ChemComm
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
COMMUNICATION
View Article Online View Journal | View Issue
Framework Al zoning in zeolite ECR-1† Cite this: Chem. Commun., 2014, 50, 1956
Jiho Shin,a Nak Ho Ahn,a Sung June Cho,b Limin Ren,c Feng-Shou Xiaoc and Suk Bong Hong*a
Received 2nd November 2013, Accepted 13th December 2013 DOI: 10.1039/c3cc48403c www.rsc.org/chemcomm
Rietveld analyses of the synchrotron X-ray diffraction data for various cation forms of zeolite ECR-1 have demonstrated framework Al zoning, which parallels the alternation of Al-rich maz and Al-poor mor layers. This can be further supported by notable differences in the average bond valence of its 10 crystallographically distinct tetrahedral sites.
Zeolites are microporous, crystalline aluminosilicates that are constructed from full corner sharing of SiO4 and AlO4 tetrahedra, yielding a three-dimensional four-connected network. While they are widely used as industrial adsorbents, ion exchangers, and catalysts, their performance depends primarily not only on the framework Al content, but also on the Al distribution over the framework tetrahedral sites (T-sites). Hence, detailed information about the local ordering of Al atoms in zeolites is essential for systematically controlling many important physicochemical and catalytic properties of these microporous materials.1 Because of the quite similar scattering factors of Si and Al for X-rays, however, it is almost impossible to directly locate Al atoms by conventional diffraction techniques. One alternative way around this difficulty is to determine the positions of extraframework cations balancing the negative charges of the zeolite framework created by Al substitution. The large-pore zeolite ECR-1 (framework type EON) was first synthesized by Vaughan and Strohmaier using bis(2-hydroxyethyl)dimethylammonium and Na+ cations as structure-directing agents (SDAs). It has been proposed to consist of strictly alternating maz and mor layers from which the well-known mazzite (MAZ) and mordenite (MOR) structures are built, respectively.2 Because of a
Centre for Ordered Nanoporous Materials Synthesis, School of Environmental Science and Engineering and Department of Chemical Engineering, POSTECH, Pohang 790-784, Korea. E-mail: [email protected] b Department of Applied Chemical Engineering, Chonnam National University, Kwangju 500-757, Korea c Department of Chemistry, Zhejiang University, Hangzhou 310028, P. R. China † Electronic supplementary information (ESI) available: Experimental procedures, final atomic coordinates, refined extraframework cation positions and occupancies, and final Rietveld plots. See DOI: 10.1039/c3cc48403c
1956 | Chem. Commun., 2014, 50, 1956--1958
the presence of heavy stacking faults in its crystals, however, the structure refinement of ECR-1 became successful only when the crystal structure of TNU-7, one of the synthetic gallosilicate zeolites discovered by us under wholly inorganic conditions, was reported.3,4 On the other hand, Rietveld analyses of the synchrotron X-ray diffraction (XRD) data for various cation forms of TNU-7 have revealed the existence of structural chemical zoning within its structure, i.e., the alteration of Ga-rich maz and Ga-poor mor layers perpendicular to the a direction.3b Thus, this behaviour is substantially different from the Al zoning observed for ZSM-5 at the single-crystal level.5 Here we show that ECR-1 follows the same trend as TNU-7, its aluminosilicate analogue: the maz and mor layers of this boundary phase are rich and poor in Al, respectively. We also show that the actual distribution of trivalent framework atoms (i.e., Al or Ga) over the 10 crystallographically distinct T-sites of the EON framework is non-random, which has been corroborated by comparing the bond valence of each T-site in ECR-1 with that in TNU-7. Although the bond valence method or sum is widely used in crystallography to validate the newly determined crystal structures,6 to our knowledge, it has not been applied to the structural characterization of zeolites yet. A Na-ECR-1 zeolite with Si/Al = 3.5 was synthesized without using any organic SDA according to the procedures reported by Song et al.7 The Sr2+- and La3+-exchanged forms (Sr-ECR-1 and LaNa-ECR-1, respectively) of ECR-1 were prepared by stirring Na-ECR-1 twice in 1.0 M Sr(NO3)2 and 0.1 M La(NO3)3 solutions at 80 1C for 6 h, respectively. The high-resolution synchrotron XRD patterns for these three different cation forms of ECR-1 in the dehydrated state were collected in capillary mode on the BL44B2 beamline at the SPring-8 (Harima, Japan) using a monochromated X-ray (l = 0.5000 Å). For comparison, the synchrotron XRD patterns for Na-MAZ and Na-MOR zeolites with Si/Al ratios of 3.2 and 5.0, respectively, were also obtained in flat plate mode on the 9B beamline at the PAL (Pohang, Korea) using monochromatic synchrotron radiation (l = 1.5472 Å). Details of the XRD measurements, as well as the Rietveld refinements, are described in the ESI.†
This journal is © The Royal Society of Chemistry 2014
View Article Online
Communication
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
Table 1
ChemComm
Compositional and crystallographic data for zeolites studied here and their agreement factors for Rietveld refinement
Zeolite
Anhydrous unit cell compositiona
Degree (%) of ion exchange
Na-MAZ Na-ECR-1 Sr-ECR-1 LaNa-ECR-1 Na-MOR
Na8.6Al8.6Si27.4O72 Na13.4Al13.4Si46.6O120 Sr6.4Na0.6Al13.4Si46.6O120 La2.4Na6.2Al13.4Si46.6O120 Na8.0Al8.0Si40.0O96
— — 95 55 —
a
Unit cell parameter
Agreement factors
Si/Al
a (Å)
b (Å)
c (Å)
V (Å3)
Rwp (%)
Rp (%)
3.2
18.19274(13) 7.5724(4) 7.5674(4) 7.5657(4) 18.09463(25)
18.19274(13) 18.1615(11) 18.1270(13) 18.0990(13) 20.35695(28)
7.61853(5) 25.8749(27) 25.9753(29) 25.8792(28) 7.49005(12)
2183.725(30) 3558.5(5) 3563.1(5) 3543.7(5) 2758.97(7)
6.05 2.79 1.92 2.49 6.52
4.09 2.02 1.27 1.80 4.59
3.5 5.0
Determined from elemental analysis.
Table 1 lists the compositional and crystallographic data for all five zeolites studied in this work, together with their agreement factors for Rietveld analyses. The structures of three different cation-exchanged forms of ECR-1 have been refined based on the framework model in the space group Pmmn proposed by Gualtieri et al.,4 and the positions of extraframework cations were derived from Fourier difference maps. Excellent final Rwp values of less than 3% were obtained for these materials, satisfactorily confirming the proposed structure model. Indeed, the ECR-1 structure is composed of maz and mor layers connected by chains of 5-rings in a regular 1 : 1 stacking sequence. It thus appears that different environments exist on the different sides near the two layers within its asymmetric 12-ring (6.6 7.4 Å) channels. Fig. 1 shows the refined extraframework cation positions in Na-ECR-1, Sr-ECR-1, and LaNa-ECR-1. Unlike the case of Na-ECR-1 reported by Gualtieri et al.,4 in which five different cation sites exist, six different sites labelled A–F were found in our Na-ECR-1. As shown in Fig. 1, sites A and B are located at the puckered 8-ring within the maz layer and the smaller interstrip channel, respectively. While site C is in the large 12-ring channel associated with the mor layer, in addition, sites D and E are within the maz strip. Position F is in the 6-ring of the distorted gme-cage of the maz layer. These locations match well with the three Na+ positions in Na-MAZ and Na-MOR, respectively (see ESI†). A similar result was obtained for Sr-ECR-1 and LaNa-ECR-1. However, site B in Sr-ECR-1 is split into two sites, which may be associated with the movement of the larger Sr2+ ion away from the puckered 8-ring and into the adjacent 8-ring (Fig. 1). We also note that site E is empty and an additional site G is close to 4-rings of the gme-cage. In LaNa-ECR-1, the partially La3+-exchanged ECR-1, on the other hand, neither La3+ ions nor Na+ ions occupy site F. Such changes in cation distribution in ECR-1 caused by
Table 2
Extraframework cation charges per T-site in zeolites studied here
Extraframework cation charge per T-sitea Zeolite
maz layer
Na-MAZ Na-ECR-1 Sr-ECR-1 LaNa-ECR-1 Na-MOR
0.247 (0.202) 0.335 (0.267) 0.144 (0.208) 0.199
mor layer 0.143 (0.090) 0.114 (0.145) 0.106 0.161 (0.134)
a
The values in parentheses are the extraframework cation charges per T-atom in gallosilicate zeolites with the corresponding framework topologies.
Sr2+ or La3+ ion exchange can be attributed to differences in the cation charge and radius (Na+, 1.02 Å; Sr2+, 1.18 Å; La3+, 1.03 Å).8 Table 2 lists extraframework cation charges per T-site of the maz and mor layers in three different cation forms of ECR-1, together with those for Na-MAZ and Na-MOR. Also, the values in the distinct layers within TNU-7, the gallosilicate analog of ECR-1, and in gallosilicate Na-MAZ and Na-MOR zeolites are given for comparison. It is clear that there are more cations in the maz layer of ECR-1 than in the mor one, as found in Na-MAZ and Na-MOR that are rich and poor in Al, respectively. This suggests the presence of ‘‘Al zoning’’ in the ECR-1 framework, like the alteration of Ga-rich maz and Ga-poor mor layers in TNU-7.3b To obtain further evidence for the framework Al zoning in ECR-1, we have calculated the bond valences of 10 distinct T-sites in Na-ECR-1 from its crystallographic data. Clearly, when Al and Si atoms are occupied in the particular T-sites of the zeolite framework the bond valences are calculated to be 3.0 and 4.0, respectively. Hence, the higher Al occupancy the T-site has, the smaller value it shows. The bond valence data presented in Table 3 reveal that the average value (3.84 vs. 3.99)
Fig. 1 Refined extraframework cation sites in dehydrated (a) Na-ECR-1, (b) Sr-ECR-1, and (c) LaNa-ECR-1, viewed along [100]. Large-coloured circles represent the extraframework cations (yellow for Na+, green for Sr2+, and purple for La3+).
This journal is © The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 1956--1958 | 1957
View Article Online
ChemComm
Communication
Table 3 Bond valences of 10 crystallographically distinct T-sites in Na-ECR-1 and Na-TNU-7 zeolites
Bond valence,b VT
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
T-sitea
Na-ECR-1(I)c,d
Na-ECR-1(II)d,e
Na-TNU-7d, f
T1(4) T2(4) T3(4) T4(8) T5(8) T6(8)
maz layer
3.92 3.80 3.76 3.54 3.88 4.14
(3.84)
3.56 3.81 4.02 3.67 3.89 3.68
(3.76)
3.83 3.96 3.76 3.45 3.78 3.66
(3.70)
T7(4) T8(4) T9(8) T10(8)
mor layer
3.97 3.81 4.00 4.07
(3.99)
4.06 4.16 3.99 4.27
(4.12)
4.01 4.06 4.00 3.76
(3.93)
3.78
(3.90)
3.81
(3.91)
3.80
VT,avgg a
(3.79) b
The values in parentheses are the multiplicity of each T-site. Calculated using the equation VSi = Sexp[(RSi–O dSi–O)/b], where VSi–O is the average Si–O bond distance between each crystallographically distinct Si atom and its four different O atoms in the zeolite structure crystallographically determined, dSi–O is the ideal Si–O bond distance in zeolites equal to 1.624 Å, and b is an empirical parameter, usually 0.37 Å for oxides.6 c Calculated from the crystallographic data of NaECR-1 with Si/Al = 3.5 obtained in this work. d The values in parentheses are the average bond valences of the available T-sites in the maz or mor layer within each EON-type zeolite. e Calculated from the crystallographic data of Na-ECR-1 with Si/Al = 4.2 reported in ref. 4. f Calculated from the crystallographic data of Na-TNU-7 with Si/Ga = 3.9 given in ref. 3b. g The average bond valence of all T-sites in each zeolite calculated from its Si/Me ratio (Me = Al or Ga) determined from elemental analysis. The values in parentheses are the average bond valances crystallographically determined.
of T-sites is rather smaller in the maz layer of Na-ECR-1(I), the ECR-1 sample prepared in our work. The same trend was observed for Na-ECR-1(II), the ECR-1 zeolite synthesized by Gualtieri et al.,4 as well as for TNU-7. This again confirms a larger Al or Ga content within the maz region, i.e., the presence of structural chemical zoning in EON-type zeolites. Another important observation obtained from Table 3 is that there are notable differences in the bond valence of T-sites in three EONtype zeolites studied here. For example, the bond valence (3.92) of site T1 in the maz region within Na-ECR-1(I) is larger by ca. 0.4 than the value (3.54) of site T4 in the same layer. Such a difference in the bond valence implies that the percentage difference between the framework Al and Si atoms in a particular T-site is about 40%, which in our view may be large enough to be considered as clear evidence for the non-random Al distribution over the available T-sites in zeolites. Similar differences in the bond valence were found when not only the values of site T1 or T6 in Na-ECR-1(I) and Na-ECR-1(II), but also those of site T6 or T10 in Na-ECR-1(I) and Na-TNU-7 are compared with each other, respectively. We should note here that the final ‘‘goodness-of-fit’’ values (Rwp’s) for the Rietveld refinements of the powder XRD data were satisfactory
1958 | Chem. Commun., 2014, 50, 1956--1958
(o9%) for all these EON-type zeolites.3,4 We also note that Na-ECR-1(I) and Na-TNU-7 were prepared without the aid of any organic SDA, but Na-ECR-1(II) was prepared using a small amount of tetramethylammonium ions together with Na+.3,4,7 Therefore, it is most likely that the spatial distribution of trivalent framework atoms (i.e., Al and Ga) in EON-type zeolites is altered according to both the synthesis method and type of heteroatoms other than Si added to the zeolite synthesis mixture. This may explain why zeolites with the same framework topology and chemical composition, which were obtained by the same synthesis procedure but from different synthesis runs, frequently show differences in their physicochemical and/ or catalytic properties. In summary, we have demonstrated that the maz and mor layers in zeolite ECR-1 are rich and poor in Al, respectively. In particular, comparison of the bond valence values of 10 distinct T-sites in this boundary phase with those in its gallosilicate analog TNU-7 reveals the non-random nature of substitution of trivalent framework atoms (i.e., Al or Ga) over the available T-sites in EON-type zeolites. This work was supported by the National Creative Research Initiative Program (2012R1A3A2048833) and the BK 21-plus program through the National Research Foundation of Korea funded by the Korea government (MSIP). We thank the beamline scientists at PAL and SPring-8 for their assistance with the powder diffraction measurements. The work at PAL and SPring-8 was supported in part by MSIP and POSTECH.
Notes and references 1 (a) J. A. van Bokhoven, J. Am. Chem. Soc., 2000, 122, 12842; (b) D. H. Olson, N. Khosrovani, A. W. Peters and B. H. Toby, J. Phys. Chem. B, 2000, ´mez, A. R. Ruiz-Salvador, M. Mistry 104, 4844; (c) N. Almora-Barrios, A. Go and D. W. Lewis, Chem. Commun., 2001, 531; (d) O. H. Han, C. S. Kim and ˇdec ˇek, S. B. Hong, Angew. Chem., Int. Ed., 2002, 41, 469; (e) S. Sklenak, J. De ´, V. Ga ´bova ´, M. Sierka and J. Sauer, Angew. Chem., Int. C. Li, B. Wichterlova Ed., 2007, 46, 7286. 2 (a) D. E. W. Vaughan and K. G. Strohmaier, US Pat., 4,657,748, 1987; (b) M. E. Leonowicz and D. E. W. Vaughan, Nature, 1987, 329, 819; (c) C. S. Chen, J. L. Schlenker and S. E. Wentzek, Zeolites, 1996, 17, 393. 3 (a) S. J. Warrender, P. A. Wright, W. Zhou, P. Lightfoot, M. A. Camblor, C.-H. Shin, D. J. Kim and S. B. Hong, Chem. Mater., 2005, 17, 1272; (b) B. Han, C.-H. Shin, S. J. Warrender, P. Lightfoot, P. A. Wright, M. A. Camblor and S. B. Hong, Chem. Mater., 2006, 18, 3023. 4 A. F. Gualtieri, S. Ferrari, E. Galli, F. Di Renzo and W. Van Beek, Chem. Mater., 2006, 18, 76. 5 (a) R. von Ballmoos and W. M. Meier, Nature, 1981, 289, 782; (b) Z. Ristanovic´, J. P. Hofmann, U. Deka, T. U. Schulli, M. Rohnke, A. M. Beale and B. M. Weckhuysen, Angew. Chem., Int. Ed., 2013, 52, 13382. 6 (a) N. E. Brese and M. O’Keeffe, Acta Crystallogr., 1991, B47, 192; (b) I. D. Brown, The Chemical Bond in Inorganic Chemistry: The Bond Valence Model, Oxford University Press, New York, 2002; (c) I. D. Brown, Chem. Rev., 2009, 109, 6858. 7 J. Song, L. Dai, Y. Ji and F.-S. Xiao, Chem. Mater., 2006, 18, 2775. 8 R. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751.
This journal is © The Royal Society of Chemistry 2014
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
COMMUNICATION
View Article Online View Journal | View Issue
Framework Al zoning in zeolite ECR-1† Cite this: Chem. Commun., 2014, 50, 1956
Jiho Shin,a Nak Ho Ahn,a Sung June Cho,b Limin Ren,c Feng-Shou Xiaoc and Suk Bong Hong*a
Received 2nd November 2013, Accepted 13th December 2013 DOI: 10.1039/c3cc48403c www.rsc.org/chemcomm
Rietveld analyses of the synchrotron X-ray diffraction data for various cation forms of zeolite ECR-1 have demonstrated framework Al zoning, which parallels the alternation of Al-rich maz and Al-poor mor layers. This can be further supported by notable differences in the average bond valence of its 10 crystallographically distinct tetrahedral sites.
Zeolites are microporous, crystalline aluminosilicates that are constructed from full corner sharing of SiO4 and AlO4 tetrahedra, yielding a three-dimensional four-connected network. While they are widely used as industrial adsorbents, ion exchangers, and catalysts, their performance depends primarily not only on the framework Al content, but also on the Al distribution over the framework tetrahedral sites (T-sites). Hence, detailed information about the local ordering of Al atoms in zeolites is essential for systematically controlling many important physicochemical and catalytic properties of these microporous materials.1 Because of the quite similar scattering factors of Si and Al for X-rays, however, it is almost impossible to directly locate Al atoms by conventional diffraction techniques. One alternative way around this difficulty is to determine the positions of extraframework cations balancing the negative charges of the zeolite framework created by Al substitution. The large-pore zeolite ECR-1 (framework type EON) was first synthesized by Vaughan and Strohmaier using bis(2-hydroxyethyl)dimethylammonium and Na+ cations as structure-directing agents (SDAs). It has been proposed to consist of strictly alternating maz and mor layers from which the well-known mazzite (MAZ) and mordenite (MOR) structures are built, respectively.2 Because of a
Centre for Ordered Nanoporous Materials Synthesis, School of Environmental Science and Engineering and Department of Chemical Engineering, POSTECH, Pohang 790-784, Korea. E-mail: [email protected] b Department of Applied Chemical Engineering, Chonnam National University, Kwangju 500-757, Korea c Department of Chemistry, Zhejiang University, Hangzhou 310028, P. R. China † Electronic supplementary information (ESI) available: Experimental procedures, final atomic coordinates, refined extraframework cation positions and occupancies, and final Rietveld plots. See DOI: 10.1039/c3cc48403c
1956 | Chem. Commun., 2014, 50, 1956--1958
the presence of heavy stacking faults in its crystals, however, the structure refinement of ECR-1 became successful only when the crystal structure of TNU-7, one of the synthetic gallosilicate zeolites discovered by us under wholly inorganic conditions, was reported.3,4 On the other hand, Rietveld analyses of the synchrotron X-ray diffraction (XRD) data for various cation forms of TNU-7 have revealed the existence of structural chemical zoning within its structure, i.e., the alteration of Ga-rich maz and Ga-poor mor layers perpendicular to the a direction.3b Thus, this behaviour is substantially different from the Al zoning observed for ZSM-5 at the single-crystal level.5 Here we show that ECR-1 follows the same trend as TNU-7, its aluminosilicate analogue: the maz and mor layers of this boundary phase are rich and poor in Al, respectively. We also show that the actual distribution of trivalent framework atoms (i.e., Al or Ga) over the 10 crystallographically distinct T-sites of the EON framework is non-random, which has been corroborated by comparing the bond valence of each T-site in ECR-1 with that in TNU-7. Although the bond valence method or sum is widely used in crystallography to validate the newly determined crystal structures,6 to our knowledge, it has not been applied to the structural characterization of zeolites yet. A Na-ECR-1 zeolite with Si/Al = 3.5 was synthesized without using any organic SDA according to the procedures reported by Song et al.7 The Sr2+- and La3+-exchanged forms (Sr-ECR-1 and LaNa-ECR-1, respectively) of ECR-1 were prepared by stirring Na-ECR-1 twice in 1.0 M Sr(NO3)2 and 0.1 M La(NO3)3 solutions at 80 1C for 6 h, respectively. The high-resolution synchrotron XRD patterns for these three different cation forms of ECR-1 in the dehydrated state were collected in capillary mode on the BL44B2 beamline at the SPring-8 (Harima, Japan) using a monochromated X-ray (l = 0.5000 Å). For comparison, the synchrotron XRD patterns for Na-MAZ and Na-MOR zeolites with Si/Al ratios of 3.2 and 5.0, respectively, were also obtained in flat plate mode on the 9B beamline at the PAL (Pohang, Korea) using monochromatic synchrotron radiation (l = 1.5472 Å). Details of the XRD measurements, as well as the Rietveld refinements, are described in the ESI.†
This journal is © The Royal Society of Chemistry 2014
View Article Online
Communication
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
Table 1
ChemComm
Compositional and crystallographic data for zeolites studied here and their agreement factors for Rietveld refinement
Zeolite
Anhydrous unit cell compositiona
Degree (%) of ion exchange
Na-MAZ Na-ECR-1 Sr-ECR-1 LaNa-ECR-1 Na-MOR
Na8.6Al8.6Si27.4O72 Na13.4Al13.4Si46.6O120 Sr6.4Na0.6Al13.4Si46.6O120 La2.4Na6.2Al13.4Si46.6O120 Na8.0Al8.0Si40.0O96
— — 95 55 —
a
Unit cell parameter
Agreement factors
Si/Al
a (Å)
b (Å)
c (Å)
V (Å3)
Rwp (%)
Rp (%)
3.2
18.19274(13) 7.5724(4) 7.5674(4) 7.5657(4) 18.09463(25)
18.19274(13) 18.1615(11) 18.1270(13) 18.0990(13) 20.35695(28)
7.61853(5) 25.8749(27) 25.9753(29) 25.8792(28) 7.49005(12)
2183.725(30) 3558.5(5) 3563.1(5) 3543.7(5) 2758.97(7)
6.05 2.79 1.92 2.49 6.52
4.09 2.02 1.27 1.80 4.59
3.5 5.0
Determined from elemental analysis.
Table 1 lists the compositional and crystallographic data for all five zeolites studied in this work, together with their agreement factors for Rietveld analyses. The structures of three different cation-exchanged forms of ECR-1 have been refined based on the framework model in the space group Pmmn proposed by Gualtieri et al.,4 and the positions of extraframework cations were derived from Fourier difference maps. Excellent final Rwp values of less than 3% were obtained for these materials, satisfactorily confirming the proposed structure model. Indeed, the ECR-1 structure is composed of maz and mor layers connected by chains of 5-rings in a regular 1 : 1 stacking sequence. It thus appears that different environments exist on the different sides near the two layers within its asymmetric 12-ring (6.6 7.4 Å) channels. Fig. 1 shows the refined extraframework cation positions in Na-ECR-1, Sr-ECR-1, and LaNa-ECR-1. Unlike the case of Na-ECR-1 reported by Gualtieri et al.,4 in which five different cation sites exist, six different sites labelled A–F were found in our Na-ECR-1. As shown in Fig. 1, sites A and B are located at the puckered 8-ring within the maz layer and the smaller interstrip channel, respectively. While site C is in the large 12-ring channel associated with the mor layer, in addition, sites D and E are within the maz strip. Position F is in the 6-ring of the distorted gme-cage of the maz layer. These locations match well with the three Na+ positions in Na-MAZ and Na-MOR, respectively (see ESI†). A similar result was obtained for Sr-ECR-1 and LaNa-ECR-1. However, site B in Sr-ECR-1 is split into two sites, which may be associated with the movement of the larger Sr2+ ion away from the puckered 8-ring and into the adjacent 8-ring (Fig. 1). We also note that site E is empty and an additional site G is close to 4-rings of the gme-cage. In LaNa-ECR-1, the partially La3+-exchanged ECR-1, on the other hand, neither La3+ ions nor Na+ ions occupy site F. Such changes in cation distribution in ECR-1 caused by
Table 2
Extraframework cation charges per T-site in zeolites studied here
Extraframework cation charge per T-sitea Zeolite
maz layer
Na-MAZ Na-ECR-1 Sr-ECR-1 LaNa-ECR-1 Na-MOR
0.247 (0.202) 0.335 (0.267) 0.144 (0.208) 0.199
mor layer 0.143 (0.090) 0.114 (0.145) 0.106 0.161 (0.134)
a
The values in parentheses are the extraframework cation charges per T-atom in gallosilicate zeolites with the corresponding framework topologies.
Sr2+ or La3+ ion exchange can be attributed to differences in the cation charge and radius (Na+, 1.02 Å; Sr2+, 1.18 Å; La3+, 1.03 Å).8 Table 2 lists extraframework cation charges per T-site of the maz and mor layers in three different cation forms of ECR-1, together with those for Na-MAZ and Na-MOR. Also, the values in the distinct layers within TNU-7, the gallosilicate analog of ECR-1, and in gallosilicate Na-MAZ and Na-MOR zeolites are given for comparison. It is clear that there are more cations in the maz layer of ECR-1 than in the mor one, as found in Na-MAZ and Na-MOR that are rich and poor in Al, respectively. This suggests the presence of ‘‘Al zoning’’ in the ECR-1 framework, like the alteration of Ga-rich maz and Ga-poor mor layers in TNU-7.3b To obtain further evidence for the framework Al zoning in ECR-1, we have calculated the bond valences of 10 distinct T-sites in Na-ECR-1 from its crystallographic data. Clearly, when Al and Si atoms are occupied in the particular T-sites of the zeolite framework the bond valences are calculated to be 3.0 and 4.0, respectively. Hence, the higher Al occupancy the T-site has, the smaller value it shows. The bond valence data presented in Table 3 reveal that the average value (3.84 vs. 3.99)
Fig. 1 Refined extraframework cation sites in dehydrated (a) Na-ECR-1, (b) Sr-ECR-1, and (c) LaNa-ECR-1, viewed along [100]. Large-coloured circles represent the extraframework cations (yellow for Na+, green for Sr2+, and purple for La3+).
This journal is © The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 1956--1958 | 1957
View Article Online
ChemComm
Communication
Table 3 Bond valences of 10 crystallographically distinct T-sites in Na-ECR-1 and Na-TNU-7 zeolites
Bond valence,b VT
Published on 16 December 2013. Downloaded by University of Illinois at Chicago on 23/10/2014 16:48:31.
T-sitea
Na-ECR-1(I)c,d
Na-ECR-1(II)d,e
Na-TNU-7d, f
T1(4) T2(4) T3(4) T4(8) T5(8) T6(8)
maz layer
3.92 3.80 3.76 3.54 3.88 4.14
(3.84)
3.56 3.81 4.02 3.67 3.89 3.68
(3.76)
3.83 3.96 3.76 3.45 3.78 3.66
(3.70)
T7(4) T8(4) T9(8) T10(8)
mor layer
3.97 3.81 4.00 4.07
(3.99)
4.06 4.16 3.99 4.27
(4.12)
4.01 4.06 4.00 3.76
(3.93)
3.78
(3.90)
3.81
(3.91)
3.80
VT,avgg a
(3.79) b
The values in parentheses are the multiplicity of each T-site. Calculated using the equation VSi = Sexp[(RSi–O dSi–O)/b], where VSi–O is the average Si–O bond distance between each crystallographically distinct Si atom and its four different O atoms in the zeolite structure crystallographically determined, dSi–O is the ideal Si–O bond distance in zeolites equal to 1.624 Å, and b is an empirical parameter, usually 0.37 Å for oxides.6 c Calculated from the crystallographic data of NaECR-1 with Si/Al = 3.5 obtained in this work. d The values in parentheses are the average bond valences of the available T-sites in the maz or mor layer within each EON-type zeolite. e Calculated from the crystallographic data of Na-ECR-1 with Si/Al = 4.2 reported in ref. 4. f Calculated from the crystallographic data of Na-TNU-7 with Si/Ga = 3.9 given in ref. 3b. g The average bond valence of all T-sites in each zeolite calculated from its Si/Me ratio (Me = Al or Ga) determined from elemental analysis. The values in parentheses are the average bond valances crystallographically determined.
of T-sites is rather smaller in the maz layer of Na-ECR-1(I), the ECR-1 sample prepared in our work. The same trend was observed for Na-ECR-1(II), the ECR-1 zeolite synthesized by Gualtieri et al.,4 as well as for TNU-7. This again confirms a larger Al or Ga content within the maz region, i.e., the presence of structural chemical zoning in EON-type zeolites. Another important observation obtained from Table 3 is that there are notable differences in the bond valence of T-sites in three EONtype zeolites studied here. For example, the bond valence (3.92) of site T1 in the maz region within Na-ECR-1(I) is larger by ca. 0.4 than the value (3.54) of site T4 in the same layer. Such a difference in the bond valence implies that the percentage difference between the framework Al and Si atoms in a particular T-site is about 40%, which in our view may be large enough to be considered as clear evidence for the non-random Al distribution over the available T-sites in zeolites. Similar differences in the bond valence were found when not only the values of site T1 or T6 in Na-ECR-1(I) and Na-ECR-1(II), but also those of site T6 or T10 in Na-ECR-1(I) and Na-TNU-7 are compared with each other, respectively. We should note here that the final ‘‘goodness-of-fit’’ values (Rwp’s) for the Rietveld refinements of the powder XRD data were satisfactory
1958 | Chem. Commun., 2014, 50, 1956--1958
(o9%) for all these EON-type zeolites.3,4 We also note that Na-ECR-1(I) and Na-TNU-7 were prepared without the aid of any organic SDA, but Na-ECR-1(II) was prepared using a small amount of tetramethylammonium ions together with Na+.3,4,7 Therefore, it is most likely that the spatial distribution of trivalent framework atoms (i.e., Al and Ga) in EON-type zeolites is altered according to both the synthesis method and type of heteroatoms other than Si added to the zeolite synthesis mixture. This may explain why zeolites with the same framework topology and chemical composition, which were obtained by the same synthesis procedure but from different synthesis runs, frequently show differences in their physicochemical and/ or catalytic properties. In summary, we have demonstrated that the maz and mor layers in zeolite ECR-1 are rich and poor in Al, respectively. In particular, comparison of the bond valence values of 10 distinct T-sites in this boundary phase with those in its gallosilicate analog TNU-7 reveals the non-random nature of substitution of trivalent framework atoms (i.e., Al or Ga) over the available T-sites in EON-type zeolites. This work was supported by the National Creative Research Initiative Program (2012R1A3A2048833) and the BK 21-plus program through the National Research Foundation of Korea funded by the Korea government (MSIP). We thank the beamline scientists at PAL and SPring-8 for their assistance with the powder diffraction measurements. The work at PAL and SPring-8 was supported in part by MSIP and POSTECH.
Notes and references 1 (a) J. A. van Bokhoven, J. Am. Chem. Soc., 2000, 122, 12842; (b) D. H. Olson, N. Khosrovani, A. W. Peters and B. H. Toby, J. Phys. Chem. B, 2000, ´mez, A. R. Ruiz-Salvador, M. Mistry 104, 4844; (c) N. Almora-Barrios, A. Go and D. W. Lewis, Chem. Commun., 2001, 531; (d) O. H. Han, C. S. Kim and ˇdec ˇek, S. B. Hong, Angew. Chem., Int. Ed., 2002, 41, 469; (e) S. Sklenak, J. De ´, V. Ga ´bova ´, M. Sierka and J. Sauer, Angew. Chem., Int. C. Li, B. Wichterlova Ed., 2007, 46, 7286. 2 (a) D. E. W. Vaughan and K. G. Strohmaier, US Pat., 4,657,748, 1987; (b) M. E. Leonowicz and D. E. W. Vaughan, Nature, 1987, 329, 819; (c) C. S. Chen, J. L. Schlenker and S. E. Wentzek, Zeolites, 1996, 17, 393. 3 (a) S. J. Warrender, P. A. Wright, W. Zhou, P. Lightfoot, M. A. Camblor, C.-H. Shin, D. J. Kim and S. B. Hong, Chem. Mater., 2005, 17, 1272; (b) B. Han, C.-H. Shin, S. J. Warrender, P. Lightfoot, P. A. Wright, M. A. Camblor and S. B. Hong, Chem. Mater., 2006, 18, 3023. 4 A. F. Gualtieri, S. Ferrari, E. Galli, F. Di Renzo and W. Van Beek, Chem. Mater., 2006, 18, 76. 5 (a) R. von Ballmoos and W. M. Meier, Nature, 1981, 289, 782; (b) Z. Ristanovic´, J. P. Hofmann, U. Deka, T. U. Schulli, M. Rohnke, A. M. Beale and B. M. Weckhuysen, Angew. Chem., Int. Ed., 2013, 52, 13382. 6 (a) N. E. Brese and M. O’Keeffe, Acta Crystallogr., 1991, B47, 192; (b) I. D. Brown, The Chemical Bond in Inorganic Chemistry: The Bond Valence Model, Oxford University Press, New York, 2002; (c) I. D. Brown, Chem. Rev., 2009, 109, 6858. 7 J. Song, L. Dai, Y. Ji and F.-S. Xiao, Chem. Mater., 2006, 18, 2775. 8 R. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751.
This journal is © The Royal Society of Chemistry 2014
