Third Worldwide Occurrence of Juangodoyite, Na2Cu(CO3)2, and Other Secondary Na, Cu, Mg, and Ca Minerals in the Fore-Sudetic Monocline (Lower Silesia, SW Poland)

Na-Cu carbonates are relatively rare secondary minerals in weathering zones of ore deposits. Hereby we describe mineral composition and crystal chemistry of the most important secondary (Na)Cu minerals and their Na- and Mg-bearing associates forming rich paragenesis in Rudna IX mine. A non-bulky Ca-rich dripstone-like paragenesis from Lubin Główny mine is also characterized, using Powder X-Ray Diffraction, Rietveld, and Electron Microprobe methods. Light blue juangodoyite (3rd occurrence worldwide) and darker chalconatronite are the most important members of the Rudna IX paragenesis, being associated with malachite, aragonite (intergrown with hydromagnesite and northupite), and probably cornwallite. Most of the minerals are chemically close to their ideal composition, with minor Mg substitution in malachite. Cu chlorides are mainly represented by clinoatacamite and probably herbertsmithite. Additional, minor phases include trace Cu minerals langite, wroewolfeite, and a lavendulan-group mineral, and monohydrocalcite. Separate halite-rich encrustations are shown to be filled with eriochalcite, ktenasite, and kröhnkite. The most likely to be confirmed coexisting species include paratacamite, wooldridgeite/nesquehonite, johillerite, melanothallite, and kipushite. The Lubin paragenesis mainly comprises aragonite, gypsum, rapidcreekite, and monohydrocalcite, with trace vaterite. Blue colouration is mainly provided by a yet unspecified Ni-, Co-, Mg-, and Mn-bearing Cu-Zn-Ca arsenate mineral close to parnauite.


Introduction
Juangodoyite, Na 2 Cu(CO 3 ) 2 (monoclinic, P2 1 /a), is an exceedingly rare mineral, known-prior to the current study-from a single locality in Chile: Santa Rosa mine, Iquique Province [1]. It is the natural equivalent of sodium di(carbonato)cuprate(II), named to honour Juan Godoy-the discoverer of the famous Chilean silver occurrence at Chañarcilllo. The synthetic equivalent of juangodoyite is obtained as small pinacoids, with dimensions~0.1 mm, slightly elongated along the c axis, and somewhat irregular [2]. The hydrated, or trihydrate, counterpart of juangodoyite-chalconatronite (monoclinic, P2 1 /n)-is more common, with about 20 localities worldwide. Its structure was analyzed

Materials and Methods
The locations of the Rudna IX and Lubin Główny mines against the Fore-Sudetic Monocline and Poland are shown in Figure 1. The sampling points of the Lubin minerals under scope are shown in Figure 2.
Samples of the Rudna IX mine come from mine field No. 23 at a site located at the depth of ca. 1100 m below ground level. Here, the chalcocite-rich mineralization under scope occurs at a point of contact between sandstones and Cu-bearing shale. In the Lubin Główny mine, mineral samples were collected from mining excavations located in a vicinity of Chamber C1B. Here, flowstones are developed in places where groundwaters outflow from laminated dolostone contacting with sandstone layers. The samples include thin blue encrustations (up to several centimeters in diameter) on sandstones and dolostones covered by gray (locally white) protrusions. The protrusions may reach~3.8 mm in height and comprise intergrown, gray crystals. Single, colorless, clearly monoclinic crystals are also found disseminated within the blue portions.
Minerals 2020, 10, x FOR PEER REVIEW 3 of 23 chlorides and sulphates or with carbonates contained in ore bearing rocks. This process leads to formation of numerous interesting mineral species. In total, about 30 supergene minerals were described from these mines [36,37]. The most common are hydroxyhalides (e.g., botallackite, atacamite, clinoatacamite, paratacamite) and Cu sulfates (e.g., chalcanthite). Secondary arsenates (annabergite, erythrite, and possibly geminite) and carbonates (cerussite, malachite, azurite, and chalconatronite) are much less common [38]. This preliminary paper approximates mineralogical composition and mineral chemistry of Na-Cu carbonates and their major associates derived from the Rudna IX mine. Attention is also paid to Ca-rich paragenesis from the Lubin Główny mine. Selected results of related to the current study were shown in a conference [39]. Further studies including field, geochemical, and a more detailed crystallochemical research are planned.

Materials and Methods
The locations of the Rudna IX and Lubin Główny mines against the Fore-Sudetic Monocline and Poland are shown in Figure 1. The sampling points of the Lubin minerals under scope are shown in Figure 2.     The determination of the mineralogical nature of the Cu-bearing samples began with Powder X-Ray Diffraction (PXRD) technique. Powdered sample was placed in a Bruker D8 ADVANCE equipped with superfast linear position-sensitive detector (LPSD) VÅNTEC-1 and kβ-filtered CoKα radiation. The apparatus is located in the Institute of Geological Sciences, Polish Academy of Sciences (Cracow section, Poland). The samples were scanned using 0.02 2θ increment and 1 s/step counting time (equivalent of 416 s in the zero-dimensional detector language), in the 3-80 2θ range. Some additional samples were analyzed using a Thermo Electron X'TRA diffractometer at the same location, using similar conditions but 0.3 • /min rate. Unit cell parameters were calculated using the Rietveld method implemented in the TOPAS v. 3.0 software. The macroscopic features of the mineral assemblages and minerals themselves were studied using a camera-equipped binocular. The photo layering (stacking) technique was involved to obtain the best-focused images. The crystal chemistry and spatial relation of the particular minerals were analyzed using electron microprobe (EPMA), using a CAMECA SX100 apparatus operating in the WDS mode and located in the Inter-Institution Laboratory of Microanalysis of Minerals and Synthetic Substances, Institute of Geochemistry, Mineralogy and Petrology, Faculty of Geology, University of Warsaw, Warsaw, Poland. The beam current and size were variable depending on the particular minerals analyzed (see tables for details); the acceleration voltage was 15 kV. Sodium, as a light element prone to fast escape during analyzing, was measured using time-zero intercept method. In this method, the sodium concentration is measured at various steps of the full analysis and then approximated to time of zero. The following standards were used: albite (Na), baryte (S), celestine (Sr), crocoite (Pb), CoO (Co), NiO (Ni), cuprite (Cu), diopside (Mg, Ca, Si), Fe 2 O 3 (Fe), GaAs (As), InSb (Sb), orthoclase (Al, K), rhodonite (Mn), sphalerite (Zn), tugtupite (Cl), and YPO 4 (P).    The azure crystals are spear-like and commonly intergrown into more or less compact masses that may surpass 4 mm in diameter. In some samples, the interiors of the azure crystals are either partially or completely replaced by the light blue phase. Various stages of chalconatronite replacement are observed ( Figure 4).   The dark green phase forms aggregates with single botryoids under 0.2 mm in diameter; no individual crystals were observed among them. This phase occurs between and beneath the two blue species (Figure 3a). Rare examples concern the coarse-botryoidal green phase associated with closely-packed azure crystals. The white socket-like aggregates also stand for an important matrix constituent, where they associate with the green phase. Their greenish parts are locally somewhat rounded (cauliflower-like in shape). In a few specimens, these sockets are stained bluish (Figure 3c). Close examination of a side part of a single specimen of dolomite allowed us to pinpoint an additional type of substance that occurs as intergrown pseudotrigonal crystals of teal colour, somewhat resembling dioptase (Figure 3d).

Mineral Composition and Habit
The PXRD analysis of the Rudna IX minerals results are juxtaposed in the Table 1. As some of the samples comprise various types of aggregate, some minerals reported in the above quantitative data may represent random accumulations of particular species. This is especially the case of herbertsmithite, which was not yet detected during the EPMA sessions. The issue addressed reflects the large level of mineralogical variability in the material studied.
The light-blue phase is usually juangodoyite. The dark blue one and the dioptase-like one is chalconatronite. However, at least a single case of the azure crystals (sample CuR4, which also bears orange baryte) were shown by the PXRD to be juangodoyite, too. As such, the precise identification of the mineralogical nature of both phases proves to be complex and blurred by the hydration-dehydration dependence of the two minerals. The green botryoidal mineral is usually malachite. However, the rare, coarser (up to~0.20 mm) botryoids associated with azure juangodoyite are cornwallite, one of the natural polymorphs of Cu 5 (AsO 4 ) 2 (OH) 4 . This phase may cover areas over 2 mm long.
The white socket-like aggregates are mainly composed of gypsum and aragonite. Hydromagnesite, Mg 5 (CO 3 ) 4 (OH) 2 ·4H 2 O, and the rare mineral northupite, Na 3 Mg(CO 3 ) 2 Cl, are present, too. The greenish and bluish parts of the socket-like aggregates are colored by a Cu-Cl-O species-with clinoatacamite and herbertsmithite, respectively. Separate clinoatacamite-dominant zones surrounded by thin halite/gypsum linings on a sandstone matrix (e.g., sample CuR6) may be as large as 2 mm in diameter.
A second batch of two, small samples, representing halite encrustations filled with some further secondary Cu minerals most likely derived from another, currently unknown section of the mine, was also studied in a preliminary manner. These samples bear eriochalcite, CuCl 2 ·2H 2 O, ktenasite, Zn(Cu,Zn) 4 (SO 4 ) 2 (OH) 6 ·6H 2 O, and kröhnkite, Na 2 Cu(SO 4 ) 2 ·2H 2 O (sample RSR9-1); or are kröhnkite-dominant (RSR9-2). A single sample (RSR9-3) of a dolomitic shale covered by compact, light-blue material, proved to bear further interesting species associated with chalconatronite-a lavendulan-group species, with best PXRD fit for likely lavendulan sensu stricto, NaCaCu 5 (AsO 4 ) 4 Cl·5H 2 O. The mixture of the two is covered by spray-forming needle-like aragonite intergrown with hydromagnesite (sample RSR9-4-a sub-sample of the RSR9-3 one). Results of a more detailed study of both these phases and cornwallite, lavendulan-group phase, kröhnkite, and eriochalcite will be published elsewhere.
A number of additional mineral phases were tentatively determined in the course of the whole PXRD data treatment, as reported in the Table 1. State of art allows us to mention those of the uncertain species which are most likely to occur in the material studied considering both the initial PXRD identification and local geochemical conditions. The list is opened by wooldridgeite (samples CuR1, CuR1A, and CuR6), a very unique Na-Cu pyrophosphate which may, however, be coincident with nesquehonite, MgCO 3 ·3H 2 O. The latter phase conjures up the identification of chemically related hydromagnesite and northupite. The same is true, e.g., for wegscheiderite, Na 5 H 3 (CO 3 ) 4 (two other Na carbonates-trona and nahcolite-were, indeed, observed in some additional samples analyzed by E.S.). Possible coexisting Cu minerals include paratacamite, atacamite; melanothallite, Cu 2 OCl 2 , kipushite, (Cu,Zn) 5 Zn(PO 4 ) 2 (OH) 6 ·H 2 O, and the langite-and wroewolfeite-related mineral posnjakite, Cu 4 (SO 4 )(OH) 6 ·H 2 O. A likely phase is rhomboclase, (H 5 O 2 )Fe(SO 4 ) 2 ·2H 2 O, is genetically related to melanterite (e.g., [40]).

Lubin Główny Minerals
The microphotographs of the Lubin Główny paragenesis are juxtaposed in Figure 4. The Lubin Główny paragenesis is much less complex than the former one (Table 2). Table 2. Results of the PXRD analysis of the Lubin Główny mine secondary minerals. R wp = 16.97%, GOF (χ 2 ) = 1.53% 1 wt.% contents reported based on Rietveld quantitative phase analysis (QPA); slight deficiency is due to initial inclusion of minor and/or uncertain species in the refinements (see main text for details); 2 for the QPA model; 3 a mixed sample (various types of aggregates).
As opposed to prior suggestions (e.g., [38]), the blue coloration of the thin encrustations (sample MSCu-1, Figure 5a) does not come from geminite, Cu(AsO 3 OH)·H 2 O. Instead, these aggregates are composed of monohydrocalcite, aragonite, gypsum, quartz, and minor calcite. Vaterite and dolomite are trace species here. The gray protrusions are mainly comprising gypsum. The sample Cu4 (Figure 5b) also bears sprays-up to~2.4 mm in diameter-composed of colorless needles of rapidcreekite. Single needles are up to 1 mm long. Such spray-forming needle-like crystals of rapidcreekite were also mentioned in [38]. The coloring agent is mainly a not yet fully analyzed CuZnNiCoCa(MnMg) arsenate. Brushite, Ca(PO 3 OH)·2H 2 O, although suggested by the Rietveld method to be present at~2 wt.% level, could not be found in the material separated for the EPMA study. A very small P impurity was only detected in a few EDS spectra of gypsum. The thick, blue to greenish-blue, bulky encrustations (sample Cu4) are also composed of aragonite, with gypsum, calcite, dolomite, and monohydrocalcite as coexisting or trace components. The identification of moolooite, Cu[(COO) 2 ] 2 ·0.4H 2 O, needs further examination.

Lubin Główny Minerals
The microphotographs of the Lubin Główny paragenesis are juxtaposed in Figure 4. The Lubin Główny paragenesis is much less complex than the former one ( Table 2).
As opposed to prior suggestions (e.g., [38]), the blue coloration of the thin encrustations (sample MSCu-1, Figure 5a) does not come from geminite, Cu(AsO3OH)•H2O. Instead, these aggregates are composed of monohydrocalcite, aragonite, gypsum, quartz, and minor calcite. Vaterite and dolomite are trace species here. The gray protrusions are mainly comprising gypsum. The sample Cu4 (Figure 5b) also bears sprays-up to ~2.4 mm in diameter-composed of colorless needles of rapidcreekite. Single needles are up to 1 mm long. Such spray-forming needle-like crystals of rapidcreekite were also mentioned in [38]. The coloring agent is mainly a not yet fully analyzed CuZnNiCoCa(MnMg) arsenate. Brushite, Ca(PO3OH)•2H2O, although suggested by the Rietveld method to be present at ~2 wt.% level, could not be found in the material separated for the EPMA study. A very small P impurity was only detected in a few EDS spectra of gypsum. The thick, blue to greenish-blue, bulky encrustations (sample Cu4) are also composed of aragonite, with gypsum, calcite, dolomite, and monohydrocalcite as coexisting or trace components. The identification of moolooite, Cu[(COO)2]2•0.4H2O, needs further examination.

Spatial Relations, Crystal Chemistry and Unit Cell Parameters of the Minerals
The BSE images of both locality samples are juxtaposed in Figure 6. Both juangodoyite and chalconatronite bear malachite and chalcocite inclusions (Figure 6a,b). Aragonite and quartz may be included in chalconatronite (Figure 6b). Malachite either occurrs as usually rounded to botryoidal, small inclusions, or larger, rounded, spherolitic aggregates are found outside the Na-Cu carbonates. The botryoids may reach 50 µm in diameter. Malachite may be interstitial (Figure 6a-c). Still larger, elongated malachite aggregates, reaching~250 µm in length, were also observed. The socket-forming aragonite is a core species, overgrown by hydromagnesite and northupite (Figure 6d). The northupite and hydromagnesite zones alone can be as large as 200 µm in width, with aragonite cores >400 µm in diameter.

The Rudna IX Minerals Data
Juangodoyite and chalconatronite are chemically very pure and have very similar empirical formulas. The crystal chemistry of the light blue phase, juangodoyite (Table 3), is described as  Table 5); (d) A fragment of white socket-like aggregate with aragonitic core overgrown by hydromagnesite (Hdm) and northupite(Nth)-bearing mixture; (e) Large spray-forming crystals of rapidcreekite (Rpc) growing on aragonite/monohydrocalcite (Ar/Mhc) with very thin ribbons of CuZnNiCoCa(MnMg) arsenate (As); calcite (Cc) crystals also visible; (f) Dolomite-rich sample overgrown by aragonite dripstone covered by the mentioned arsenate (As).

The Rudna IX Minerals Data
Juangodoyite and chalconatronite are chemically very pure and have very similar empirical formulas. The crystal chemistry of the light blue phase, juangodoyite (Table 3), is described as Na 2.41 Cu 1.00 (CO 3 ) 2.21 (n = 7, i.e., first aggregate). The corresponding formula for a second aggregate (n = 3) is Na 2.55 Cu 1.00 (CO 3 ) 2.28 . The two formulas bear average 0.48 surplus Na, in spite of the use of the time-zero intercept measurement method. After the stoichiometric normalization of wt.% Na 2 O content to the ideal content, the formulas take the form of Na 2.00 Cu 1.00 (CO 3 ) 2.00 ·3H 2 O. Chalconatronite analyses are juxtaposed in Table 4. The corresponding formula for the azure, spear-like crystals of chalconatronite is Na 2.00 Cu 1.00 (CO 3 ) 2.00 (n = 10; anhydrous part due to almost complete dehydration due to beam energy) after the normalization of wt.% Na 2 O to the ideal composition of the species. Un-normalized analysis gives mean 0.33 apfu Na excess. The analysis of two greenish-blue dioptase-like crystals of chalconatronite gave the following empirical formula (using 3-cation-basis, i.e., Σ(Na,Cu) = 3, recasting basis): Na 1.93-1.96 Cu 1.07-1.04 [(CO 3 ) 2.03-2.00 (SiO 3 ) 0.01-0.02 ] Σ2.04-2.02 (anhydrous basis due to mineral destruction). Similar wt.% data for juangodoyite and chalconatronite have two sources: (1) the dehydration of chalconatronite under the electron beam, leading to an equivalent of juangodoyite, and (2) pseudomorphic relation of the two species, constricting macroscopic species distinction. The deviation of the listed wt.% data from comparative data of the Handbook of Mineralogy is, again, due to mineral's instability in the microprobe column. The Na-Cu carbonates are closely associated with slightly magnesian malachite (Table 5). Large chalconatronite crystals were observed to be intergrown with quartz (which may include the latter) and bearing inclusions of chalcocite ("type 1", Figure 6b). Some chalcocites bear numerous, tiny inclusions of chemically pure pyrite or marcasite. Chalcocite may also enclose strontian aragonite with single inclusions of a Cu-O-Cl phase (non-analytical). Malachite is either found to be directly overgrowing chalconatronite, including it, or residing between quartz and chalcocite inclusions in chalconatronite (Figure 6c). The first type bears less Mg (analyses 1-5, Table 5); its empirical formula based on a larger analytical representation (n = 9) is (Cu 1.92 Mg 0.04 Na 0.04 ) Σ2.00 (CO 3 ) 1.00 (OH) 1.96 . This, when omitting Na, corresponds to the mean end-member representation of Mal 98 Mgc 2 where Mal stands for malachite and Mgc for mcguinnessite. Second-type malachite, forming much larger, ovoid-shaped but also typical, rounded aggregates ("type 2", Figure 6b), is more magnesian: (Cu 1.85 Mg 0.10 Na 0.05 ) Σ2.00 (CO 3 ) 1.00 (OH) 1.95 . This formula corresponds to normalized (Na-devoid) representation of Mal 95 Mcg 5 . An additional, very minor phase was observed as thin (up to ca. 9 µm thick and ca. 45 µm in length), lobate, curved accumulation at a malachite-chalcocite interface (Figure 6c, phase "U"). Only a single analysis could be obtained. Its chemical composition (analysis no. 12 in Table 5) does not fit to any known mineral species.  (1) 17.09 2.14 1 Analyses 1-5: low-Mg variety; analyses 6-11: higher-Mg variety (analyses 6-8 and 9-11 for two separate aggregates); 2 Al, Mn, Zn, and K were analyzed but not detected; 3 backward calculation from amount of CO 3 2set as ideal content; 4 by difference (100-Total1-wt.% CO 2 ); 5 occupancy-normalized amounts; 6 by charge balance; 7 normalized to Cu-and Mg-dominant end-members only.
A very interesting, though relatively subordinate, Cu-bearing K-Na-Ca-Mg(Fe) aluminosilicate phase is occasionally found as an inclusion in juangodoyite (Figure 6a). The results of the EPMA analyses of this phase are reported in Table 6. This phase is somewhat reminiscent of the Cu-bearing glauconite of [42] which, however, has >7 wt.% MgO, >6 wt.% Na 2 O, and just up to 2.02 wt.% CuO. Recasting the first analysis to muscovite stoichiometry gives the empirical formula of (K 0.67 Cu 0. 30  formula. The X site occupancy in both cases is calculated by stoichiometry, charge balance, and according to the current nomenclature of the mica group. The second analysis fits to the Al-dominant analogue of oxyphlogopite-a potentially new mineral. This mineral is, however, another case needing further study to confirm its suggested mica-group membership, site occupancy and Cu and Fe valency. This mica-like mineral shows somewhat phengitic or aluminoceladonite-like composition; indeed, [43] also described a Cu-bearing phengite from the Monocline area, suggesting ideal composition of K 2 Mg 2 AlCu 2+ (Fe 3+ AlSi 6 O 20 )(OH) 4 . However, our species do not recast to such stoichiometry.  Table 7) corresponding to Arg 95 Ntr 3 Srt 1 Mgs 1 where Arg is aragonite, Str is strontianite, Mgs is magnesite end-members, and Ntr is natrite (Na 2 CO 3 ) equivalent.
The occurrence of Na admixture in the aragonite seems somewhat aberrant, but the areas and aragonite individuals analyzed are large and devoid of visible inclusions. Also, Na + distribution in calcites and aragonites is mentioned by [44]. Some separate, relatively large grains of core-forming aragonite (Figure 6d) are locally intergrown with BSE-dark phase-hydromagnesite, Mg 5 (CO 3 ) 4 (OH) 2 ·4H 2 O-and a medium-dark external (rim-forming) phase of a composition suggesting northupite, Na 3 Mg(CO 3 ) 2 Cl. The latter is probably a mixture or northupite and magnesite. The crystal chemistry of hydromagnesite (n = 3, derived from data in Table 8) may be expressed as (Mg 4.90 Cu 0.08 Ca 0.02 ) Σ5.00 (CO 3 ) 4.00 (OH) 2.00 ·4.80H 2 O. This leads to 98% of the hydromagnesite member content, the remainder being attributed to unknown Cu-and Ca-dominant analogues. The water content deviation from the ideal content may be related to both material destruction under the electron beam and the porous nature of the aggregates. The latter is also confirmed by analyses no. 4-6 in the same table, with apfu data calculated the same way as for hydromagnesite. The identity of the phase related to the latter three analyses is, however, unclear, especially as no higher-hydrate counterparts of hydromagnesite, like giorgiosite, were found in course of the PXRD analysis.
The following ratios were calculated and compared for the northupite-like substance under study (n = 8; Table 9) and northupite data from [41] 1.08 . The latter procedure is preceded by the calculation of the positive charge, the subtraction of the Cl − charge, the balancing of the M 2+ site to occupancy of 1.00 (1 − apfu(Cu) − apfu(Ca) = Mg northupite ), the balancing of the remaining charge by the CO 3 2− groups, and finally relocating the remaining CO 3 2− and Mg 2+ to MgCO 3 . This is equal to Mg distribution derived from ideal Na/M ratio for northupite, calculated by proportion. The carbonate ion inclusion in the above formulas is argued by close association with aragonite. The only other anion that could exist in place and would explain large wt.% deficiency is OH − , which is unlikely (Na-Mg-OH-Cl phase is unknown). Halite is also unlikely to exists as a part of the mixture studied due to non-congruent Na/Cl ratio. The northupite-related analyses were also recast with one (Mg + Cu + Ca) cation basis, which does not give good results.   (9) 8.372 (9) 114.54 (25) 11.79 1.04 1 S, Al, Fe, Mn, Zn, Sr, Na, K, and Cl were analyzed but not detected; 2 backward calculation from amount of CO 3 2− set as ideal one; 3 calculated after backward wt.% CO 2 calculation, as wt.% H 2 O = 100-total-(wt.% CO 2 ); 4 occupancy-normalized ideal amounts; 5 by difference, after subtracting apfu(H) attributed to OH − .
The generally white accumulations are locally greenish due to inclusions of either single microcrystals or veinlets of a Cu-Cl-O phase. According to the PXRD data, this phase is represented by herbertsmithite (or an isostructural substance). However, no Zn-or Mg-bearing Cu-Cl-O species were found in the material portions selected for the EPMA study. Instead, the EPMA analyses (Table 10) recast to clinoatacamite stoichiometry, which is in compliance to the PXRD data for other portions of the material. The recasted EPMA data, indeed, give the OH:Cl ratio of 2.99:0.99. Clinoatacamites analysed in two separate samples, CuR1B and CuR3, are chemically very similar and their chemistry may be expressed as (Cu 1.95 Mg 0.05 ) Σ2.00 (OH) 3.00 Cl 1.00 (n = 11), and (Cu 1.97 Mg 0.03 ) Σ2.00 (OH) 3.03 Cl 0.97 (n = 6), respectively. The formulas suggest a respective 2% and 1% mean part of "anhydrous korshunovskite counterpart", Mg 2 (OH) 3 Cl. The unit cell parameters of juangodoyite from different samples vary only by up to 1%. Just in a single case-also in terms of the whole unit cell parameters calculated within the current study-the difference reaches 7%. The unit cell edges calculated for the different samples of chalconatronite are both similar to each other and to the synthetic material data. A maximum difference for the β angle is 0.14% (inter-sample comparison) or 0.57% (sample-literature data comparison). The malachite crystallography data also do not stand out. The unit cell parameters of aragonite show very low to low differences in both the intra-sample and sample-literature comparisons-the variations are expected due to the below-described substitutions. The data obtained for hydromagnesite and northupite are also very similar to those available in the literature ( [40]). Although pure (as shown below), clinoatacamite shows slightly higher intra-sample variation ranges, being 1.1% for the parameter β. The value derived for halite present-together with possible traces of eriochalcite-in the matrix of chalconatronite and juangodoyite is a = 5.9429(7) Å; and that for the material of the sample CuR6 is 5.6416(2) Å. The second-batch halite unit cell parameter a is equal to 5.6400(2) (sample RSR9-1) and 5.637(2) Å (sample RSR9-2).

The Lubin Główny Minerals
The Lubin Główny rapidcreekite is chemically very pure (Table 11). Its empirical formula, based on 16 datapoints, is Ca 2.00 (SO 4 ) 0.97 (CO 3 ) 1.03 ·4.23H 2 O. The main component of the Lubin Główny assemblage-monohydrocalcite-is, as opposed to rapidcreekite, enriched in Cu, Zn, and possibly As. (Table 12). Most of the analyses were derived from a profile of datapoints, and the profile line was set in a microarea free of visible inclusions of the mentioned CuZnNiCoCa(MnMg) arsenate. However, juxtaposing As 2 O 5 -CuO, As 2 O 5 -ZnO, and CuO-ZnO (but not As 2 O 5 -SO 3 ) in a Pearson-correlation manner suggests strong positive trends, although not that evident for the As 2 O 5 -ZnO pair. As such, we suspect the monohydrocalcite itself of being devoid of As, and bearing small amounts of Zn, Cu, and S-as shown in Table 12. The analyses listed here concern spots with either lacking or minor As amounts.
The unit cell parameters calculated for rapidcreekite resemble those found in the literature (both References [31] and [40]). The same is true for monohydrocalcite. Slight difference in the latter case may be related to minor substitutions, as shown below. The relatively BSE-bright aragonite has the following unit cell parameters: a   The arsenate phase either forms very thin, ribbon-like, curved veinlets within monohydrocalcite-aragonite aggregates, or tiny (up to~60 µm in length) crystals included in rapidcreekite (Figure 6e). However, compact aggregates, up to 20 µm thick, covering aragonite ( Figure 6f) are more frequent. The analyses of the arsenate phase are to be found in Table 13. The manipulation of the measurement conditions, both the beam current and size, does not impact any larger changes in the wt.% results. Two slightly different crystallochemical types of the species were identified, coming from two separate fragments of the dripstone material. The first one (analyses 1-5) is slightly more cuprian (31.59 wt.% versus 26.13 wt.% CuO) and arsenian (20.23 wt.% versus 18.59 wt.% As 2 O 5 ), but less calcian (4.02 versus 6.58 wt.% CaO) than the second variety. The latter, coming from a dolomite-rich sample, is also characterized by a slightly elevated average SiO 2 content-0.38 wt.% as compared to 0.25-and may bear more Na. It is also slightly more magnesian (mean 0.59 wt.% MgO against 0.44) and clearly more calcian (mean 6  soda glass. The Rudna IX Na-Cu carbonates were likely formed via an interaction of saline, CO 2 -rich (and thus possibly also Na 2 CO 3 -saturated) mine waters with chalcocite, tennantite, and other Cu ore minerals undergoing weathering. The holotype rapidcreekite is not the only product of surface weathering. The other minerals of this genesis include aragonite, gypsum, hydromagnesite, nesquehonite, and jarosite-group species [25]. The Diana Cave rapidcreekite occurs in a fault line with a hot spring (51 • C on average). The spring waters are almost pH-neutral and rich in SO 4 2− , Cl − , Na + , and Ca 2+ ions [31].
Such a description in reminiscent of the seepage phenomenon observed in the Lubin Główny mine. The Diana Cave rapidcreekite is slightly impure, with an empirical formula given as (Ca The second reaction concerns equilibrium between CO 3 2− (solution) , SO 4 2− (solution) and CO 3 -SO 4 -bearing solid(s), derived from [44]. The [44] team also cited [49,50] who dealt with rapidcreekite crystallization conditions, e.g., during nanofiltration of saturated CaSO 4 solution in the presence of CO 3 2− ions. According to these authors, rapidcreekite, as a phase more soluble than calcite, needs lower aCO 3 2− to precipitate. Example activity conditions favouring rapidcreekite crystallization were described with a CO 3 2− /SO 4 2− ratio of 4.5 × 10 −3 . Rapidcreekite crystallization at the Diana Cave is suggested to take place at pH > 6.4 (i.e., above gypsum precipitation conditions) and CO 3 2− /SO 4 2− ratio of 0.06-0.46. As such, the stability range of rapidcreekite is narrow, as also confirmed by the abundance of gypsum and calcite in the Lubin samples, suggestive of variations in the local conditions. The rapidcreekite crystallization window is located between the preceding gypsum and postdating CaCO 3 crystallization stages. Interestingly, the "CaCO 3 " phase of the third reaction was interpreted by Onac's team [44] as aragonite or vaterite-both also present in Lubin. Indeed, these two CaCO 3 polymorphs are stabilized by Mg 2+ ions contained in solutions. Such ions may have easily been mobilized from the Lubin dolostones. The Na-Cu carbonate paragenesis from the Rudna IX mine most likely formed due to an interaction of saline and/or sodic and carbonaceous mine waters with the copper ores. The rapidcreekite-and CaCO 3 -bearing dripstones of the Lubin Główny mine undoubtedly resulted from an interaction of the local, warm mine water seepage, with wall-forming rocks. Such outflows' waters were also studied by [51] who reported them as weakly mineralized, SO 4 2− -CO 3 2− ones, with dissolved NaCl content increasing with depth. Calcium and carbonate ions may have been either derived from the seepage or from dolostone, while sulfate ions may have been derived from seepage or Cu sulfide ore weathering. Copper, nickel, cobalt, and arsenic were mobilized from arsenide and sulfarsenide ores disseminated within the dolostone. The most probable source of Mn is both the dolostone and local calcite fillings. We have confirmed the Mn enrichment of this calcite. Although the stochiometry and structural nature of the rapidcreekite-associated arsenate species remain unknown, the proposed CO 3 2− substitution has an important argument: the simultaneous Na enrichment of the mineral. The inclusion of Na + induces a drop in positive charge. The sole occupancy of the T site by AsO 4 3− and SiO 4 4− group would cause surplus negative charge. The presented microchemical data for malachite are similar, as the Na presence reported is repetitive. The charge may be balanced by the CO 3 2− . Coupled Na + and CO 3 2− substitution in the mentioned apatite group is reported by [45]. This, however, concerns the easily substituted Na + -Ca 2+ pair. As opposed to that, Na-Cu diadochy is almost unknown among minerals, with exceptions in channel sites in piypite (e.g., [52]) and to some extent in barahonaite-group minerals (e.g., [53]). The same is true for K-Cu pair, which is only confirmed in channels of aleutite [54]. A (Zn,Na,Cu,Mg) diadochy is also known in majzlanite [55]. Nevertheless, possible Na-Cu and K-Cu substitution systems in our minerals need further studies to be confirmed or disproved.