inorganic compounds
High-pressure synthetic (Na0.97Mg0.03)(Mg0.43Fe0.173+Si0.40)Si2O6, with six-coordinated silicon, isostructural with P2/n omphacite
aDepartment of Geosciences, University of Arizona, 1040 E. 4th Street, Tucson, AZ 85721-0077, USA, bInstitut für Mineralogie und Petrographie, Universität Innsbruck, Innsbruck, Austria, and cBayerisches Geoinstitut, Universität Bayreuth, Bayreuth, Germany
*Correspondence e-mail: posnere@email.arizona.edu
The title compound, (sodium magnesium) [magnesium iron(III) silicon] disilicate, (Na0.97Mg0.03)(Mg0.43Fe0.173+Si0.40)Si2O6, is isotypic with ordered P2/n omphacite. Its structure is characterized by single chains of corner-sharing SiO4 tetrahedra, extending along the c axis, which are crosslinked by bands of edge-sharing octahedra (site symmetry 2), statistically occupied by (Mg2+ + Fe3+ + Si4+). Between the bands built up of the octahedra are two non-equivalent highly distorted six-coordinated sites (site symmetry 2), statistically occupied by (Na + Mg). In contrast to omphacites, the great differences in size and charge between Mg2+ and Si4+ result in complete, rather than partial, ordering of Mg and Si into two distinct octahedral sites, whereas Fe3+ is disordered between the two sites. The octahedron filled by (Mg + Fe) is larger and markedly more distorted than that occupied by (Si + Fe). The average (Mg + Fe)—O and (VISi + Fe)—O bond lengths are 2.075 and 1.850 Å, respectively.
Related literature
For structures of high-pressure synthetic clinopyroxenes with six-coordinated Si, see: Angel et al. (1988); Yang & Konzett (2005); Yang et al. (2009). For structures of ordered P2/n omphacites, see: Curtis et al. (1975); Matsumoto et al. (1975); Rossi et al. (1983). For background on the stability of clinopyroxenes at high pressures and temperatures, see: Gasparik (1989); Konzett et al. (2005). For general background on materials with six-coordinated silicon, see: Finger & Hazen (1991). For the geologic occurrence of clinopyroxene with six-coordinated Si, see: Wang & Sueno (1996). For spectroscopic measurements on P2/n clinopyroxenes, see: Boffa Ballaran et al. (1998); Yang et al. (2009). For general information on polyhedral distortion and ionic radii, see: Robinson et al. (1971) and Shannon (1976), respectively.
Experimental
Crystal data
|
Data collection: APEX2 (Bruker, 2004); cell SAINT (Bruker, 2004); data reduction: SAINT; program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: XtalDraw (Downs & Hall-Wallace, 2003); software used to prepare material for publication: publCIF (Westrip, 2010).
Supporting information
10.1107/S1600536812002966/fj2504sup1.cif
contains datablocks I, global. DOI:Structure factors: contains datablock I. DOI: 10.1107/S1600536812002966/fj2504Isup2.hkl
The specimen used in this study was synthesized in a multi-anvil apparatus at 15 GPa and 1500 °C for 23.2 h (run #JKB2006–12) and then rapidly quenched (< 5 s) to ambient conditions. The crystal chemistry was determined with a Jeol electron microprobe on the same single-crystal used for the
data collection. The average composition of nine analysis points yielded a chemical formula (normalized on the basis of 6 oxygen atoms and 4 cations while maintaining charge-balance) (Na0.97Mg0.03)(Mg0.43Fe0.17Si0.40)Si2O6.The Raman spectrum of (Na0.97Mg0.03)(Mg0.43Fe0.17Si0.40)Si2O6 was collected from a randomly oriented crystal at 100% power on a Thermo Almega microRaman system, using a 532 nm solid-state laser, and a thermoelectrically cooled CCD detector. The laser is partially polarized with 4 cm-1 resolution and a spot size of 1 µm.
Throughout the structure refinements, the chemical composition of the crystal was fixed to that determined from electron microprobe analysis. From crystal-chemical considerations, all Mg and VISi were assigned to the M1 and M1(1) sites, respectively, with the rest filled by Fe3+ for both sites. Because the average distance of M2—O is shorter than that of M2(1)-O, we assigned 0.03 Mg apfu to the M2 site. The highest residual peak in the difference Fourier maps was located at (0.7991, 0.0789, 0.2448), 0.54 Å from M2, and the deepest hole at (0.2602, 0.1856, 0.1853), 0.48 Å from M1(1).
Data collection: APEX2 (Bruker, 2004); cell
SAINT (Bruker, 2004); data reduction: SAINT (Bruker, 2004); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELXL97 (Sheldrick, 2008); molecular graphics: XtalDraw (Downs & Hall-Wallace, 2003); software used to prepare material for publication: publCIF (Westrip, 2010).(Na0.97Mg0.03)(Mg0.43Fe0.17Si0.40)Si2O6 | F(000) = 409 |
Mr = 206.39 | Dx = 3.360 Mg m−3 |
Monoclinic, P2/n | Mo Kα radiation, λ = 0.71073 Å |
Hall symbol: -P 2yac | Cell parameters from 1071 reflections |
a = 9.4432 (8) Å | θ = 3.3–35.6° |
b = 8.6457 (7) Å | µ = 1.69 mm−1 |
c = 5.2540 (5) Å | T = 293 K |
β = 108.003 (6)° | Cube, pale gray |
V = 407.95 (6) Å3 | 0.06 × 0.05 × 0.05 mm |
Z = 4 |
Bruker APEXII CCD area-detector diffractometer | 1481 independent reflections |
Radiation source: fine-focus sealed tube | 980 reflections with I > 2σ(I) |
Graphite monochromator | Rint = 0.033 |
ϕ and ω scan | θmax = 32.5°, θmin = 3.3° |
Absorption correction: multi-scan (SADABS; Sheldrick 2005) | h = −14→14 |
Tmin = 0.906, Tmax = 0.920 | k = −13→10 |
6586 measured reflections | l = −7→7 |
Refinement on F2 | Primary atom site location: structure-invariant direct methods |
Least-squares matrix: full | Secondary atom site location: difference Fourier map |
R[F2 > 2σ(F2)] = 0.036 | w = 1/[σ2(Fo2) + (0.0264P)2 + 0.6629P] where P = (Fo2 + 2Fc2)/3 |
wR(F2) = 0.085 | (Δ/σ)max = 0.001 |
S = 1.07 | Δρmax = 0.51 e Å−3 |
1481 reflections | Δρmin = −0.66 e Å−3 |
96 parameters | Extinction correction: SHELXL97 (Sheldrick, 2008), Fc*=kFc[1+0.001xFc2λ3/sin(2θ)]-1/4 |
3 restraints | Extinction coefficient: 0 |
(Na0.97Mg0.03)(Mg0.43Fe0.17Si0.40)Si2O6 | V = 407.95 (6) Å3 |
Mr = 206.39 | Z = 4 |
Monoclinic, P2/n | Mo Kα radiation |
a = 9.4432 (8) Å | µ = 1.69 mm−1 |
b = 8.6457 (7) Å | T = 293 K |
c = 5.2540 (5) Å | 0.06 × 0.05 × 0.05 mm |
β = 108.003 (6)° |
Bruker APEXII CCD area-detector diffractometer | 1481 independent reflections |
Absorption correction: multi-scan (SADABS; Sheldrick 2005) | 980 reflections with I > 2σ(I) |
Tmin = 0.906, Tmax = 0.920 | Rint = 0.033 |
6586 measured reflections |
R[F2 > 2σ(F2)] = 0.036 | 96 parameters |
wR(F2) = 0.085 | 3 restraints |
S = 1.07 | Δρmax = 0.51 e Å−3 |
1481 reflections | Δρmin = −0.66 e Å−3 |
Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. |
Refinement. Refinement of F2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F2, conventional R-factors R are based on F, with F set to zero for negative F2. The threshold expression of F2 > σ(F2) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F2 are statistically about twice as large as those based on F, and R- factors based on ALL data will be even larger. |
x | y | z | Uiso*/Ueq | Occ. (<1) | |
M2 | 0.7500 | 0.04980 (16) | 0.2500 | 0.0129 (3) | 0.9400 (1) |
M2MG | 0.7500 | 0.04980 (16) | 0.2500 | 0.0129 (3) | 0.0600 (1) |
M2(1) | 0.7500 | 0.45601 (17) | 0.7500 | 0.0169 (3) | |
M1 | 0.7500 | 0.65310 (10) | 0.2500 | 0.00495 (18) | 0.8600 (1) |
M1Fe | 0.7500 | 0.65310 (10) | 0.2500 | 0.00495 (18) | 0.1400 (1) |
M1(1) | 0.7500 | 0.84745 (9) | 0.7500 | 0.00855 (17) | 0.8000 (1) |
M11Fe | 0.7500 | 0.84745 (9) | 0.7500 | 0.00855 (17) | 0.2000 (1) |
Si1 | 0.04344 (7) | 0.84714 (7) | 0.22988 (13) | 0.00748 (14) | |
Si2 | 0.03811 (7) | 0.66398 (7) | 0.73666 (13) | 0.00731 (14) | |
O1 | 0.86242 (18) | 0.8417 (2) | 0.1106 (4) | 0.0133 (4) | |
O2 | 0.85803 (18) | 0.6879 (2) | 0.6524 (4) | 0.0129 (4) | |
O3 | 0.1201 (2) | 0.0136 (2) | 0.3071 (4) | 0.0130 (4) | |
O4 | 0.1012 (2) | 0.4949 (2) | 0.7949 (4) | 0.0141 (4) | |
O5 | 0.10966 (18) | 0.7652 (2) | 0.0132 (3) | 0.0105 (3) | |
O6 | 0.09535 (18) | 0.74992 (19) | 0.5087 (3) | 0.0112 (3) |
U11 | U22 | U33 | U12 | U13 | U23 | |
M2 | 0.0162 (7) | 0.0106 (7) | 0.0093 (7) | 0.000 | 0.0002 (5) | 0.000 |
M2MG | 0.0162 (7) | 0.0106 (7) | 0.0093 (7) | 0.000 | 0.0002 (5) | 0.000 |
M2(1) | 0.0210 (7) | 0.0107 (7) | 0.0128 (8) | 0.000 | −0.0037 (6) | 0.000 |
M1 | 0.0060 (4) | 0.0045 (4) | 0.0037 (4) | 0.000 | 0.0006 (3) | 0.000 |
M1Fe | 0.0060 (4) | 0.0045 (4) | 0.0037 (4) | 0.000 | 0.0006 (3) | 0.000 |
M1(1) | 0.0094 (3) | 0.0085 (4) | 0.0073 (4) | 0.000 | 0.0020 (3) | 0.000 |
M11Fe | 0.0094 (3) | 0.0085 (4) | 0.0073 (4) | 0.000 | 0.0020 (3) | 0.000 |
Si1 | 0.0081 (3) | 0.0074 (3) | 0.0065 (3) | −0.0002 (2) | 0.0016 (2) | −0.0005 (2) |
Si2 | 0.0075 (3) | 0.0076 (3) | 0.0065 (3) | 0.0000 (2) | 0.0018 (2) | 0.0004 (2) |
O1 | 0.0103 (7) | 0.0128 (9) | 0.0165 (9) | 0.0007 (6) | 0.0039 (7) | −0.0023 (7) |
O2 | 0.0074 (7) | 0.0173 (9) | 0.0132 (9) | −0.0013 (6) | 0.0018 (6) | 0.0035 (7) |
O3 | 0.0200 (9) | 0.0074 (8) | 0.0106 (9) | −0.0014 (6) | 0.0033 (7) | −0.0009 (6) |
O4 | 0.0188 (9) | 0.0104 (9) | 0.0132 (9) | 0.0006 (7) | 0.0050 (7) | 0.0002 (7) |
O5 | 0.0098 (7) | 0.0136 (9) | 0.0078 (8) | 0.0009 (6) | 0.0024 (6) | −0.0018 (6) |
O6 | 0.0125 (8) | 0.0129 (9) | 0.0086 (8) | −0.0017 (6) | 0.0037 (6) | 0.0017 (6) |
M2—O1i | 2.319 (2) | M1—O2 | 2.0658 (18) |
M2—O1ii | 2.319 (2) | M1—O2xi | 2.0658 (18) |
M2—O3iii | 2.3356 (18) | M1—O1xi | 2.1917 (19) |
M2—O3iv | 2.3356 (18) | M1—O1 | 2.1917 (19) |
M2—O6v | 2.3651 (19) | M1(1)—O3v | 1.8071 (19) |
M2—O6vi | 2.3651 (19) | M1(1)—O3vii | 1.8071 (19) |
M2—O5vii | 2.7127 (19) | M1(1)—O1xii | 1.8645 (18) |
M2—O5viii | 2.7127 (19) | M1(1)—O1xi | 1.8645 (18) |
M2(1)—O2ix | 2.376 (2) | M1(1)—O2ix | 1.8788 (19) |
M2(1)—O2 | 2.376 (2) | M1(1)—O2 | 1.8788 (19) |
M2(1)—O4x | 2.4073 (19) | Si1—O3xiii | 1.6058 (19) |
M2(1)—O4vi | 2.4073 (19) | Si1—O5 | 1.6216 (18) |
M2(1)—O5vii | 2.438 (2) | Si1—O6 | 1.6276 (18) |
M2(1)—O5v | 2.438 (2) | Si1—O1xiv | 1.6294 (18) |
M2(1)—O6vii | 2.8944 (19) | Si2—O4 | 1.5725 (19) |
M2(1)—O6v | 2.8944 (19) | Si2—O2xiv | 1.6326 (18) |
M1—O4vi | 1.9678 (19) | Si2—O6 | 1.6366 (18) |
M1—O4v | 1.9678 (19) | Si2—O5xii | 1.6510 (18) |
O4vi—M1—O4v | 98.88 (11) | O3v—M1(1)—O2ix | 170.80 (8) |
O4vi—M1—O2 | 96.72 (7) | O3vii—M1(1)—O2ix | 89.37 (8) |
O4v—M1—O2 | 94.15 (7) | O1xii—M1(1)—O2ix | 83.67 (8) |
O4vi—M1—O2xi | 94.15 (7) | O1xi—M1(1)—O2ix | 94.09 (8) |
O4v—M1—O2xi | 96.72 (7) | O3v—M1(1)—O2 | 89.37 (8) |
O2—M1—O2xi | 163.26 (11) | O3vii—M1(1)—O2 | 170.80 (8) |
O4vi—M1—O1xi | 90.36 (7) | O1xii—M1(1)—O2 | 94.09 (8) |
O4v—M1—O1xi | 164.04 (7) | O1xi—M1(1)—O2 | 83.67 (8) |
O2—M1—O1xi | 71.74 (7) | O2ix—M1(1)—O2 | 85.52 (11) |
O2xi—M1—O1xi | 95.54 (7) | O3xiii—Si1—O5 | 109.12 (10) |
O4vi—M1—O1 | 164.04 (7) | O3xiii—Si1—O6 | 104.39 (10) |
O4v—M1—O1 | 90.36 (7) | O5—Si1—O6 | 109.41 (9) |
O2—M1—O1 | 95.54 (7) | O3xiii—Si1—O1xiv | 117.40 (10) |
O2xi—M1—O1 | 71.74 (7) | O5—Si1—O1xiv | 107.69 (9) |
O1xi—M1—O1 | 83.83 (10) | O6—Si1—O1xiv | 108.64 (10) |
O3v—M1(1)—O3vii | 96.67 (12) | O4—Si2—O2xiv | 118.08 (10) |
O3v—M1(1)—O1xii | 89.08 (8) | O4—Si2—O6 | 111.93 (10) |
O3vii—M1(1)—O1xii | 92.94 (8) | O2xiv—Si2—O6 | 107.21 (9) |
O3v—M1(1)—O1xi | 92.94 (8) | O4—Si2—O5xii | 107.11 (10) |
O3vii—M1(1)—O1xi | 89.08 (8) | O2xiv—Si2—O5xii | 106.27 (9) |
O1xii—M1(1)—O1xi | 176.97 (12) | O6—Si2—O5xii | 105.39 (9) |
Symmetry codes: (i) −x+3/2, y−1, −z+1/2; (ii) x, y−1, z; (iii) x+1/2, −y, z−1/2; (iv) −x+1, −y, −z+1; (v) −x+1, −y+1, −z+1; (vi) x+1/2, −y+1, z−1/2; (vii) x+1/2, −y+1, z+1/2; (viii) −x+1, −y+1, −z; (ix) −x+3/2, y, −z+3/2; (x) −x+1, −y+1, −z+2; (xi) −x+3/2, y, −z+1/2; (xii) x, y, z+1; (xiii) x, y+1, z; (xiv) x−1, y, z. |
Experimental details
Crystal data | |
Chemical formula | (Na0.97Mg0.03)(Mg0.43Fe0.17Si0.40)Si2O6 |
Mr | 206.39 |
Crystal system, space group | Monoclinic, P2/n |
Temperature (K) | 293 |
a, b, c (Å) | 9.4432 (8), 8.6457 (7), 5.2540 (5) |
β (°) | 108.003 (6) |
V (Å3) | 407.95 (6) |
Z | 4 |
Radiation type | Mo Kα |
µ (mm−1) | 1.69 |
Crystal size (mm) | 0.06 × 0.05 × 0.05 |
Data collection | |
Diffractometer | Bruker APEXII CCD area-detector diffractometer |
Absorption correction | Multi-scan (SADABS; Sheldrick 2005) |
Tmin, Tmax | 0.906, 0.920 |
No. of measured, independent and observed [I > 2σ(I)] reflections | 6586, 1481, 980 |
Rint | 0.033 |
(sin θ/λ)max (Å−1) | 0.756 |
Refinement | |
R[F2 > 2σ(F2)], wR(F2), S | 0.036, 0.085, 1.07 |
No. of reflections | 1481 |
No. of parameters | 96 |
No. of restraints | 3 |
Δρmax, Δρmin (e Å−3) | 0.51, −0.66 |
Computer programs: APEX2 (Bruker, 2004), SAINT (Bruker, 2004), SHELXS97 (Sheldrick, 2008), SHELXL97 (Sheldrick, 2008), XtalDraw (Downs & Hall-Wallace, 2003), publCIF (Westrip, 2010).
Acknowledgements
The authors gratefully acknowledge support of this study by the Arizona Science Foundation.
References
Angel, R. J., Gasparik, T., Ross, N. L., Finger, L. W., Prewitt, C. T. & Hazen, R. M. (1988). Nature (London), 335, 156–158. CrossRef CAS Google Scholar
Boffa Ballaran, T., Carpenter, M. A., Domeneghetti, M. C. & Tazzoli, V. (1998). Am. Mineral. 83, 434–443. CAS Google Scholar
Bruker (2004). APEX2 and SAINT. Bruker AXS Inc., Madison, Wisconsin, USA. Google Scholar
Curtis, L., Gittins, J., Kocman, V., Rucklidge, J. C., Hawthorne, F. C. & Ferguson, R. B. (1975). Can. Mineral. 13, 62–67. Google Scholar
Downs, R. T. & Hall-Wallace, M. (2003). Am. Mineral. 88, 247–250. CAS Google Scholar
Finger, L. W. & Hazen, R. M. (1991). Acta Cryst. B47, 561–580. CrossRef CAS IUCr Journals Google Scholar
Gasparik, T. (1989). Contrib. Mineral. Petrol. 102, 389–405. CrossRef CAS Google Scholar
Konzett, J., Yang, H. & Frost, D. J. (2005). J. Petrol. 46, 749–781. CrossRef CAS Google Scholar
Matsumoto, T., Tokonami, M. & Morimoto, N. (1975). Am. Mineral. 60, 634–641. CAS Google Scholar
Robinson, K., Gibbs, G. V. & Ribbe, P. H. (1971). Science, 172, 567–570. CrossRef PubMed CAS Web of Science Google Scholar
Rossi, G., Smith, D. C., Ungaretti, L. & Domeneghetti, C. (1983). Contrib. Mineral. Petrol. 83, 247–258. CrossRef CAS Google Scholar
Shannon, R. D. (1976). Acta Cryst. A32, 751–767. CrossRef CAS IUCr Journals Web of Science Google Scholar
Sheldrick, G. M. (2005). SADABS. University of Göttingen, Germany. Google Scholar
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122. Web of Science CrossRef CAS IUCr Journals Google Scholar
Wang, W. & Sueno, S. (1996). Mineral. J. 18, 9–16. CrossRef CAS Google Scholar
Westrip, S. P. (2010). J. Appl. Cryst. 43, 920–925. Web of Science CrossRef CAS IUCr Journals Google Scholar
Yang, H. & Konzett, J. (2005). Am. Mineral. 90, 1223–1226. CrossRef CAS Google Scholar
Yang, H., Konzett, J., Frost, D. J. & Downs, R. T. (2009). Am. Mineral. 94, 942–949. CrossRef CAS Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.
The coordination of silicon with oxygen in crystalline materials is crucial for our understanding of the structure and composition of the Earth's interior because together they account for 63% of the atoms in the planet. In general, silicon is four-coordinated (IVSi) in the Earth's crust and upper mantle, but six-coordinated (VISi) in the lower mantle (e.g., Finger & Hazen, 1991). In the Earth's transition zone (between depths of 410 and 670 km), minerals are found to contain both IVSi and VISi. Phase transitions that involve a change in the Si coordination may affect many important physical and chemical properties of materials, such as density, bulk moduli, and elasticity, which, when coupled with seismic observations, can provide vital information on the complex constituents of the Earth's mantle.
Clinopyroxenes, one of the major rock-forming minerals of the Earth's upper mantle, were long assumed to contain IVSi only. Studies of pyroxenes synthesized at high temperatures and pressures, however, have revealed their capacity to accommodate both IVSi and VISi (Angel et al. 1988; Konzett et al. 2005; Yang & Konzett 2005; Yang et al., 2009), pointing to their possible stabilities at higher pressures. In particular, Angel et al. (1988) reported a high-pressure Na(Mg0.5Si0.5)Si2O6 clinopyroxene (designated as NaPx hereafter), which is isostructural with ordered P2/n omphacite, with VISi and Mg fully ordered into two distinct octahedral sites. Later studies of NaPx-CaMgSi2O6 (diopside) and NaPx-NaAlSi2O6 (jadeite) solid solutions uncovered a symmetry transition from an ordered P2/n to a disordered C2/c structure as VISi content decreases (Yang & Konzett, 2005; Yang et al., 2009). The natural occurrence of a clinopyroxene containing VISi, (Na0.16Mg0.84)(Mg0.92Si0.08)Si2O6, was reported by Wang & Sueno (1996) as an inclusion in a diamond from a kimberlite in China. According to the phase stability relations for the NaPx-Mg2Si2O6 (enstatite) join (Gasparik, 1989), this inclusion crystallized at pressures greater than 16.5 GPa, or at a depth within the Earth's transition zone (~500 km). The foremost implications of this finding include that (1) some portions of the Earth's upper mantle may contain a greater ratio of Na/Al than previously inferred from the xenolith chemistry, and (2) clinopyroxenes may be one of potential candidates as a silica-rich phase in the Earth's mantle. To gain more insights into the systematics on the crystal chemistry and stability field of pyroxenes with VISi, we conducted a structure refinement of a high-pressure synthetic NaPx-NaFeSi2O6 (aegirine) solid solution (designated as NaPxFe hereafter) based on the single-crystal X-ray diffraction data.
NaPxFe is isotyic with P2/n omphacite (Matsumoto et al. 1975; Curtis et al. 1975; Rossi et al. 1983) and NaPx (Angel et al., 1988; Yang et al., 2009). Its structure is characterized by a distorted closest-packed array of oxygen atoms with a layer of single silicate chains, extending along c, formed by Si1O4 and Si2O4 tetrahedra, alternating with a layer containing two distinct, edge-sharing octahedra M1 and M1(1), occupied by (Mg2+ + Fe3+ + Si4+). Also in the octahedral layer are two nonequivalent, considerably distorted six-coordinated sites M2 and M2(1), occupied by (Na + Mg) (Fig. 1). Note that Na in clinopyroxenes has been regarded to be eight-coordinated in all previous studies. However, our electron-density and bond-valence sum calculations indicate that Na in NaPxFe is actually six-coordinated. In contrast to omphacites, the great differences in size and charge between Mg2+ and Si4+ result in complete, rather than partial, ordering of Mg at M1 and Si at M1(1), while Fe is disordered between the two sites. An inspection of structure data for pyroxenes containing VISi shows that, as the respective contents of Mg and Si in the M1 and M1(1) sites decreases from NaPx to Na(Mg0.45Al0.10Si0.45)Si2O6 (Angel et al., 1988; Yang et al., 2009) and NaPxFe, the average M1—O bond length decreases from 2.092 to 2.084 and 2.075 Å, respectively, whereas the average M1(1)-O bond distance increases from 1.807 Å to 1.813 and 1.850 Å, respectively. This observation is evidently a direct consequence of the differences in ionic radii among Mg2+ (0.72 Å), Fe3+ (0.645 Å), Al3+ (0.535 Å), and VISi4+ (0.40 Å) (Shannon, 1976). The coupled changes in the average M1—O and M1(1)-O bond distances from NaPx to Na(Mg0.45Al0.10Si0.45)Si2O6 and NaPxFe lead to a substantial reduction in the size mismatch between the two edge-shared octahedra, as well as the degree of distortion of the M1 octahedron in terms of the octahedral angle variance (OAV) and octahedral quadratic elongation (OQE) (Robinson et al., 1971). The OAV and OQE values are 107.2 and 1.035, respectively, for the M1 octahedron in NaPx, 101.4 and 1.033 in Na(Mg0.45Al0.10Si0.45)Si2O6, and 88.2 and 1.029 in NaPxFe. Interestingly, as the geometric differences between the M1 and M1(1) octahedra decreases with decreasing VISi content from NaPx to Na(Mg0.45Al0.10Si0.45)Si2O6 and NaPxFe, the difference between the mean M2—O and M2(1)-O bond distances is also reduced, which is 0.123, 0.113, and 0.096 Å for NaPx, Na(Mg0.45Al0.10Si0.45)Si2O6, and NaPxFe, respectively. Conceivably, as the VISi content is further decreased, the P2/n structure will eventually transform to the C2/c structure, in which M1 and M1(1) become identical, so do M2 and M2(1), and Si1 and Si2.
Plotted in Figure 2 are Raman spectra of NaPxFe and two clinopyroxenes in the NaPx-jadeite system for comparison, one with P2/n symmetry and the other C2/c (Yang et al., 2009). The detailed assignment of Raman bands for clinopyroxenes containing VISi has been discussed by Yang et al. (2009). Generally, the Raman spectra of pyroxenes can be classified into four regions. Region 1 includes the bands between 800 and 1200 cm-1, which are assigned as the Si—O stretching vibrations in the SiO4 tetrahedra. Region 2 is between 630–800 cm-1, which includes the bands attributable to the Si—O—Si vibrations within the silicate chains. Region 3, from 400 to 630 cm-1, includes the bands that are mainly associated with the O—Si—O bending modes of SiO4 tetrahedra. The bands in Region 4, which spans from 50 to 400 cm-1, are of a complex nature, chiefly due to lattice vibration modes, polyhedral librations and M—O interactions, as well as possible O—Si—O bending. As noted by Boffa Ballaran et al. (1998) and Yang et al. (2009), the C2/c-to-P2/n transformation is characterized by the splitting of many observable Raman bands in the C2/c structure into doublets in the P2/n structure, consistent with the doubled number of independent atomic sites in the P2/n structure relative to that in the C2/c structure. Compared to the Raman spectrum of P2/n Na(Mg0.45Al0.10Si0.45)Si2O6 (Sample J2, Yang et al., 2009), most Raman bands for NaPxFe are noticeably broader, indicating the increased positional disorder of atoms in our sample, which further suggests that the chemistry of our sample is closer to the P2/n-to-C2/c transition than that of Na(Mg0.45Al0.10Si0.45)Si2O6 (Yang et al., 2009), in agreement with the conclusion we derived above from the structure refinement.