural data. was found to be fairlygood forthe orange-red crystals, whereas reflections areweaker and broader forpaler orange Fig. 2. As4S4 (a) and As4S5 (b)moleculesin alacraniteviewedalong crystals. For some of these latter,no diffraction effectswere the c axis. detected, although long-time exposures wereused. When possible, the unit-cell dimensions weredetermined by means of least-squares refinements using the sameset of 25 Å3).Although these observations pointed towards a P2/c (or reflections (24 2 30°) measured with aCAD4 single- Pc)ratherthan C2/c (or Cc)symmetry,the number of the crystal diffractometer.Surprisingly ,awide variety of pa- observed reflections with h + k =2n+1 was too scarceto at- rameterswere obtained (Table 1). Inparticular, orange-red tempt astructure determination using aprimitive unit-cell. crystals (ALA11, ALA12) exhibit unit-cell parameterssim- Therefore,the C2/c structure of the synthetic -phase (Por- ilarto those of asynthetic -phase, whereas orange and ter& Sheldrick, 1972) was assumed as starting model. 2 pale-orange crystals show greaterunit-cell volumes (up to Structure refinementswere performed on F o using the 840.6 Å3 for ALA 6). SHELXL-97program (Sheldrick, 1997). Isotropic full-matrix Due to the brittleness of the mineral,it was not possible to least-squares cycles wereinitially run assuming the atomsites obtain polished surfaces forcrystals embedded in resin; as fully occupied, although the unusually high value of the therefore, semi-quantitative analyses wereperformed by isotropic displacement factorfor the S2 atom strongly sug- means of an EDS-EDAXsystem. Acrystal of realgarwas gested partial occupancy atthis site. An examination of the used asstandard. As shown in Table 2, the As/Sratio ranges F-Fourier maprevealed the presence of residual peaks from8:8.0 to 8:8.9 forthe crystalline material(ALA11, clearlyindicating split positions forboth As atoms (As1b and ALA12, ALA7, ALA6);amorphous materialexhibits wider As2b) and an additional position fora sulphur atom(S4). The variation of the As/Sratio (ranging from8:8.5 up to 8:14). convergence was quickly achieved by adding these peaks to Suspiciously high sulphur contents werelocally found for the atom arraywithout constraint on their occupancy factors. single crystals having poor diffraction quality ( e.g., ALA1), For both ALA15 and ALA2 crystals the occupancy factorsof probably due to intergrowth of alacranite and an amorphous As1, As2 and S2 resulted close to acommon value (k), while S-richphase. the occupancy factorfor As1b, As2b and S4 resulted approxi- Two crystals of relatively high diffraction quality matelyequal to 1-k. For both crystals, S1 and S3 appeared as (ALA15 and ALA2)were chosen forthe structural study. fully occupied. An attemptto refineindependently the split Intensity data werecollected up to 2 MoK =50°and subse- positions S1-S1b and S3-S3b did not lead any significant im- quently corrected forLorentz-polarization. Absorption cor- provement due to the low number of observed reflections (Ta- rection was performedusing the semi-empiricalmethod of ble 3). Furthermore, the coexistence of As1b +As2b with S2 North et al. (1968). Experimental details aregiven in Table and of As1 +As2 with S4 in the samemolecule had to be ex- 3. (This table can be obtained fromthe authors or through cluded on the basis of the bond distances. Therefore,in order the E.J.M. Editorial Office– Paris.) to reduce the number of freevariables and to obtain areliable model, only one parameter(k) was refined to constrain occ.(As1) =occ.(As2) =occ.(S2) =k, and occ.(As1b) = Structuresolution occ.(As2b) =occ.(S4) =1-k. Successive cycles wererun with an anisotropic model. Final Robs were6.12 %forALA15 and For all the crystals tested (Table1) the h0l reflections with 6.41% forALA2, respectively.Fractional atomiccoordinates l =2n +1 weresystematically absent. Moreover, most inten- and atomicdisplacement parametersare given in Table 4and sities wereconsistent with the C-centered lattice, although Table 5, respectively.(They can be obtained fromthe authors some hkl reflections with h + k =2n+1 wereobserved. In or through the E.J.M. Editorial Office– Paris.) particular, the most evident violations arethe following: 50 0, 3 2 1, 6 1 1, 3 2 2, 1 42, 304. Itwas also noted that the greaterthe unit-cell volume, the stronger and sharper the re- Results flections violating the C latticesymmetry ,thus suggesting a gradual change of the translation symmetry fromthe - The models obtained forboth ALA15 and ALA2 areconsis- phase (C2/c; V = 803(2) Å3)to alacranite( P2/c; V = 860(3) tent with the coexistence in the structure of two kinds of 286 P.Bonazzi,L. Bindi,F. Olmi,S. Menchetti

Table6. Selected interatomic distances (Å )andangles (° ) forthe alacranites examined.

ALA15 ALA2

As4S4 mol. As4S4 mol. As1 - S12.239(5) As1 - S1 2.252(6) S3 2.292(5) S3 2.359(6) As22.593(7) As2 2.58(1) As2 - S32.177(5) As2 - S3 2.124(7) S2 2.227(6) S2 2.238(8) As12.593(7) As1 2.58(1) S1– As1- S394.6(2) S1 – As1- S393.2(2) S1– As1- As299.8(2) S1 – As1- As2100.1(2) S3– As1- As297.8(2) S3 – As1- As297.5(3) S3- As2- S295.4(2) S3 - As2- S296.6(3) S3- As2- As199.4(2) S3 - As2- As1100.1(3) S2- As2- As199.1(1) S2 - As2- As199.5(2)

As4S5 mol. As4S5 mol. As1b - S32.00(1) As1b - S3 2.046(8) S1 2.09(1) S1 2.136(7) S4 2.17(4) S4 2.19(2) As2b - S42.21(2) As2b - S4 2.19(2) S3 2.34(1) S3 2.313(9) As2b2.45(3) As2b2.48(2) S3- As1b- S1109.0(6) S3 - As1b- S1106.5(3) S3- As1b- S496.4(6) S3 - As1b- S496.1(4) S1- As1b- S4107.8(8) S1 - As1b- S4105.4(5) S4- As2b- S3100.6(6) S4 - As2b- S3101.2(5) S4- As2b- As2b101.1(8) S4 - As2b- As2b101.7(5) S3- As2b- As2b95.8(6) S3 - As2b- As2b95.6(3) Fig. 3. The a, b, and c-sin parametersplotted against unit-cell vol- umes.Filled circles refer to alacranites from KateÏrinaMine (this study).Empty symbols refer to data from literature. Down triangle: cage-like molecules. The firstone (2As1 +2As2 +S1 +S2 synthetic -As4S4 (Porter& Sheldrick,1972); square: synthetic As S (Kothiyal& Ghosh,1976); circle: alacranite from Kamchat- +2S3) is identical to the As 4S4 molecule found in the struc- 2 2.15 tures of both realgarand -phase (Mullen &Nowacki, ka (Popova et al., 1986). 1972; Porter& Sheldrick, 1972), in that each As atom links one As and two Satoms (Fig. 2a). The other molecule (2As1b +2As2b +S1 +2S3 +2S4) is chemically and struc- -phase (2.593(6) Å;Porter& Sheldrick; 1972) and in the turallyidentical to that found in the As 4S5 compound (Whit- natural -phase fromPapua NewGuinea (2.596(2) Å; field, 1973). Itcan be derived fromthe As 4S4 molecule by Burns &Percival, 2001). removing S2 (4 e Wyckoff position) and adding two S4 at- The As4S5 molecule closely resembles that observed in oms (8f Wyckoff position). In such amolecule, two As at- the structure of synthetic As 4S5 (Whitfield, 1973). In both oms link one As and two Satoms, whereas the other two As molecules, the As-S bond distances arein the range 2.00 to atoms link three Satoms (Fig. 2b). In both the molecules, 2.36 Å,noticeably wider than that usually observed forin- the As atoms lie atthe vertices of adisphenoid. In the As 4S4 tramolecularAs-S distances. In particular, the As-S dis- molecule, two edges of the disphenoid correspond to As-As tances involving S2 and S4 areconsistent with the sum of bonds, while the other four edges arebridged by Satoms, the covalent radii (1.21 +1.02 Å,as reported in the Cam- which forma square parallel to (102). The As 4S4 molecule bridge Structural Database), while those involving S1 and possess a D2d symmetry within the experimental errors.In S3 deviate fromthe expected values. This is reasonably due the As4S5 molecule, five disphenoidic edges arebridged by to the factthat S1 and S3 areaverage positions forboth Satoms, thus causing alowering of the molecular symmetry As4S4 and As4S5 molecules; as pointed out above, in fact,an to C2v (approximately). Four Satoms lie atthe vertices of a attempt to refinethe split positions S1-S1b and S3-S3b was square parallel to (010). not successful. Table 6reports selected intramolecular distances forboth The structural models obtained forALA15 and ALA2 As4S4 and As4S5 molecules forALA15 and ALA2, respec- differfrom each other mainly in the percentage of the As 4S5 tively.In the As 4S4 molecule, the As-As distances (2.593(7), molecule (21 and 35 %forALA15 and ALA2, respective- 2.58(1) Å)arevery similarto those observed in the synthetic ly). Accordingly,the following chemicalformulae can be How manyalacranites do exist? 287

written: As8S8.42 (ALA15) and As 8S8.70 (ALA2).The differ- ence in the chemicalcomposition estimated fromthe refine- ment is in accordance with the differenceobserved in the unit-cell volume forALA15 ( V = 825.0(3) Å3) and ALA2 (V = 837.3(3) Å3).

Discussion Structural evidences (ALA15 and ALA2)together with chemicaldata and unit-cell parameters(ALA11, ALA12, ALA7, ALA6)indicate that alacranites from KateÏrina Mine exhibit non-stoichiometric chemicalcomposition ranging from As8S8 to As8S9.As the Scontent increases, the unit- cellvolume increases accordingly.The expansion of the unit cellappears considerably anisotropic: in particular, a lengthening of c-sin and, to alesser extent, of b is ob- served, whereas a remainsalmost unchanged (Fig. 3). In Figure 4the unit-cell volume is plotted against the chemical Fig.4. Unit-cellvolume plotted against the As S percentage.Filled composition forthe crystals fromKate Ïrina Mine together 4 5 circlesrefer to alacranites from KateÏrinaMine (this study). Empty with data fromliterature. The data fitthe regression line V = 3 symbolsrefer to data from literature. Down triangle: synthetic - 801(2) +1.04(5) [%As 4S5] (Å )(r= 0.989), thus confirming As4S4 (Porter& Sheldrick,1972); square: synthetic As 2S2.15 (Kothi- acontinuous series between -As4S4 (unnamed mineral) yal& Ghosh,1976); circle: alacranite from Kamchatka (Popova et and alacranite(As 8S9).The doubled unit-cell volume of al.,1986);up triangle:synthetic As S (Whitfield,1973). 3 4 5 synthetic As 4S5 (2 x 453.0 Å for4 molecules; Whitfield, 1973), despite its differentcrystal structure (space group P21/m),is consistent with the model obtained (Fig. 4). Alac- named. On the other hand, the namealacranite was ap- ranite fromPapua NewGuinea should be close to astoi- proved fora mineralwith the chemicalformula As 8S9 (Po- chiometric -As4S4,as confirmed by the structural investi- pova et al.,1986). On the basis of their structural study, gation (Burns &Percival, 2001). However, acontent of however, Burns &Percival (2001) stated that this formula 3 As4S5 = 10 % (V = 811.4 Å )in this mineralcan be predicted had been incorrectly determined. Itis now evident that min- using the linear regression obtained in our study.Indeed, ac- eralswith chemicalcomposition ranging continuously from cording to Burns &Percival (2001), the largerresidual As8S8 to As8S9 can crystallize. peaks in the final difference-Fouriermap (3.21 e/Å 3) were located within 1Åof the Aspositions. Added in proofs: Inthe Strunz Mineralogical Tables re- cently published by Strunz &Nickel (2001), alacraniteis re- ported with the chemicalformula As 4S4 and the space group Conclusions C2/c,according to the results obtained by Percival et al. (1999). Itappears evident that the namealacranite has to be Chemical and crystallographic data obtained for alacranites referredto the mineraloriginally described by Popova et al. from KateÏrina colliery strongly suggest the existence of a (1996). continuous series between natural -As4S4 and alacranite s.s. As8S9.Fromthe results of the structure refinements, it Acknowledgments: This work was funded by M.U.R.S.T., appears that non-stoichiometric compounds crystallize as cofinanziamento 2001, project “Structural complexity and disordered mixtures of As 4S4 and As4S5 molecules packed mineralproperties: microstructures, modularity,modula- in the sameway as in the -As4S4 phase. To establish wheth- tions”. erthis is along- or ashort-range disorder is an arduous task. The paper benefited fromconstructive reviews by T.Ba- Ï However, ahypothesis can be made. Because the As 8S9 stoi- li´c-Zuni´cand P.Berlepsch. chiometry (alacranite s.s.)seems to be the upper limitof the compositional range, one can speculate that alacranite( P latticesymmetry) consists of an ordered sequence of As 4S4 References and As4S5 molecules. The [½½ 0] translation vector can be Blachnik,R., Hoppe,A., Wickel,U. (1980):Die Systeme Arsen- eliminated by ordering As 4S5 and As4S4 molecules along Schwefelund Arsen-Selen und die thermodynamisch enDaten this direction. The simultaneous presence of As 4S4 (C2/c) ihrerV erbindungen. Z.anorg.allg. Chem. , 463, 78–90. and As8S9 (P2/c)microdomains accounts forthe observed gradual change of the translation latticesymmetry fromthe Bonazzi,P .,Menchetti,S., Pratesi,G. (1995):The crystal structure ofpararealgar,As S . Am. Mineral. , 80, 400–403. -phase to alacranite s.s.. 4 4 Finally,anadditional remarkon nomenclature is due. De- Bonazzi,P .,Menchetti,S., Pratesi,G., Muniz-Miranda,M., Sbrana, spite the firstoccurrence of natural -As S phase is that G.(1996):Light-induced variations in realgar and -As4S4: X- 4 4 raydiffraction and Raman studies. Am. Mineral. , 81, 874–880. from Alacr`an Mine (Clark, 1970), this mineralremained un- 288 P.Bonazzi,L. Bindi,F. Olmi,S. Menchetti

Bryndzya,L.T .&Kleppa,O.J. (1988): Standard molar enthalpies of rocksand sediments from the Lihir area, Papua New Guinea. In

formationof realgar( -AsS)and orpiment (As 2S3)byhigh-tem- Claysfor our future (H. Kodama, A.R. Mermur,J.K. Torrance, peraturedirect-synthesis calorimetry. J.Chem.Thermodyn. , 20, eds.).Proc. 11 th Int.Clay Conf. (Ottawa, 1997). ICC97 Organiz- 755–764. ingCommittee, Ottawa, Canada, 689– 696.

Burns,P .C.& Percival,J.B. (2001): Alacranite, As 4S4:anewoccur- Popova,V .I.,Popov,V .A.,Clark, A., Polyakov,V .O.,Borisovski, rence,new formula, and determination of the crystal structure. S.E. (1986):Alacranite As 8S9;anewmineral. Zap.Vses. Miner- Can.Mineral. , 39, 809–818. al.Obshchest. , 115,360–368 (in Russian). Clark,A.H. (1970):Alpha-arsenic sulfide, from Mina Alacr an,` Pam- Porter,E.J. & Sheldrick,G.M. (1972): Crystal structure of a new paLarga,Chile. Am. Mineral. , 55,1338–1344. crystallinemodification of tetra-arsenictetrasulphide (2,4,6,8-te- Douglass,D.L., Shing, C., Wang,G. (1992):The light-induced alter- trathia-1,3,5,7-tetra-arsatricyclo[3,3,0,0 3,7]-octane). J. Chem. ationof realgarto pararealgar. Am. Mineral. , 77,1266–1274. Soc., DaltonTrans. , 13,1347–1349. Hall,H.T. (1966):The system Ag-Sb-S, Ag-As-S, and Ag-Bi-S: Roberts,A.C., Ansell, H.G., Bonardi, M. (1980):Pararealgar, a new Phaserelations and mineralogical significance. Ph. D. Thesis, polymorphof AsS, fromBritish Columbia. Can.Mineral. , 18, BrownUniversity, Providence, Rhode Island. 525–527. Hawthorne,F .C.,Burke, E.A.J., Ercit, T.S., Grew,E.S., Grice,J.D., Roland,G.W. (1972): Concerning the -AsS realgarinversion. Can. Jambor,I.L., Puziewicz, J., Roberts,A.C., V anko,D.A. (1988): Mineral., 11, 520–525. New mineralnames. Am. Mineral. , 73, 189–199. Sheldrick,G.F. (1997):SHELXL-97. Aprogramfor crystal struc- Jambor,I.L. & Roberts,A.C. (2002):New mineralnames. Am. Min- turerefinement. University of Gö ttingen, Germany. eral., 87, 355–367. Strunz,H. &Nickel,E.H. (2001):Strunz Mineralogical T ables. Kothiyal,G.P .&Ghosh,B. (1976):Preparation of non-stoichiomet- Chemical-StructuralMineralClassification System. 9 th ed.

ricarsenic sulphide crystals As 2S2.15,andmeasurement of their SchweizerbartPubl., Stuttgart, Germany. electricalconductivity. J.Cryst.Growth , 32, 29–32. Whitfield,H.J. (1973):Crystal and molecular structure of tetra-ar- Mullen,D.J.E. & Nowacki,W .(1972):Refinement of the crystal senicpentasulphide. J.Chem.Soc. , DaltonTrans. ,1740–1742. Ï structuresof realgar, AsS andorpiment, As 2S3. Z. Kristallogr. , Z´acek,Ï V .&Ondru s,Ï P.(1997):Mineralogy of recentlyformed subli- 136, 48–65. matesfrom Kate Ïrinacolliery in Radvanice, Eastern Bohemia, Muniz-Miranda,M., Sbrana,G., Bonazzi,P .,Menchetti,S., Pratesi, CzechRepublic. Bull.Czech Geol. Survey , 72(3),289– 302. G.(1996):Spectroscopic investigation and normal mode analysis

of As4S4 polymorphs. Spectr.Acta , A52,1391–1401. North,A.C.T., Phillips,D.C., Mathews, F .S.(1968):A semiempiri- calmethodof absorptioncorrection. Acta Cryst., A24, 351–359. Received7 January2002 Percival,J.B., Hannington, M.D., Herzig, P .M.,Jonasson,I.R. Modifiedversion received 16 April2002 (1999):Clay mineral associations in sulphide-bearin gvolcanic Accepted24 May2002