Na6Si2O7 – the missing structural link among alkali pyrosilicates Volker Kahlenberg, Thomas Langreiter, Erik Arroyabe
To cite this version:
Volker Kahlenberg, Thomas Langreiter, Erik Arroyabe. Na6Si2O7 – the missing structural link among alkali pyrosilicates. Journal of Inorganic and General Chemistry / Zeitschrift für anorganische und allgemeine Chemie, Wiley-VCH Verlag, 2010, 636 (11), pp.1974. 10.1002/zaac.201000120. hal- 00552471
HAL Id: hal-00552471 https://hal.archives-ouvertes.fr/hal-00552471 Submitted on 6 Jan 2011
HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. ZAAC
Na6Si2O7 – the missing structural link among alkali pyrosilicates
Journal: Zeitschrift für Anorganische und Allgemeine Chemie
Manuscript ID: zaac.201000120.R1
Wiley - Manuscript type: Article
Date Submitted by the 13-Apr-2010 Author:
Complete List of Authors: Kahlenberg, Volker; University of Innsbruck, Institute of Mineralogy and Petrography Langreiter, Thomas; University of Innsbruck, Institute of Mineralogy and Petrography Arroyabe, Erik; University of Innsbruck, Institute of Mineralogy and Petrography
Keywords: Na6Si2O7, sodium pyrosilicate, sorosilicate, twinning
Wiley-VCH Page 1 of 21 ZAAC
1 2 3 4 NaNaNa 666SiSiSi 222OOO777 ––– the missing structural link among alkali pyrosilicates 5 6 7 8 9 a a a 10 V. Kahlenberg , T. Langreiter and E. Arroyabe 11 12 13 14 15 16 a 17 Innsbruck/Austria, Institute of Mineralogy and Petrography, University of Innsbruck, 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 *Prof. Dr. Volker Kahlenberg 43 44 Institut für Mineralogie und Petrographie 45 46 Leopold-Franzens-Universität Innsbruck 47 Innrain 52 48 49 A – 6020 Innsbruck 50 51 Tel.: +43(0)5125075503 ; FAX: +43(0)5125072926 52 53 E-Mail: [email protected] 54 55 56 57 58 59 60
Wiley-VCH ZAAC Page 2 of 21
1 2 3 4 Abstract 5 6 The crystal structure of sodium pyrosilicate (Na 6Si 2O7) has been solved from single crystal 7 8 diffraction data and refined to an R index of 0.051 for 17034 independent reflections. The 9 10 compound is triclinic with space group P 1 (a = 5.8007(8) Å, b = 11.5811(15) Å, c = 11 3 12 23.157(3) Å, α = 89.709(10)°, β = 88.915(11)°, γ = 89.004(11)°, V = 1555.1(4) Å , Z = 8, D x = 13 3 1 14 2.615 g/cm , (Mo Kα) = 7.94 cm ). A characteristic feature of the crystals is a twinning by 15 16 reticular pseudo merohedry, simulating a much larger monoclinic C centered lattice ( V’ = 17 18 6220 Å3, Z = 32). The twin element corresponds to a twofold rotation axis running parallel 19 20 to the [0 2 1] direction of the triclinic cell. The compound belongs to the group of 21 22 sorosilicates, i.e. it is based on [Si 2O7] groups, which are arranged in layers parallel to 23 24 (100). Charge compensation within the structure is accomplished by monovalent Na 25 26 cations distributed among 24 crystallographically independent positions. They are 27 28 coordinated by four to six nearest oxygen neighbors. Most of the coordination polyhedra 29 30 can be approximately described as distorted tetrahedra or tetragonal pyramids. An 31 32 alternative understanding of Na 6Si 2O7 can be gained if the tetrahedrally coordinated 33 34 sodium atoms are considered for the construction of a framework. Actually, each four of 35 36 the dimers within a single slice are linked by a more or less distorted [NaO 4] tetrahedron. 37 38 The resulting structural motif is similar to the one that can be observed in melilites, where 39 40 linkage between the T 2O7 (T:Al, Si) moieties is provided by [MgO 4] (as in akermanite, 41 42 Ca 2Mg[Si 2O7]) or [AlO 4] tetrahedra (as in gehlenite, Ca 2Al[AlSiO 7]). By sharing common 43 44 edges, the [NaO 4] tetrahedra in Na 6Si 2O7 are forming columns running parallel to [100]. 45 46 The resulting framework contains tunnels in which the more irregularly coordinated 47 48 sodium cations are incorporated. 49 50 51 52 53 54 55 56 Keywords : Na 6Si 2O7, sodium pyrosilicate, sorosilicate, twinning 57 58 59 60
2 Wiley-VCH Page 3 of 21 ZAAC
1 2 3 4 Introduction 5 6 Sodium silicates have been studied intensively in the past since they are of special interest 7 8 for certain areas of industrial inorganic chemistry and technical mineralogy. Fields of 9 10 applications include, for example, production of water glass solutions, ion exchangers and 11 12 builders in washing powders, making of acid resistant enamel frits or components of 13 14 refractory cements as well as inorganic binders just to mention a few. Therefore, it is not 15 16 surprising that the phase diagram Na 2O SiO 2 has been investigated frequently. 17 18 Concurrently, Kracek [1] as well as D’Ans & Löffler [2] reported results on the phase 19 20 relationships in the alkali rich part of this binary system. However, their observations 21 22 concerning the number of crystalline phases and their melting behaviour were 23 24 contradictory. Whereas D’Ans & Löffler mentioned the occurrence of a congruently 25 26 melting silicate with composition 3Na 2O 2SiO 2 (melting point: 1115°C), Kracek could not 27 28 find any evidence for this so called 3:2 phase. However, several later investigations [3 7] 29 30 undoubtedly proved the existence of the compound Na 6Si 2O7. 31 32 First basic structural data including a powder diffraction pattern as well as a proposal for 33 34 the unit cell parameters for sodium pyrosilicate were given by Kautz, Müller & Schneider 35 36 [8]. Moreover, the same authors reported that Na 6Si 2O7 has a lower stability of about 37 38 620°C where it decomposes into Na 2Si 2O5 and Na 4SiO 4 but that it can be preserved at 39 40 ambient conditions by quenching from higher temperatures. 41 42 In summary one can say, that eighty years after its first description in this journal and 43 44 forty years after a preliminary crystallographic characterization a detailed structural 45 46 investigation of sodium pyrosilicate is still missing. In the course of a long term project 47 48 aiming on the elucidation of the phase equilibria and the crystal chemistry of alkali 49 50 silicates, we decided to study the crystal structure of Na 6Si 2O7 in more detail. 51 52 53 54 Experimental details 55 56 So far, successful synthesis experiments of sodium pyrosilicate were based on the 57 58 following approaches: (a) conversion of Na 2SiO 3 NaOH mixtures [2,8] (b) reaction 59 60 between Na 2O2 and SiO 2 [8] and (c) thermal decomposition of Na 3(HSiO 4) 5H 2O [9]. Since
Na 6Si 2O7 has been reported to melt congruently at about 1115°C [2] we decided to grow
3 Wiley-VCH ZAAC Page 4 of 21
1 2 3 4 single crystalline material directly from the melt. Starting materials for our own 5 6 preparations were Na 2CO 3 (Merck, p.a.) and SiO2 (quartz, Alfa Aesar, 99.995%). The 7 8 educts for 2g of a stoichiometric mixture were homogenized in a planetary mill for 45 9 10 minutes under ethanol, dried at 60°C, placed into a platinum crucible, covered with a lid 11 12 and transferred to a resistance heated furnace. The samples were fired from 300°C to 13 14 1200°C with a heating rate of 100°C/h, held at the final temperature for 2 hours, 15 16 subsequently cooled with 10°C/h to 700°C and finally quenched in air. A first inspection 17 18 of the run product was based on light microscopy as well as X ray powder diffraction. As 19 20 known from previous investigations, sodium pyrosilicate is very hygroscopic. Therefore, 21 22 after removing of the crucible from the furnace, the part of the sample which was 23 24 intended to be used for the microscopic studies and the selection of single crystals was 25 26 immediately covered with inert oil (Paratone N, Hampton Research). For the same 27 28 reason, the preparation for the X ray powder diffraction samples including grinding of the 29 30 material and filling of the glass capillaries (0.3 mm diameter) was performed in a glove bag 31 32 under an argon atmosphere. The optical investigations using a polarizing microscope 33 34 proved the yield to be a mixture of two crystalline phases. A smaller amount of platy, low 35 36 birefringent crystals with well developed faces (phase 1) occurred along with large 37 38 irregularly shaped crystals of high birefringence showing polysynthetic twinning (phase 39 40 2). The powder diffraction pattern collected on a STOE STADI MP diffractometer 41 42 operated in transmission geometry indicated Na 6Si 2O7 (Powder Diffraction File (PDF 2), 43 44 entry no. 27 0784) to be the main phase (= phase 2) of the synthesis run. Furthermore, 45 46 sodium metasilicate (PDF 2 entry 16 0818) could be detected (corresponding to phase 1). 47 48 The presence of Na 2SiO 3 was surprising because the initial mixture had a Na 2O:SiO 2 ratio 49 50 of 3:2. We attribute this finding to a potential evaporation loss of sodium during the 51 52 crystal growth experiment. 53 54 Structural investigations of single crystalline Na 6Si 2O7 were performed on an Oxford 55 56 Diffraction Gemini R Ultra single crystal diffractometer using Mo Kα radiation. Therefore, 57 58 optically pre selected crystals were mounted on the tip of a glass fiber with nail hardener. 59 60 Almost twenty crystals were screened by short data collections revealing the low overall diffraction quality of the samples. In addition to the already mentioned twinning
4 Wiley-VCH Page 5 of 21 ZAAC
1 2 3 4 phenomena almost all crystals exhibited radial smearing of the reflections. For the final 5 6 data collection a twinned crystal with comparatively sharp diffraction spots was selected. 7 8 Acquisition of the intensities was accomplished at 25°C using a nitrogen stream generated 9 10 by an Oxford Cryostream 700 series cooler. Flushing the crystal in dried nitrogen 11 12 successfully prevented the decay of the hygroscopic material during the data collection. 13 14 Processing of the data with the CrysAlisPro software package indicated the “single crystal” 15 16 to be an intergrowth of one prominent (I) and two minor (II,III) individuals which all 17 18 three could be indexed with a pseudo tetragonal, monoclinic C centered cell, similar to 19 20 the one that had been already proposed by Kautz et al. [8]: a’ = 32.84 Å, b’ = 32.67 Å, c’ = 21 22 5.80 Å, β’ = 91.47°. Due to the low absorption coefficient of the material for Mo Kα 23 24 radiation ( = 0.794 mm 1) no absorption correction has been performed. Scattering curves 25 26 for neutral atoms, together with anomalous dispersion corrections, were taken from the 27 28 International Tables for Crystallography, Volume C [10]. However, structure solution by 29 30 direct methods with the non overlapping data belonging to component I in all possible 31 32 monoclinic C centered space groups corresponding to the extinction symbol 2/m C1 1 33 34 failed (program SIR2002 [11]). 35 36 A thorough investigation of precession type reconstructions of reciprocal space disclosed 37 38 that the diffraction pattern of the large C centered cell can be explained by 39 40 superposition/twinning of two much smaller triclinic cells with the following lattice 41 42 parameters: a = 5.8007 Å, b = 11.5811 Å, c = 23.157 Å, α = 89.709°, β = 88.915°, γ = 89.004° 43 44 (V = 1555 Å 3). The twin element corresponds to a twofold rotation axis running parallel to 45 46 the [0 2 1] direction of the triclinic cell. The primed and unprimed basis vectors of the 47 48 small and the large cells are related via the following transformation: a’a’a’ = 2b2b2b+c2b ccc, b’b’b’ = 2b2b2b c2b ccc 49 50 and c’c’c’= c’ aaaa. In summary one can say that each of the three domains within the “single 51 52 crystal” in turn consists of two twin individuals related by reticular pseudo merohedry. In 53 54 the next step, the data of the largest domain I were re processed in order to produce two 55 56 data sets Ia and Ib including only the non overlapping reflections (for structure solution) 57 58 as well as a third data set containing the specific twin information, i.e. the overlapping as 59 60 well as the non overlapping reflections (so called HKLF5 format) for the subsequent refinement. Structure determination with data set Ia using direct methods resulted in a
5 Wiley-VCH ZAAC Page 6 of 21
1 2 3 4 crystallochemically reasonable model showing the expected number of [SiO 4] tetrahedra 5 6 as well as most of the sodium atoms. After completing the structure by difference Fourier 7 8 calculations (program SHELX97 [12]), subsequent least squares refinements with isotropic 9 10 displacement parameters converged to a residual of R(|F|)=0.208. Taking the twin model 11 12 into consideration improved the calculations substantially (R(|F|)=0.059, 242 parameters). 13 14 The introduction of anisotropic displacement parameters for all atoms, however, lowered 15 16 the residual index only slightly (R(|F|=0.039, 543 parameters) and resulted in a non 17 18 positive definite temperature factor of the oxygen atom O5. A re examination of the 19 20 diffraction data as well as of the crystal structure using the MISSYM algorithm 21 22 implemented in the PLATON program suite [13] did not reveal any indication that a 23 24 wrong space group symmetry had been chosen nor did we detect any evidence for a 25 26 systematic error in the data reduction. We attribute the problem with the thermal motion 27 28 of the oxygen atom O5 to the generally lower quality of the twinned data set relative to 29 30 the diffraction data which can be gained from a good single crystal. The final atomic 31 32 coordinates of the calculations with anisotropic temperature factors for the Si and Na 33 34 atoms as well as selected bond distances and angles are given in Table 2, 3 and 4 35 36 respectively. Crystallographic data for the structure reported here have been deposited 37 38 with the Fachinformationszentrum Karlsruhe, D 76344 Eggenstein Leopoldshafen, 39 40 Germany ( crysdata@FIZ Karlsruhe.de ), and are available on quoting the deposition 41 42 number 421643. 43 44 Using the structure models for Na 6Si 2O7 as well as for Na 2SiO 3 [14] a quantitative phase 45 46 analysis based on the Rietveld method was performed. The two phase refinement 47 48 confirmed that sodium pyrosilicate is the dominant crystalline compound (67.3 wt.%) in 49 50 the sample. 51 52 53 54 Results and discussion 55 56 As may be anticipated from the chemical formula, Na 6Si 2O7 belongs to the group of 57 58 sorosilicates, i.e. the material is based on [Si 2O7] groups. In more detail, a total of four 59 60 crystallographically independent bitetrahedral units occupying general positions have to be distinguished. As shown in Figure 1a, the tetrahedra are arranged in layers parallel to
6 Wiley-VCH Page 7 of 21 ZAAC
1 2 3 4 (100). Charge compensation within the structure is accomplished by monovalent Na 5 6 cations, which are distributed among a total of 24 crystallographically independent 7 8 positions. 9 10 The Si O bond distances within the dimers show a considerable spread, ranging between 11 12 1.590 1.685 Å. However, the observed variation follows the expected trend for [Si 2O7] 13 14 groups having one bridging and three terminal atoms: the distances between the silicon 15 16 atoms and the terminal oxygens are considerably shorter (average: 1.614 Å) than the 17 18 corresponding bond length to the bridging oxygens O(4), O(11), O(18) and O(25) 19 20 (average: 1.673 Å). The shortening of the Si Oterm bond lengths (by an average of 0.059 Å) 21 22 results from the stronger attraction between O and Si than between O and the sodium 23 24 cations in the structure. The distortion is also reflected in the O Si O angles ranging from 25 26 101.9° to 115.7°, respectively. Nevertheless, the mean