The selectivity of four small-pore silicoaluminophosphate molecular sieves, including , , , and SAPO-18 (AEI), toward light olefins in general and ethylene in particular has been investigated for the methanol-to-olefins reaction using gas chromatography. This study was prompted by earlier electron spin resonance and electron spin-echo modulation results on nickel-modified SAPO materials in which Ni(I) was incorporated into both framework and ion-exchanged sites. These seemingly similar materials behaved significantly differently with respect to reducing agents and adsorbates. Attention was focused on whether the catalyst performance is influenced by the structural type, the presence of a transition metal ion (Ni) either in the framework or at ion-exchanged positions, or the amount of incorporated transition metal ion. Our results show that these factors indeed play an important role in the catalytic behavior. Among the protonated H-SAPO-n materials, the highest combined distribution of ethylene, propylene, and butenes (C 2 -C 4 olefins) was obtained with H-SAPO-34, and the lowest with H-SAPO-35, which also had the shortest lifetime for catalytic activity. H-SAPO-18 turned out to be the best catalyst in terms of lifetime for catalytic activity. Incorporation of Ni(II) into the framework increased the lifetime, the overall distribution of C 2 -C 4 olefins (in the case of NiAPSO-34), and the selectivity of the catalysts toward ethylene (in the cases of , whereas incorporation of Ni(II) by means of solid-state ion exchange (NiH-SAPO-n) increased only the selectivity toward ethylene. Also, the increase in ethylene selectivity was more prominent in synthesized than in ion-exchanged samples. Among the Ni-loaded samples, NiAPSO-34 was found to be the best catalyst in terms of both ethylene selectivity and lifetime, whereas NiH-SAPO-18 exhibited the worst ethylene yield and NiH-SAPO-35 the shortest lifetime. Finally, the effect of the amount of Ni was investigated in NiAPSO-34. It appears that the selectivity toward ethylene does not increase linearly with NiAPSO samples prepared with an increasing amount of Ni in the reaction gel. In fact, there seems to be an optimum Ni concentration for which ethylene selectivity reaches a maximum and above which ethylene selectivity decreases. This optimum concentration is the same as was found in earlier studies to yield the strongest Ni(I) signal, as observed by ESR.

The formation of monovalent nickel in where nickel is believed to be incorporated into the framework of SAPO-34, and its interaction with several adsorbates are compared to Ni(I) species formed in NiH-SAPO-34, where Ni(II) is incorporated by solid-state ion exchange into known extraframework sites using electron spin resonance (ESR) and electron spin-echo modulation (ESEM) spectroscopies. Dehydration at temperatures above 573 K and hydrogen treatment at 573 K as well as γ-irradiation at 77 K produce one nickel species assigned by ESR as isolated Ni(I) in the two samples. Even though the ESR parameters of isolated Ni(I) species are similar after reduction, NiAPSO-34 and NiH-SAPO-34 show noticeable differences in their ESR characteristics after adsorption of various adsorbates, suggesting that Ni(I) in these two materials is in different sites. As a supplement to this, ESEM studies of 31 P and 27 Al, used to ascertain the location of the incorporated paramagnetic transition metal ion, also show significant differences in the modulation patterns. Simulation of the 31 P modulation observed for NiH-SAPO-34 shows two nearest-neighbor phosphorus atoms at 3.9 Å and three next-nearest-neighbor phosphorus atoms at 6.5 Å, indicating that Ni(I) is at site II′ in the chabazite cage near a six ring window after reduction. The 31 P simulation for NiAPSO-34 shows three nearestneighbor phosphorus atoms at a distance of 4 Å and two next-nearest-neighbor phosphorus atoms at 5.3 Å. This is consistent with Ni(I) ions substituting into a framework phosphorus site. In as-synthesized NiAPSO-34, the nickel ion is possibly also coordinated to additional waters to give distorted octahedral coordination. On dehydration, tetrahedrally coordinated nickel in a framework site of SAPO-34 is formed.

Electron spin resonance (ESR) and electron spin echo modulation (ESEM) spectroscopies were used to study the reducibility, location, and adsorbate interactions of Ni(I) in NiH-SAPO-34, in which Ni(II) was introduced into extraframework sites of SAPO-34 by partial ion exchange of H(I) by Ni(II). After dehydration at temperatures above 573 K as well as after γ-irradiation at 77 K, one nickel species assigned as isolated Ni(I) by ESR is observed. Along with the isolated Ni(I) species, a second species assigned to Ni(I)-(H 2 ) n is also observed after hydrogen reduction at 573 K. Adsorption of water into reduced NiH-SAPO-34 forms an axially symmetric Ni(I)-(O 2 ) n complex indicating water decomposition by Ni(I). Adsorption of methanol into reduced NiH-SAPO-34 forms two Ni(I)-methanol complexes suggested to be located at two different sites in the chabazite structure. Similarly, two Ni(I)-(C 2 D 4 ) n complexes, the predominant one exhibiting axial symmetry and the other one having rhombic symmetry, are formed after ethylene adsorption onto reduced NiH-SAPO-34. Analysis of the 13 C hyperfine structure obtained after 13 CO adsorption showed a Ni(I)-(CO) 3 species. The location of Ni(I) in NiH-SAPO-34 was determined qualitatively by 27 Al ESEM which suggested that even though Ni is situated within 5 Å of the framework aluminum, the two species are not in close proximity. A more quantitative analysis using 31 P ESEM showed that Ni(I) is at site II′ in the chabazite cage near a six-ring window after both thermal and hydrogen reduction.

Erionite-like silicoaluminophosphate molecular sieve SAPO-17 and levyne-like SAPO-35, in which Ni ions were incorporated via solid-state ion-exchange into known extraframework sites, have been studied by electron spin resonance (ESR) and electron spin echo modulation (ESEM). The Ni ion reducibility, location, and interaction with several adsorbates have been investigated. Among these adsorbates, the interaction with nitric oxide was emphasized and compared to that of Ni ion with NO in the previously studied chabazite-like SAPO-34. Room-temperature adsorption of C 2 D 4 on NiH-SAPO-17 after dehydration at 573 K, oxygen treatment at 823 K, evacuation, and subsequent hydrogen treatment at 573 K produces two Ni-ethylene complexes. Carbon monoxide adsorption gives rise to a Ni(I)-(CO) n complex with unresolved 13 C hyperfine lines. Following the kinetics of nitric oxide adsorption on NiH-SAPO-17 shows that initially, a Ni(I)-(NO) + complex, a NO radical, and a new species which appears to be another NO species are generated. After a reaction time of 24 h, NO 2 is observed. As the adsorption time further increases, NO 2 becomes stronger while Ni(I)-(NO) + decays, and after 5 days only NO 2 remains. NO adsorption on NiH-SAPO-35 shows different features. Initially, two Ni(I)-(NO) + complexes along with a NO radical are seen. As the adsorption time increases, one of the Ni(I)-(NO) + complexes decreases in intensity while the other one increases, and after a few days only one Ni(I)-(NO) + complex remains. Simulation of the 31 P ESEM spectrum, supplemented by 27 Al modulation, suggests that, upon dehydration, Ni ions in NiH-SAPO-17 migrate from the erionite supercage to the smaller cancrinite cage. In dehydrated NiH-SAPO-34 and NiH-SAPO-35, Ni ions remain in the large chabazite and levyne cages, respectively. As a consequence, Ni(II) in NiH-SAPO-17 is less sensitive to reduction by hydrogen than it is in NiH-SAPO-34 and NiH-SAPO-35.