Received 9 February 2011
Synthetic ferric sulfate trihydrate, Fe2(SO4)3·3H2O, a new ferric sulfate salt
Ferric sulfate trihydrate has been synthesized at 403 K under hydrothermal conditions. The structure consists of quadruple chains of [Fe2(SO4)3(H2O)3] parallel to . Each quadruple chain is composed of equal proportions of FeO4(H2O)2 octahedra and FeO5(H2O) octahedra sharing corners with SO4 tetrahedra. The chains are joined to each other by hydrogen bonds. This compound is a new hydration state of Fe2(SO4)3·nH2O; minerals with n = 0, 5, 7.25-7.75, 9 and 11 are found in nature.
Ferric sulfate trihydrate belongs to a group of ferric sulfates sharing the general formula Fe2(SO4)3·nH2O, which includes mikasaite (n = 0; Miura et al., 1994), lausenite (n = 5; Majzlan et al., 2005), kornelite (n = 7.25-7.75; Ackermann et al., 2009; Robinson & Fang, 1973), paracoquimbite (n = 9; Robinson & Fang, 1971), quenstedtite (n = 11; Thomas et al., 1974) and a synthetic Fe2(SO4)3 (n = 0), which is a polymorph of mikasaite (Christidis & Rentzeperis, 1975). These ferric sulfate minerals are usually found as efflorescence in acid mine drainage (AMD) areas, formed as oxidation products of pyrite (FeS2) and other sulfide minerals (Jambor et al., 2000). Other commonly found ferric sulfates in AMD regions include coquimbite [(Fe,Al)2(SO4)3·9H2O], jarosite [KFe3(SO4)2(OH)], rhomboclase [(H5O2)Fe(SO4)2·2H2O] and copiapite [Fe2+Fe3+4(SO4)2(OH)2·20H2O]. Dissolution of these minerals greatly increases water acidity because of ferric ion hydrolysis. Plus, these ferric sulfates usually absorb or co-precipitate with toxic metals such as Cr and Pb, and act as a secondary source of these toxic metal pollutants in local waters as they dissolve (Jambor et al., 2000). To better control and reduce the adverse environmental effects of ferric sulfates require a thorough understanding of the behavior and stability of these minerals as functions of environmental factors, such as pH, relative humidity and temperature. Understanding the transformations of the ferric sulfates as a function of environmental conditions has been the focus of recent studies (Ackermann et al., 2009; Majzlan, 2010; Tosca et al., 2007; Xu et al., 2009, 2010).
One difficulty in delineating phase stability relationships in the FeIII-SO4-H2O system is the need for correct and complete crystal structure models for all phases in the system. For example, paracoquimbite and coquimbite had for a long time been considered as polymorphs (Fang & Robinson, 1970) until a recent study showing the amount of aluminium in the mineral determines the structure type: paracoquimbite is pure Fe2(SO4)3·9H2O, while coquimbite has an Al/Fe ratio close to 1:3, with Al predominantly occupying the [M(H2O)6]3+ metal site (Majzlan et al., 2010). A third structure type in this coquimbite series was recently found in a mineral with a 1:1 Al/Fe ratio (Demartin et al., 2010). Further, the synthesis of ferric sulfates often produces new phases (Chipera et al., 2007; Freeman et al., 2009; Majzlan et al., 2005; Peterson et al., 2009; Xu et al., 2009). Structures of these new phases are mostly unresolved because of their occurrence only as fine-grained mixed-phase powders. During the course of our survey of the FeIII-SO4-H2O system, we discovered a new phase, namely ferric sulfate trihydrate, Fe2(SO4)3·3H2O.
The structure of the trihydrate contains identical quadruple chains of [Fe2(SO4)3(H2O)3] parallel to , as shown in Fig. 1. Each quadruple chain consists of four single chains of alternating FeO6 octahedra and SO4 tetrahedra extending along the b axis. Of the four single chains, two are symmetrically independent as `-Fe1-S1-Fe1-' and `-Fe2-S2-Fe2-'; the other two are generated through an inversion center. A third sulfate group S3O4 connects to Fe1O6 and Fe2O6 by corner sharing. Of the three unique water molecules, one (O1W) is coordinated to the Fe1 site, and the other two (O2W and O3W) are coordinated to the Fe2 site.
The quadruple chains are linked to one another by water-sulfate O-HO hydrogen bonds (Table 2). All hydrogen bonds involve the terminal O atoms of SO4 groups as the acceptor, except the O1W-H1BO3iv [symmetry code: (iv) -x, y - , -z + ] hydrogen bond, where the acceptor O atom bridges Fe1 and S1. There is also one intramolecular hydrogen bond, viz. O1W-H1BO8.
The quadruple chain structure of ferric sulfate trihydrate has no similar counterparts in any known sulfates. Though unique, the trihydrate structure shares similarities with other phases in the Fe2(SO4)3·nH2O group. For example, the FeO6 octahedra do not directly connect with each other but are linked via corner-sharing SO4 tetrahedra. It should be mentioned that FeO6 octahedra do share corners in some other ferric sulfates such as copiapite and jarosite.
Including the trihydrate described here, there are a total of six hydration states of Fe2(SO4)3·nH2O; namely n = 0, 3, 5, 7.25-7.75, 9 and 11. Anhydrous ferric sulfate has two polymorphs, a monoclinic form and a trigonal form. The trigonal form occurs in nature as the mineral mikasaite, while the synthetic monoclinic form has no mineral equivalent (Christidis & Rentzeperis, 1975; Miura et al., 1994). Both forms consist of frameworks of connected FeO6 octahedra and SO4 tetrahedra. The structure of lausenite, Fe2(SO4)3·5H2O, is composed of corrugated slabs of [Fe2(SO4)3(H2O)5] (Majzlan et al., 2005). The five water molecules shown in the formula are all coordinated to the two Fe sites, two coordinated to one and three coordinated to the other. The structure of kornelite, Fe2(SO4)3·7.25-7.75H2O, consists of slabs similar to those in lausenite, with a formula [Fe2(SO4)3(H2O)6]. Each of the two Fe sites is bonded to three water molecules. The remaining 1.25 to 1.75 water molecules shown in the formula are located between neighboring slabs as isolated water (Robinson & Fang, 1973). The paracoquimbite structure contains isolated clusters of [Fe(H2O)6]3+ and [Fe3(SO4)6(H2O)6]3-, together with six uncoordinated water molecules (Robinson & Fang, 1971). The structure of quenstedtite, Fe2(SO4)3·11H2O, consists of isolated clusters of [Fe(SO4)(H2O)5]+ and [Fe(SO4)2(H2O)4]-, together with two uncoordinated water molecules (Thomas et al., 1974).
As the hydration state increases, FeO6 octahedra and SO4 tetrahedra tend to be disassociated, as more water molecules coordinate to Fe to form simple clusters. The hydration state of the trihydrate lies between that of anhydrous ferric sulfate and lausenite. Like lausenite, the structure of the trihydrate does not have uncoordinated water molecules as found in kornelite and higher hydration states. As opposed to limited clusters in the higher hydration states, both the trihydrate and lausenite have infinite clusters, viz. one-dimensional chains for the trihydrate and two-dimensional slabs for lausenite. Further, the quadruple chain of the trihydrate has three terminal O-atom sites for water molecules, while the slab of lausenite has five. This appears to be counterintuitive: a common first impression would be that a chain arrangement has more terminal O-atom sites than two-dimensional sheets. It is easy to understand this if one considers the fact that a quadruple chain has far fewer terminal sites than four single chains; also, the slab in lausenite is less like a dense sheet such as the MnO2 layer in birnessite, and closer to a net of interlaced chains.
Therefore, the trihydrate, though having a unique quadruple chain structure, shares the basic structural features present in other phases in the Fe2(SO4)3·nH2O system. The trihydrate structure also fits in the trend of structural changes set by hydration levels.
| || Figure 1 |
The ferric sulfate trihydrate structure, viewed down the b axis. FeO6 octahedra are dark (brown in the electronic version of the paper), SO4 tetrahedra are light (yellow) and H atoms are small spheres. Dotted lines denote hydrogen bonds. Atom labels are generic and do not correspond exactly to the asymmetric unit defined in the CIF.
-Fe2O3 [1.000 (1) g] and sulfuric acid [1.939 (1) g] with a nominal concentration of 95.9 wt% H2SO4 were mixed in a 23 ml Teflon-lined vessel, sealed in a Parr stainless steel autoclave and then stored in an isotemp oven set at 403 K for 9 d. The product was an inhomogeneous yellow-pink solid block. Examination with optical microscopy revealed transparent yellow crystals embedded in a pink powdery matrix. The yellow crystals are cuboids or short rectangular prisms, 100 to 300 µm on the longest edge and 40 to 100 µm on the shortest edge. The structure of the yellow crystal was determined from single-crystal X-ray diffraction to be the title ferric sulfate trihydrate. The pink matrix was a combination of lausenite and rhomboclase identified from powder X-ray diffraction (see Supplementary material ).
H atoms were found in difference Fourier maps and subsequently placed in idealized positions with restrained O-H distances of 0.90 (3) Å, and with Uiso(H) = 1.2Uiso(O).
Data collection: SMART (Bruker, 1998); cell refinement: SAINT-Plus (Bruker, 1998); data reduction: SAINT-Plus; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: CrystalMaker (Palmer, 2010); software used to prepare material for publication: SHELXL97.
Supplementary data for this paper are available from the IUCr electronic archives (Reference: QS3002 ). Services for accessing these data are described at the back of the journal.
The synthesis of ferric sulfates at Stony Brook by Wenqian Xu was supported by NASA grant No. MFRP07-0022 (JBP). The single-crystal diffractometer and software used to determine the trihydrate structure were funded by grant No. NSF-DMR-0800415 (JBP).
Ackermann, S., Lazic, B., Armbruster, T., Doyle, S., Grevel, K. D. & Majzlan, J. (2009). Am. Mineral. 94, 1620-1628.
Bruker (1998). SMART and SAINT-Plus. Bruker AXS Inc., Madison, Wisconsin, USA.
Bruker (2001). SADABS. Bruker AXS Inc., Madison, Wisconsin, USA.
Chipera, S. J., Vaniman, D. T. & Bish, D. L. (2007). Lunar and Planetary Science Conference XXXVIII, March 12-16, League City, Texas. Abstract No. 1409.
Christidis, P. C. & Rentzeperis, P. J. (1975). Z. Kristallogr. 141, 233-245.
Demartin, F., Castellano, C., Gramaccioli, C. M. & Campostrini, I. (2010). Can. Mineral. 48, 323-333.
Fang, J. H. & Robinson, P. D. (1970). Am. Mineral. 55, 1534-1540.
Freeman, J. J., Wang, A. & Ling, Z. C. (2009). Lunar and Planetary Science Conference XXXX, March 23-27, Houston, Texas. Abstract No. 2284.
Jambor, J. I., Nordstrom, D. K. & Alpers, C. N. (2000). Metal-Sulfate Salts from Sulfide Mineral Oxidation. Washington, DC: Mineralogical Society of America, Geochemical Society.
Majzlan, J. (2010). Chimia, 64, 699-704.
Majzlan, J., Botez, C. & Stephens, P. W. (2005). Am. Mineral. 90, 411-416.
Majzlan, J., Dordevic, T., Kolitsch, U. & Schefer, J. (2010). Mineral. Petrol. 100, 241-248.
Miura, H., Niida, K. & Hirama, T. (1994). Mineral. Mag. 58, 649-653.
Palmer, D. (2010). CrystalMaker. CrystalMaker Software Ltd, Yarnton, Oxfordshire, England.
Peterson, R. C., Valyashko, E. & Wang, R. Y. (2009). Can. Mineral. 47, 625-634.
Robinson, P. D. & Fang, J. H. (1971). Am. Mineral. 56, 1567-1572.
Robinson, P. D. & Fang, J. H. (1973). Am. Mineral. 58, 535-539.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112-122.
Thomas, J. N., Robinson, P. D. & Fang, J. H. (1974). Am. Mineral. 59, 582-586.
Tosca, N. J., Smirnov, A. & McLennan, S. M. (2007). Geochim. Cosmochim. Acta, 71, 2680-2698.
Xu, W. Q., Parise, J. B. & Hanson, J. (2010). Am. Mineral. 95, 1408-1412.
Xu, W. Q., Tosca, N. J., McLennan, S. M. & Parise, J. B. (2009). Am. Mineral. 94, 1629-1637.