Surprising Intrazeolitic Chemistry of Silver - The Journal of Physical

Feature Article pubs.acs.org/JPCC

Surprising Intrazeolitic Chemistry of Silver Nam Ho Heo,† Yang Kim,‡ Jong Jin Kim,† and Karl Seff*,§ †

Laboratory of Structural Chemistry, Department of Applied Chemistry, School of Applied Chemical Engineering, Kyungpook National University, Daegu 41566, Korea ‡ Department of Chemistry, College of Natural Science, Pusan National University, Busan 46241, Korea § Department of Chemistry, University of Hawaii, 2545 The Mall, Honolulu, Hawaii 96822, United States ABSTRACT: New small molecules and cations have been synthesized within silver-containing zeolites, and their structures were determined using single-crystal crystallography. Some of them are the hydronitrogens N3H5 and cyclo-N3H3, octahedral Ag6, and the silver halide clusters Ag4Cl4, Ag4Br4, and Ag4I4. The silver halide clusters within their zeolites are ordered three-dimensional arrangements of identical quantum dots. The (C3O3H3)22+ cation is a reduced dimer of C3O3H33+ (tripyrylium); tripyrylium is planar and isoelectronic with benzene. The two rings bond together facially via their π* orbitals in an eclipsed manner with a very short interplanar distance, 2.43 Å. When a hydrated silver-exchanged zeolite is heated (to dehydrate it), Ag+ ions react with framework oxygen atoms to give silver atoms and O2(g); this leads to crystal damage and the complete loss of crystalline diffraction. This damage, however, can be repaired and the diffraction fully restored by heating in O2(g). To avoid crystal damage altogether, silver-exchanged zeolites can be dehydrated at elevated temperatures in flowing O2(g).

1. INTRODUCTION This article aims to gather together in one place several remarkable aspects of the chemistry of silver-exchanged zeolites. Here are three reasons why this chemistry is remarkable: because silver is a noble metal, Ag+ is easily reduced, even (upon heating) by the oxide ions of the zeolite’s aluminosilicate framework, liberating O2(g). Also, perhaps because Ag+, like Pd0 and Pt0, has a d10 configuration, it promotes reactions within the zeolite that are reminiscent of the behavior of those metals. Finally, with its unique complexation properties, enhanced by the inadequate coordination that the zeolite provides to it, Ag+ can capture and stabilize new molecules, ions, clusters, and perhaps reaction intermediates within the zeolite. The void space of a zeolite is a medium, like space or a solvent, in which a chemistry can occur. Unlike those media, however, because a zeolite is a crystalline solid, the detailed structure of the product of a reaction or process (such as sorption) can be observed crystallographically to relatively high precision. Accordingly, crystallography, especially single-crystal crystallography, is a powerful method for the study of intrazeolitic chemistry. Unfortunately, few zeolites have been crystallized to date with crystals large enough for single-crystal work, so the number of zeolite framework types that can be studied crystallographically for this purpose is limited. Additional considerations govern the choice of zeolite to be studied. If it is the situation of the exchangeable cations within zeolites that is to be learned (their positions, their populations at each position, and the structures of their complexes), the results are clearer if these ions are present at higher concentrations, so zeolites with a high ion exchange capacity © XXXX American Chemical Society

are preferred. Large rings (windows) are preferred because they facilitate the movement of entering and leaving ions and molecules and allow larger molecules to be studied. Finally, zeolites with greater void volume are preferred because they offer more space (a larger vessel) for chemistry to occur. Thus, some zeolites are better than others for the study of intrazeolitic chemistry. Accordingly, only two zeolite framework types were used in the work to be presented, usually zeolite A (LTA, Na12Si12Al12O48·nH2O) but sometimes zeolite X (FAU, Na92Si100Al92O384·mH2O in this report). These are the unit cell formulas of the zeolites as they are synthesized hydrothermally from aqueous solution. (The actual unit cell of zeolite A, the one that encompasses the Si and Al ordering in its framework, is twice as long in each dimension. However, the exchangeable cations, often at positions of partial occupancy, irregularly distort the framework, preventing this lesser feature (the small differences in both size and X-ray scattering power of Al and Si) from being resolved. This report will discuss only the smaller (pseudo) unit cell of zeolite A and refer to it as the unit cell.) Because Ag+ has a particular affinity for zeolites, it exchanges easily and quantitatively into them from aqueous solution, usually replacing Na+ ions. For a variety of reasons including uptake of anions, incomplete exchange, and cation hydrolysis (which can lead to partial H+ exchange and, in turn, crystal Received: November 24, 2015 Revised: February 4, 2016

A

DOI: 10.1021/acs.jpcc.5b11490 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Feature Article

The Journal of Physical Chemistry C damage), this cannot be said of most cations. Ag+ has none of these problems. Shorter names, Ag12-A and Ag92-X or simply Ag-A and Ag-X, will be used for these fully silver-exchanged zeolites. The general fully silver-exchanged zeolite is Ag-Z. These names can be far from definitive. For example, the actual formula of each dehydrated sample depends profoundly upon the conditions of dehydration and subsequent treatment if any. Issues of disorder and partial occupancy plague the crystallography of zeolites. This is pervasive, often beginning with the zeolite frameworks themselves. For all of the results presented here for zeolite A, the identities of the Si and Al ions in the zeolite frameworks (the T (tetrahedral) atoms) have not been resolved, so only averaged positions have been found for them, and only averaged T−O distances and T−O−T angles can be reported. The situation is much better for zeolite X, but still not all atoms are resolved. The partial occupancies of the exchangeable cations introduce even greater disorder as they locally distort the zeolite framework. Resolving all of this is generally beyond the reach of the available data, so a measure of chemical sense must be interjected to handle them. Accordingly, the results presented here are necessarily somewhat less reliable than those achievable by conventional crystallography where all or nearly all positions are fully occupied and there is little or no disorder. Zeolites have vast economic importance.1 For example, they are the cracking catalysts in the petrochemical industry. They have many other applications as catalysts and are heavily used as selective sorbing agents and ion exchangers. Accordingly most of the research on zeolites has been application driven. Much more effort has been invested in finding better zeolite catalysts and in perfecting new applications than has been spent exploring the chemistry that can occur within their channels and cavities. This article presents some of the more surprising aspects of that chemistry, that which occurs in Ag+-exchanged zeolites.

H+ ions introduced into the zeolite had destroyed the zeolite, i.e., that H12-A, the zeolite product of reaction 1, is unstable. Perhaps, at the elevated temperatures used, the zeolite decomposed because water was being lost. The total reaction would then be Ag12Si12Al12O48 + 6H 2 → “Si12Al12O42 ” + 6H 2O + 12Ag (2)

Again, most chemists and crystallographers, seeing that their sample had decomposed, would have stopped studying it. These workers, however, continued. They heated the “decomposed” zeolite in O2(g) and found that its original crystalline powder diffraction pattern had fully returned. That was a startling result. Perhaps H12Si12Al12O48 + 3O2 + 12Ag → Ag12Si12Al12O48 + 6H 2O (3)

or “Si12Al12O42 ” + 3O2 + 12Ag → Ag12Si12Al12O48

had occurred. This suggests that the structure of the zeolite, its general topology, whether H12-A or “Si12Al12O42”, had not been lost at all, even though its crystalline diffraction pattern had. Zeolites are metastable with respect to denser phases where atoms are packed much more closely together. Accordingly, if a zeolite is destroyed, the crystalline diffraction pattern of one or more product phases should be seen. Zeolite A, for example, loses its diffraction pattern at about about 973 K, and the diffraction pattern of a crystalline product, β-cristobalite (a form of SiO2), can be seen at about 1073 K.2 Those are good indications that the zeolite was destroyed. That those temperatures are so much higher, about 400 K higher, than those that are needed to dehydrate Ag-A and were used for treatment with H2 and O2 (two previous paragraphs) suggests again that the zeolite had not at all been destroyed, as does the absence of a diffraction pattern of a decomposition product. Yet, because no diffraction peaks could be observed, there must have been some substantial damage. Perhaps the peaks had become very broadened, in a manner akin to particle size broadening,3 and thus unobservable above background. Treatment with O2(g), then, by oxidizing Ag0 to Ag+ (reactions 3 and 4), had produced oxide ions whose absence must have been responsible for this damage. As they rejoined the zeolite framework, it was fully restored. The direct observation of intermediate stages in the loss of diffraction pattern supports this conclusion.4 Normally, by our methods at that time, ca. 1978, an undamaged zeolite A crystal would have yielded several hundred sharp unique peaks. However, only eight very broad low-angle reflections could be seen above background for a single crystal of Ag-A that had been heated under vacuum at 748 K for 7 days. Also indicative of damage and collapse was its cubic unit cell edge length. It was 11.42(2) Å, astoundingly less than the values of ca. 12.300(1) Å seen for crystals treated similarly but at temperatures below 698 K. Between 698 and 748 K, much smaller monotonic decreases in the unit cell edge length were seen; they ranged from 12.295(1) to 12.148(2) Å, respectively.4 We repeated the Hungarian work using a single crystal of zeolite A. After Ag+ exchange at 294 K, dehydration under vacuum at 673 K, and treatment with H2(g) at 603 K, the crystal had lost its diffraction pattern entirely. Diffraction lines from silver metal powder were seen, and the crystal had become dark brown to black with finely divided silver metal on

2. LOSS OF CRYSTALLINITY AND ITS FULL RESTORATION Crystallographers often encounter crystal damage. Some crystals, for example, decompose in the X-ray beam as they are being studied. This damage causes their diffraction patterns to lose their resolution: the diffraction peaks broaden and become weaker, and the number of them that can be detected above background decreases. When they can no longer be seen, crystallographers easily say that the crystal has been destroyed. It is time to study another crystal. With that in mind, the findings of researchers at the Central Research Institute of Chemistry, Hungarian Academy of Sciences, communicated informally to me by Hermann K. Beyer, must be viewed as remarkable. They found that, when H2(g) was allowed to come in contact with a dehydrated powder of Ag-A, Ag+ ions were reduced. That does not seem very remarkable; it is in accord with the aqueous reduction potentials with which we are familiar. Even though a zeolite is a very different medium from water, there should be some similarities; e.g., Ag+ should be relatively easy to reduce. The reaction was envisioned to be, per unit cell Ag12Si12Al12O48 + 6H 2 → H12Si12Al12O48 + 12Ag

(4)

(1)

They also observed, however, that the product no longer had a crystalline diffraction pattern. One might conclude that the B


Feature Article

The Journal of Physical Chemistry C its surface. Then, after exposure to O2(g) at 603 K, we saw that its diffraction pattern was fully restored, indicating that the integrity of the entire single crystal had been restored!5 The color did not change, however, and only 11 Ag+ ions per unit cell were found within the zeolite. It was clear that 11 of the original 12 Ag+ ions, and therefore many oxide ions, by stoichiometry 5.5 of the six, had returned to the body of the zeolite. A likely reason for the return of only 11 Ag+ ions per unit cell is that the zeolite A framework offers inadequate coordination to a 12th ion. Eight Ag+ ions can find reasonable coordination (3-coordinate trigonal in 6-rings), three more Ag+ ions possibly acceptable coordination (3-coordinate but toward one side of 8-rings), but only inadequate coordination (opposite 4-rings with all four oxygen atoms very much to one side) to a 12th Ag+ ion. This spoke even more loudly that the zeolite had not been destroyed. Far less positive results could have been seen: the original crystal could have returned as a crystalline powder or could have been fractured into a number of pieces. The singlecrystal diffraction pattern could have been diminished in some way, or some amount of a new phase, the result of partial decomposition, might have been seen. That the framework of the entire macroscopic crystal had been fully reconstituted meant that it had never been destroyed at all. It had only been injured enough to have lost its single-crystal diffraction pattern. Let us consider more carefully how heating hydrated Ag-A (Ag12Si12Al12O48) gave rise to Ag0. If Ag+ was reduced, what was oxidized? Well, neither Al3+ nor Si4+ can reasonably be expected to lose electrons. Only oxide ions could provide these electrons, either the oxide ions of water molecules remaining in the zeolite during the dehydration procedure or oxide ions of the zeolite framework. Texter et al. were able to observe directly that it was framework oxide ions that were being lost.6 If instead water oxygens had been lost, the zeolite product would have been H-A, and there is evidence that H-A would have immediately lost water at the elevated temperatures used,7 to give the same neutral product, Si12Al12O42. This is the same conclusion reached above in paragraph 5 of this section. The result becomes clearer if we rewrite the unit cell formula of Ag12Si12Al12O48 as (Ag2O)6(SiO2)12(Al2O3)6. The Ag2O component of the composition is decomposing to the elements. Jacobs et al. observed the evolution of O2(g) directly upon heating Ag-A. It began with the evolution of small amounts O2(g) at a surprisingly low temperature, about 400 K.8 This is about 75 K below the temperature at which Ag2O(s) decomposes to the elements, 473 K.9 This low volume of O2(g) being emitted at such a low temperature might be attributable to the early reaction of the least satisfactorily coordinated Ag+ ion (the 12th ion, see paragraph 7 of this section) with the zeolite framework. Note that no H2(g) was involved in this work. It appears that the use of H2(g) in the earlier experiments was unnecessary; Ag+ could be reduced to Ag0 simply by heating, as a direct consequence of the attempted dehydration of Ag-A. It is surprising that oxide ions bridging between Al3+ and Si4+ ions (one might think that these are among the most stable oxide ions in chemistry) were being oxidized to O2(g), to leave oxide vacancies in the structure of the zeolite. Facilitating this result, presumably, is the low coordination number of these oxide ions, only two, and the high mobility of the oxide ions in zeolite frameworks.10 The central area of Figure 1 shows the kind of damage that a zeolite framework suffers when it loses some of its oxide ions

Figure 1. Large central area of this square net illustrates the kind of damage that a zeolite framework would suffer if it lost some of its oxide ions. Each intersection (representing a T (Si or Al) atom) of the original (undamaged) two-dimensional net had four bonds (to four oxide ions), like the T atoms of a zeolite framework. Similarly, each oxide ion (near the center of each line segment) bridged between two T atoms. The removal of a line corresponds to the loss of an oxide ion. The oxide ions selected for removal were generated randomly by a numerical procedure, and their density corresponds to the loss of 6 of each 48 oxide ions, as occurs when the Ag2O component of Ag-A, (Ag2O)6(SiO2)12(Al2O3)6, decomposes to the elements. For a onedimensional sequence, any one break (loss of an oxide ion) would destroy the string (the structure). Accordingly, it may be expected that a three-dimensional structure would be more robust than this twodimensional diagram appears to be. The magnitude of the distortions shown is arbitrary.

(due to oxidation to O2(g)). The nature of the relaxations that are responsible for the diminution of the zeolite’s crystallinity and crystalline diffraction pattern is apparent.

3. HEXASILVER 3.1. Reduction of Ag+ by Oxide Ions of the Zeolite Framework. To do an experiment in zeolite chemistry, the first step is usually to replace the exchangeable cations, usually Na+, in the zeolite as it was synthesized with other cations, whichever ones one wants to study. That is generally done from aqueous solution. The second step (unless the objective is just to study the hydrated zeolite) is to dehydrate the zeolite to get the water out of the way. It is generally the dehydrated zeolite that has utility in catalysis and sorption. Also, other molecules of interest can be sorbed into an empty zeolite, and those sorption structures can be learned. If those other molecules react, the reaction products can be seen. More complex steps of ion exchange, dehydration, and the sorption of more substances, all for various times and at various temperatures, may be used. Here is a very simple example of that: two single crystals of zeolite A were exchanged with Ag+ and dehydrated under vacuum at 603 K. The resulting crystal structures showed that some silver atoms approached oxygen atoms of the anionic zeolite framework as cations should and were therefore Ag+ as anticipated. Others, however, did not and were therefore not Ag+.11 The zeolite had been fully dehydrated as intended (no water molecules could be seen in the structure), but something else had happened. The word “dehydration”, then, when used in a discussion of Ag-Z, accurately describes the experimental C


Feature Article

The Journal of Physical Chemistry C

Figure 2. Hexasilver molecule, Ag60, in a sodalite cavity.

Figure 3. Ag60 coordinated by eight Ag+ ions in a sodalite cavity. The greater formula is Ag148+.

eight Ag+ ions, one in each face of the octahedron. Accordingly, the larger cluster is Ag+8Ag60, or Ag148+ (Figure 3). The loss of diffraction pattern at the higher temperatures had apparently prevented us from observing Ag60 (Ag148+) at full occupancy. Ethylene vapor was sorbed onto this zeolite (containing Ag+8Ag60) at 295 K.12 The resulting structure showed that two Ag+ ions had left Ag+8Ag60 to form two 1:1 lateral Ag+−C2H4 complexes (Figure 4) leaving behind Ag+6Ag60 or Ag126+

process (heating under vacuum, for example) but falls far short of describing the product. To almost three significant figures, these other “silvers”, located at a special position (0,0,z) in sodalite cavities, had nearly the same interatomic distances as those in silver metal, indicating that clusters of silver atoms had been produced.11 The greater the time or temperature of “dehydration”, the more silver atoms appeared until a maximum was reached at 3.6 silver atoms per sodalite cavity.4 (The capacity at this special position is 6.) As the temperature was further increased, that value began to decrease. Unfortunately, the size of the diffraction data sets also decreased continuously with increasing temperature (due to the production of silver atoms and therefore of oxygen gas (paragraph 9 of section 2)) until no diffraction peaks could be seen.4 The crystals had become too damaged for crystallography to continue. This was the same result seen before when Ag-A was “dehydrated” and then treated with H2(g), but this time the zeolite had only been “dehydrated”. If this value, 3.6, had been 6.00, an octahedron of silver atoms, perfect by symmetry, would exist in each sodalite cavity, filling them (Figure 2). At 3.6, an octahedron of silver atoms could exist in 60% of those cavities, but many other arrangements, averaging to 3.6 per sodalite cavity, were possible. For example, 60% of the sodalite cavities could contain Ag40 and the remainder Ag30. More generally, a broader mix of complete and incomplete octahedra could be present. We judged that octahedra with one or more atoms missing were not acceptable at positions of 4mm (C4v) symmetry. Furthermore, the bonding in these partial clusters would not be sensible. In addition, it would be expected that the remaining atoms would have relaxed away from their special positions, but the rather normal thermal parameter at this silver position, 0.023 Å2, did not indicate that such relaxation had occurred. Finally, we noted that silver metal itself (FCC) contains octahedral Ag60 subunits. Thus, we concluded, tentatively, that Ag60 had formed in varying amounts in up to 60% in the sodalite cavities of the crystals studied. Because the natural growth form of silver crystals is octahedral, it follows that hexasilver is the smallest possible fully developed single crystal of silver.4,11 Surrounding each Ag60 molecule was a cube of

Figure 4. Ethylene molecule coordinating to a Ag+ ion in a 6-ring of the large cavity of zeolite A. Two of its four hydrogen atoms hydrogen bond to oxygen atoms of the zeolite framework.

(Figure 5). Nicely, with fewer Ag+ ions to draw bonding electron density from Ag60, its bond lengths decreased, from slightly more (2.928(4) Å) than that in bulk silver metal (2.889 Å) to slightly less (2.850(4) Å). With more bonding electron density per silver atom and ion in Ag+6Ag60 than in Ag+8Ag60, the Ag0−Ag+ distances also decreased from 3.33(1) to 3.26(2) Å. The simplicity of these results lent some further support to the conclusion that Ag60 is present in Ag-A dehydrated under vacuum at an elevated temperature, e.g., 600 K. Further support for the formation of Ag60 in the sodalite cavities of zeolite A came quickly from esr and hyperfine NMR experiments13,14 and most recently by direct imaging.15 Finally, to learn something more of the chemistry of Ag60, two crystals of Ag12-A were dehydrated under vacuum at 673 K for 4 days. Each had a hexasilver molecule in about 3/8 of its sodalite cavities. When one was treated with Cl2(g) at 297 K, each hexasilver molecule was fully oxidized to six Ag+ ions.16 However, when the second was treated with Br2(g) at the same D


Feature Article


Figure 5. Ag126+, the result of the removal of two Ag+ ions from Ag148+.

temperature, a redox reaction did not occur. Instead, Ag+ ions left the Ag148+ group, partly to complex to Br2. The atoms of the hexasilver molecules, lacking coordinative support, migrated out of the zeolite to its surface.17 Cl−Cl···O and Br−Br···O charge transfer complexes were seen in the two structures, respectively; such complexes were first seen in zeolites when I2 was sorbed into Ca4Na4-A.18 3.2. Reduction of Ag+ by Cs(g) and Rb(g). Can silver clusters be generated without damaging the zeolite structure? That is, is there a reducing agent other than the oxide ions of the zeolite framework that can reduce Ag+? Cs(g) was tried and later Rb(g). Two crystals of Ag5.6K6.4-A were dehydrated at 633 K. That “dehydration” caused some crystal damage, and their structures showed a low population of Ag148+ in their sodalite cavities; however, about half of the diffraction data remained, so the work could continue. They were then treated with Cs(g) at 523 K. The resulting structures showed that all of the exchangeable cations in the structure, all Ag+ and all K+ ions, had been reduced by Cs0 and that all K0 and some Ag0 atoms had exited the zeolite structure.19 The product zeolites contained enough Cs+ ions to balance the charge of the zeolite framework. They also contained the same Ag60 clusters that had been found before at the centers of sodalite cavities (at (0,0,0), the atomic positions were (0,0,z)), but this time they centered the large cavities (at (1/2, 1/2, 1/2)) instead. The atomic positions are (1/2, 1/2, z), also at positions of 4mm (C4v) symmetry and therefore also of order six (Figure 6). More Ag0 atoms than

much larger than Ag+8Ag60, and these Cs+ ions, because of their size, fit even more poorly than Ag+ ions in the 6-rings of the zeolite framework and therefore extend farther from their planes, so CsnAg6n+ fits the large cavity nicely (Figure 6). Also seen were Cs64+ and Cs43+, the result of the sorption of additional cesium atoms. Altogether, the approximate unit cell composition is Cs13.6Ag4.4-A.19 Lower populations of the corresponding Ag60 cluster (2.4 instead of 4.4) were seen in the large cavity when Ag5.6K6.4 was exposed to Rb(g).20 The greater cluster is RbmAg6m+, where m is again at least 11. As with Cs+, eight Rb+ ions surround each Ag60 cluster, and up to six more can approach Ag60 as the Cs+ ions did. The Ag0−Ag0 bond lengths in CsnAg6n+ in the two crystals studied (they had been prepared somewhat differently) are 2.65(3) and 2.75(4) Å.19 This distance in RbmAg6m+ is 2.68(3) Å.20 These are all much shorter than those in Ag148+, 2.928(4) Å,4 and Ag126+, 2.850(4) Å,12 and therefore also much shorter than those in silver metal, 2.889 Å. Clearly, the Cs+ and Rb+ ions are far less effective than Ag+ at delocalizing bonding density away from the hexasilver cluster.19,20 Thus, it appears that the bonding density in Ag148+, and therefore its charge, is substantially delocalized over that entire cluster. This provides some explanation for the noticeably short 2.775(11) Å distance between each silver atom of Ag60 and four oxygen atoms each of the zeolite framework in “dehydrated” Ag-A.4 This was also seen in Ag126+.12 We may therefore conclude that hexasilver is relatively stable on its own and does not require substantial electron delocalization into a surrounding shell of cations to exist. When the dehydrated zeolites Ag4Ca4-A, Ag6Ca3-A, and Ag8Ca2-A were exposed to Rb(g) at 523 K, their resulting large cavities were 37%, 40%, and 80% filled with Ag60 clusters21 (Figure 7). Each was held in place by coordination to Rb+ ions as before. All Ca2+ ions had been reduced and had exited the zeolite. A small number of Ag+ ions remained, and some

Figure 6. Hexasilver held in place by coordination to 14 Cs+ ions in the large cavity of zeolite A.

before, 4.4 per unit cell, were seen; the maximum number seen in sodalite cavities had been 3.6.4 Hexasilver clusters, then, centered 4.4/6 → 73% of the large cavities. As before, each was surrounded by a cube of eight cations, Cs+ this time instead of Ag+, and some additional Cs+ ions to give CsnAg6n+, where n is at least 11 and perhaps as much as 14. Eight Cs+ ions surround each Ag60 cluster, and up to six more Cs+ ions approach Ag60 through the faces of the cube of the first eight. CsnAg6n+ is

Figure 7. Hexasilver held in place by coordination to 14 Rb+ ions in the large cavity of zeolite A. E


Feature Article


Figure 8. 92 Ag+ ions in fully dehydrated fully oxidized zeolite Ag92-X occupy eight equipoints.

additional Rb0 atoms were retained to form complexes of the form Rbnm+, m < n. This time the Ag0−Ag0 bond lengths in RbmAg6m+ were 2.74(4), 2.77(5), and 2.92(3) Å, respectively. The last distance is longer than the others and those in CsmAg6m+ (previous paragraph), indicating, perhaps, that Ag+ ions, present at a low occupancy equipoint, are bonding to hexasilver in that structure.

paragraph, 17 instead of 27, were seen. The crystal, however, had become yellow (no longer dark gray), indicating that the Ag(s) on its surface had been oxidized to Ag+ which had returned to the body of the zeolite crystal.27 With the addition of these Ag+ ions to the zeolite, anions are needed for charge balance. This had been seen before when only Ag0 and O2 were present (reaction 4), and oxide ions accompanied Ag+ into the zeolite to occupy lattice vacancies. This time only Ag0 and C2H4 are present. Thus, it appears that the oxide vacancies have been filled by the carbanions H2C2− (isoelectronic with O2−) or possibly H2CC2−. No other atoms were seen in the crystal structure. One or both of these reactions must have occurred.

4. DEHYDRATING AG-Z WITHOUT DECOMPOSITION When Ag-X was dehydrated in flowing O2(g) at 673 K, the resulting diffraction pattern was strong, indicating that no crystal damage had occurred. The crystal was colorless, indicating that no Ag(s) had deposited on its surface. Thereafter, all 92 Ag+ ions were readily located crystallographically; they all approached oxide ions of the zeolite framework as cations normally do.22 Nothing more than simple dehydration had occurred, and the structure, although it was complex (Ag+ ions distributed among eight equipoints, Figure 8), was simply Ag+92-X. In contrast, the same zeolite dehydrated under vacuum at the same temperature was dark gray, indicating that Ag(s) was present on its surface. Other Ag+ ions that had been reduced to Ag0 by the zeolite framework had remained in the structure, bonding to Ag+ to form cationic clusters of Agmn+, n < m ≤ 8.23 Weak Ag+−Ag+ bonding with distances generally between 3.0 and 3.3 Å was seen in both hydrated21 and fully dehydrated Ag+92-X.22 This intercationic bonding is common in compounds of Ag+.24 An example is Ag2CO3 whose structure25 contains layers of carbonate ions separated by wavy sheets of silver ions. Previous workers, who had learned of the partial reduction of Ag+ and the production of O2(g) upon heating Ag-Z,8 had been careful, as a final step in the preparation of dehydrated Ag-Z, to heat their “dehydrated” silver zeolites in oxygen or dry air to fully oxidize their Ag0 atoms and reconstitute their frameworks. The ethylene complex of fully oxidized (nondecomposed) Ag-X was studied.26 The crystal was simply exposed to 4 × 104 Pa of C2H4(g) at 294 K. Again all 92 Ag+ ions per unit cell were found, verifying that no reduction of Ag+ had occurred. The crystal color, by being dark yellow and not dark gray, supported this observation. Of the 92 Ag+ ions, 27 had formed 1:1 lateral π complexes with ethylene in the large cavity (Figure 4). Weak Ag+−Ag+ bonding was again seen.

4Ag + C2H4 → 4Ag + + 2H 2C2 −

(5)

2Ag + C2H4 → 2Ag + + H 2 + H 2CC2 −

(6)

The stoichiometry of both of these reactions is the same as that of reaction 4. For both reactions, all framework vacancies would have been filled with carbanions.

6. HYDRONITROGENS N3H5 AND CYCLO-N3H3 With the simple objective of seeing the Ag+−NH3 complexes that should form when NH3(g) is sorbed onto “dehydrated” Ag-A and how this would impact the Ag+8Ag60 cluster4 (as C2H4(g) had12 or differently), Ag12-A dehydrated under vacuum at 623 K was treated with NH3(g) at 294 K. Surprisingly, neither Ag60 nor NH3 was present in the product. Instead, high concentrations of two new hydronitrogen compounds, N3H5 (triazane, Figure 9) and N3H3 (cyclotriazane or triaziridine, Figure 10), were found.28 Both were stabilized by coordination to Ag+ ions. Each sodalite cavity contained a Ag+−(NH2−NH−NH2)3−Ag+ complex. In each large cavity, four cyclo-triazane molecules coordinated laterally to Ag+ ions. Peaks corresponding to both were seen by mass

5. INTRODUCTION OF CARBANIONS INTO A ZEOLITE FRAMEWORK A single crystal of Ag92-X was dehydrated under vacuum at 673 K, thus generating silver atoms and framework oxide vacancies. Its characteristic dark gray color indicated that some of this silver had deposited on its surface. Upon exposure to ethylene gas, fewer Ag+−C2H4 complexes than in the previous

Figure 9. Three triazane molecules, N3H5, bridge between two Ag+ ions in the sodalite cavity of zeolite A. F


Feature Article


7. TRIPYRYLIUM ION, C3O3H33+, AND ITS REDUCED DIMER, (C3H3O3)22+ The tripyrylium ion is a planar six-membered conjugated ring, isoelectronic with benzene (Figure 11). Formally, the C−H groups (seven electrons) at the 1,3,5 positions of benzene have been replaced by O+ (seven electrons).

Figure 10. cyclo-Triazane (triaziridine) molecule, N3H3, coordinating to a Ag+ ion at a 6-ring site in the large cavity of zeolite A.

spectrometry as a sample was slowly warmed from ambient temperature.28 Two reactions can be envisioned for this synthesis. Per unit cell they are 21NH3 → 3N3H5 + 4N3H3 + 18H 2

Figure 11. Tripyrylium ion, C3H3O33+, is isoelectronic with benzene. Like benzene it is planar, but its axis is only 3-fold. The ring angles are O−C−O = 90(6)o and C−O−C = 150(5)o. Although the positive charge formally resides on the oxygen atoms, calculations show that it is rather evenly distributed among the carbon and hydrogen atoms.33 Accordingly, the distortion to the ring may be attributed to repulsion among the carbon atoms.

(7)

and 36Ag + + 21NH3 → 3N3H5 + 4N3H3 + 36H+ + 36Ag 0

The corresponding pyrylium ion, C5H5O+, does not exist. The pyrylium ring has only been seen as a component of larger molecules where it conjugates with adjacent unsaturation. Diand tripyrylium rings have never been reported before in any context. It was not the tripyrylium ring, but its reduced dimer, (C3O3H3)22+, that was found in Ag-A (Figure 12).33 The two

(8)

Reaction 8 is simply an oxidation of ammonia by Ag+. It cannot be the only reaction occurring because there are never more than 12 Ag+ ions per unit cell, and 8 are seen in the final structure. The crystal was black, consistent with the color of finely divided Ag(s) on its surface. At least some of that silver came from the Ag60 that formed during the dehydration procedure; it must have exited the zeolite as NH3 stripped Ag+ ions from Ag148+. Some may have also formed by reaction 8. Therefore, triazane and cyclo-triazane formed primarily, and perhaps entirely, by reaction 7, an oxidative condensation of ammonia. Previous ab initio quantum mechanical calculations had shown that triazane has sufficient stability to be “a promising candidate for synthesis”.29 A subsequent more precise calculation affirmed that finding.30 Triazane was subsequently prepared for the second time (again not in bulk, again without intent) by irradiating a thin layer (500 nm) of NH3(s) deposited on Ag(s) with energetic electrons at 5.5 K. It was detected in situ upon warming by reflection time-of-flight mass spectrometry. To confirm their result, these authors also prepared and detected D5-triazane, N3D5.31 It remains to be seen whether the Ag(s) substrate, chosen for its ability to conduct heat and electrons away from the thin sample, played a role in this synthesis. Calculations indicate that five isomers of N3H3 are stable; one of them is cyclo-triazane, but it is not the most stable.32 The reaction sequence responsible for the synthesis of triazane and cyclo-triazane in Ag-A is not known. Somehow Ag0, Ag+, and the zeolite framework have cooperated to form these two new hydronitrogen compounds from ammonia. This is a remarkable synthesis, reminiscent of that of palladium and platinum which, like Ag+, have d10 configurations. Furthermore, Ag+ and Pd0 are isoelectronic (46 electrons). It appears that the zeolite, perhaps by the quite generally inadequate coordination that it provides to its exchangeable cations and by the electron density that it might contribute to a formal Ag3+ ion that may be an intermediate in an oxidative addition reaction, has enhanced the ability of Ag+ to participate in this oxidative trimerization of ammonia.

Figure 12. Reduced tripyrylium dimer, (C3H3O3)22+, in a sodalite cavity of zeolite A.

rings of the dimer are parallel and eclipsed but are rotated 60° with respect to one another. A Ag+ ion is atop one ring, and another may be below the other, both on the 3-fold axis of the dimer (symmetry 3̅m = D3d). Each C3O3H33+ ion had been reduced by two electrons. The resulting four electrons σ bond between the rings via their π* orbitals; thus the rings are eclipsed. This bonding is also responsible for the very short interplanar spacing that is seen, 2.43 Å. It is substantially shorter than any seen before between conjugated rings, including the cyclophanes where the shortest interplanar spacing, seen in superphane, C24H24 , [26][2.2.2.2.2.2](1,2,3,4,5,6)-cyclophane,34 is 2.624 Å.35 As in section 6, two reactions can be envisioned for this synthesis. Per unit cell, they are 1.1Ag + + 3.3CH3OH → 0.55(C3H3O3)2 2 + + 4.95H 2 + 1.1Ag 0

(9)

and G


Feature Article


We responded by saying that organic chemists would not be finding this chemistry for 100 years and that this work was for us physical chemists to do. G.N. Lewis’s definition of physical chemistry comes to mind: “Anything that is interesting.”37

11Ag + + 3.3CH3OH → 0.55(C3H3O3)2 2 + + 9.9H+ + 11Ag 0

(10)

+

Except for the 1.1 Ag ions per unit cell, reaction 9 is similar to reaction 7 in section 6; it is again an oxidative trimerization. Reaction 10 is simply the oxidation of methanol by Ag+, albeit to give a novel product. This time, both are possible because reaction 10 does not require more than 12 Ag+ ions. It does, however, require 11, but only about 8 are available because ca. 3 Ag+ ions are found per unit cell in the final structure. Furthermore, the color of the crystal after reaction, dark reddish brown, is not inconsistent with the result that nine silver atoms per unit cell have exited the zeolite to be on its surface in finely divided form. (This crystal also contains clusters of Ag4Cl4 (section 8) which may contribute color to the crystal.) So, once again, reaction 10, the oxidation of a sorbed molecule by Ag+, cannot be the only reaction that is occurring. Reaction 9, however, may be the only reaction that is occurring. Reaction 9 shares some similarities with the commercial reaction for the production of formaldehyde. Finely divided silver metal has traditionally been the catalyst for this.36 2CH3OH + O2 (g) → 2CH 2O + 2H 2O

8. NEUTRAL AG4X4 CLUSTERS, X = CL−, BR−, AND I− In an attempt to introduce photosensitive centers into zeolites, crystals of Ag-A were treated with dilute methanolic solutions of KCl, KBr, and KI. (Aqueous KCl had led to a loss of diffraction pattern as discussed in section 7.) It was hoped that silver halide nanoparticles would form in the cavities of the zeolite. Indeed, the intended nanoparticles were found in the crystal already described in section 7.33 They were seen to be Ag4Cl4, neutral tetrahedrally distorted cubes, occupying sodalite cavities (Figure 13). These Ag4Cl4 molecules are structural units of the bulk material, AgCl(s), as hexasilver was with regard to Ag(s).

(11)

In this reaction methanol is oxidized to formaldehyde. (Like reaction 11, reaction 9 is also an oxidation reaction.) The transformation of methanol to tripyrylium may begin with reaction 11 and continue with further steps of oxidation, the loss of one more hydrogen atom per formaldehyde molecule, cyclic trimerization, and at some point in its synthesis, the oxidation of the tripyrylium molecule to its 1+ cation by Ag+. Note how easily this reaction took place. A crystal of hydrated Ag-A was simply placed in a flowing stream of dilute KCl in methanol at 294 K, and its structure was determined. No attempt was made to dry the methanol; the temperature was never raised; and at the end no attempt was made to remove the solvent molecules from within the zeolite. It was done because attempts to add chloride ions to a Ag-A crystal (with the objective of forming AgCl clusters within the zeolite) using aqueous KCl had repeatedly resulted in the loss of its crystalline diffraction pattern; the next paragraph offers an explanation. So it was done using KCl in methanol. The final structure did contain clusters of silver chloride (45% of the sodalite cavities contained Ag4Cl4 (section 8), so the initial objective was at least partially realized), but it also contained (C3O3H3)22+ ions which occupied the remaining 55% of the sodalite cavities, many (about 14.4) K+ ions, and another Ag+ ion per unit cell. Additional Cl− ions bridged between K+ ions. (The attempted introduction of Cl− ions into hydrated Ag-A using aqueous KCl may have failed (led to loss of crystalline diffraction pattern) because the concentration of water in the aqueous exchange solution was too high. The concentration of water in undried methanol solution should be less by more than 3 orders of magnitude. As Cl− ions enter the zeolite to bond to Ag+, additional cations must also enter the zeolite for charge balance. Both K+ and, at a far lower concentration, H+ from the dissociation of water are available; H+ ions may have selectively exchanged into the zeolite and thus have led to the loss of diffraction pattern. Reaction 10, if it had proceeded to some degree, could also have been the source of some H+ ions.) When this result was first submitted for publication to a journal that publishes papers in physical chemistry, it was rejected because it was judged to be about organic synthesis.

Figure 13. Ag4Cl4 cluster in a sodalite cavity of zeolite A.

Although the Ag4Cl4 clusters are nanoparticles, they differ profoundly from most particles studied in current nanoscience. First of all, they are the size of small molecules and are therefore 1 to 3 orders of magnitude smaller in each dimension. Second, they are monodisperse: all have the same formula and structure, so they are all exactly the same size. Finally, there may be no orientational disorder. They may all have exactly the same orientation throughout the crystal (Figure 14). These clusters within the zeolite crystal are an ordered threedimensional array of quantum dots. Clusters of Ag4Br438 and Ag4I439 were also prepared in zeolite A (Figures 15a and 15b, respectively). Their concentrations were higher than had been seen for Ag4Cl4 (see Table 1), and the tripyrylium dimer was not seen. Recall that the photosensitive material in photographic film is finely divided AgBr(s).

Figure 14. Ag4Cl4 clusters in zeolite A. Only 45% of the sodalite cavities host one, and their arrangement in this drawing is arbitrary. This is an ordered array of identical quantum dots, with vacancies. Two orientations (only one is shown) are possible for each cluster. The sodalite cavities in the corresponding Ag4Br4 and Ag4I4 structures are 75% and 50% filled, respectively. H


Feature Article


10. EPILOGUE Molecules formed within a zeolite may be stable only within the zeolite; it may not be possible to isolate them. While such isolation should be possible for N3H5 and perhaps cyclo-N3H3, we do not easily foresee this to be the case for the other species discussed in this article. The reader will note that most of the results discussed above were unexpected. It appears that the further straightforward study of Ag+-exchanged zeolites will yield many more surprises. We judge that single-crystal crystallography will continue to be the most definitive method for the establishment of new species. However, it is clear that many of the substances described herein require further characterization by other methods; that work is also likely to yield new surprises. The new molecules and ions described in this report were all found in silver zeolites that had been dehydrated under vacuum. That work was done before we realized that dehydration in flowing oxygen would yield a fully oxidized zeolite. If fully oxidized zeolites had been used, the results might have been different, and some could have been simpler. The occupancies (concentrations, populations) of the products found may have been sensitive to variations, intentional or unintentional, in the experimental procedures and may not be fundamental. A discussion of the experimental considerations that should be kept in mind in the preparation of zeolite samples is available.44 For example, keeping a very small amount (a single crystal) of a dehydrated zeolite (a very strong desiccant) from becoming partially hydrated (a few molecules per unit cell) upon cooling after bake out deserves some attention.

Figure 15. (a) Ag4Br4 and (b) Ag4I4 clusters in the sodalite cavities of zeolite A. As can be seen, the Ag−Br and Ag−I distances are not very different (see also Table 1). The conventional ionic radii of Br− (1.96 Å) and I− (2.20 Å) differ more.40 For comparison, the conventional ionic radius of Cl− is 1.81 Å.40 All three of these large halide anions are soft, and their effective radii in molecules vary widely.

Table 1 shows that the Ag−O distances are all about the same, 2.50 Å, in all three Ag4X4 clusters. They are close to the sum, 2.58 Å, of the conventional radii40 of Ag+ and O2−, 1.26 and 1.32 Å, respectively. The Ag−X distances, however, display a sharp irregularity as compared to those in their bulk phases, AgX(s). It can be seen in Table 1 that the Ag−Cl distance in its cluster is about 0.3 Å too long. This shows that Ag4Cl4 is too small for the sodalite cavity. Perhaps it is for that reason that fewer Ag4Cl4 molecules than Ag4Br4 and Ag4I4 were found per unit cell. Ag4Cl4 is either sharply stretched, or the chloride ions are disordered, approaching two of the three nearest Ag+ ions more closely. The thermal parameter at this Cl− position does not, however, indicate disorder.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: seff@hawaii.edu.

9. APPARENT REDUCTION OF CO Due to an experimental error, water was present in a crystal of dehydrated Ag−A that had been exposed to CO(g). The C−O unit bonds to a Ag+ ion via its carbon atom in a bent manner; the Ag−C−O angle is 105(3)°.41 The C−O bond, 1.51(8) Å, although not very precise, is much longer than that in CO (1.131 Å) and H2CO (1.209 Å) and closer to that in CH3OH (1.427 Å). The C−O bond length was found to be relatively constant in the linear molecules OC···AgX, X = F, Cl, Br, and I,42 approximately unchanged from (never more than 0.009 Å less than) its value in free CO(g). It was 1.07(16) Å in its complex with Co2+; this complex was also linear.43 Thus, it appears that CO was reduced by water in that Ag−A crystal, perhaps to Ag+−CH2OH. In further support of this conclusion is the short Ag+−C distance, 2.41(4) Å. Although it is longer than the Co2+−C distance in the CO complex of Co2+, 2.29(16) Å,43 the difference is more than offset by the large difference in the radii of these two cations (Ag+, 1.26 Å; Co2+, 0.72 Å).40

Notes

The authors declare no competing financial interest. Biographies Nam Ho Heo received his PhD in chemistry with Karl Seff at the University of Hawaii in 1987. He is currently a Professor in the Department of Applied Chemistry at Kyungpook National University (KNU) in Daegu, Korea where he heads the Laboratory of Structural Chemistry. Current foci of his group include the preparation and characterization of zeolites with luminescent properties. He is attempting to find and develop new inorganic scintillators by introducing extraframework rare-earth elements into zeolites by various vapor-phase ion-exchange methods. Yang Kim received his PhD in chemistry with Karl Seff at the University of Hawaii in 1978. He worked as a Professor in the Department of Chemistry at Pusan National University in Korea; he is now an Emeritus Professor there. He is also an Emeritus Fellow at the Korean Academy of Science and Technology. He is interested in the structures and chemical bonding of molecules within zeolites.

Table 1. Geometries of the Silver Halide Clusters in Zeolite A Ag4Cl4 Ag4Br4 Ag4I4

a

fraction of sodalite cavities occupied

Ag−O (Å)

Ag−Xa (Å)

Ag−Xa (Å) in AgX(s) (NaCl structure)

Xa−Ag−X (deg)

45% 75% 50%

2.493(5) 2.53(3) 2.478(11)

3.105(17) 2.93(3) 2.973(21)

2.778 2.888 3.04 (high P) 2.81 (1 atm)b

76.47(7) 60.5(16) 61(3)

In this table, X represents a halogen, not a zeolite. bDoes not have the NaCl structure. I


Feature Article


Decomposed Fully Ag+-Exchanged Zeolite A. J. Am. Chem. Soc. 1978, 100, 175−180. (13) Morton, J. R.; Preston, K. F. Paramagnetic Silver Clusters in Irradiated Ag-A Molecular Sieves. J. Magn. Reson. 1986, 68, 121−128. (14) Morton, J. R.; Preston, K. F. Superhyperfine Interactions of Ag6+ Clusters in Silver Loaded 4A Zeolite. Zeolites 1987, 7, 2−4. (15) Mayoral, A.; Carey, T.; Anderson, P. A.; Lubk, A.; Diaz, I. Atomic Resolution Analysis of Silver Ion-Exchanged Zeolite A. Angew. Chem., Int. Ed. 2011, 50, 11230−11233. (16) Kim, Y.; Seff, K. A Dichlorine to Oxide Ion Charge-Transfer Complex. Oxidation of Hexasilver to Silver Chloride by Dichlorine. Crystal Structure of H2.25Ag12Cl2.25Si12Al12O48·6Cl2, a Chlorine Sorption Complex of Partially Dehydrated Fully Ag+-Exchanged Zeolite A. J. Am. Chem. Soc. 1978, 100, 3801−3805. (17) Kim, Y.; Seff, K. The Crystal Structure of a Bromine Sorption Complex of Dehydrated Fully Ag+-Exchanged Zeolite A. J. Phys. Chem. 1978, 82, 925−929. (18) Seff, K.; Shoemaker, D. P. The Structures of Zeolite Sorption Complexes. I. The Structures of Dehydrated Zeolite 5A and Its Iodine Sorption Complex. Acta Crystallogr. 1967, 22, 162−170. (19) Jeong, M. S.; Kim, Y.; Seff, K. Crystal Structures of Dehydrated Zeolite Ag5.6K6.4-A and of the Product of Its Reaction with Cesium: Cs13.5Ag4.5-A, Silver and Cationic Cesium Clusters. J. Phys. Chem. 1993, 97, 10139−10143. (20) Kim, Y.; Jeong, M. S.; Seff, K. Crystallographic Studies of Dehydrated Ag+ and K+ Exchanged Zeolite A Reacted with Alkali Metal Vapor. Bull. Korean Chem. Soc. 1993, 14, 603−610. (21) Song, S. H.; Kim, Y.; Seff, K. Formation of Hexasilver at the Center of the Large Cavity. Three Crystal Structures of Dehydrated Ag+- and Ca2+-Exchanged Zeolite A, Ag12−2xCax-A (X = 2, 3, and 4) Treated with Rubidium Vapor. J. Phys. Chem. 1991, 95, 9919−9924. (22) Lee, S. H.; Kim, Y.; Seff, K. Weak Ag+-Ag+ Bonding in Zeolite X. Crystal Structures of Ag92Si100Al92O384 Hydrated and Fully Dehydrated in Flowing Oxygen. Microporous Mesoporous Mater. 2000, 41, 49−59. (23) Kim, S. Y.; Kim, Y.; Seff, K. Two Crystal Structures of Fully Dehydrated, Fully Ag+-Exchanged Zeolite X. Dehydration in Oxygen Prevents Ag+ Reduction. Without Oxygen, Ag8n+ (Td) and cyclo-Ag4m+ (near S4) Form. J. Phys. Chem. B 2003, 107, 6938−6945. (24) Jansen, M. Homoatomic d10-d10 Interactions: Their Effects on Structure and Chemical and Physical Properties. Angew. Chem., Int. Ed. Engl. 1987, 26, 1098−1110. (25) Masse, R.; Guitel, J. C.; Durif, A. Structure of Silver Carbonate. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1979, 35, 1428−1429. (26) Choi, E. Y.; Kim, S. Y.; Kim, Y.; Seff, K. Crystal Structure of an Ethylene Sorption Complex of Fully Dehydrated, Fully Oxidized, Fully Ag+-Exchanged Zeolite X. Microporous Mesoporous Mater. 2003, 62, 201−210. (27) Lee, Y. M.; Choi, S. J.; Kim, Y.; Seff, K. Crystal Structure of an Ethylene Sorption Complex of Fully Vacuum-Dehydrated Fully Ag+Exchanged Zeolite X (FAU). Silver Atoms Have Reduced Ethylene to Give CH22‑ Carbanions at Framework Oxide Vacancies. J. Phys. Chem. B 2005, 109, 20137−20144. (28) Kim, Y.; Gilje, J. W.; Seff, K. Synthesis and Structures of Two New Hydrides of Nitrogen, Triazane (N3H5) and Cyclotriazane (N3H3). Crystallographic and Mass Spectrometric Analyses of Vacuum-Dehydrated Partially Decomposed Fully Ag+-Exchanged Zeolite A Treated with Ammonia. J. Am. Chem. Soc. 1977, 99, 7057. (29) Schlegel, H. B.; Skancke, A. Thermochemistry, Energy Comparisons, and Conformational Analysis of Hydrazine, Triazane, and Triaminoammonia. J. Am. Chem. Soc. 1993, 115, 7465−7471. (30) Richard, R. M.; Ball, D. W. G2, G3, and Complete Basis Set Calculations on the Thermodynamic Properties of Triazane. J. Mol. Model. 2008, 14, 29−37. (31) Förstel, M.; Maksyutenko, P.; Jones, B. M.; Sun, B. J.; Chen, S. H.; Chang, A. H.; Kaiser, R. I. Detection of the Elusive Triazane Molecule (N3H5) in the Gas Phase. ChemPhysChem 2015, 16, 3139− 3142.

Jong Jin Kim received his PhD in applied chemistry with Nam Ho Heo at KNU in 2015. He is currently working as a postdoctoral student with Prof. Hong Joo Kim in the Department of Physics at KNU on the development and crystallographic characterization of various inorganic single-crystal scintillators. With an eye toward applications, he has prepared various In- and Ga-exchanged zeolites, determined their structures, and measured their physical properties. Karl Seff grew up on a chicken farm in Penngrove, California. He received his BS in chemistry at UC Berkeley in 1959 and his PhD in physical chemistry with David P. Shoemaker at MIT in 1964. After working with Kenneth N. Trueblood at UCLA for three years, he began his academic career at the University of Hawaii. He retired from teaching and departmental administration in 2006 but continues his research as an Emeritus Professor of Chemistry, working with his former students and their former students in Korea. Trained as a crystallographer, he now thinks of himself as a zeolite chemist.

■

ACKNOWLEDGMENTS We gratefully acknowledge the support of the Central Laboratory at KNU, the Pohang Accelerator Laboratory (PAL) at the Pohang Institute of Science and Technology (POSTECH), Korea, and the Photon Factory (PF) at the High Energy Accelerator Research Organization (KEK), Japan for the use of their diffractometers and computing facilities. This work was partly supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (NRF-2014R1A2A1A11054075).

■

REFERENCES

(1) Van Bekkum, H.; Flanigen, E. M.; Jacobs, P. A.; Jansen, J. C. Introduction to Zeolite Science and Practice; Elsevier: Amsterdam, 2001; Vol. 137. (2) Breck, D. W. Zeolite Molecular Sieves; Wiley: New York, 1974; p 449. (3) International Tables for X-Ray Crystallography; Vol. III, pp 318− 323. (4) Kim, Y.; Seff, K. The Octahedral Hexasilver Molecule. Seven Crystal Structures of Variously Vacuum-Dehydrated Fully Ag+Exchanged Zeolite A. J. Am. Chem. Soc. 1978, 100, 6989−6997. (5) Kim, Y.; Seff, K. The Crystal Structure of Dehydrated Fully Ag+Exchanged Zeolite A Reduced by Hydrogen and Reoxidized by Oxygen, Both at 330 °C. The Loss Range Order and Its Subsequent Return. J. Phys. Chem. 1978, 82, 921−924. (6) Texter, J.; Kellerman, R.; Gonsiorowski, T. Formation of Charged Silver Clusters and Their Reversible Silver Ion Desorption in Zeolite A. J. Phys. Chem. 1986, 90, 2118−2124. (7) Lim, W. T.; Seo, S. M.; Wang, L.; Qing, G.; Heo, N. H.; Seff, K. Single-Crystal Structures of Highly NH4+-Exchanged, Fully Deaminated, and Fully Tl+-Exchanged Zeolite Y (FAU, Si/Al = 1.56), All Fully Dehydrated. Microporous Mesoporous Mater. 2010, 129, 11−21. (8) Jacobs, P. A.; Uytterhoeven, J. B.; Beyer, H. K. Some Unusual Properties of Activated and Reduced AgNaA Zeolites. J. Chem. Soc., Faraday Trans. 1 1979, 75, 56−64. (9) Handbook of Chemistry and Physics, 80th ed.; Lide, D.R., Ed.; CRC Press: Boca Raton FL, 1999−2000; pp 4−84. (10) Gensse, C.; Anderson, T. F.; Fripiat, J. J. Study of Oxygen Mobility in Some Synthetic Faujasites by Isotopic Exchange with CO2. J. Phys. Chem. 1980, 84, 3562−3567. (11) Kim, Y.; Seff, K. Structure of a Very Small Piece of Silver Metal. The Octahedral Ag6 Molecule. Two Crystal Structures of Partially Decomposed Vacuum-Dehydrated Fully Ag+-Exchanged Zeolite A. J. Am. Chem. Soc. 1977, 99, 7055−7057. (12) Kim, Y.; Seff, K. The Hexasilver Molecule Stabilized by Coordination to Six Silver Ions. The Structure of (Ag+)6(Ag6). The Crystal Structure of an Ethylene Sorption Complex of Partially J


Feature Article

The Journal of Physical Chemistry C (32) Magers, D. H.; Salter, E. A.; Bartlett, R. J.; Salter, C.; Hess, B. A.; Schaad, L. J. Do Stable Isomers of N3H3 Exist? J. Am. Chem. Soc. 1988, 110, 3435−3446. (33) Kim, S. H.; Thi, T. N. N.; Heo, N. H.; Kim, G. H.; Hong, S. B.; Head, J. D.; Seff, K. Single Crystal Structure of Zeolite A (LTA) Containing Ag4Cl4 Nanoclusters and Reduced 1,3,5-Tripyrylium Dimers with Remarkably Short 2.43 Å Interplanar Spacings. J. Phys. Chem. C 2008, 112, 11181−11193. (34) Sekine, Y.; Boekelheide, V. A Study of the Synthesis and Properties of [26](1,2,3,4,5,6)Cyclophane (Superphane). J. Am. Chem. Soc. 1981, 103, 1777−1785. (35) Hanson, A. W.; Cameron, T. S. Crystal Structures of, and Strain in, Some Adjacent-Bridged [2n] Cyclophanes. J. Chem. Res. Synop. 1980, 336−337. (36) Reuss, G.; Disteldorf, W.; Gamer, A. O.; Hilt, A. Formaldehyde; Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH: Weinheim, 2002. (37) Lewis, E. S. A Biography of Distinguished Scientist Gilbert Newton Lewis; Edward Mellen Press: Lewiston NY, 1998. (38) Lim, W. T.; Choi, S. Y.; Kim, B. J.; Kim, C. M.; Lee, I. S.; Kim, S. H.; Heo, N. H. Synthesis and Crystal Structure of Ag4Br4 Nanoclusters in the Sodalite Cavities of Fully K+-Exchanged Zeolite A (LTA). Bull. Korean Chem. Soc. 2005, 26, 1090−1096. (39) Heo, N. H.; Kim, H. S.; Lim, W. T.; Seff, K. Synthesis and Crystal Structure of Ag4I4 Nanoclusters in the Sodalite Cavities of Fully K+-Exchanged Zeolite A. J. Phys. Chem. B 2004, 108, 3168−3173. (40) Handbook of Chemistry and Physics, 72nd ed.; Lide, D.R., Ed.; CRC Press: Boca Raton FL, 1991−1992; p 12−18. (41) Kim, Y.; Song, S. H.; Seff, K. Crystallographic Evidence for the Reduction of CO in Partially Dehydrated Silver Zeolite A. Bull. Korean Chem. Soc. 1989, 10, 230−234. (42) Stephens, S. L.; Mizukami, W.; Tew, D. P.; Walker, N. R.; Legon, A. C. Molecular Geometry of OC···AgI Determined by Broadband Rotational Spectroscopy and Ab Initio Calculations. J. Chem. Phys. 2012, 136, 064306/1−064306/9. (43) Riley, P. E.; Seff, K. Crystal Structures of Dehydrated Partially Cobalt(II)-Exchanged Zeolite A and of Its Carbon Monoxide Adduct. Inorg. Chem. 1974, 13, 1355−1360. (44) Seff, K. Doing Chemistry in a One-Nanometer Test Tube (in a Zeolite). Comments Inorg. Chem. 2007, 28, 173−179.

K


Surprising Intrazeolitic Chemistry of Silver - The Journal of Physical

Recommend Documents