American Mineralogist, Volume 67, pages 1039-142, l9E2

Petersite,a REE and phosphateanalog of mixite

DoNer-o R. PBecon Department of Geological Sciences University of Michigan Ann Arbor, Michigan 48109

nNn PBIB J. DUNN Department of Sciences Smithsonian Institution Washington, D.C.20560

Abstract

The new mineral petersite (Y,REE,Ca)Cuo@Oa)I(OH)6.3H2O)occurs as a supergene mineral at the traprock quarry at Laurel Hill in Secaucus,New Jersey. It occurs in a brecciatedand mineralizedhornfels near a diabasecontact, in associationwith opal and .Prismatic crystals less than 0. I mm in lengthwith forms {1010}and {0001}occur as radiatingsprays. It is optically uniaxial, positive,with or : 1.666(4) and e: 1.747(4). The measureddensity is 3.41g/cm3. Petersite is hexagonal,probable space group PQlm or P63,with a : 13.288(5)c : 5.877(5)4,V : 898.6(8)43,and Z:2.The strongestlines in the powderdiffraction pattern are: (d, intensity,index) I1.6, 100,100; 4.36, 50, 210;3.49,40, 2ll;2.877,40, 400;2.433, 60,212. The nameis in honorof Thomasand JosephPeters.

Introduction under catalog # NMNH 148973at the Smithsonian Institution. The new mineral describedherein was sent to us for examinationby Mr. ThomasPeters of the Pater- Morphology son Museum,who had obtainedit from Mr. Nicho- Petersiteoccurs as prismatic hexagonalcrystals las Facciolla, who found it in early 1981. Mr. of simplemorphology. The only forms presentare Peters'examination by SEM techniquessuggested the prism {1010}, and the pinacoid {0001}. The a hexagonal morphology for the mineral and our crystals are usually euhedral and occur in radiating subsequentX-ray diffraction study showedit to be clustersand sprayswhich are somewhatisolated on hexagonaland isostructural with members of the the matrix (Fig. 1).Radial aggregates, which consist mixite group. Preliminary electron microprobe data of a large number of prismatic crystals (Fig. 2), showed that phosphorus is present as a major appear hemispherical and slightly dull due to a dull elementbut that arsenicis not present.A detailed luster on the exposedpinacoids, as comparedwith examinationhas confirmedthat it is a new member the highly vitreous luster on the lessexposed prism of the mixite group, and that it is the first phosphate faces.The habit of petersiteis in markedcontrast to analogueof this group of arsenates. that of the arsenatemembers of the series, gou- We take greatpleasure in namingthis new miner- deyite, mixite, and ,which usually occur in al petersitein honor of Messers.Thomas and Jo- wispy aggregatesof extremely acicular crystals. sephPeters, brothers, in recognitionof their contri- The maximum size of crystals examinedwas less butions to the mineralogyof New Jerseyand their than 0.1 mm in length, and most are considerably devotion to curating at the Paterson Museum in smaller. The diameter of radiating clusters ap- Paterson,New Jersey,and the American Museum proaches0.1 mm (Fig. 2). of Natural History in New York, respectively.The mineral and the name were approved by the IMA X-ray crystallography Commissionon New and Mineral Names, Single crystals were studied using Weissenberg prior to publication. Type material is preserved techniques. The resulting photographs confirmed 0003--004>v8?109l 0- l 039$02. 00 1039 1040 PEACOR AND DUNN: PETERSITE

Table l. X-ray powder ditrraction data for petersite, obtained usinga 114.6mm diameterGandolfi camera

d(Obs) d(Calc.) hkl' t/Io* d(Obs.) d(Calc.) hk'l !llo

11.6 11.s 100 100 L.767 I.769 332 10 4.36 4.35 210 50 I.742 I.745 303 20 4.12 4.11 201 10 r .687 1.691 512 I5 3,84 3.84 300 i0 1.687 223 3.49 3.50 2tr 40 1.659 1.661 440 2

3.32 3.32 220 t0 1.603 1.606 602 c 1 10 1 lO 310 5 I.598 441 2.877 2,877 400 40 1.583 1.583 531 2.806 2.805 311 10 1.583 701 2.687 2.697 tLz 10 1.559 l.561 522 2.634 2.640 320 15 1.523 L.524 710 2.509 2.511 410 30 1.505 1.507 6L2 2.433 2.435 2L2 60 1.466 1.469 004 2.411 2.408 321 1 L467 JJJ 2.197 2.201 222 5 I.444 1.445 442

z, lot z. loz 'La 2 1.403 1.406 720 2.051 2.162 510 2 1.402 622 2.056 402 1.394 t.397 80r Fig. l. Scanning electron microscope photomicrograph t.961 1.964 i22- 2_0 L.39?- 2r4 1.888 1.892 430 1 1.265 r.268 722 15 showingdivergent radiating spray of petersitecrystals. I.268 414 L.ZOI 443 1.264 731 that petersiteis hexagonal,with point group sym- metry 6lm. Extinctions occurredfor 00/ reflections *Intensities visually estimated. 'Indexed having I : 2n * l However, as all crystals are on^thebasis of a hexagonalunit cell with a=13.228(5)and exceptionally small and the reflections are very c=5.877(5)X. weak even for very long exposures,this extinction rule is not as well defined as we would like. Assuming that it is present, however, the space agreementwith other membersof the mixite group. group must be PQlm or P63. For example, Wise (1978) reports the probable Powder diffraction data were obtained using a spacegroup of goudeyiteas P63lmor P63, with a : : polycrystallinesample in a Gandolfi camera(114.6 13.472(l)and c 5.902U1A. mm diameter),CuKa radiation, and Si as an inter- properties nal standard.The data are presentedin Table 1. Physicaland optical Unit cell parameters(a : 13.288(5),c : 5.877(5)A Petersiteis bright yellowish green in color, quite V : 898.6(8)43)were obtained through leasr- similar in hue to cuprosklowdowskite.The hard- squaresrefinement of the powder diffraction data. nesscould not be determineddue to the very small These crystallographic parameters are in general crystal size and a high degree of brittleness.The density, obtainedusing heavy liquid techniques,is 3.41glcm3, in excellentagreement with the calculat- ed value of 3.40 g/cm3 based on the measured composition.The luster of surfacescould not be accurately observed, but the luster of the prism faces is vitreous. Petersiteis readily soluble in 1:1 hydrochloric acid, coloring the solution light greenishyellow. Optically,petersite is uniaxial (+) with indicesof refraction e : 1.747(4)and r,,l : 1.666(4).The birefringenceis similar to that of other membersof the mixite group. Pleochroismis strong, E > O, with E : green,O : light yellowishgreen. Calcula- tion of the Gladstone-Dalerelationship, using the constantsof Mandarino (1976)yields Kc : 0.191 andKp : 0.203,providing good agreementbetween Fig. 2. Scanning electron microscope photomicrograph physicaland chemicaldata. There is no discernible showing semispherulesof tightly packed petersite crystals. fluorescenceunder ultraviolet radiation. PEACOR AND DUNN: PETERSITE lMl

Chemistry group, and having rare earth elements dominant Petersitewas chemically analyzedusing an ARL- over Ca. snuq electron microprobe utilizing an operating The substitutionof Ca2* in a site occupiedprinci- voltageof 15kV and a samplecurrent of 0.025pA, pally by cations having a charge of +3 requires standardizedon brass. The standardsused were somediscussion, as the charge-balanced,end-mem- synthetic olivenite (Cu), fluorapatite (P), horn- ber formularequires that no cationsof charge*2 be blende(Fe, Ca), and the Drake rare earth glasses in that site.The mutual solid solutionof Ca for REE for REE. The data were correctedusing a modified occurs in many phases.However, it requires cou- versionof the uecrc-4 programof the Geophysical pling with some other substitution mechanismin Laboratory. The resultant analysis yields: P2O5 order to maintain charge balance. One possible ---> 23.1,CuO 52.1, CaO 2.4,FeO 0.8, Y2O3 2.5, Ce2O3 substitutionis 3Ca2* 2M3+ * n. However, the 2.5,Nd2O3 1.2, Sm2O3 1.2,La2O31.0 with HzO 12.3 crystal structureof mixite and isostructuralphases weightpercent by difference,sum : 100.0percent. is not known, so we cannot hypothesizethe exis- (Pr,Gd, Eu, Tm, Ho werepresent as traces; Pb, Sr, tence of a site for such an addition solid solution Er, Dy, Ba, Yb and S were soughtbut not found). mechanism.There is some indication of such a The accuracyof the data is estimatedto be t 3Voof mechanismin that there is a slight excess(0.15) of the amount presentfor Ca, Fe, Cu and P; -+7Vaof Y, REE, Fe and Ca over the value of 2.0 per unit the amount present for REE. Fe was assignedto cell. A second possible substitution mechanism Fe2* on the basisof a microchemicaltest. The unit couldbe Ca2* + ! (or HzO)- M3* (OH)t-. There cell contentsfor petersite,calculated on the basisof is also some indication for this mechanismin the P:6atoms,are: observeddeficiency (0.4) of OH relativeto the ideal structure,coupled with an excessof H2O. The two Ys.a1Ce6. 2sNde.3 Sm6. 13lae. 1 1 Fefr j6Cas n)>2.I s postulated substitution mechanisms account for Cu12.s7(POa)o.oo(OH) r | 60'6.78H2Omuch of the Caz+ and Fe2*. However, such inter- pretationsare beyond the limits of the precisionof or, ideally, (Y,REE,Ca)zCurz(POa)6(OH)r2.6H2Othe available chemical data. Definition of the actual with the sum of REE > Ca and with Y as the substitutionmechanisms requires additional analyt- dominant rare earth. A wavelength-dispersivemi- ical data on other phases,preferably those with a croprobe scan indicated the absenceof any other highercontent of calcium. elementswith atomic numbergreater than 9 includ- Occurrence ing Al, As and Bi, elementsfound in other members of the mixite group. There is too little materialfor Petersitewas found by Mr. Nicholas Facciollain the direct determinationof water. It was therefore early 1981at the traprock quarry operated by the calculatedby difference.The resultingOH and HzO Gallo Asphalt Company at Laurel Hill (known contentsare in reasonableagreement with thosefor locally as big Snake Hill), in Secaucus,Hudson other membersof the mixite group. Co., New Jersey. When the significanceof the Becausethe chemical analysisindicates that no mineral was known, Mr. Facciolla, together with singleelement occupies more than half of the site Mr. Thomas Peters and Mr. Russell Titus of the occupied in part by rare earth elements,there is PatersonMuseum, revisited the locality and ob- someproblem with nomenclature.Peter Modreski tained additional material for study. of the United States Geological Survey, Denver, The geology of Laurel Hill was described by (personalcommunication and paper in preparation) Darton et al. (1908),who noted that it consistsof a has consideredsuch factors in rare earth arsenate plug of diabasecutting Triassic sediments,which equivalentsof petersite and notes that the name originally were ferruginous sediments,now meta- agardite-(La) has been approved for such a mineral morphosedinto a biotite-plagioclasehornfels. A havingonly 0.23La as the dominantelement in that review of the geology of Laurel Hill and a discus- site. Similar considerationsmay eventuallyhave to sion of the diabase and included magnetite veins be applied to phosphatesolid-solution equivalents was given by Puffer and Peters (1974).A detailed of petersite,but until such minerals have actually review of the minerals of Laurel Hill has been been discoveredwe need not be concerned with privately publishedby Facciolla (1981). suchfactors. Petersiteis definedsimply as a phos- Petersiteoccurs wholly within the hornfelslocat- phate isostructural with members of the mixite ed approximately 30-60 m from the northern con- PEACOR AND DT]NN: PETERSITE tact with the diabasestock. The hornfels has been desiccatedappearance of the previously described extensivelybrecciated in places and subjectedto assemblages. secondarymineralization. The occurrencefor the Petersiteappears to be a supergenemineral with type material has been "mined-out" and thus copper derived from the altered chalcopyrite.The chances of finding additional samples are poor. origin of the rare earth elementsand phosphorus is Petersiteoccurs in minute cavitiesbetween brecci- problematical. However, allanite and apatite are ated hornfels fragments where approximately a known to occur at the locality and alteration of dozen hand-sizedsamples were recovered. peter- these minerals may well have provided the rare site must be considered a rare mineral becauseof earths. the limited occurrence. Petersiteoccurs in severalassemblages. The orig- Acknowledgments inally discovered material occurs on a matrix of The authors expresstheir gratitude to Mr. Christian H. Grube severelyaltered chalcopyrite, which is coatedwith of the PatersonMuseum and Ms. Joan Whelan of the American an amorphous dark red mineral similar to pitticite, Museumof Natural History, in New York, for their assistancein obtaining SEM photomicrographs. but more likely The contributions of Mr. near diadochite in composition, Nicholas Facciolla and Mr. Russell Titus in obtaining additional given the geochemistryof the occurrence.This in sampleswere very significant.Dr. M. J. Modreski kindly made turn is coated with abundant malachite and chryso- availablemuch data relating to rare earth membersof the mixite colla, coated with hematite and a thin, desiccated group, including discussionsof the nomenclatural problems. layer of light bluish amorphous opal. The final References mineralization was of malachite, in small mammi- lary aggregates,coated sparingly with Darton, N. H., Bayley, W. S., Salisbury,R. D. and Kummel, tiny clusters (190E) of petersite H. B. PassaicFolio, New Jersey-NewYork. In Geo- crystals. Although petersite occurs in logic Atlas of the United States. U.S. GeologicalSurvey Folio other assemblages,the best formed and largest 557. crystals occur within this altered sulfide assem- Facciolla,N. W. (l9El) Mineralsof Laurel Hill, Secaucus,New blage. Petersite encrusts both the opal and the Jersey. Privately published by Nicholas Facciolla. malachite. Mandarino, J. A. (1976)The Gladstone-Dale relationship-Part I: Derivation of new constants. Canadian Mineralogist, 14, A secondassemblage consists of white albitewith 498-502. or without chlorite, malachite and anatase, the Pufer, J. H. and Peters,I. J. (1974)Magnetite veins in diabase latter in bright euhedralcrystals. These minerals are of Laurel Hill, New Jersey. Economic Geology, 69, 1294- coated with light blue opal and, lastly, petersite. 1299. (1978) Petersitemay form on any of theseminerals but it is Wise, W. S. Parnauiteand goudeyite,two new copper arsenateminerals usuallyfound from the Majuba Hill Mine, PershingCoun- on the light blue opal. Other samples ty, Nevada.American Mineralogist, 63, 704-708. occur in slightly varied assemblageswith cloudy albite which gives a grayish appearance to the matrix. In these samples, the opaline coating is Manuscript received, March 23, 1982; more glassyand chalcedonicas comparedwith the acceptedfor publication, June 9, 1982