And Pb-Rich Willemite, Tres Marias Mine, Mexico

minerals

Article Chemical Environment of Unusually Ge- and Pb-Rich Willemite, Tres Marias Mine, Mexico

Bernhardt Saini-Eidukat 1,*,†, Frank Melcher 2,†, Jörg Göttlicher 3,† and Ralph Steininger 3,†

1 Department of Geosciences, North Dakota State University, Fargo, ND 58108, USA 2 Chair of Geology and Economic Geology, Peter-Tunner-Straße 5, Leoben 8700, Austria; [email protected] 3 Karlsruhe Institute of Technology, Institute for Synchrotron Radiation, Hermann-von-Helmholtz-Platz 1, Eggenstein-Leopoldshafen D-76344, Germany; [email protected] (J.G.); [email protected] (R.S.) * Correspondence: [email protected]; Tel.: +1-701-231-8785 † These authors contributed equally to this work.

Academic Editor: Raymond M. Coveney Jr. Received: 30 December 2015; Accepted: 23 February 2016; Published: 9 March 2016

Abstract: The Tres Marias carbonate-hosted Zn-Ge deposit in Chihuahua, Mexico contains willemite [Zn2SiO4] with unusually high concentrations of minor and trace elements (e.g., Pb, Ge, As, P, V); Pb concentrations are as high as 2 wt %, and Ge may reach 4000 ppm (average 900 ppm). Electron microprobe analyses and synchrotron X-ray fluorescence maps show that Zn and Ge, as well as Zn and Pb are negatively correlated, whereas Ge and Pb are positively correlated across zoned willemite crystals. In cathodoluminescence (CL) images, those areas of willemite having high trace element concentrations have no, or low CL intensities, whereas zones low in trace elements (except for P) display bright blue CL colors. X-ray absorption fine structure (XAFS) spectroscopy was used to characterize the chemical nature of Ge and Pb in willemite. Comparisons to reference spectra of natural and artificial substances points to the presence of Ge4+ and Pb2+ in Tres Marias willemite. No evidence for Pb4+ was detected. Oscillatory zonation reflects trace element incorporation into willemite from the oxidation of primary Ge-bearing sphalerite and galena (PbS) by siliceous aqueous fluids.

Keywords: XAFS; XANES spectroscopy (Ge, Pb); germanium; lead; willemite; Zn2SiO4; Tres Marias

1. Introduction Germanium (Ge) is a high technology metal in great demand for use in the manufacture of optics and sensors, computer chips, solar cells, and as a catalyst in plastics manufacturing. Most Ge is recovered as a byproduct from processing zinc sulﬁde ore, or from combustion products of Ge-bearing lignites [1–3]. Investigating the chemical behavior of trace elements in willemite is of interest not only because of its use as an ore of non-sulﬁde zinc [4] but also for understanding chemical substitution mechanisms, and the controls on its luminescence and other electronic properties [5,6]. While the geochemistry of germanium has been reviewed [7,8], very few detailed chemical data on Ge in willemite, and in related minerals such as hemimorphite, are available in the literature. Willemite from Franklin, N.J. and New Mexico contains up to 350 ppm Ge [9]. Willemite in the Tsumeb deposit contains up to 1280 ppm Ge [10]. Zinc silicates in concentrate leach residues that are stoichiometrically equivalent to willemite contain 800 to 1500 ppm Ge and from 4000 to 8100 ppm Pb [11]. Ore from the Beltana deposit in the Flinders Range of Australia contains up to 50 ppm Ge in willemite, and a range of 0.4–2.5 wt % Pb [12]. Willemite from the Bou Arhous deposit [13] in Morocco contains over 1000 ppm Ge [14].

Minerals 2016, 6, 20; doi:10.3390/min6010020 www.mdpi.com/journal/minerals Minerals 2016, 6, 20 2 of 13

The substitution of Pb for Zn raises the interrelated questions of the oxidation state, ionic radius, and substitution mechanism of Pb in the crystal structure. Tetravalent, tetrahedrally coordinated lead, IVPb4+, has an ionic radius of 0.65 Å, similar to 0.60 Å for IVZn2+ [15]. In this case, Pb would exist in a sp3 hybridized form, and following arguments concerning Fe [16] and Mn [17] substitution into willemite, it might preferentially enter the slightly larger Zn(2) position. Charge balance, perhaps in the form of vacancies, would also be required. If the Pb exists as the much larger Pb2+, with radius 1.19–1.49 Å, its s2 lone-electron pairs would likely lead to strong distortion of the site. The present work provides the ﬁrst X-ray absorption ﬁne structure (XAFS) studies of the oxidation state of Ge and Pb in willemite which is naturally highly enriched in these elements. Understanding the oxidation state of these elements in the willemite structure is critical for understanding trace element substitution mechanisms for geochemical and technological applications.

2. Analytical Methods Quantitative wavelength dispersive analyses of minerals were performed at the German Federal Institute for Geosciences and Natural Resources (BGR), Hanover, on polished thin sections using a CAMECA SX100 electron microprobe (CAMECA, Gennevilliers, France). For the analyses of willemite, the following lines were measured at 30 kV using the listed minerals as standards, for the number of seconds listed, and with average detection-limits given in ppm: Fe Kα, magnetite, 10 s, 300 ppm; Ca Kα DICR (chromian diopside) 10 s, 210 ppm; Mn Kα, rhodochrosite, 10 s, 300 ppm; Cu Kα, Cu˝, 10 s, 370 ppm; Zn Kα, willemite, 10 s, 600 ppm; Si Kα, willemite, 10 s, 270 ppm; Al Kα, plagioclase, 10 s, 280 ppm; S Kα pentlandite, 10 s, 180 ppm; P Kα, apatite, 10 s, 360 ppm; V Kα,V˝, 10 s, 300 ppm; As Lα, GaAs, 70 s, 480 ppm; Pb Mα, galena, 30 s, 620 ppm; Ga Kα, GaAs, 80 s, 200 ppm; Ge Kα, Ge˝, 90 s, 230 ppm; Cd Lα, Cd˝, 80 s, 400 ppm; Mo Lα, Mo˝, 40 s, 260 ppm; Mg Kα, kaersutite, 10 s, 380 ppm; Ti Kα, kaersutite, 10 s, 100 ppm; K Kα, biotite, 10 s, 200 ppm; Cl Kα, tugtupite, 20 s, 100 ppm. The Pb Mα and the Ge Kα lines were also used to produce X-ray fluorescence maps. A defocused beam was used to analyze hemimorphite. Detection-limits for the hemimorphite and carbonate mineral analyses are provided in the tables. Powder X-ray diffraction analysis was carried out at the BGR. Whole-rock chemical analyses were carried out at the BGR using X-ray fluorescence (XRF) on a Philips PW-2400 (Cr and Rh targets) (Philips, Amsterdam, The Netherlands), and with LECO methods (LECO, St. Joseph, MO, USA). Additional whole-rock analyses were carried out by Activation Laboratories of Canada using instrumental neutron activation analysis (INAA), and inductively coupled plasma-mass spectrometry (ICP-MS) following total digestion. XAFS spectroscopy was carried out at the Synchrotron Umwelt-Labor X-ray SUL-X beamline of ANKA with a wiggler as synchrotron radiation source on the X-ray absorption near-edge spectroscopy XANES part of the Pb L3 edge (Pb L3 edge at « 13.035 keV) and Ge K edge («11.103 keV) absorption spectra [18]. In both cases, Si(111) crystals were used as a monochromator. The prefocused beam was subsequently focused with a Kirkpatrick–Baez mirror system to the sample position. Spot sizes at the sample position were adjusted to about 20 µm (vert.) ˆ 30 µm (hor.) for the Ge K-edge and to about 40 µm (vert.) ˆ 70 µm (hor.) for the Pb L3-edge spectra by collimating the beam with slits at the intermediate focus. When measuring the Pb L3 edge XANES, the Ge Kα fluorescence emission is excited. The difference between the Pb Lα and the Ge Kα emission lines is about 600 eV. Effective energy resolution was adequate to distinguish these lines, even at higher count rates and shorter detector peaking times. In order to select suitable positions for the XAFS measurements, a mapping of fluorescence intensities of the emission lines for Pb (Pb Lα, about 10.5 keV) and Ge (Kα, about 9.9 keV) was carried out. Ore samples were prepared as thin or polished sections, and Kirkpatrick–Baez mirrors allowed final focusing of beam sizes sufficient to select different parts of the zoned crystals. Because of low element concentrations (Ge < 1 wt %; Pb about 2 wt %), the XAFS measurements of the samples were performed in fluorescence mode using a seven-element Si(Li) solid state detector (SGX Sensortech, formerly Gresham Scientific Instruments Ltd., Corcelles-Cormondreche, Switzerland). Minerals 2016, 6, 20 3 of 13

Ge K-edge XAFS measurements suffer from the overlap of the Ge Kα emission («9.9 keV) with the strong Zn Kβ emission at about 9.57 keV (high Zn content of the minerals). To overcome this drawback, separation of the emission peaks and improvement of the signal to noise ratio could be achieved by choosing a long peaking time at a large sample to detector distance which reduces the total count rates that are dominated by Zn Kβ emission. Reference materials containing the following Pb silicates (identification number) were acquired from the Smithsonian Institution, Washington, D.C. and measured: margarosanite [PbCa2Si3O9] (R18305), esperite [PbCa2(ZnSiO4)3] (C6175), and larsenite [PbZnSiO4] (C6174). Crystallites were separated under binocular microscope. Samples were mounted in two ways. First, approximately 10 to 18 mg of material was mixed with 100 mg cellulose and pressed into a 13 mm diameter pellet. Pb L3 XANES spectra were measured in transmission and fluorescence mode. A second mounting method was to place pulverized material on Kapton tape. X-ray diffraction (XRD) and XRF measurements were carried out on a number of crystallites to assure their identity. Reference compounds for Ge were GeO2, GeS2, briartite [Cu2(Fe,Zn)GeS4] and argyrodite [Ag8GeS6]. Transmission data were preferred because fluorescence data suffer from self absorption effects. Energies were calibrated for the Pb L3 spectra to 13.035 eV (maximum of 1st derivative of the spectrum) with a foil of elemental Pb, and for the Ge K-edge at the Pt L3 edge to 11.564 eV (maximum of 1st derivative of the spectrum) at a Pt foil. Because of the thickness of the samples, measurements for energy calibration could not be performed parallel to sample measurements. Energy calibration was checked and corrected before and after a series of XANES/extended x-ray absorption fine-structure spectroscopy (EXAFS) scans. Two to four scans were measured on three sample spots at the Pb L3 edge, four scans at three samples’ positions for the Ge K edge spectra of willemite and Ge containing sphalerite each. For the reference spectra, one to two scans were sufficient for collecting data with a good signal to noise ratio. At the Pb L3 edge, only XANES spectra were recorded. At the Ge K-edge, EXAFS data up to k 13 are still noisy for the Ge measurements of willemite, and Fourier transformation is limited to k < 10. XAFS data were pre-edge and post-edge background corrected using a linear and a polynomial function, respectively, and were normalized to an edge jump of one using the ATHENA program of the IFFEFIT software package [19]. In addition, the EXAFS functions and Fourier transforms have been extracted and calculated with the ATHENA program. EXAFS data have only been evaluated on a comparative basis and no fits have been performed.

3. Sample Descriptions Samples were collected from the Tres Marias ore body which occurs in a solution collapse breccia in Cretaceous age carbonate rocks [20]. In the deposit itself, two major ore types exist: a type consisting primarily of zinc and lead sulfides, and a type referred to as “oxidized ore”, i.e., nonsulfide ore, consisting of Zn-silicates, carbonates plus oxide minerals. The willemite reported here is from two samples, TM06-15 and TM06-18, taken from mine level 6 and 6.5, respectively. Whole rock chemical analyses show these samples contain significant zinc, lead, and germanium, with subordinate Cd, As, and related elements (Table1). In hand sample, ore sample TM06-15 is porous, dense and cryptocrystalline. Powder X-ray diffraction analysis identified willemite, smithsonite, calcite, hemimorphite, quartz, and cerussite. Willemite occurs as fine-grained matrix material, and as well-crystallized white to tan vein fillings containing crystals up to approximately 1 mm across. In polished section, willemite crystals in sample TM06-15 are transparent with high pink to blue interference colors (Figure1). In some cases, distinct zoning is apparent in transmitted light. Similarly, using backscattered electron (BSE) imaging, lighter and darker areas indicate either oscillatory or irregular zoning. The brightness of these areas corresponds to measured variation in mean atomic weight. Minerals 2016, 6, 20 44 ofof 1313

Table 1. Whole rock chemical analyses of samples TM06-15 and TM06-18 from the Tres Marias Mine. Table 1. Whole rock chemical analyses of samples TM06-15 and TM06-18 from the Tres Marias Mine. TM06-15 TM06-18 TM06-15 TM06-18 Analyte Method 1 Analyte Method 1 TM06-15wt % TM06-18 wt % TM06-15 ppm TM06-18 ppm Analyte Method 1 Analyte Method 1 SiO2 16.03wt % wt 28.68 % 1 Ag ppm 0.8 ppm 6.8 6 TiO2 0.004 0.044 1 As 1200 685 7 SiO2 16.03 28.68 1 Ag 0.8 6.8 6 Al2O3 0.07 16.97 1 Au <2 <2 7 TiO2 0.004 0.044 1 As 1200 685 7 Fe2O3 6.9 21.16 1 Ba 219 13 1 Al2O3 0.07 16.97 1 Au <2 <2 7 MnO 0.007 0.007 1 Bi <12 6 1 Fe2O3 6.9 21.16 1 Ba 219 13 1 MgOMnO 0.0070.03 0.007 0.15 1 1 Cd Bi 179<12 80.4 6 15 MgOCaO 4.03 0.03 0.15 0.34 1 1 Ce Cd 179 <3 80.4 18 57 NaCaO2O <0.01 4.03 0.34<0.01 1 1 Cr Ce <8<3 18137 71 NaK2O2O <0.010.01 <0.01 0.26 1 1 Cs Cr <1<8 137 7 17 PK22OO5 0.028 0.01 0.260.137 1 1 Cu Cs 32<1 719 7 5 (SOP2O35) 0.0284.22 0.137 0.95 1 5 Ga Cu 32 5 19 58 5 6 (SO(F)3 )<0.10 4.22 0.95<0.10 5 1 Ge Ga 1070 5 58304 6 6 (F) <0.10 <0.10 1 Ge 1070 304 6 LOI 3.2 14.7 3 Ge 1248 344 1 LOI 3.2 14.7 3 Ge 1248 344 1 Pb 21.4 21.4 0.7280.728 4 4 In In 0.2 0.2 6.5 6.5 6 Zn 33.5 33.5 12.4 12.4 1 1 Mo Mo 44 44 99 99 7 C org 0.23 0.23 0.30 0.30 2 2 Nb Nb 20 20 <5 <5 1 C total 2.03 2.03 0.70 0.70 2 2 Ni Ni 6 6 <5 <5 1 CO2 8.998.99 2.00 2.00 diff diff Sb Sb 28 28 22 22 1 1 Sum 98.75 98.75 98.84 98.84 Sc Sc <2 5 5 1 1 SrSr 252 252 478478 11 VV 438 438 168168 6 W <30 <18 1 W <30 <18 1 Zr <9 14 1 Zr <9 14 1 1 1 : Analytical Methods: diff = by difference; 1: BGR-XRF; 2: BGR-LECO; 3: BGR-LOI; 4: Actlabs-ICP-OES; : Analytical5: Actlabs-TD-ICP; Methods: 6: Actlabs-FUS-MS; diff = by difference; 7: Actlabs-INAA. 1: BGR-XRF; 2: BGR-LECO; 3: BGR-LOI; 4: Actlabs-ICP-OES; 5: Actlabs-TD-ICP; 6: Actlabs-FUS-MS; 7: Actlabs-INAA.

Figure 1. (a) Backscattered scanning electron micrograph, micrograph; and ( b) transmitted optical micrograph of zoned willemite in sample TM-06-15. Numbers Numbers indica indicatete location location of of analyses in Table S1. Note the chemical zonation is not observable optically.

The occurrence of willemite in sample TM06-18 di differsffers markedly from that in TM06-15. In hand sample, the clay-likeclay-like materialmaterial isis red-brownred-brown colored colored and and friable. friable. XRD XRD analysis analysis indicated indicated the the presence presence of ofan an amorphous amorphous phase, phase, plus plus goethite, goethite, hematite, hematite, quartz quartz and and halloysite. halloysite. SEMSEM andand electron microprobe analyses show thatthat muchmuch ofof thethe samplesample consists consists of of blades blades of of hemimorphite hemimorphite crystals crystals with with lengths lengths up up to toapproximately approximately 100 100µm, µm, which which surround surround willemite willemite crystallites crystallites of of 20–30 20–30µ µmmsize size (Figure(Figure2 2).). Thin Thin rimsrims ofof smithsonite surroundsurround hemimorphitehemimorphite and and porous porous smithsonite. smithsonite. Cerussite Cerussite and and rare rare descloizite descloizite were were also alsoobserved observed based based on electron on electron microprobe microprobe analysis. analysis. Minerals 2016, 6, 20 5 of 13 Minerals 2016, 6, 20 5 of 13

Minerals 2016, 6, 20 5 of 13

Figure 2.2.Backscattered Backscattered scanning scanning electron electron micrograph micrograph showing showing textural textural relationships relationships between willemitebetween willemite (bright, W), hemimorphite (intermediate, H) and smithsonite (dark, S) in sample TM06-18. (bright,Figure W), hemimorphite2. Backscattered (intermediate, scanning electron H) and micrograph smithsonite showing (dark, textural S) in sample relationships TM06-18. between Willemite Willemite analysis #40, Table S1 is from this sample. analysiswillemite #40, Table (bright, S1 W), is from hemimorphite this sample. (intermediate, H) and smithsonite (dark, S) in sample TM06-18. Willemite analysis #40, Table S1 is from this sample. 4. Results Results 4. Results 4.1. X-Ray Mapping and Microprobe Analyses 4.1. X-Ray Mapping and Microprobe Analyses X-ray mapping (Figure3 3)) shows shows a a distinct distinct coincidence coincidence of of high high Pb Pb and and Ge Ge in in zoned zoned crystals crystals (and(and X-ray mapping (Figure 3) shows a distinct coincidence of high Pb and Ge in zoned crystals (and vice versa ).). Quantitative Quantitative electron electron microprobe microprobe analyses analyses were were made made along along two two profiles proﬁles in in single single willemite vice versa). Quantitative electron microprobe analyses were made along two profiles in single willemite crystals from sample TM06-15TM06-15 (Figures(Figures4 4 and and5 ;5; Table Table S1). S1). The The width width ofof thethe growthgrowth zones varies from from a crystals from sample TM06-15 (Figures 4 and 5; Table S1). The width of the growth zones varies from a few to approximately 50 µm.µ a fewfew to approximatelyto approximately 5050 µm.m.

(a) (b) (a) (b)

(c)

FigureFigure 3. (a3.): (a Backscattered): Backscattered electron electron image;image; ( b):): Ge Ge X-ray X-ray fluorescence ﬂuorescence map map and and (c): (Pbc): X-ray Pb X-ray fluorescence map of zoned willemite in sample TM-06-15.(c) ﬂuorescence map of zoned willemite in sample TM-06-15. Figure 3. (a): Backscattered electron image; (b): Ge X-ray fluorescence map and (c): Pb X-ray fluorescence map of zoned willemite in sample TM-06-15. Minerals 2016, 6, 20 6 of 13 Minerals 2016, 6, 20 6 of 13 Minerals 2016, 6, 20 6 of 13

Figure 4. Backscattered scanning electron micrograph of zoned willemite, sample TM06-15. FigureFigure 4. Backscattered4. Backscattered scanning scanning electron electron micrograph micrograph of zoned of zoned willemite, willemite, sample sample TM06-15. TM06-15. Numbers Numbers represent analyses, reported in Table S1. representNumbers analyses, represent reported analyses, in reported Table S1. in Table S1.

Figure 5. Backscattered scanning electron micrograph of zoned willemite, sample TM06-15. Figure 5. Backscattered scanning electron micrograph of zoned willemite, sample TM06-15. FigureNumbers 5. Backscattered represent analyses, scanning reported electron in micrographTable S1. of zoned willemite, sample TM06-15. Numbers representNumbers analyses, represent reported analyses, in reported Table S1. in Table S1. Results of electron microprobe analyses show the following trends (Figure 6): Results of electron microprobe analyses show the following trends (Figure 6): (a)Results Pb and of Ge electron are positively microprobe correlated analyses (r = show0.77). theThe following Pb and Ge trends concentrations (Figure6): range from below (a) Pb and Ge are positively correlated (r = 0.77). The Pb and Ge concentrations range from below the detection limits to over 1.5 wt % and 1400 ppm, respectively. Concentrations of both (a) Pbthe and detection Ge arepositively limits to correlatedover 1.5 wt (r %= 0.77).and 1400 The Pbppm, and respectively. Ge concentrations Concentrations range from of both below elements increase from core to rim, overprinting tandem oscillations between extreme values. theelements detection increase limits tofrom over core 1.5 to wt rim, % andoverprinti 1400 ppm,ng tandem respectively. oscillations Concentrations between extreme of both values. elements (b) Pb and Zn are negatively correlated, with a slope near −1.0 (r = −0.94). The slope of this line is (b)increase Pb and from Zn are core negatively to rim, overprinting correlated, with tandem a slope oscillations near −1.0 between(r = −0.94). extreme The slope values. of this line is similar if data are plotted in wt % or in formula units. similar if data are plotted in wt % or in formula units. ´ ´ (b)(c) Pb Zn and and Zn Ge are are negatively also negatively correlated, correlated with (r a = slope −0.66). near 1.0 (r = 0.94). The slope of this line is (c) similar Zn and if dataGe are are also plotted negatively in wt correlated % or in formula (r = −0.66). units. (c) Zn and Ge are also negatively correlated (r = ´0.66). Minerals 2016, 6, 20 7 of 13 Minerals 2016, 6, 20 7 of 13

Minerals 2016, 6, 20 7 of 13

(a) (b) (c)

FigureFigure 6. 6.(a ():a): Ge Ge wt wt % %vs. vs.Pb Pb wtwt %;%; ((bb)) ZnZn wt % vs. PbPb wt wt %; %; and and (c ()c )Ge Ge vs.vs. ZnZn wt wt % %for for the the zoned zoned (a) (b) (c) willemitewillemite shown shown in Figurein Figure4. Closed 4. Closed symbols symbols plotting plotting below below the dashed the dashed line, which line, indicateswhich indicates detection detection limit, were not used in plotting regression line nor in calculating r value. limit,Figure were not 6. ( useda): Ge inwt plotting % vs. Pb regressionwt %; (b) Zn line wt nor% vs. in Pb calculating wt %; and r(cvalue.) Ge vs. Zn wt % for the zoned willemite shown in Figure 4. Closed symbols plotting below the dashed line, which indicates In additiondetection limit,to analyses were not usedof the in plottingzoned regressioncrystals, lineanalyses nor in calculatingof fine-grained r value. matrix willemite were conducted.In addition When to compositions analyses of thefrom zoned both samples crystals, are analyses plotted in of terms ﬁne-grained of atom % matrix (Figure willemite 7), two main were conducted.trends becomeIn When addition apparent: compositions to analyses a low Ge-high fromof theboth zoned Pb samplestrend crystals, and are aanalyses high plotted Ge-low of in fine-grained terms Pb trend. of atom matrixTwo % measurements (Figurewillemite7), were two with main trendshighconducted. becomeGe and intermediate apparent: When compositions a lowPb contents Ge-high from bothlie Pb outside trendsamples andthe ar emain a plotted high trends, Ge-low in terms and Pb of analyses atom trend. % Two(Figurefrom measurements sample 7), two TM06-18main with trends become apparent: a low Ge-high Pb trend and a high Ge-low Pb trend. Two measurements with highlie below Ge and the intermediate low Ge-high Pb Pbcontents trend. Concentrations lie outside the ofmain P, V, trends,Fe and As and were analyses sporadically from sample detected TM06-18 and high Ge and intermediate Pb contents lie outside the main trends, and analyses from sample TM06-18 liemay below reach the up low to Ge-high1 wt % P2 PbO5, trend.As2O3 Concentrationsand FeO, and 0.15 of wt P, % V, As Fe2 andO3. As were sporadically detected and lie below the low Ge-high Pb trend. Concentrations of P, V, Fe and As were sporadically detected and may reach up to 1 wt % P O , As O and FeO, and 0.15 wt % As O . may reach up to 1 wt %2 P25O5, As22O33 and FeO, and 0.15 wt % As2O3. 2 3

Figure 7. Pb vs. Ge (atom %) in willemite. Circles: TM06-15; squares: TM06-18. Dotted line outlines Figure 7. Pb vs. Ge (atom %) in willemite. Circles: TM06-15; squares: TM06-18. Dotted line outlines Figure 7. Pb vs. Ge (atom %) in willemite. Circles: TM06-15; squares: TM06-18. Dotted line outlines lowlow Ge-high Ge-high Pb Pbtrend, trend, while while solid solid line line outlines outlines ththe high Ge-lowGe-low Pb Pb trend. trend. Open Open symbols symbols are arebelow below low Ge-high Pb trend, while solid line outlines the high Ge-low Pb trend. Open symbols are below detectiondetection limits, limits, which which are are indicated indicated by by dashed dashed lines. detection limits, which are indicated by dashed lines. HemimorphiteHemimorphite in insample sample TM06-18 TM06-18 waswas studiedstudied foforr itsits tracetrace element element content, content, but but no noGe Gewas was detected,Hemimorphitedetected, even even in incases in cases sample where where TM06-18the the hemimorphitehemimorphite was studied is inin for contactcontact its trace withwith element willemite willemite content, crystals crystals butcontaining containing no Ge was detected,900 900ppm ppm evenGe. Ge. Of in Ofthe cases the trace trace where elements elements the measured, hemimorphitemeasured, onlyonly Fe is was in contact aboveabove the the with detection detection willemite limit, limit, with crystals with an averagean containingaverage concentration of 400 ppm. Smithsonite and cerussite in sample TM06-18 were also analyzed, and 900concentration ppm Ge. Of of the 400 trace ppm. elements Smithsonite measured, and cerussite only Fe wasin sample above TM06-18 the detection were limit,also analyzed, with an average and found to contain Ge at below detection limits. Cerussite contains between 0.6 and 5.8 wt % Zn, concentrationfound to contain of 400 Ge ppm. at below Smithsonite detection and limits. cerussite Cerussite insample contains TM06-18 between were 0.6 and also 5.8 analyzed, wt % Zn, and whereaswhereas smithsonite smithsonite contains contains an an average average of of 440440 ppm Pb, 0.390.39 wt wt % % Fe, Fe, 0.18 0.18 wt wt % %Mg, Mg, and and 0.18 0.18 wt %wt % found to contain Ge at below detection limits. Cerussite contains between 0.6 and 5.8 wt % Zn, whereas Ca. Ca.Cadmium Cadmium content content was was below below the the limit limit of of detection.detection. smithsonite contains an average of 440 ppm Pb, 0.39 wt % Fe, 0.18 wt % Mg, and 0.18 wt % Ca. 4.2. XAFS and XANES Analyses Cadmium4.2. XAFS content and XANES was belowAnalyses the limit of detection. Ge K XANES spectra measured at three different positions of the willemite part of the sample differ 4.2. XAFSinGe their K and XANES shape XANES (Figure spectra Analyses A1) measured due to atcrystal three orientat differention positionseffects on of the the polarized willemite synchrotron part of the beam.sample For differ in their shape (Figure A1) due to crystal orientation effects on the polarized synchrotron beam. For Gefurther K XANES discussion, spectra the averaged measured spectrum at three will differentbe used, and positions is designated of the as willemite the Ge K XANES/EXAFS part of the sample furtherspectrum discussion, of willemite. the averaged spectrum will be used, and is designated as the Ge K XANES/EXAFS differ in their shape (Figure A1) due to crystal orientation effects on the polarized synchrotron spectrumThe of willemite.positions of the white line and of the maximum of the first derivative of the Ge K XANES beam. For further discussion, the averaged spectrum will be used, and is designated as the Ge spectrumThe positions of the of willemite the white is linevery and close of tothe the maximu one ofm a of GeO the2 firstreference derivative spectrum, of the whereas Ge K XANES the K XANES/EXAFS spectrum of willemite. spectrumpositions of ofthe Ge willemitein sulfidic environmentis very close are to loca theted one at about of a 2.7GeO eV2 lowerreference energies spectrum, (Figure 8).whereas the positionsThe positions of Ge in ofsulfidic the white environment line and are of theloca maximumted at about of 2.7 the eV ﬁrst lower derivative energies (Figure of the Ge 8). K XANES spectrum of the willemite is very close to the one of a GeO2 reference spectrum, whereas the positions of Ge in sulﬁdic environment are located at about 2.7 eV lower energies (Figure8). Minerals 2016, 6, 20 8 of 13 Minerals 2016, 6, 20 8 of 13

(A) (B)

Figure 8. (A) Ge K XANES spectra of Ge in (a) willemite (solid line) compared to GeO2 (dashed line) Figure 8. (A) Ge K XANES spectra of Ge in (a) willemite (solid line) compared to GeO2 (dashed 4+ lineindicating) indicating Ge as GeGe as in Ge the4+ oxidicin the environment oxidic environment of willemite, of willemite; and (b) sphalerite and (b) sphalerite (solid line ()solid compared line) to briartite [Cu2(Fe,Zn)GeS4] (dashed line), argyrodite [Ag8(GeS4)S2] (dash-dotted line), GeS2 (dotted compared to briartite [Cu2(Fe,Zn)GeS4](dashed line), argyrodite [Ag8(GeS4)S2](dash-dotted line), line) and GeS (dash double dotted line). The shift of the white line from Ge monosulfide to GeS2 (dotted line) and GeS (dash double dotted line). The shift of the white line from Ge monosulfide todisulfides disulfides and and oxides oxides is isindicated indicated by by vertic verticalal markers markers at at about about 1108.0, 1108.0, 1109.6, 1109.6, and and 1112.3 eV, respectively.respectively. Energy Energy has has been been calibrated calibrated at the at Ptthe L3 Pt edge L3 toedge 11564 to eV. 11564 Sample eV. spectra Sample were spectra measured were inmeasured fluorescence in fluorescence mode, reference mode, spectra reference in transmission. spectra in transmission. (B) 1st derivative (B) 1st ofderivative Ge K XANES of Ge spectraK XANES of spectra of Ge in (a) willemite (solid line) compared to GeO2 (dashed line) indicating Ge4+ as Ge4+ in the Ge in (a) willemite (solid line) compared to GeO2 (dashed line) indicating Ge as Ge in the oxidic oxidic environment of willemite, and (b) sphalerite (solid line) compared to briartite environment of willemite; and (b) sphalerite (solid line) compared to briartite [Cu2(Fe,Zn)GeS4] [Cu2(Fe,Zn)GeS4] (dashed line), argyrodite [Ag8(GeS4)S2] (dash-dotted line), GeS2 (dotted line) and (dashed line), argyrodite [Ag8(GeS4)S2](dash-dotted line), GeS2 (dotted line) and GeS (dash double dottedGeS (dash line ).double The shift dotted of the line maxima). The shift of the of firstthe maxima derivatives of the from first Ge derivatives monosulfide from to Ge disulfides monosulfide and oxidesto disulfides is indicated and byoxides vertical is indicated markers atby about vertical 1106.5, markers 1108.0, at and about 1110.7 1106.5, eV, respectively. 1108.0, andEnergy 1110.7 haseV, beenrespectively. calibrated Energy at the has Pt L3 been edge calibrated to 11564 at eV. the Pt L3 edge to 11564 eV.

The reference GeO2 has low quartz structure where Ge is coordinated by four oxygens in The reference GeO2 has low quartz structure where Ge is coordinated by four oxygens in distances distances of 1.734 Å ([21], recalculated from Inorganic Crystal Structure Database (ICSD) No. 59624) of 1.734 Å ([21], recalculated from Inorganic Crystal Structure Database (ICSD) No. 59624) in contrast in contrast to the six-fold oxygen coordination of GeO2 with rutile structure with four Ge-O to the six-fold oxygen coordination of GeO2 with rutile structure with four Ge-O distances of 1.872 Å distances of 1.872 Å and two of 1.902 Å ([22], recalculated from ICSD No. 9162). With an estimated and two of 1.902 Å ([22], recalculated from ICSD No. 9162). With an estimated phase shift of 0.44 Å, phase shift of 0.44 Å, the Ge-O distance of Ge in willemite from the EXAFS Fourier Transform results the Ge-O distance of Ge in willemite from the EXAFS Fourier Transform results to about 1.73 Å and, to about 1.73 Å and, hence, is very close to the Ge-O distance in GeO2 with low quartz structure hence, is very close to the Ge-O distance in GeO2 with low quartz structure (Figure9). (Figure 9). Ge-S distances in Ge-containing sphalerite are similar to the Ge-S distances in briartite, argyrodite, Ge-S distances in Ge-containing sphalerite are similar to the Ge-S distances in briartite, and GeS2 (Figure9). Ge is also four-fold coordinated in these structures. Average Ge-S distances argyrodite, and GeS2 (Figure 9). Ge is also four-fold coordinated in these structures. Average Ge-S calculated from several ICSD entries are for briartite 2.24 Å ([23], ICSD No. 627304; [24] ICSD distances calculated from several ICSD entries are for briartite 2.24 Å ([23], ICSD No. 627304; [24] No. 47165; [25] ICSD No. 627306), for argyrodite 2.2 Å ([26]; ICSD No. 100079), and for GeS2 2.21 Å ICSD No. 47165; [25] ICSD No. 627306), for argyrodite 2.2 Å ([26]; ICSD No. 100079), and for GeS2 ([27] ICSD No. 31685; [28] ICSD No. 16951). With an estimated phase shift of 0.39 Å the Ge-S distance 2.21 Å ([27] ICSD No. 31685; [28] ICSD No. 16951). With an estimated phase shift of 0.39 Å the Ge-S of Ge in sphalerite is about 2.2 Å. distance of Ge in sphalerite is about 2.2 Å. A comparison of the Pb L3 XANES spectrum of Pb in Tres Marias willemite shows that Pb is A comparison of the Pb L3 XANES spectrum of Pb in Tres Marias willemite shows that Pb is present as Pb2+. No evidence for Pb4+ was detected. A reference spectrum of the Pb2+ mineral larsenite 2+ 4+ 2+ present as Pb . No evidence for Pb was detected. A reference spectrum of the Pb mineral2+ larsenite [PbZnSiO4] is very similar to the sample spectrum (Figure 10a). In addition, spectra of the Pb -bearing [PbZnSiO4] is very similar to the2+ sample spectrum (Figure 10a). In addition, spectra of the minerals margarosanite [Pb(Ca,Mn )2(Si3O9)] and esperite [PbCa2(ZnSiO4)3] are similar to those of Pb2+-bearing minerals margarosanite [Pb(Ca,Mn2+)2(Si3O9)] and esperite [PbCa2(ZnSiO4)3] are similar the sample and to the larsenite spectrum from its edge position. The inset (b) of Figure 10 shows to those of the sample and to the larsenite spectrum from its edge position.4+ The inset (b) of Figure 10 the lower edge position of elemental Pb and the edge position of a Pb reference [PbO2] at higher shows the lower edge position of elemental Pb and the edge position of a Pb4+ reference [PbO2] at energies, and the good match of the spectrum of the sample with that of larsenite. higher energies, and the good match of the spectrum of the sample with that of larsenite. Minerals 2016, 6, 20 9 of 13 Minerals 2016, 6, 20 9 of 13 Minerals 2016, 6, 20 9 of 13

Figure 9.9. Fourier Transformations (F.T.), notnot phasephase shiftshift corrected,corrected, ofof GeGe KK EXAFSEXAFS spectraspectra (weight(weight Figure 9. Fourier Transformations (F.T.), not phase shift corrected, of Ge K EXAFS spectra (weight k3, kk rangerange upup toto 11.5)11.5) forfor ((aa)) willemitewillemite ((solidsolid lineline)) comparedcompared toto GeOGeO22 (dashed line),); and for (b) k3, k range up to 11.5) for (a) willemite (solid line) compared to GeO2 (dashed line), and for (b) sphaleritesphalerite ((solidsolid lineline)) comparedcompared toto briartite briartite [Cu [Cu2(Fe,Zn)GeS2(Fe,Zn)GeS44](] dashed(dashed lineline),), argyrodite argyrodite [Ag [Ag8(GeS8(GeS44)S)S22]] sphaleritedash-dotted (solid line line) compareddotted lineto briartite [Cudash2(Fe,Zn)GeS double4] dotted(dashed line line), argyrodite [Ag8(GeS4)S2] ((dash-dotted line), GeS2 ((dotted line) and GeS ( dash double dotted line). The markers are locatedlocated atat (a)dash-dotted Ge-OGe-O ((aa)) andand line ((b),b) )GeS Ge-SGe-S2 ( dotted distances,distances, line indicatingindicating) and GeS Ge-O(Ge-Odash distancesdistancedouble dotteds in willemite line). The are markers similar toare the located onesones inatin a)GeO Ge-O(low (a) quartzand (b structure),) Ge-S distances, and Ge-S indicating distances Ge-O in Ge distance containings in sphaleritewillemite areare closesimilar to Ge-Sto the distances ones in GeO22 (low quartz structure), and Ge-S distances in Ge containing sphalerite are close to Ge-S distances 3 GeOin briartite,2 (low quartz argyrodite structure), and GeS and. Ge-S Ge K distances EXAFS spectra in Ge containing (k3 chi(k)) aresphalerite presented are inclose Figure to Ge-S A2. distances in briartite, argyrodite and GeS2. Ge K EXAFS spectra (k chi(k)) are presented in Figure A2. in briartite, argyrodite and GeS2. Ge K EXAFS spectra (k3chi(k)) are presented in Figure A2.

Figure 10. (a) Comparison of the averaged Pb L3 XANES spectrum for Pb containing Tres Marias Figure 10. (a) Comparison of the averaged Pb L3 XANES spectrum for Pb containing Tres Marias Figurewillemite 10. compared(a) Comparison with the spectra of the of averaged elemental Pb Pb, L3 ofXANES larsenite spectrum [PbZnSiO for4] (Smithsonian Pb containing C6174), Tres willemite compared with 2+the spectra of elemental Pb, of larsenite [PbZnSiO4] (Smithsonian C6174), Mariasmargarosanite willemite [Pb(Ca,Mn compared)2(Si with3O9)] the (Smithsonian spectra of elemental C6175), and Pb, esperite of larsenite [PbCa [PbZnSiO2(ZnSiO44)]3] (Smithsonian(Smithsonian 2+ margarosanite2+ [Pb(Ca,Mn )2(Si3O9)]2+ (Smithsonian4+ C6175), and esperite [PbCa2(ZnSiO4)3] (Smithsonian C6174),R18305) margarosaniteas Pb references, [Pb(Ca,Mn and PbO2 )as2(Si Pb3O9 reference.)] (Smithsonian The inset C6175), (b) demonstrates and esperite the [PbCa edge2(ZnSiO shift from4)3] R18305) as Pb2+ references, and2+ PbO2 as Pb4+ reference. The4+ inset (b) demonstrates the edge shift from (Smithsonianelemental Pb R18305)via Pb2+as minerals Pb references, to Pb4+ in and PbO PbO2. 2Sampleas Pb andreference. reference The spectra inset (b )were demonstrates measured the in 2+ 2+ 4+ 4+ elementaledgefluorescence shift Pb from and via elemental transmission Pb minerals Pb viamode, to Pb Pb respectively.minerals in PbO to2. Sample Pb in PbOand 2reference. Sample andspectra reference were spectrameasured were in fluorescencemeasured in and ﬂuorescence transmission and mode, transmission respectively. mode, respectively. 5. Discussion 5. Discussion Results presented here provide evidence for the chemical environment of Ge and Pb in natural willemite,Results and presented provide hereconstraints provide on evidence the substitution for the mechanismschemical environment for minor elementsof Ge and in Pbthe in willemite natural willemite, and provide constraints on the substitution mechanisms for minor elements in the willemite Minerals 2016, 6, 20 10 of 13

5. Discussion Results presented here provide evidence for the chemical environment of Ge and Pb in natural willemite, and provide constraints on the substitution mechanisms for minor elements in the willemite structure. The chemical nature of Ge in silicates can be derived from Ge K-edge XANES and EXAFS spectra (Ge K-edge 11.103 keV) [18]. Ge-O distance, oxygen coordination number and comparison of the XANES resonances with reference compounds enable the distinction between Ge in tetrahedral and octahedral coordination. From Ge K XANES and EXAFS measurements, it is likely that Ge in willemite occurs as Ge4+ and is four-fold coordinated with oxygen, as expected because the ionic radius of Si4+ (0.39 Å) is close to that of Ge4+ (0.44 Å) and isomorphism of Ge4+ with Si4+ is most probable [9]. Lattice parameters of Zn2SiO4 and Zn2GeO4 are also similar (see below). Changes in lattice parameters are in the range of 0.00034 Å for a, b and 0.00023 Å for c (R-3H setting) for 1000 ppm Ge assuming a linear behavior. These small values explain why changes in lattice constants with substitution are not or are barely detectable. Measurement of the XANES spectra of Pb in willemite indicates it is present as Pb2+. One interpretation of the positive correlation of concentrations of Pb with Ge in willemite is that these elements may be simultaneously accommodated in the willemite structure. The cosubstitution could be due to overall expansion of the lattice with increased Ge in the tetrahedral site. Support for this idea is the increase in cell parameters from a = 13.948 Å and c = 9.315 Å for willemite [16,29] IV to a = 14.284 Å and c = 9.547 Å for Zn2GeO4 [30]. Similarly, average M-O bond length increases from 1.6319 Å in synthetic Zn2SiO4 to 1.7628 Å in Zn2GeO4 [31]. However, as discussed in Section1, Pb2+ is unlikely to simply replace Zn2+ at its tetrahedrally coordinated site due to its much larger size. This is also indicated by the similarities with the Pb L3 XANES spectrum of larsenite, where Pb is coordinated very differently. Following [4], an interstitial location of the larger Pb2+ ion in the willemite structure, charge compensated by a vacant Zn-site is more favorable. Another possibility is that Pb is accommodated as microdomains of larsenite or a similar mineral. Further work using TEM or Pb EXAFS might clarify the crystal structural setting of Pb in willemite. Textural relationships from ore body to thin section scale indicate alteration of primary sulﬁdes released Ge and Pb for incorporation into willemite, in the presence of quartz. Reference [32] discussed models of willemite formation from hydrothermal ﬂuids and indicated that, at low temperatures, 2´ SO4 is the dominant sulfur species at greater than neutral pH and oxidizing conditions with quartz saturation. A possible reaction at Tres Marias is:

p q ` ` “ p q ` 2´ ` ` 2 Zn, Ge, Pb S H4SiO4paqueousq 4O2 Zn2´xPbx Si, Ge O4 2SO4paqueousq 4Hpaqueousq (1) with additional Pb potentially being supplied to crystallizing willemite by reaction of subordinate galena in the ore. Analyses of willemite from sample TM06-18 fall slightly below the main low Ge/Pb trend (Figure7). These grains are enclosed in hemimorphite associated with smithsonite, whereas cerussite also occurs in this sample. The process of hydrothermal alteration that formed the hemimorphite-smithsonite-cerussite assemblage could have led to germanium loss from the willemite as aqueous germanic acid [33]), resulting in even lower Ge/Pb ratios. This released germanium may reside adsorbed onto oxide or other minerals elsewhere in the deposit, as it occurs in the oxidation zones of Tsumeb type Ge deposits [34] and at the Apex deposit, Utah [35]. In coals and related organic material, organic compounds account for the concentration of Ge [7]. Likewise in the Tres Marias deposit, the original source of Ge may have been migrating petroleum, now represented by extensive bitumen present in disseminated form in and around the orebody. Willemite in the Tres Marias mine contains up to 1.5 wt % Ge substituting for IVSi; concomitantly, up to 2 wt % Pb substitutes for Zn. To our knowledge, these are the highest reported Ge contents in natural willemite. The mechanism of substitution may involve interstitial Pb2+ in the willemite structure. Some Ge-rich willemite contains little to no Pb, so the converse is not true. Willemite at Minerals 2016, 6, 20 11 of 13

Tres Marias incorporated Ge as primary Ge-bearing zinc sulﬁde was dissolved and reprecipitated by oxidizing siliceous aqueous ﬂuids.

Supplementary Materials: The following are available online at www.mdpi.com/2075-163X/6/1/20/S1, Table S1: Minerals 2016, 6, 20 11 of 13 Compositions of willemite from the Tres Marias deposit.

Acknowledgments:head of the electron Thismicroprobe research laboratory was begun at whilethe BGR. Bernhardt Thomas Saini-Eidukat Oberthür’s support was a visiting is greatly senior appreciated. researcher We at thethank BGR Dave funded Hackman by the and German Roberto American Chavez Fulbright of War Commission;Eagle Mining Frankfor their Melcher support was in a the senior field. researcher We acknowledge and head of the electron microprobe laboratory at the BGR. Thomas Oberthür’s support is greatly appreciated. We thank Davethe Smithsonian Hackman and Institution Roberto for Chavez reference of War material Eagle Miningand the for ANKA their supportÅngströmquelle in the field. Karlsruhe We acknowledge for allocating the Smithsonianbeamtime. Bob Institution Martin for commented reference material on an andearlier the ANKAversion Ångströmquelle of this article. Karlsruhe We thank for allocatingThomas Malcherek beamtime. Bob(University Martin commentedHamburg) for on discussions an earlier version about Pb of thisin the article. willemite We thank structure. Thomas Malcherek (University Hamburg) for discussions about Pb in the willemite structure. Author Contributions: Bernhardt Saini-Eidukat and Frank Melcher carried out the geologic, petrographic Authorand geochemical Contributions: analyses;Bernhardt Jörg Göttlicher Saini-Eidukat and andRalph Frank Steininger Melcher carried carried out out the the XAFS/XANES geologic, petrographic analyses and geochemical analyses; Jörg Göttlicher and Ralph Steininger carried out the XAFS/XANES analyses and interpretations. and interpretations. ConflictsConflicts ofof Interest: Interest:The The authors authors declare declare no no conflict conflict of interest. of interest The. foundingThe founding sponsors sponsors had no had role no in role the designin the ofdesign the study; of the instudy; the collection,in the collection, analyses, analyses, or interpretation or interpretation of data; of data; in the in writing the writing of the of manuscript, the manuscript, and and in the in decisionthe decision to publish to publish the results.the results.

Appendix AA

X-Ray AbsorptionAbsorption SpectroscopySpectroscopy atat thethe GeGe K-EdgeK-Edge Ge K-edgeK-edge X-rayX-ray absorptionabsorption spectra have been recorded fromfrom threethree differentdifferent samplesample positionspositions (Figure(Figure A1A1).). VariationsVariations inin thethe shapeshape ofof thethe spectraspectra areare mostmost probablyprobably duedue toto orientationorientation effectseffects ofof thethe crystallitescrystallites to thethe polarizedpolarized synchrotronsynchrotron beam.beam. ForFor further data evaluation, the three spectraspectra havehave been merged.merged. One shouldshould keep inin mindmind thatthat thethe mergedmerged spectrumspectrum isis mostmost probablyprobably notnot identicalidentical withwith thethe oneone ofof anan assemblageassemblage ofof randomlyrandomly orientatedorientated crystallites, but it can bebe usedused asas anan approximateapproximate spectrumspectrum for Ge in willemite here.

Figure A1.A1. Ge K XANESXANES spectra from three differentdifferent positions on the willemitewillemite sample (dashed, dotted, dash-dotted). Variations Variations in the shape of the spectra areare most probably due to orientationorientation effectseffects of thethe crystallitescrystallites toto thethe polarizedpolarized synchrotronsynchrotron beam.beam. The merged spectrum ( solid line) isis usedused forfor furtherfurther datadata interpretationinterpretation inin thethe manuscript.manuscript. Minerals 2016, 6, 20 12 of 13 Minerals 2016, 6, 20 12 of 13

3 Figure A2. Ge K EXAFS spectra (k3chi(k))chi(k)) for ( a) willemite willemite ( solid line ) compared to GeO22 (low quartz

structure) (dashed line),), andand forfor ((bb)) sphaleritesphalerite ((solidsolid lineline)) comparedcompared toto briartitebriartite [Cu[Cu22(Fe,Zn)GeS(Fe,Zn)GeS44]

(dashed lineline),), argyroditeargyrodite [Ag[Ag88(GeS4)S)S22]]( (dotteddotted line line),), GeS GeS22 ((dashed-dotteddashed-dotted line line)) and GeS (dash double dotted line).

References References 1. Butterman, W.C.; Jorgenson, J.D. Mineral Commodity Profiles: Germanium; U.S. Geological Survey Open-File 1. Butterman, W.C.; Jorgenson, J.D. Mineral Commodity Profiles: Germanium; U.S. Geological Survey Open-File Report 2004–1218; U.S. Geological Survey: Reston, CA, USA, 2004. Report 2004–1218; U.S. Geological Survey: Reston, CA, USA, 2004. 2. Melcher, F.; Buchholz, P. Germanium. In Critical Metals Handbook; Gunn, G., Ed.; Wiley-Blackwell: 2. Melcher, F.; Buchholz, P. Germanium. In Critical Metals Handbook; Gunn, G., Ed.; Wiley-Blackwell: Hoboken, Hoboken, NJ, USA, 2014; pp. 177–204. NJ, USA, 2014; pp. 177–204. 3. U.S. Geological Survey. Mineral commodity summaries, 2015. Available online: 3. U.S. Geological Survey. Mineral commodity summaries, 2015. Available online: minerals.usgs.gov/ minerals.usgs.gov/minerals/pubs/mcs/2015/mcs2015.pdf (accessed on 5 November 2015) minerals/pubs/mcs/2015/mcs2015.pdf (accessed on 5 November 2015). 4. Hitzman, M.W.; Reynolds, N.A.; Sangster, D.F.; Allen, C.R.; Carman, C.E. Classification, genesis and 4. Hitzman, M.W.; Reynolds, N.A.; Sangster, D.F.; Allen, C.R.; Carman, C.E. Classification, genesis and exploration guides for nonsulfide zinc deposits. Econ. Geol. 2003, 98, 685–714. exploration guides for nonsulfide zinc deposits. Econ. Geol. 2003, 98, 685–714. [CrossRef] 5. Wang, S.F.; Lü, M.K.; Gu, F.; Qi, Y.X.; Xu, D.; Yuan, D.R.; Cao, D.X. Effect of Pb2+ ions on the 5. Wang, S.F.; Lü, M.K.; Gu, F.; Qi, Y.X.; Xu, D.; Yuan, D.R.; Cao, D.X. Effect of Pb2+ ions on the 2+ photoluminescence characteristics of Mn 2+-doped Zn2SiO4. Appl. Phys. A 2005, 80, 871–874. photoluminescence characteristics of Mn -doped Zn2SiO4. Appl. Phys. A 2005, 80, 871–874. [CrossRef] 6. Yang, P.; Lü, M.K.; Song, C.F.; Liu, S.W.; Gu, F.; Wang, S.F. Photoluminescence of Pb2+ ions in sol-gel 6. Yang, P.; Lü, M.K.; Song, C.F.; Liu, S.W.; Gu, F.; Wang, S.F. Photoluminescence of Pb2+ ions in sol-gel derived derived Zn2SiO4. Inorg. Chem. Commun. 2004, 7, 268–270. Zn2SiO4. Inorg. Chem. Commun. 2004, 7, 268–270. [CrossRef] 7. Bernstein, L.R. Germanium geochemistry and mineralogy. Geochim. Cosmochim. Acta 1985, 49, 2409–2422. 7. Bernstein, L.R. Germanium geochemistry and mineralogy. Geochim. Cosmochim. Acta 1985, 49, 2409–2422. 8. Höll, R.; Kling, M.; Schroll, E. Metallogenesis of germanium—A review. Ore Geol. Rev. 2007, 30, 145–180. [CrossRef] 9. Sheffer, H.W. The occurrence of germanium in willemite. Geochim. Cosmochim. Acta 1966, 30, 837–838. 8. Höll, R.; Kling, M.; Schroll, E. Metallogenesis of germanium—A review. Ore Geol. Rev. 2007, 30, 145–180. 10. Lombaard, A.F.; Günzel, A.; Innes, J.; Krüger, T.L. The Tsumeb lead-copper-zinc-silver deposit, South [CrossRef] West Africa/Namibia. In Mineral Deposits of Southern Africa; Anhaeusser, C.R., Maske, S., Eds.; Geological 9. Sheffer, H.W. The occurrence of germanium in willemite. Geochim. Cosmochim. Acta 1966, 30, 837–838. Society of South Africa: Johannesburg, South Africa, 1986; Volume 2, pp. 1761–1782. [CrossRef] 11. Dutrizac, J.E.; Chen, T.T.; Longton, R.J. The mineralogical deportment of germanium in the Clarksville 10. Lombaard, A.F.; Günzel, A.; Innes, J.; Krüger, T.L. The Tsumeb lead-copper-zinc-silver deposit, South West electrolytic zinc plant of Savage Zinc Inc. Metall. Mater. Trans. B 1996, 27, 567–576. Africa/Namibia. In Mineral Deposits of Southern Africa; Anhaeusser, C.R., Maske, S., Eds.; Geological Society

12. Groves,of South I.M.; Africa: Carman, Johannesburg, C.E.; Dunlap, South W.J. Africa, Geology 1986; Volumeof the Beltana 2, pp. 1761–1782.willemite deposit, Flinders Range, South 11. Australia.Dutrizac, J.E.;Econ. Chen, Geol. 2003 T.T.;, Longton,98, 797–818. R.J. The mineralogical deportment of germanium in the Clarksville

13. Choulet,electrolytic F; zincBuatier, plant M.; of Richard, Savage Zinc J.; Barbanson, Inc. Metall. L.; Mater. Guéguan, Trans. R.; B 1996Ennaciri,, 27, 567–576. A. Formation [CrossRef of supergene] zinc 12. silicatesGroves, in I.M.; carbonates: Carman, New C.E.; advances Dunlap, from W.J. the Geology Bou Ar ofhous the deposit Beltana (Morocco). willemite In deposit, Proceedings Flinders of the Range, 13th BiennialSouth Australia. SGA Meeting,Econ. Geol.Nanc2003y, France,, 98, 797–818. 24–27 August [CrossRef 2015.]

13.14. Choulet, F. F.; Personal Buatier, M.;communication. Richard, J.; Barbanson, 2015. L.; Guéguan, R.; Ennaciri, A. Formation of supergene zinc

15. Shannon,silicates in R.D. carbonates: Revised Neweffective advances ionic radii from and the Bousystematic Arhous studies deposit of (Morocco). interatomic In distances Proceedings in halides of the 13thand chalcogenides.Biennial SGA Meeting, Acta Crystallogr. Nancy, France, 1976, A32 24–27, 751–767. August 2015. 16. Ericsson, T.; Filippidis, A. Cation ordering in the limited solid solution Fe2SiO4-Zn2SiO4. Am. Mineral. 1986, 71, 1502–1509. Minerals 2016, 6, 20 13 of 13

14. Choulet, F. Personal communication. 2015. 15. Shannon, R.D. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr. 1976, A32, 751–767. [CrossRef]

16. Ericsson, T.; Filippidis, A. Cation ordering in the limited solid solution Fe2SiO4-Zn2SiO4. Am. Mineral. 1986, 71, 1502–1509. 17. Harper, K.; Griffen, D.T. Crystallographic consequences of Mn substitution in willemite. In Proceedings of the Abstracts, Geological Society of America Annual Meeting, Salt Lake City, UT, USA, 16–19 October 2005; Volume 37, p. 287.

18. Gusken, E.; Nunes, C.C.; Dalmon, D.L.; Ono, E.; Santos, J.S.; Suzuki, C.K. Study of GeO2 addition in silica-germania glasses by XAS; Photon Factory Activity Report 2006 #24 Part B, Users’ Report 169, 9A/2006P020; The State University of Campinas, Faculty of Mechanical Engineering, Laboratory of Photonic Materials & Devices: Campinas, Brazil, 2007. 19. Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: Data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537–541. [CrossRef][PubMed] 20. Saini-Eidukat, B.; Melcher, F.; Lodziak, J. Zinc-germanium ores of the Tres Marias Mine, Chihuahua, Mexico. Miner. Depos. 2009, 44, 363–370. [CrossRef] 21. Haines, J.; Cambon, O.; Philippot, E.; Chapon, L.; Hull, S. A neutron diffraction study of the thermal stability of the alpha-quartz-type structure in germanium dioxide. J. Solid State Chem. 2002, 166, 434–441. [CrossRef]

22. Baur, W.H.; Khan, A.A. Rutile-type compounds. IV. SiO2, GeO2 and a comparison with other rutile-type structures. Acta Crystallogr. Sect. B 1971, 27, 2133–2139. [CrossRef]

23. Nitsche, R.; Sargent, D.F.; Wild, P. Crystal growth of quaternary 122464 chalcogenides by iodine vapor transport. J. Cryst. Growth 1967, 1, 52–53. [CrossRef]

24. Wintenberger, M. Etude de la structure cristallographique et magnetique de Cu2FeGeS4 et remarque sur la structure magnetique de Cu2MnSnS4. Mater. Res. Bull. 1979, 14, 1195–1202. (In French) [CrossRef] 25. Guen, L.; Glaunsinger, W.S. Electrical, magnetic, and EPR studies of the quaternary chalcogenides II IV Cu2A B X4 prepared by iodine transport. J. Solid State Chem. 1980, 35, 10–21. [CrossRef] 26. Eulenberger, G. Die Kristallstruktur der Tieftemperaturmodiﬁkation von Ag8GeS6. Monatshefte Chem. 1977, 108, 901–913. (In German) [CrossRef] 27. Zachariasen, W.H. The crystal structure of germanium disulphide. J. Chem. Phys. 1936, 4, 618–619. [CrossRef] 28. Prewitt, C.T.; Young, H.S. Germanium and silicon disulﬁdes: Structure and synthesis. Science 1965, 149, 535–537. [CrossRef][PubMed] 29. Klaska, K.H.; Eck, J.C.; Pohl, D. New investigation of willemite. Acta Crystallogr. 1978, B34, 3324–3325. [CrossRef]

30. Oribe, A.; Tanaka, K.; Morikawa, H.; Marumo, F. Reﬁnement of Crystal Structure of Zn2GeO4; Report of the Research Laboratory on Engineering Materials; Tokyo Institute of Technology: Tokyo, Japan, 1987; pp. 7–12. 31. Chichagov, A.V.; Belonozhko, A.B.; Lopatin, A.L.; Dokina, T.N.; Samokhvalova, O.L.; Ushakovskaya, T.V.; Shilova, Z.V. Information-calculating system on crystal structure data for minerals (MINCRYST). Kristallograﬁya 1990, 35, 610–616. [CrossRef] 32. Brugger, J.; McPhail, D.C.; Wallace, M.; Waters, J. Formation of willemite in hydrothermal environments. Econ. Geol. 2003, 98, 819–835. [CrossRef] 33. Wood, S.A.; Samson, I.M. The aqueous geochemistry of gallium, germanium, indium and scandium. Ore Geol. Rev. 2006, 28, 57–102. [CrossRef] 34. Melcher, F. The Otavi Mountain Land in Namibia: Tsumeb, germanium and snowball earth. Mitt. Österreichischen Mineral. Ges. 2003, 148, 413–435. 35. Dutrizac, J.E.; Jambor, J.L.; Chen, T.T. Host minerals for the gallium-germanium ores of the Apex mine, Utah. Econ. Geol. 1986, 81, 946–950. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons by Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).