8626
J . Phys. Chem. 1991, 95, 8626-8632
Electron Spin Resonance and Electron Spin Echo Modulation Studies of Cupric Ion Location and Adsorbate Interactions in the Cu*+-Exchanged H-SAPO-11 Molecular Sieve Chul Wee Lee, Xinhua Chen, and Larry Kevan* Department of Chemistry, University of Houston, Houston, Texas 77204-5641 (Received: February 26, 1991; In Final Form: May 20, 1991)
The locations and interactions of cupric ion exchanged into H-SAPO-] 1 molecular sieve with water, methanol, ethanol, ammonia, ethylene, and pyridine have been studied by electron spin resonance (ESR)and electron spin echo modulation spectroscopies. There is a gradual transition upon dehydration of CuH-SAPO-I 1 from a Cuz+ species with ESR parameters gll = 2.371 and A,, = 159 X lo4 cm-' to a Cuz+species with g, = 2.302, g, = 2.071, All = 185 X lo4 cm-l, and A, = 31 X IO4 c d . All of the samples with adsorbates show a major Cu2+species with three different g values, which means that the Cu2+ions lie in a distorted environment. Except for the ammonia-adsorbed sample, they show a weak reversed g value component in the ESR spectra with g, si 1.94, g, z 2.15, and All 106 X lo4 cm-l. Equilibration with ethylene was slow compared to the cases of other adsorbates. Adsorption of ammonia produces a complex containing four molecules of ammonia based on resolved superhyperfine. However, no resolved superhyperfine between pyridine and Cu2+is seen. Adsorption of polar molecules such as water, methanol, ethanol, ammonia, and pyridine leads to the migration of Cu2+ions to a position where adsorbates can coordinate with the Cuz+ions. In the hydrated sample, the Cu2+ions directly coordinate to two water molecules. In the methanol-adsorbed sample, the Cu2+ions directly coordinate to two methanols and indirectly coordinate to one methanol. In the ethanol-adsorbed sample, the Cuz+ ions directly coordinate to two ethanol molecules.
Introduction One remarkable discovery in the past decade has been the synthesis of AlPO, molecular sieves containing P5+ and A13+ in the framework.'J Further advances have been achieved by the introduction of Si4+and other elements into the AlP04 framework. The addition of silica to the AlP04 structure introduces both ion-exchange capacity and catalytic a c t i ~ i t y . ~ -These ~ silicoaluminophasphate (SAPO) molecular sieves crystallize with several zeolite structures as well as with novel structures not found in the zeolite (aluminosilicate) system. SAPO is defined as a silicoaluminophosphate with the general chemical formula mR.(Si,AI,,P,)02.bH20, where m represents the moles of R present per mole of (SixAl,,P,)O2, R represents an organic templating agent in the intracrystalline pore system, b represents the moles of water, and x + y + z = 1. These structures have a threedimensional microporous crystal framework arragement of S O 4 , PO4, and AlO, tetrahedra connected through the shared oxygen atoms. This arrangement results in an open structure containing channels and cages of molecular size. It is known that the character of the templating agent correlates with the shape and size of the channels and cages generated. For the SAPO-I 1 structure, x + z is usually greater than 0.5 so that the SAPO-I 1 framework is slightly negatively ~ h a r g e d ,which ~ , ~ is compensated by exchangeable cations which are usually protons in the usual synthesis. Hence, the designation is H-SAPO-11. X-ray diffraction analysis can be employed to determine the location of cations.' All of the SAPO molecular sieves have very low framework negative charge? So only a small amount of Cu2+ ions are exchanged. This is typically at the limits of X-ray analysis. However, electron spin resonance (ESR)and electron spin echo ( I ) Wilson, S. T.; Lok, B. M.; Messina, C. A,; Cannan, T. R.; Flanigen, E. M. J . Am. Chem. Soc. 1982,104, 1146. (2) Wilson, S.T.; Lok, B. M.; Flanigen, E. M. US. Patent 4 310 440, 1982. (3) Flanigen, E. M.; Lok, B. M.; Patton, R. L.; Wilson, S. T. Pure Appl. Chem. 1986, 58, 1351. (4) Szostak, R. Molecular Sieves: Principles of Synthesis and Identification; Van Nostrand Reinhold: New York, 1989; Chapter 4. ( 5 ) Lok, B . M.;Messina, C. A.; Patton, R. L.; Gajek, R. T.; Cannan, T. R.;Flanigen, E. M. US. Patent 4 400 871, 1984. (6) Lok, B. M.; Messina, C. A.; Patton, R. L.; Gajek, R. T.; Cannon, T. R.; Flanigen, E. M. J . Am. Chem. SOC.1986, 106,6092. (7) For example, see: Gallezot, P. Cafal.Rev. Sci. Eng. 1979, 20, 121. (8) For example, SAPO-I I has a negative charge of -0.02 to -0.04 per framework (Si,AI,P,)02; see ref 3.
0022-365419112095-8626$02.50/0
modulation (ESEM) techniques can be used to determine the location of paramagnetic cations even though present in very low concentration. Thus, ESR and ESEM were used to probe the location of Cu2+ in H-SAPO-11. Comprehensive studies have been made to characterize the locations and adsorbate interactions of paramagnetic transitionmetal ions in several kinds of zeolite^.^ ESR studies of the interactions of Cu2+with various adsorbates in SAPO-I 1 provide qualitative information about the stereochemistry of a paramagnetic species, and ESEM enables quantitative analysis of the number, distance, and orientations of ligands that interact with cuz+.
Experimental Section The molecular sieve H-SAPO-11 was synthesized according to the Union Carbide patentqs Since the patent did not give detailed procedures, we developed the following method, which gives good reproducibility. A solution was prepared by stirring 9.13 mL of 85 wt % H3P04 and 40 mL of water for 10 min. Then, 27.27 g of aluminum isopropoxide (finely ground in a mortar) was added gradually to this solution, which was stirred for 1 h. To this solution were added 3.4 mL of 30 wt % SiOz and 3.4 mL of H 2 0 successively drop by drop with mixing for 30 min. Then, 9.26 mL of i-Pr2NH and 4.0 mL of water were added drop by drop, and the solution aged a t room temperature for 24 h to form a gel. The gel was sealed in a stainless steel pressure vessel lined with Teflon and heated in an oven at 200 OC at autogeneous pressure for 24 h. This solid reaction product was collected by filteration, washed with distilled water, and dried in air a t 100 OC. This product is called yassynthesized" H-SAPO-11. To remove the organic templating agent (diisopropylamine), the as-synthesized product was heated in air at 550 OC for 48 h to form calcined H-SAPO-11 with the composition Ho,03(Sio.07A10,48P0,45)0z~50H20. The samples were exchanged with Cu2+ by adding 10 mL of lW3 M C U ( N O ~ ) ~ . ~solution H ~ O and 90 mL of distilled water to 1 g of calcined SAPO-1 1 and stirring overnight. The samples were then filtered, washed with hot distilled water in order to remove excess Cuz+ ions on the surface of the sample, and dried in air a t room temperature to form CuHSAPO- 1 1 with the composition C U ~ . ~ ~ H O . O ~ ~ (45)-S ~ O . ~ ~ A ~ O (9) Kevan, L. Rev. Chem. Inrermed. 1987, 8, 5 3 .
0 1991 American Chemical Society
The Journal of Physical Chemistry, Vol. 95, No. 22, 1991 8627
Cu2+-Exchanged H-SAPO- 1 1 Molecular Sieve
TABLE 11:
TABLE I: X-my Data of As-Syntbesized SAPO.l1° ref 5 this work
28 8 -05-8.3 9.4-9.65 13.1-13.4 15.6-1 5.85 16.2-16.4 18.95-19.2 20.3-20.6 21 W21.3 22.1-22.35 22.5-22.9*
IWI/IrJ)
23.15-23.35 24.5-24.9* 26.4-26.8 *
64-74 7-10 11-19
27.2-27.3 28.3-28.5** 28.6-28.85** 29.0-29.2 29.45-29.65 3 1.45-3 1.7 32.8-33.1 34.1-34.4 35.7-36.0 36.3-36.7 37538.0'
0-1 11-17
20-42 36-58 12-16 23-38 3-5 5-6 36-49 100 47-59 55-60
18.99 20.50 21.05 22.17 22.57 22.76 23.23 24.74 26.35 26.68
treatment hvdrated
1@ w / I O ) 48 59 20 35
evacuated evacuated evacuated evacuated
at at at at
RT' 153 OC 245 OC 350 "C
IO evacuated at 420 OC evacuated at 440 OC evacuated at 480 OC +D2O
42 100
56 57 53 69 13 17 18
+CH,OD +CD,OH +CzH,OD
28.65
17
29.14 29.53 31.55 32.96 34.23
7
2-3 0-2 0-2 3-6 3-5
45.9-46. I 46.8-47.1 48.7-49.0 50.5-50.8 54.6-54.8 55.4-55.1
0-2 0-1 2-3 3-4 2-3 0-2
11 15 10
5
36.65 37.53 37.88
IO
42.83 44.71 45.07
6 4 5
48.92 50.70 54.73
+pyridine
IO
13
4 4 5
*, doublet; **, shoulder.
O2.50H20. The degree of crystallinity of our sampled0 is reduced by about 18% and 33% when they are calcined and ion-exchanged, respectively. For the various ESR and ESEM experiments, the samples were treated as follows. (a) The ESR samples were loaded into 3mm-0.d. X 2-mm4.d. Suprasil quartz tubes and evacuated to a pressure of IO4 Torr for 12 h at various temperatures. (b) All samples dehydrated at temperatures higher than 150 "C were heated at the same temperature with 500 Torr of dry oxygen for about 6 h to reoxidize any reduced cupric ions. Then, the oxygen was pumped off for 2 h at room temperature. (c) After dehydration, adsorbates such as DzO, CD,OH, CH30D, CH3CH20D, I5NH3,ND,, CzH4,and pyridine from Stohler Isotope Chemicals and Aldrich Chemical Co. were adsorbed at ambient vapor pressure. The samples were frozen in liquid nitrogen and sealed. . ESR spectra were measured at 77 K on a Varian E-4 spectrometer. ESEM spectra were recorded at 4 K with a home-built spectrometer described elsewhere." Three-pulse echoes were recorded by using a 9O0-r-9Oo-T-9O0 pulse sequence, where the echo intensity is measured as a function of the time T between the second and third pulses. Both the theory and simulation of ESEM data are described in detail elsewhere.12 (IO) Since we do not have a standard sample for determining the degree of crystallinity, we assume that the as-synthesized sample has 100% crystallinity. The degree of crystallinity was determined relatively by using summed areas of five peaks corresponding to a 20 range of 20.0-25.0° in the X-ray diffraction pattern. (1 I ) Ichikawa, T.; Kevan, L.; Narayana, P. A. J . Phys. Chem. 1979,83, 3378. ( 1 2 ) Kevan. L. In Time Domain Electron Spin Resonance; Kevan, L., Schwartz, R. N., Eds.; Wiley-Interscience: New York, 1978; Chapter 8.
g,'
gia
+C2H4 0-3 5-7 7-9 11-14 7-9 0-3 3-4 10-13
39.3-39.55 40.3 42.2-42.4 42.8-43.1 44.8-45.2*
'Key:
28 8.13 9.48 13.23 15.71
ESR Parameters of Cu2+ in H-SAPO-11 at 77 K 2.371' 2.063 1.94od 2.342 2.334 2.327 2.3OSc 2.354 2.306 2.304 2.302 2.372' 2.065 1.939d 2.378' 2.065 1 .938d 2.386' 2.071 1.944d 2.374' 2.068 1 .942d 2.291c9f 2.377's 2.357'sh 2.298'~~ 2.352'*' 2.059 1.933d*' 2.287'-1 2.057
2.147 2.070 2.068 2.068 2.073 2.073 2.07 1 2.07 1 2.149
A,b ALb 2.025 159 106 170 171 175 29 185 31 163 185 32 184 31 185 31 2.024 157 106 2.023 140 111
2.145 2.027
142 107 2.151 2.021 152 108 2.152 2.072 182 150 2.067 180 2.051 157 2.012 152 2.142 100 2.022 200
28 28
are IO4 cm-I; estimated 'Estimated uncertainty is f0.004. uncertainty for A, is f5 X IO-' cm-' and for A , is fl X IO-' cm-l. Major component. Minor component. Room temperature. /Adsorption for 25 h. gAdsorption for 3 days. hAdsorption for 12 h. 'Adsorption for 37 h. 'Adsorption for 37 h at room temperature and 8 h at 110 OC.
Results X-ray Powder Mmction. The structures of the as-synthesized and calcined H-SAPO- 1 1 were characterized by X-ray powder diffraction. The X-ray diffraction patterns of these samples agree reasonably with the patent literature5to identify the material. For the as-synthesized H-SAPO-11, the peak positions agree well but the relative intensities vary somewhat from the values given in the patent (see Table I). For the calcined H-SAPO-11, both the peak positions and the relative intensities vary somewhat from those in the patent. Khouzami et al." have reported on the structural effects of water adsorption on SAPO-11; as-synthesized SAPO- 1 1 shows a space group change for rehydrated SAPO- 11 so the degree of water incorporated seems critical. We suggest that the small deviation of the X-ray diffraction pattern of calcined H-SAPO-11 from the patent values is related to differing water content. Thermogravimetric Analysis. The thermogravimetric analysis of as-synthesized H-SAPO-11 reveals three distinct weight losses at 20-200,200-400, and 400-600 "C.The first loss is assigned to water desorption. The second and third losses are assigned to the decomposition of the organic templating agent diisopropylamine. The thermogravimetric analysis of calcined H-SAPO-11 shows that there is only one stage of weight loss at 20-150 OC. This is assigned to the desorption of physically adsorbed water and confirms that the diisopropylamine template has been removed by calcination. Electron Spin Resonance. Table I1 shows gvalues and hyperfine splitting constants for samples after several different treatments. Figure 1 shows the changes in the ESR spectra at 77 K after dehydration of CuH-SAPO- 11 at various temperatures. There is a gradual transition upon dehydration from a species with ESR parameters gll = 2.371, g, = 2.063, 2.025, and A, = 159 X lo4 cm-I to anothec species with gll = 2.302, g, = 2.071, All = 185 X lo-" cm-I, and A, = 31 X lo-" cm-I. The second species becomes prominent after dehydration at 480 "C. In the freshly ~~
~~~
~
~~~
~
(13) Khouzami, R.;Coudurier, G.; Lefebvre, F.; Vedrin, J. C.; Mentzen. B. F. Zeolites 1990, IO, 183.
8628 The Journal of Physical Chemistry, Vol. 95, No. 22, 1991 2
IF
z
CUH-SAPO-I1
A
d
a
l
x4
.-
Hydrated
1 gll= 1.940
41' 2.147
Lee et al.
TABLE III: ESR Parameters for CuH-SAPO-11with Adsorbed Ammonia at 77 K
treatment
g(
+ND3 +ISNH3
2.225 2.226
AIb
g,'
192 189
2.033 2.033
A,(N)C 16 23
A'(N)d 14
Ohtimated uncertainty is *0.004. *Units are IO4 cm-I; estimated uncertainty is f5 X IO-' cm-I. CSuperhyperfineconstant due to nitrogen; units are lo4 cm-l; estimated uncertainty is *I X IO-' cm-l. Extraneous superhyperfine constant due to nitrogen (see text); units are IO-' cm-I; estimated uncertainty is & I X IO4.
-
CUH-SAPO-I IlPyridine a
-
x4
12h
J
Evac. 7 4 8 0 ° C
\ I
L_i_u gI1= 2.298
gL=Z 071
Figure 1. (b) after
-ESR -
-
spectra at 77 K of CuH-SAPO-I 1: (a) hydrated sample,
12-h evacuation at 153 OC, (c) after 12-h evacuation at 350 O C , and (d) after 12-h evacuation at 480 OC.
x4
-
8h
at Il0"C
gll=2.287
Figure 3.
ESR spectra at 77 K of CuH-SAPO-I 1 with pyridine: (a) adsorbed for 12 h, (b) adsorbed for 37 h, and (c) adsorbed for 37 and for 8 hat 110 O C .
7
w
u
u
L
3100
I
I
3200
I
I
3300
I
I
I
3400
H,G
ESR spectra at 77 K of CuH-SAPO-I 1 after adsorption of (a) I5NH3and (b) I4ND3. In (c) the perpendicular region of (b) is expanded.
Figure 2.
hydrated sample at room temperature, there is a minor species showing reversed g values (gll < g,), with gll = 1.940 and g, = 1.940 and g, = 2.147. Although this reversed g value species is very weak, the Cu hyperfine on the gll region is clearly resolved and similar to that observed in zeolites. This species disappears rapidly upon evacuation at room temperature. Adsorption of water resulted in the recovery of the ESR spectrum observed for the freshly hydrated sample, and the same magnetic parameters were obtained. This result indicates that the dehydration process is reversible. Table I1 shows g values and hyperfine splittings of CuHSAPO-] 1 with several adsorbates. In the samples fully adsorbed with D20, CH30D, CD30H, and CH3CH20D,commonly there are also minor species showing reversed g values similar to that in hydrated CuH-SAPO- 1 1. When I5NH3was adsorbed at room temperature on a dehydrated sample of CUH-SAPO-I1, a new Cu" spectrum developed with ESR parameters of gll = 2.226, All = 189 X IO-" cm-I, and g, = 2.033 (Figure 2a). An additional superhyperfine structure due to IsN consisting of five lines with a separation of 20.4 G (22.8 X lo4 cm-I) is seen in the g, region of Cu2+.
Adsorption of ND3 (Figure 2b.c) produced a species with ESR parameters of gll = 2.225, All = 192 X IO4 cm-', and g, = 2.033, which are almost same as for ISNH3. Also an additional superhyperfine structure due to I4N consisting of nine lines with a separation of 14.4 G (16 X IO4 cm-I) is observed in the g, region of Cu2+. These results are summarized in Table 111. In order to adsorb ethylene, a dehydrated sample was exposed to ethylene for 25 h at room temperature, but the ESR spectrum of this sample shows almost the same magnetic parameters of gH = 2.291, g, = 2.072, and A, = 182 X 1 P cm-'as that of a sample without adsorbed ethylene. After exposure to ethylene for at least 3 days, another species is produced. This new species has gl = 2.377, g, = 2.067, and A, = 150 X IP cm-I. For the other polar molecules adsorbed, it took 3-4 h to generate a new Cu2+species. Adsorption of pyridine for 12 h (Figure 3a) shows two different species. One has g,, = 2.357 and All = 180 X IO-" cm-I, and the other has gll = 2.298 and A, = 157 X lo4 cm-I. Prolonged adsorption of pyridine for 25 h produced different ESR patterns (Figure 3b) compared with the above. The major species has gu = 2.352, g, = 2.059,2.012, and All = 152 X lo4 cm-I, while the minor species has gll = 1.933, g, = 2.142, and A, = 100 X IO-" cm-'. In order to detect the ESR spectrum change for a pyridine-adsorbed sample, adsorption was continued for 8 h a t 110 OC after 25-h adsorption a t room temperature. This sample has ESR parameters of gu= 2.287, g, = 2.057, 2.022, and A, = 200 X lo4 cm-' (Figure 3c). In all of the pyridine-adsorbed samples, no indications of superhyperfine splitting due to nitrogen were seen. Electron Spin Echo Modulation. According to the three-pulse ESEM spectrum of CuH-SAPO-I 1 hydrated with D20, which has almost same ESR spectrum as a freshly hydrated sample, Cu2+ is directly coordinated to two water molecules, Cu2+-(H20)2, situated at a distance (Cu2+-D) of 0.28 nm with At, = 0.29 MHz. The three-pulse ESEM spectrum that was obtained from CuHSAPO-I 1 after C2HSOD adsorption was simulated with two deuteriums interacting with Cu2+ at a distance of 0.27 nm with Aim= 0.24 MHz. Figure 4 shows the three-pulse ESE modulation
The Journal of Physical Chemistry, Vol. 95, No. 22, 1991 8629
Cu2+-Exchanged H-SAPO- 1 1 Molecular Sieve CUH-SAPO-I I + CDSOH
I
a
0
n
-CALC
6
4
O
I
I
3
2
4
5
T, PS
Experimental and simulated (solid line) three-pulse ESEM spectrum recorded at 4 K of CuH-SAPO-11 with adsorbed CD30H. TABLE IV: Simulated ESEM Parameters for Cu2+in H-SAPO-11 adsorbate shell Nu d (nm) Aimc(MHz) D20 1 4 0.28 0.29 CDyOH 1 6 0.43 0.04 2 3 0.34 0.21 CZHc,OD 1 2 0.21 0.24
b
Number of deuterium atoms. *Distance between Cu2+ and deuterium; estimated uncertainty is iO.01 nm. CIsotropichyperfine coupling constant; estimated uncertainty is *IO%. (I
of CuH-SAPO-11 with adsorbed CD30H. In this case, Cu2+ interacts with three deuteriums, Le., one molecule of methanol, at 0.34 nm with an isotropic hyperfine coupling of 0.21 MHz and with six more deuteriums at 0.43 nm with an isotropic hyperfine coupling of 0.04 MHz. The ESEM results are summarized in Table 1V.
Discussion Silicon can be considered to incorporate into the AIP04 structure to form SAPO materials. However, the location of the substituted ion, Si4+,in the AIP04 structural framework is not clear." It is known that SAPO-I 1 has the AIP04-I 1 structure and that the structure of SAPO-I 1 is composed of AIO,, PO4, and Si04tetrahedra. Four vertices of each tetrahedron are linked to form 4-, 6-, and IO-rings. These rings are interconnected by 6-ring windows. The three-dimensional framework (Figure 5) is produced by linking the remaining vertices that project alternately up and down normal to these sheets. Columns of 4-, 6-, and 10-rings are formed parallel to the c-axis. The 10-rings have an elliptical cross section of 0.63 X 0.39 13m.I' In freshly hydrated CuH-SAPO-11, the ESR spectrum at 77 K shows two species: The dominant species has gll > g,, and the minor species has reversed g values (gll < g,). Several papers have discussed the Cu2+ symmetry consistent with reversed g values.18-22 In zeolite A such Cu2+ species have a trigonal-bipyramidal configuration, where the Cu2+ is located in the plane of the O3oxygens of the 6-ring and two adsorbates are coordinated above and below the O3 plane. After evacuation at room temperature, the ESR spectrum no longer shows the reversed g value (14) Jahn, E.; Muller, D.; Becker, K. Zeolites 1990, 101, 151. (15) Bennett, J. M.; Richardson, J. W.; Pluth, J. J.; Smith, J. V. Zeolites 1987, 7, 160. (16) Richardson, J. W.; Pluth, J. J.; Smith, J. V. Acra Crystallogr. 1988, E l l , 361. (17) Meiet, W.M.; Olson, D. H.Arlas ofZeolite Structure Types; Butterworths: London, 1987; p 15. (18) Anderson, M.W.; Kevan, L. J . Phys. Chem. 1986, 90, 6452. (19) Narayana, M.; Zhan, R. Y.;Kevan, L. J. Chem. Soc.,Faraday Tram. I 1986, 82, 213. (20) Narayana, M.; Zhan, R. Y.;Kevan, L. J. Chem. Soc.,Faraday Tram. I 1987. 91, 1850. (21) Narayana, M.; Kevan, L. J . Chem. Phys. 1983, 78, 3573. (22) Herman, R. G . Inorg. Chem. 1979, 18, 995.
Figure 5. (a) View along the elliptical ]O-ring channel axis of SAPO-11 where the dashed lines show edges of 6-ring windows that form the surfaces of a 10-ring channel. (b) Simplified structure of SAPO-II, showing possible cation positions. See text for description of the cation
positions.
component and indicates only one species. This is indicative of Cuz+ losing some water ligands. Near 350 OC another species is generated, which becomes the major species at 420 OC. The spectral parameters do not change as the dehydration temperature is increased from 420 to 480 "C, and four components in the g, region are clearly seen. Possible Locations of Cu2+ in SAPO-11. The locations of chargebalancing cations are important to understand the catalytic and adsorptive properties of molecular sieves. Possible cation locations are close to the walls of the channels adjacent to 6-rings because the 4-, 6-, and 10-rings are interconnected by 6-ring windows. Also there are possible Cu2+sites that are recessed into the 10-ring channels, because the 10-rings have an elliptical cross section. By analogy to cation sites in zeolite XZoand H-SAPO-5?3 we propose possible Cuz+ sites for H-SAPO-11 as follows (see Figure 5). Site U is at the center of the double 10-ring channel. Site I is at the center of the double 6-rings that form 6-ring channels. Site I' is in the plane of the O3oxygens of the 6-ring. Site I1 is at the center of the hexagonal window on the surface of the IO-ring channel. When we carefully examine the surface of the 10-ring elliptical channels, which are composed of five 6-ring windows, we realize that there are two different kinds of site 11: Le., 111 and I12. Sites 111and 112 have different environments. Site 111 is located at a more recessed position relative to site I12. Sites 11' and 11* correspond to displacement from site I1 into and out of the 10-ring channel, respectively. The line shapes and magnetic parameters of Cuz+in dehydrated CuH-SAPO-11 are different from those in CuH-SAPO-11 with adsorbates. The anisotropic line shape of dehydrated CuHSAPO-1 1 corresponds to Cu2+ions subject to an axial crystal field. (23) Chen, Xinhua; Kevan, L. J . Am. Chem. SOC.1991, 113, 2861.
8630 The Journal of Physical Chemistry, Vol. 95, No. 22, 1991
Lee et al.
The adsorption of adsorbates at room temperature at which Cu2+ does not migrate may induce the migration of Cu2+ ions to positions where adsorbates can coordinate with them. Adsorbate Effects. Although ethylene is a small molecule (kinetic diameter 0.39 nm), the rate of ethylene adsorption was very slow when compared with that of polar adsorbates. Similar effects were observed in Cu-ZSM-5 by Anderson and K e ~ a n . ~ ~ In dehydrated H-SAPO-I 1, Cu2+is assumed to locate in positions inaccessible to ethylene. Site I is a coordinatively saturated position. If the Cu2+ is located in site I and coordinated with six lattice oxygens, the interaction between Cu2+and ethylene will I be small. This can explain why the adsorption of ethylene is slow. %PPH Adsorption of pyridine produced a definite ESR spectral change Figure 6. First-derivative ESR spectrum at 77 K of hydrated CuHcompared with that of a dehydrated sample. Naccache et al.,25 SAPO-1 1. The expanded part shows a second-derivative spectrum inLeith et a1.,26and Sendoda et have reported a square-planar dicating the third g value component more clearly. complex of Cu2+-pyridine in zeolites. Superhyperfine splittings due to N are typically observed if Cu2+ is coordinated directly to the nitrogen atom in pyridine. There is no evidence of this ZSM-5,27Cu2+ is most probably located at site U (see Figure 5 ) interaction between Cu2+and pyridine in H-SAPO-11. This can coordinated to four molecules of ammonia. Tetracoordinated be explained as follows. It is known that kinetic diameter of cupric complexes such as CuX4 (X = CI, NH3, etc.) are known to prefer the more stable square-planar c o n f i g ~ r a t i o n . ~This ~-~~ pyridine is almost the same as that of the 10-ring channels. behavior was different for a non-oxygen-treated sample. When Although pyridine can enter the 10-ring channel, since Cu2+ is already coordinated with six lattice oxygens (site I in Figure 5 ) , CuH-SAPO-11, which was evacuated at 450 OC, was exposed to ammonia (ND,) for 2 days, another new complex, [Cu(ND3),12+, the distance between the nitrogen atom in the pyridine molecule is observed, indicating that the location of Cu2+ions before and and Cu2+is too far to resolve superhyperfine splittings. In order after oxygen treatment is different. This spectrum (g,l = 2.214, to demonstrate the coordination of pyridine more clearly, we All = 199 X lo4 cm-l, A, = 15.4 X IO4 cm-l) also shows carried out pyridine adsorption at 110 “ C for 8 h. Although no superhyperfine structure due to 14N. The perpendicular part of superhyperfine structure was clearly seen, a third g value component appears in Figure 3c. In dehydrated H-SAPO-11, Cu2+ the spectrum shows only seven lines, showing that the complex is surrounded by six oxygens and has axial symmetry with glland is [Cu(ND3),I2+. Cu2+ remaining after activation at 450 OC is g, values. After pyridine adsorption, Cu2+is suggested to move expected to be in more constrained positions within the lattice and toward the surface of the IO-rings to interact with polar pyridine. thus is less easily reduced. The lower adsorbate coordination number is consistent with this view. This gives rhombic symmetry for Cu2+ with three g values. A comparison of the spectra before and after adsorption of Adsorbate Coordination Geometry. On close inspection of the ammonia, I5NH3,and I4ND3(see Figure 2) indicates the coorESR spectra of all the adsorbed samples, we notice two common dination of ammonia to Cu2+. The presence of five hyperfine lines characteristics. One is that all the spectra contain a “reversed in the ESR spectrum of the Cu2+-I5NH3 complex (nuclear spin g value” component (g, > gll)as a minor species; the other is that quantum number of ISN is can be assigned to the coupling they show three different g values, indicating rhombic symmetry. of four equivalent nitrogen atoms. Also a nine-line superhyperfine This is clearly observed in the second-derivative ESR spectra. structure is seen in the ESR spectrum of Cu2+-I4ND3 (nuclear Figure 6 exhibits typical first- and second-derivative ESR spectra spin quantum number of 14N is I), additionally confirming that of hydrated CUH-SAPO-11. In evacuated samples there is no the complex is [CU(ND,)~]~+. Previously, Turkevich et aI.%and indication of a third g value even in the second-derivative ESR Anderson et al.24obtained similar spectra for a Cu2+-ammonia spectra. complex in zeolite Y and ZSM-5, respectively, and have analyzed The contribution in these ESR spectra of the reversed gvalue it in terms of a square-planar complex. In H-SAPO-11, the component, which has a trigonal-bipyramidal configuration, is [ C U ( N H ~ ) ~ ]complex *+ is suggested to have a square-planar so weak that it cannot be analyzed by ESEM. The Cu2+ is configuration because the square-planar ion should have easier suggested to be bound to either three framework oxygens and two access to the ion sites in the elliptical channelI7 than a tetrahedral molecules of water or to two framework oxygens and three water complex and hydrogen from ammonia can hydrogen bond to lattice molecules, producing a trigonal-bipyramidal configuration. oxygens. One g value for Cu2+ is expected if Cu2+is in a totally symVedrine et al.29have reported a temperature dependence of the metric or cubic environment. In axial and rhombic crystal fields, hyperfine coupling for copper complexes with ammonia and two and three different g values are expected, respectively. Silver35 pyridine produced in Na-Y zeolite. In CuH-SAPO-I 1, no such has measured three different g values from Cu2+in Tutton’s salts, temperature dependence was seen. Other evidence of this suin which the copper ion shows three different bond distances to perhyperfine structure appears in the spectral feature a t about oxygen atoms.36 They interpreted this in terms of a rhombically 3325 G (see Figure 2c). This “extraneous” or “overshoot” line distorted octahedron. For five- and four-coordinate Cu2+comis often observed in X-band ESR spectra of C U ~ + We . ~ can see p l e x e ~ there , ~ ~ are also three different g values when the Cu2+ a nine-line pattern at higher magnetic field, because the maximum complex has lower symmetry. As shown in Table IV, the best intensity can be observed at the fifth line centered at 3315 (AN, = 12.7 G (14 X IO4 cm-I)). Although we do not have direct evidence of the [ C U ( N H ~ ) ~location ] ~ + in SAPO-1 1, considering (31) Hanig, V. F.; Cakajdova, I. A. Acta. Crystollogr. 1958, 11, 610. that the cross section of a ]O-ring in SAPO-I 1 is 0.63 X 0.39 nm, (32) Distler, T.; Vaughan, P. A. Inorg. Chem. 1967.6, 126. the Cu-N distances in [ C U ( N H ~ ) ~and ] ~ +[Cu(NH3),J2+ are (33) Willett, R. D. J . Chem. Phys. 1964, 41, 2243. 0.203,’ and 0.21 1 respectively, and in [ C U ( N H ~ ) ~in] ~ + (34) Felsenfeld, G. Proc. R. Soc. London 1956, A236, 506. (35) Silver, B. L.;Getz, D. J . Chem. Phys. 1974, 61, 638. (36) Robinson, D. J.; Kennard, C. H. L. Crysrol Struct. Commun. 1972, I , 185. (24) Anderson, M.W.; Kevan, L. J . Phys. Chem. 1987, 91, 4174. (37) Kokoszka. G.F.; Allen, H. C., Jr.; Gordon, G.J . Chem. Phys. 1985, (25) Naccache, C.; Taarit, Y . 9. Chem. Phys. Lctr. 1971, 11, 11. 42, 3693. (26) Leith, 1. R.; Leach, H. F. Proc. R. Soc. London 1972, ,4330, 247. (38) kncini, A.; Bertini, I.; Gatteschi, D.; Scozzafava, A. Inorg. Chem. (27) Sendoda, Y . ;Ono, Y . Zeolites 1986, 6, 209. 1978.17, 3194. (28) Turkevich, J.; Ono, Y.;Soria, J. J . Cotal. 1972, 25, 44. (29) Vedrine, J. C.; Derouane, E. G.; Taarit, Y . Ben J . Phys. Chem. 1974, (39) Kokoszka, G. F.; Reimann, C. W.; Allen, H. C., Jr. J . Phys. Chem.
v
78, 53 1. (30)
Neiman, R.; Kivelson, D. J . Chem. Phys. 1961, 35, 156.
1967, 71, 121.
(40) Sharnoff,
M.J . Chem. Phys. 1965, 42, 3383.
Cu2+-Exchanged H-SAPO- 1 1 Molecular Sieve
a n
n
n
b
The Journal of Physical Chemistry, Vol. 95, No. 22, I991 8631
a
b
Ha
H
Ha Figure 7. Schematic representation for Cuz+ coordinated to water in
H-SAPO-11, showing two directly coordinated water molecules in one 10-ring channel and three lattice oxygen atoms from the six-ring window around Cu2+,giving a five-coordinate complex, where the position of Cu" is suggested to be HI* (see Figure 5): (a) relative positions of Cuz+ ions (0)and water (A);(b) detailed view of the geometry. From ESEM data, the Cuz+-H, distance is 0.28 nm. H
I Ha
TH ' H
Figure 8. Schematic representation for Cuz+coordinated to ethanol in H-SAPO-I 1, showing two ethanol molecules and three lattice oxygen atoms from the 6-ring window around Cu2+,giving a five-coordinate
complex where the position of Cuz+ is suggested to be Ill* (see Figure 5). From ESEM data, the Cu2+-H, distance is 0.27 nm. fit for the ESEM data is for interaction with two water molecules situated a t a Cu2+-D distance of 0.28 nm. So we suggest that Cu2+is located at site 111* in which Cu2+ is coordinated with three lattice oxygens and two oxygens from the water, resulting in a five-coordinate complex (see Figure 7). For C2H50Dadsorption, the best fit is for interaction with two ethanol molecules at a Cu2+-D distance of 0.27 nm (see Table IV). Figure 8 shows a schematic of two ethanol molecules and three lattice oxygens from the six-ring window around Cu2+giving a five-coordinate complex. So we can suggest that Cu2+is also located at site Ill*. When we compare the ESR parameters of the ethanol- or water-adsorbed CuH-SAPO- 11 with the methanol-adsorbed sample, we find some notable differences. The All value of the methanol-adsorbed sample (( 129-140) X 10" cm-I) is smaller than that of the ethanol- or water-adsorbed one ((152-157) X lo4 cm-I). and the gll value is somewhat larger for the methanol-adsorbed sample. These differences suggest that the methanol-Cu2+ complex has a geometry in CuH-SAPO-11 different from that of the ethanol-Cu2+ complex. As shown in Table IV, the ESEM spectrum of the CD,OH-adsorbed sample fits a two-shell model with three deuteriums located a t 0.34 nm and
Figure 9. Schematic representation of Cu2+coordinated to methanol in H-SAPO-11, showing direct coordination to two methanols and three lattice oxygen atoms from the 6-ring window in the same 10-ring channel and indirect coordination to one other methanol molecule in an adjacent IO-ring channel: (a) relative positions of Cu2+ions (0)and methanol (A);(b) detailed view of the coordination, where the position of Cu2+is suggested to be 111* (see Figure 5). From ESEM data, the Cu2+-H, distance is 0.43nm and the Cu-Hb distance is 0.34 nm.
six at 0.43 nm from Cu2+. These data indicate that two methanols are directly coordinated and one methanol is indirectly coordinated with Cu2+ as shown in Figure 9. The framework structure of SAPO- 11 is closely related to that of SAPO-5. The sheets of SAPO-5 are composed of 4-, 6-, and 12-rings. The 4- and 6-rings in SAPO-5 alternate around the 12-ring channel of SAPO-5. The most important structural difference between SAPO-5 and SAPO-1 1 is that whereas the 1O-ring in SAPO- 11 has an elliptical cross section, the 12-ring in SAPO-5 has a circular cross section. Due to this difference, there is a more recessed cation position in SAPO-1 1, site I1 in the 10-ring, for a cupric ion so that it is expected that the coordination number of Cu2+ with adsorbates may be less than in SAPO-5. Site I1 in SAPO-SZ3might be equivalent to site 112 in SAPO- 11. Indeed, it is found that Cuz+ coordinates to three water molecules in SAP0-Sa but only to two water molecules in S A W 1 l. Because site 111* in SAPO-1 1 is located at a more recessed position than site I1 in SAPO-5, there is room to only accommodate two coordinated water molecules. For methanol adsorbate the coordination number with Cuz+ is the same in both SAPO-5 and SAPO-1 1. But the distance between Cu2+ and D of C D 3 0 H is greater in SAPO-1 1. Due to the more bulky size of methanol compared to water, coordination to Cu2+shifts the Cu2+site more from the 6-ring window in SAPO-5 to the 10-ring in SAPO-1 1. This increases the coordination distance in SAPO- 11.
Conclusions The majority of Cu2+ions in fully dehydrated H-SAPO-11 are situated in site I, the center of the double &rings that form a &ring channel. This position is characterized by an axial crystal field. Nonpolar molecules such as ethylene are slow a t inducing the migration of Cu2+ions while polar molecules complex Cu2+much more rapidly. Cu2+is found to complex with four molecules of ammonia in a square-planar configuration with resolved hyperfine coupling to the nitrogens of ammonia. No superhyperfine structure due to nitrogen is seen in a pyridine-adsorbed sample. Adsorption of polar molecules such as water, methanol, and ethanol induces
J. Phys. Chem. 1991, 95, 8632-8634
8632
the migration of Cu2+ions to site 111*, close to a 6-ring that forms a window of the IO-ring channel where adsorbates can directly coordinate with Cu2+. With D20, Cu2+directly coordinates to two D20. With CD'OH, Cu2+directly coordinates to two CD30H and indirectly coordinates to one CD'OH. With C2H50D,Cu2+ directly coordinates to two C2H50D.
Acknowledgment. This research was supported by the National Science Foundation, Robert A. Welch Foundation, and Texas Advanced Research Program. Registry NO. CU", 15158-11-9; H20, 7732-18-5; CHjOH, 67-56-1; C2HsOH,64-17-5; NH3, 7664-41-7; C2H4,74-85-1; pyridine, 110-86-1.
Gas-Phase Molecular Structure and Conformation of 3-Iodo-I-propene As Determined from Electron Dlffractlon and Microwave Spectroscopic Data Kolbjam Hagen,**+Quang She&*and Reidar Stcalevikt Department of Chemistry, University of Trondheim, AVH, N-7055 Trondheim, Norway, and Department of Chemistry, Colgate University, 13 Oak Drive, Hamilton, New York 13346 (Received: March 19, 1991; I n Final Form: May 17, 1991)
3-Iodo-I-propenehas been investigated by using data from gas-phase electron diffraction and microwave spectroscopy. Only one conformer was observed at 298 K, a nonplanar gauche form with C - C - C - I torsion angle of 9 = 111 (2)O [I$ = Oo when C-I is eclipsing MI.The values obtained for the bond distances (rg) and valence angles (L,) are r(C-H) = 1.081 (12) A, r(C=C) = 1.348 (10)A, r(C-C) = 1.478 (13) A, r(C-I) = 2.186 (8) A, L C - C - C = 123.0 ( 1 1 ) O , LC-C-I = I 1 1.7 (12)O, LC=C-H = 120° (assumed), LH-C,--H = 109.5O (assumed), and fC2-C3--H = 110.6 (21)O. Error limits are given as 2a (a includes estimates of uncertainties in voltage/height measurements and correlation in the experimental data).
Introduction The structural and conformational properties of many halosubstituted propenes have been studied by using microwave spectroscopy' (MW) and electron diffraction2 (ED). Of these the 3-halo- 1-propenes (allyl halides, CH2=CH-CH2X) have been found to exist as a mixture of two conformers, syn and gauche. These studies have revealed that the population of the gauche form increased as the size of the halogen atom increased. A diagram of the gauche form of the molecule is shown in Figure 1. The microwave spectra of allyl iodide have been analyzed, and a nonplanar gauche form was found.Id Three rotational constants were reported, and a complete structure could therefore not be obtained. Vibrational spectroscopic investigations' of the vapor, liquid, and solid phases indicated the presence of two conformers with the second form present only in small amounts. An electron diffraction investigation2g using the visual method has also been reported. To obtain a complete structure and to investigate the conformational composition of allyl iodide, we decided to initiate a gas-phase electron diffraction study. By including the available rotational constants for allyl iodide into the data analysis, the accuracy of the structure determination could be improved. Experimental and Data Reduction Allyl iodide (98%) was obtained from Aldrich Chemical Co. and was used without further purification. Diffraction photographs were collected in the Balzers Eldigraph KDG-2 apparatus at the university of Oslo with a nozzle-tip temperature of 298 K, using Kodak Electron Image plates. The voltage/distance calibration was made with benzene as the reference. The nozzle-to-plate distances were 497.19 and 247.25 mm for the long and the short camera experiments, respectively, and the electron wavelength was 0.059 78 A. Three plates from the long and five plates from the short distance experiments were selected for analysis. Optical densities were measured with a microdensitometer and the data were reduced in the usual way." University of Trondheim. * Colgate University.
A calculated background' was subtracted from the data for each plate to yield experimental molecular intensity curves in the form SI,&). The average experimental intensity curves are shown in Figure 2. The ranges of the intensity data were 2.00 Is/A-' I14.50 and 6.00 ISI.&-' I25.00; the data interval was As = 0.25 A-'. Figure 3 shows the final experimental radial distribution (RD) curve calculated in the usual way from the modified molecular intensity curve I'(s) = sI,,,(s)ZcZI(Ac-'A{') exp(-0.00259), where A = s2Fand F is the absolute value of the complex electronscattering amplitude. The scattering amplitudes and phases were taken from tables.*
Structure Analysis Calculations of vibrational quantities were made using a valence force field. The initial force constants were transferred from a force field developed to fit the vibrational frequencies observed for allyl chloride and allyl bromide.*' Small changes were made (1) (a) Hirota, E. J. Chem. Phys. 1%5,42,2071. (b) Hirota, E. J . Mol. Specrrosc. 1970, 35, 9. (c) Niide, Y.; Takano, M.; Satoh, T.; Sasada, Y. J. Mol. S p r r m c . 1976,63, 108. (d) Sasada, Y.; Niide, Y.; Takano, M.; Satoh, T. J . Mol. Specrrosc. 1917, 66, 421. (2) (a) Trongmo, 0.;Shen, Q.; Hagen, K.; s i p , R. J . Mol. Srrucr. 1981. 71, 185. (b) Schei, H.; Shen, Q.J. Mol. Srrucr. 1982.81, 269. (c) Sovik, 0. I.; Schei, S. H.; Stolevik, R.; Hagen. K.; Shen, Q. J. Mol. Struct. 1984, 116, 239. (d) Schei, S.H.; Hagen, K. J . Mol. Srrucr. 1984, 116, 249. (e) Sovik, 0.1.;Trongmo, 0.;Hagen, K.; Schei, S. H.; Stolevik, R.; Shcn, Q.J. Mol. Struct. 1984, 118, 1. (f) Schei, S. H.; Shen, Q. J . Mol. Srrucr. 1985, 128, 161. (g) Bowen, H. J. M.; Gilchrist, A.; Sutton, L. E. Truns. Furaduy Soc. 1955.51, 1341. (3) (a) Radcliffe, K.; Wood,J. L. Truns. Furuduy Soc. 1966,62, 2038. (b) McLachlan. R. D.; Nyquist, R. A. Spectrochim. Acta 1967, 2 4 4 103. (4) Hagen, K.; Hedbrg, K. J . Am. Chem. Soc. 1973, 95, 1003. ( 5 ) Gundcrsen, G.;Hedberg, K. J . Chem. Phys. 1%9,51, 2500. (6) Andersen, B.; s i p , H. M.; Strand, T.; Stolevik, R. Acru Chem. Scud. 1%9, 23, 3224. (7) Hedberg, L. Absrruct ofPupers; 5th Austin Symposium on Gas-Phase Molecular Structure, Austin, TX, March 1974; p 37. (8) Schafer, L.; Yates, A. C.; Bonham, R. A. J . Chem. Phys. 1971.56, 3056.
0022-365419112095-8632$02.50/0 0 1991 American Chemical Society
