Eur. J. Mineral. 1997, 9, 43-51
Pb2Fe3+Cb(OH)4-H20, a newly discovered natural phase from Tuscany, Italy: physico-chemical data, crystal structure, and OD character MARCO PASERO*1'**, NATALE PERCHIAZZI^, SIMONA BIGI (2) , MARCO FRANZINI^ and STEFANO MERLINO (1)
(1)
(2)
Dipartimento di Scienze della Terra, Università di Pisa, Via S. Maria 53, 1-56126 Pisa, Italy
Dipartimento di Scienze della Terra, Università di Modena, Via S. Eufemia 19, 1-41100 Modena, Italy
Abstract: A natural phase with chemical formula Pb2Fe Cl3(OH)4-H2O, monoclinic, space group P2\, a = 8.033(5), b = 6.253(3), c = 9.221(6) Å, ß = 102.98(8)°, was recently discovered within the Etruscan iron slags of Baratti. That phase, denoted as PFC, falls in the category of "geologically modified anthropogenic substances" and therefore cannot be regarded as a new mineral. Full chemical and physical data are given for PFC. Twinning on (001) is ubiquitous. The structure of PFC was solved and refined up to R = 0.082 for 1119 independent reflections collected with MoKa radiation on a twinned crystal. PFC displays a novel structural type, in which layers of face-sharing [PbXçj] polyhedra alternate with layers containing columns of edge-sharing [Fe(OH)ô] octahedra and structural voids. PFC belongs to a family of OD structures formed by equivalent layers. The OD character of PFC is fully described. Key-words: PFC, crystal structure, OD-structures, hydroxychlorides, Baratti (Italy).
Introduction The mineralogical locality of Baratti (southern Tuscany, Italy) has been recently discovered and thoroughly described (Franzini et al, 1992; Franzini & Perchiazzi, 1992). In that locality 43 different mineralogical species were found, mainly rare lead and copper oxy- and hydroxyhalides, sulfates and carbonates. For a number of them Baratti represents the first Italian occurrence, and the second in the world, the first being Laurion (Greece). There are striking analogies between Baratti and the renowned mineralogical locality of
Laurion. At both occurrences the alteration minerals are hosted within ancient metallurgical slags, smelted by the Etruscans at Baratti and by the Athenians at Laurion. The minerals represent the result of the interaction between the ancient slags and the sea-water. During a systematic inspection, by X-ray powder diffraction, of the crystals collected within the slags of Baratti, we found some crystals with a striking lemon-yellow colour, whose powder pattern was quite different from that of all known minerals. Single-crystal X-ray photographs exhibited monoclinic symmetry and unit cell parameters never reported so far. A qualitative
*e-mail:
[email protected] 0935-1221/97/0009-0043 $ 2.25 © 1997 E. Schweizerbart'sche Verlagsbuchhandlung. D-70176 Stuttgart
44
M. Pasero, N. Perchiazzi, S. Bigi, M. Franzini, S. Merlino
EDS chemical analysis revealed the presence of lead, iron and chlorine among the elements with Z > 10. It was clear we were dealing with a new phase, namely a chloride or hydroxychloride of lead and iron. A proposal for the definition of this phase from Baratti as a new mineralogical species was there fore submitted to the IMA Commission on New Minerals and Mineral Names. The proposal has been rejected, in keeping with the guidelines re cently stated by the CNMMN itself (Nickel, 1995), because of the role played by humans in producing the slags, which causes our phase to fall within the category of "geologically modified anthropogenic substances". Nevertheless, the amount of information we obtained on our new phase could be of some interest for the miner alogical science for the following main reasons: i) this phase is the first example of lead and iron hydroxychloride [among minerals, only the lead and iron arsenite-chloride nealite shows loose chemical resemblances with it (Giuseppetti et aU 1993)]; ii) it has a previously unreported structural type; iii) it belongs to a family of OD structures formed by equivalent layers. The OD character so far observed within a lot of important families of silicate minerals [ex amples are reported in Merlino (1990), Bonaccorsi et al (1990), Pasero & Reinecke (1991)] is now being recognized as a very common and uni fying feature also in lead hydroxyhalides from Baratti and Laurion (Merlino et ai, 1993, 1994). Therefore we think it should be useful to de scribe it as it is normally done for new minerals. In the following description we shall refer to this phase using the acronym PFC.
Occurrence PFC was sampled within the slags of Baratti beach, near Piombino, southern Tuscany, Italy. Such slags date back to the Etruscan times and represent by-products of the iron metallurgy. PFC was found associated with goethite, fiedlerite-lΛ, fiedlerite-2M. It was originated by the interaction between sea-water and sporadic lead fusion spherules hosted in the slags.
Physical and optical data PFC occurs as small yellow pseudorhombic crystals, tabular {001}, elongated [010] with rec tangular outline, and dimensions up to 0.3 mm
long and 0.05 mm across. The dominant forms are {001}, {011}, {101}, with minor {100}, {201}, {210}; twinning on (001) is widespread. Crystals are transparent, not fluorescent, with white streak and adamantine lustre. The Mohs' hardness is 2÷3. Cleavage was not observed. Parting along (010) is not perfect. Optically, PFC is biaxial (+), with αmeas = 1.93(5), ßcak =1.95, γcak = 1.98 (in Na light, λ = 589 nm). 2Vmeas = 73°. Dispersion r > v strong. Orientation X = b, Y ~ a, Z normal to (001), Z:c = +13°. PFC is pleochroic, with Y light yellow, X straw yellow. The α value has been measured by prism minimum deviation method, ß and γ were obtained from measure ments of α, 2V, and birefringence. The X-ray powder diffraction pattern of PFC is reported in Table 1. It has been obtained with a Gandolfi camera (diameter 114.6 mm, CuKa radiation, λ = 1.54178 A). The intensities were estimated by eye. The indexing of the lines has
Table 1. X-ray powder pattern of PFC (Gandolfi camera, diameter 114.6 mm, CuKa radiation, λ = 1.54178 Å). Intensities are given as follows: S = strong, mS = medium strong, m = medium, mw = medium weak, w = weak, vw = very weak.
I
4*>s
w S w vw mw mS S w w mw m m
7.84 5.13 4.49 3.965 3.655 3.368 3.322 2.999 2.952 2.702 2.467 2.447
mw
2.287
mw m w
2.166 2.115 2.022
w w w vw w
1.997 1.883 1.862 1.649 1.559
4alc
7.83 5.13 4.49 3.926 3.649 3.347 3.318 2.995 2.951 2.701 2.469 2.446 2.443 2.285 2.278 2.170 2.114 2.030 2.030 2.003 1.879 1.868 1.647 1.558
hkl
100 011 '002
201 012 202 210 003 212 013 213 221 220 222 221 204 014 031 222 320 132 410 232 025
Crystal structure of a lead and iron hydroxychloride
45
Table 2. Electron microprobe analysis (average of 9) for PFC
Constituent
wt. %
Range
Standard
Pb Fe Cl F
61.48 7.97 15.95 0.02
61.05-62.48 7.66 - 8.29 15.71- 16.14 0.00-0.15
PbS/Laurionite FeS2 Sodalite Fluorite
Total:
85.42 (*)
(*) The ideal formula would imply 11.87 wt % O, and 0.9 wt % H, leading to a total of 98.19 wt % of elements. Atomic ratios:
Pb2Feo.90Cl3.03Fo.01
(calculated on the basis Pb = 2)
3+
Simplifiedformula: Pb2Fe Cl3(OH)4 H 2 0
been carried out with the help of the single-crys tal intensity data collection.
Chemical data Quantitative EPMA data were obtained with a wavelength dispersive ARL-SEMQ electron mi croprobe. The working conditions were: acceler ating voltage 15 kV, sample current 20 nA, beam size 20 µm. The interference between PbLα and C\Ka analytical lines was treated according to Donovan et al. (1993). The scarcity of material prevented us from obtaining analytical data on the water content. The results of the microprobe analysis are re ported in Table 2. These data, together with the indications of the structural analysis, suggest Pb2Fe3+Cl3(OH)4 H2O as the correct chemical formula for PFC. As for the valence state of iron, whether Fe 3+ or Fe 2+ , the trivalent state has been inferred as the correct one, on the basis of crystal-chemical reasonings, namely the Fe-O bond distances and the electrostatic valence balance of the oxygens involved in the [Fe(OH)6] coordination polyhe dron (see below). Moreover, it is worth noting that iron assumes its trivalent state also in the associated mineral goethite, as well as in all other iron-bearing minerals found at Baratti (jarosite, plumbojarosite, and delafossite). Crystal structure analysis All the PFC crystals we tested were twinned, having (001) as twin plane. The reciprocal lat tices of the two twin individuals are almost ex
actly superimposed for reflections with h = In. We tried to evaluate the twin fraction, by meas uring at the four-circle diffractometer the inten sities of a set of reflections with h - In + 1 for both individuals, and we obtained always equal percentages of the twins. Least-squares refinement of 29 values of a set of 30 reflections accurately centered at the fourcircle diffractometer gave the unit cell parame ters: a = 8.033(5), b = 6.253(3), c = 9.221(6) Å, ß = 102.98(8)°, V = 451.3(5) A3. A crystal (dimensions 0.17 x 0.13 x 0.05 mm) was eventually chosen for the intensity data collection, which was carried out with an Ital Structures four-circle automatic diffractometer, equipped with graphite-monochromated MoKa radiation. The operating conditions were: 48 kV, 28 mA, ω-29 scan mode, scan range 5°-50°, scan width ± (0.6 + 0.15 tan G)°, minimum scan speed 1.57min, proportionally raised on the basis of the intensity of a pre-scan of _the peaks. A total of 1709 intensities (0 < h < 9, 7 < k < 7,10 < / < 10, plus a selection of equivalent reflections with h < 0) were measured. The reflections with / > 3α(7) were considered as observed, and were used in the least-square calculations after reduc tion for Lorentz and polarization factors. An empirical correction for the absorption effects was performed through DIFABS (Walker & Stu art, 1983); the transmission factors on F were in the range 0.67-1.54. Equivalent reflections were merged (# eq = 0.114 and 0.079 before and after the absorption correction, respectively), thus ob taining a set of 1119 F o 's. The crystal structure of PFC was solved by direct methods, using the SHELXS-86 package
46
M Pasero, N. Perchiazzi, S. Bigi, M. Franzini, S. Merlino
(Sheldrick, 1986). As the Laue symmetry and the systematic absences pointed to P2\/m and P2\ as possible space groups, direct methods were applied in both of them. In both cases the best Emap showed three maxima which, on the basis of the electron density values, were interpreted as due to two Pb and one Fe atoms. Later on, a few least-square refinement cycles (SHELXL-93 package; Sheldrick, 1993) indicated that the noncentric space group P2\ better described the structure. Therefore P2\ was definitely assumed as the correct space group for PFC. In fact, Pbl and Pb2 are related by a pseudotranslation of a/2, thus defining a pseudo sub-cell with halved a axis. This is the reason for all reflections with h = 2n + 1 being much weaker than those with h = 2n. The latter are the so-called family reflections in the OD terminology, and are independent of the stacking sequence (see below). Therefore they are also independent of the twinning. Indeed, we were lucky enough to find at once the position of iron, which "desymmetrized" the structure. Bearing in mind that reflections (hkl) with h = 2ft of the first twin individual are almost exactly superimposed with reflections (h, -k, -l-h/2) of the second twin individual, we corrected each observed structure amplitude F0(hkl) by a multiplication factor C depending on the calculated structure factors: [Fc(M/)]2 C =( )Vi [Fc(hkl)¥ + [Fc(A,-M-Ä/2)]2
Such a correction was first performed with a partial model (only leads and iron), and repeated at the end of the refinement, when the structure was completed. By alternating least-square cycles and Fourier synthesis calculations we found reliable positions for two chlorines, namely Cl 1 and C12, which formed part of the coordination polyhedra of Pbl and Pb2. At this stage, as the position of the iron atoms and the length of the b axis suggested that iron was six-fold coordinated and that octahedra formed edge-sharing chains running along [010], we found the positions of 4 oxygens (Ol to O4, all belonging to hydroxyl groups) by means of a rigid geometrical simulation (DLS76; Baerlocher et al, 1978). The octahedral chains were allowed to rotate, with iron and lead fixed, until a set of reliable Pb-O distances was obtained. The oxygen positions were then refined, by gently lowering the strength of the distance constraints, and finally by removing them at all. A difference Fourier synthesis then revealed the positions of C13 and O5, the oxygen belonging to a water molecule, which completed the coordination polyhedra for Pbl and Pb2 atoms. At this point the anions gave rise to a chemically reasonable and self-consistent set of O • Cl distances, involved in O - H • Cl bonds. Obviously, due to the presence of atoms with much greater scattering power, we did not try to find the hydrogen positions. The final reliability indices were R\ (conventional R factor) = 0.0821, wRi (weighted R factor computed on squared F's) = 0.2131, S (goodness
Table 3. Final fractional coordinates and U parameters (Å2) for PFC. U is the equivalent isotropic displacement parameter converted from anisotropic ones for Pb, Fe, and Cl [Ueq = 1/3 ΣΣ Uij en aj (az-a/)], and the isotropic displacement parameter for O.
X
Pbl Pb2 Fe Cll C12 CΓ3 Ol 02 03 04 05
0.8462(2) 0.3181(3) -0.002(2) 0.137(2) 0.649(2) 0.463(2) -0.063(4) 0.056(5) 0.165(3) 0.169(4) 0.436(8)
(*) heldfixedto define the origin
y 0.2133(*) 0.2555(4) 0.484(1) 0.229(3) 0.226(3) 0.465(2) 0.487(5) 0.442(7) 0.713(6) 0.247(7) 0.950(10)
z
0.8084(2) 0.8119(2) 0.505(1) 0.058(1) 0.065(2) 0.359(2) 0.298(4) 0.732(6) 0.495(3) 0.512(4) 0.284(8)
U
0.0198(7) 0.0227(7) 0.017(1) 0.020(3) 0.022(3) 0.032(4) 0.001(7) 0.024(10) 0.011(6) 0.022(7) 0.053(17)
47
Crystal structure of a lead and iron hydroxychloride Table 4. Bond lengths and selected interatomic distances (Å) for PFC.
Pbl
-cn -Cll -Cll -C12 -C13 -01 -02 -03 -05
2.89(1) 3.26(2) 3.44(2) 3.13(1) 3.04(2) 2.60(3) 2.43(4) 2.78(3) 2.67(6)
Pb2
-C12 -C12 -C12 -Cll -C13 -01 -02 -04 -05
2.92
3.13(1) 3.14(2) 3.49(2) 2.96(1) 3.18(2) 2.67(3) 2.37(4) 2.75(4) 2.62(6) 2.92
Fe
-01 -02 -03 -03 -04 -04
1.86(4) 2.06(5) 1.98(3) 2.14(4) 2.02(4) 2.10(4) 2.03
Ol 01 01
••• C l l
• Cll
3.22(4) 3.41(3) 3.54(4)
02 02 02
•C12 •• Cll • Cll
3.20(5) 3.27(4) 3.22(5)
03 ••• C13 03 ••• C13 03 05
3.32(3) 3.38(3) 3.53(7)
04 - C13 04 • C13 04 05
3.31(4) 3.42(4) 3.53(7)
05 • C13 05 • C13 05 • C13 05 •• Cll 05 - C 1 2 05 •C12
3.08(6) 3.21(8) 3.32(6) 3.31(7) 3.41(6) 3.43(7)
•C12
of fit) = 1.122. The weighting scheme was. w = l/[α2(Fo2) + (0.1412P) 2 + 59.95P)], where P = [Max (F 0 2 , 0) + 2 F c 2 ]/3. Neutral atomic scatter ing factors were those incorporated in SHELXL-93. A list of observed (corrected) and calculated structure factors may be obtained from one author (MP) upon request (or through the EJ.M. Editorial Office).
Description of the structure The final fractional coordinates and U equiv alent isotropic parameters are given in Table 3.
Bond lengths and selected interatomic distances are given in Table 4. The crystal structure of PFC, as seen along [010], is schematically shown in Fig. 1. It is a peculiar kind of structure, and it does not show any resemblance with known structures, neither among lead hydroxychlorides nor among other structural families. A striking feature of PFC is the alternation, along [001], of structural slices with "heavy" and "light" electron density: a) the "heavy" slice, approximately -0.25 to 0.25 in z coordinate, includes lead and chlorine atoms. b) the "light" slice, approximately 0.25 to 0.75
48
M. Pasero, N. Perchiazzi, S. Bigi, M. Franzini, S. Merlino
Fig. 1. The crystal structure ofPFC as seen along [010].
in z, is characterized by the presence of [Fe3+(OH)6] octahedra alternated with structural voids, which are the seat of a complex hydrogen bond system. The two independent lead atoms display nine fold coordination. Both are connected to five chlorines, three oxygens belonging to hydroxyl groups, and one oxygen belonging to a water molecule. Such a coordination for lead - the poly
hedron is commonly referred to as tri-capped trigonal prism - is not so common. For instance, in most of the lead hydroxyhalides from Baratti and Laurion [laurionite (Cannillo et al, 1969), paralaurionite (Merlino et al, 1993), fiedlerite\A and fiedlerite-2M (Merlino et al, 1994), penfieldite (Merlino et al, 1995)] lead is in eight fold coordination (square antiprism, or bi-capped trigonal prism), or in distorted nine-fold (7+2) coordination [cotunnite (Nozik et al, 1976)]. A somewhat different coordination for lead oc curs in phosgenite (Giuseppetti & Tadini, 1974) and in matlockite (Pasero & Perchiazzi, 1996), where lead is nine-fold coordinated in a monocapped square antiprism. Very recently, tricapped trigonal prisms as lead coordination polyhedra were described in Jaurelite (Merlino et al, 1996). In PFC the lead polyhedra share the triangular bases, giving rise to columns running along b. Moreover such columns are also linked laterally, by sharing of faces parallel to b. Despite their high coordination number, Pbl and Pb2 atoms do not show significant supersaturation with respect to the bond valence. Their bond valence sums, computed according to Brese & O'Keeffe (1991), are 2.13 and 2.07 v.u. for Pbl and Pb2, respec tively (cf. Table 5).
Table 5. Bond valence balance for PFC. H-bonds are not considered. Pbl
Pb2
Cll
0.368 0.143 0.083
0.304
0.898
C12
0.198
0.198 0.192 0.077
0.665
C13
0.252
0.173
0.425
Fe
Σcv
01
0.267
0.215
0.803
1.285
02
0.447
0.485
0.431
1.363
03
0.169
0.550 0.377
1.096
0.507 0.408
1.093
04
0.178
05
0.204
0.253
Σav
2.131
2.075
0.457 3.076
Crystal structure of a lead and iron hydroxychloride
Fig. 2. The basic OD layer in PFC as seen along [100] (the a and b axes are interchanged; see text).The layer group symmetry is P2\ma.
Iron is six-fold coordinated in a [Fe3+(OH)6] octahedron. Octahedra give rise to columns which develop along b by sharing of O3-O4 edges. It is worth noting that the shared edge is
49
not the shortest one, as O1-O3 edge has the same length (2.68 Å). The bond valence sum over Fe fairly matches the theoretical value (3.08 v.u., which incidentally confirms the trivalent state of iron). In Table 4 the list of O C\ distances which are possibly involved in the hydrogen bonding is given. It is difficult to unambiguously determine a complete hydrogen bond system, as the occurrence of atoms with strong scattering power and the twinned nature of PFC caused high uncertainties on the positions of oxygens, and consequently on the O ••• Cl distances. Therefore the bond valence balance (Table 5) was computed without the hydrogen bond contributions. Anyway, on the basis of the atomic positions, it seems reasonable that both Ol and O2 form trifurcated hydrogen bonds towards Cll and C12, and that both O3 and O4 form bifurcated hydrogen bonds towards C13. The valence balance for
~Λ ~A t
MDOl P2.ll
MDO2 P21cα
Fig. 3. The two MDO structures in PFC as seen along [100]: (a). MDOl, stacking sequence ; (b). MDO2, stacking sequence . On the right of each structure, the stacking operator between neighbour layers is indicated.
50
M. Pasero, N. Perchiazzi, S. Bigi, M. Franzini, S. Merlino
both oxygens and chlorines should settle down to reasonable values, with a possible small residual undersaturation for Cl3 only.
OD character of PFC PFC belongs to a family of Order-Disorder (OD) structures formed by equivalent layers (Dornberger-Schiff, 1966, 1979). As stated above, OD character is a very common peculiarity among lead hydroxychlorides, as it has been described in other structural families (laurionite paralaurionite, Merlino et al, 1993; fiedlerite- IA - fiedlerite-2M, Merlino et a/., 1994). The symmetry properties of the whole family are embodied in the OD groupoid family symbol, which gives the symmetry operations of the single layer and those relating a pair of adjacent layers. To obtain an OD groupoid family symbol consistent with that indicated in the basic compilation of Dornberger-Schiff & Fichtner (1972), the a and b axes of the single layer must be interchanged with respect to those so far assumed in the crystal structure description. With this interchange the basic OD layer in PFC (Fig. 2) is periodic in the two directions a = 6.25 Å, and b = 8.03 Å (γ = 90°), the width of the layer being co (= c sin oc) = 8.99 A and the layer group sym metry P2\ma. The OD-groupoid family symbol reads: P 2i m (a) {πi/2,2 2i/2(22)} Owing to the presence of a mirror in the sym metry of the layer, there exist two possibilities of stacking of the layers, differing only in the direc tion of the translational component associated with the n glide normal to a. The glide can be either «i/2,2 (translation +b/4 + co; in the following "+") or nJ/2 (translation -b/4 + co; in the following "-"). An infinite number of different OD-structures (polytypes) may exist within this family, depending on the stacking sequence. Among them, there are two Maximum Degree of Order (MDO) structures: a) MDO1, according to the stacking sequence < — > . This sequence gives rise to a structure with monoclinic symmetry and space group P 2 i l l , corresponding to the above-described structure of PFC in all but the orientation of the axes (Fig. 3a). The stacking sequence gives rise to a twin-related structure (MDOΓ).
b) MDO2, according to the stacking sequence . This sequence gives rise to a structure with orthorhombic symmetry and space group Plica (Fig. 3b). Following the polytype notation, the two MDO structures should be called PFC-1M and PFC-20, respectively. Acknowledgements: We thank the mineral col lector A. Turini for making us available the crys tals of PFC used in this study. F. Marchetti helped us in writing a computer program for the treatment of twinning. The Consiglio Nazionale delle Ricerche is also acknowledged for financ ing the electron microprobe laboratory in Modena University, whose facilities were used in the present work. Finally, we wish to thank two anonymous referees for their suggestions.
References Baerlocher, C, Hepp, A., Meier, W.M. (1978): DLS76, a program for the simulation of crystal struc tures by geometric refinement. Institute of Crystal lography and Petrography, ETH Zurich, Switzerland. Bonaccorsi, E., Merlino, S., Pasero, M. (1990): Rhonite: structural and microstructural features, crystal chemistry and polysomatic relationships. Eur. J. Mineral, 2, 203-218. Brese, N. & O'Keeffe, M. (1991): Bond-valence para meters for solids. Acta Cryst., B47, 192-197. Cannillo, E., Giuseppetti, G., Tadini, C. (1969): Riesame della struttura della laurionite PbOHCl. Per. Mineral, 38, 395-402. Donovan, J.J., Snyder, D.A., Rivers, M.L. (1993): An improved interference correction for trace element analysis. Microbeam Analysis, 2, 23-28. Dornberger-Schiff, K. (1966): Lehrgang iiber ODStrukturen. Akademie-Verlag, Berlin, 135 p. — (1979): OD-structures - A game and a bit more. Krist. Techn., 14, 1027-1045. Dornberger-Schiff, K. & Fichtner, K. (1972): On the symmetry of OD-structures consisting of equiv alent layers. Krist. Techn., 7, 1035-1056. Franzini, M. & Perchiazzi, N. (1992): I mineral! delle scorie ferrifere etrusche di Baratti (Livorno). Atti Soc. Tosc. Sc. Nat., Ser. A, 99, 43-77. Franzini, M., Perchiazzi, N., Bartoli, M.L., Chiappino, L. (1992): Baratti, una nuova località mineralogica simile al Laurion. Riv. Mineral It, 23, 1-14, 67-75. Giuseppetti, G. & Tadini, C. (1974): Reexamination of
Crystal structure of a lead and iron hydroxychloride
the crystal structure of phosgenite, Pb2Cl2(CO3). Tschermak Mineral. Petrogr. Mitt., 21, 101-109. Giuseppetti, G., Mazzi, F., Tadini, C. (1993): The crystal structure of nealite, Pb4Fe(As03)2Cl4 2H2O. N. Jb. Mineral. Mh., H. 6, 278-288. Merlino, S. (1990): OD structures in mineralogy. Per. Mineral., 59, 69-92. Merlino, S., Pasero, M , Perchiazzi, N. (1993): Crystal structure of paralaurionite and its OD relationships with laurionite. Mineral. Mag., 57, 323-328. Merlino, S., Pasero, M., Perchiazzi, N. (1994): Fiedlerite: revised chemical formula [Pb3CUF(OH)H2O], OD description and crystal structure refinement of the two MDO poly types. Mineral. Mag., 58, 69-78. Merlino, S., Pasero, M., Perchiazzi, N., Gianfagna, A. (1995): X-ray and electron diffraction study ofpenfieldite: average structure and multiple cells. Mineral. Mag., 59, 341-347. Merlino, S., Pasero, M., Perchiazzi, N., Kampf, A.R. (1996): Laurelite: its atomic structure and relationship to cc-PbF2. Am. Mineral., 81, 1277-1281. Nickel, E.H. (1995): Definition of a mineral. Eur. J. Mineral., 7, 1213-1215.
51
Nozik, Yu.Z., Fykin, L.E., Muradyan, L.A. (1976): Crystal structure of cotunnite PbCb determined more precisely by application of the neutron diffraction method. Sov. Phys. Crystallogr., 21, 38-40. Pasero, M. & Perchiazzi, N. (1996): Crystal structure refinement of matlockite. Mineral. Mag., 60, 833836. Pasero, M. & Reinecke, T. (1991): Crystal chemistry, HRTEM analysis and polytypic behaviour of ardennite. Eur. J. Mineral, 3, 819-830. Sheldrick, G. M. (1986): SHELXS-86. Program for the solution of crystal structures. Univ. of Gottingen, Germany. — (1993): SHELXL-93. Program for the refinement of crystal structures. Univ. of Gottingen, Germany. Walker, N. & Stuart, D. (1983): An empirical method for correcting diffractometer data for absorption effects. Acta Cryst., A39, 158-166.
Received 23 March 1996 Accepted 16 September 1996
