Variable pore size, variable chemical functionality, and an example of reactivity within porous phenylacetylene silver salts

Article abstract of DOI:10.1021/ja991100b

Investigations on molecular variants of the 3-fold symmetric 1,3,5- tris(4-ethynylbenzonitrile)benzene crystallized with silver triflate revealed a nearly invariant pseudohexagonal porous structure type. Modifications involved the attachment of pendant groups to the central aromatic ring of the parent molecule. Pendant groups include the vinyl group, stilbene, the chiral group myrtanol, and groups with different chemical functionalities such as alcohols, ethers, and esters. Modifications also included the addition of elongated spacer units between the central benzene ring and the peripheral nitrile groups. In these molecules the acetylene bridges of 1,3,5-tris(4- ethynylbenzonitrile)benzene were replaced with diacetylene, ethynylbenzene, and diethynylbenzene type units. Single-crystal refinements for pentoxy- 2,4,6-tris(4-ethynylbenzonitrile)benzene·AgOTf and 1,3,5-tris(4-(4- ethynylbenzonitrile)phenyl)benzene·AgOTf as well as powder data on 12 crystalline phases showed the consistent formation of pseudohexagonal channels, demonstrating that the parent porous architecture is stable both to functional modification of the interior of the channel as well as to enlargement of the pores. Pentoxy-2,4,6-tris(4- ethynylbenzonitrile)benzene·AgOTf refined in the monoclinic space group Am. 1,3,5-Tris(4-(4-ethynylbenzonitrile)phenyl)benzene·AgOTf was found to be triclinic with space group Pl. These crystals have pseudohexagonal channels respectively 15 and 25 A? in diameter. Cell constants based on powder data are compatible with channel diameters ranging from 10 to 30 A?. The latter channel diameters are among the largest known for organic porous solids. The introduction of the chiral myrtanol unit led to the preparation of a chiral porous solid. The thermal and chemical stabilities of these phases were investigated. The pseudohexagonal structure proved stable to complete solvent loss from the channel. It was found in the case of a host with alcohol functionality that an acid anhydride guest, trifluoroacetic anhydride, reacted with the host to form an ester with retention of the porous structure type.

Full text of DOI:10.1021/ja991100b

8204
J. Am. Chem. Soc. 1999, 121, 8204-8215
Variable Pore Size, Variable Chemical Functionality, and an Example
of Reactivity within Porous Phenylacetylene Silver Salts
Y.-H. Kiang, Geoffrey B. Gardner, Stephen Lee,* Zhengtao Xu, and Emil B. Lobkovsky
Contribution from the Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell
UniVersity, Ithaca, New York 14853-1301
ReceiVed April 7, 1999
Abstract: Investigations on molecular variants of the 3-fold symmetric 1,3,5-tris(4-ethynylbenzonitrile)benzene
crystallized with silver triflate revealed a nearly invariant pseudohexagonal porous structure type. Modifications
involved the attachment of pendant groups to the central aromatic ring of the parent molecule. Pendant groups
include the vinyl group, stilbene, the chiral group myrtanol, and groups with different chemical functionalities
such as alcohols, ethers, and esters. Modifications also included the addition of elongated spacer units between
the central benzene ring and the peripheral nitrile groups. In these molecules the acetylene bridges of 1,3,5-
tris(4-ethynylbenzonitrile)benzene were replaced with diacetylene, ethynylbenzene, and diethynylbenzene type
units. Single-crystal refinements for pentoxy-2,4,6-tris(4-ethynylbenzonitrile)benzene‚AgOTf and 1,3,5-tris-
(4-(4-ethynylbenzonitrile)phenyl)benzene‚AgOTf as well as powder data on 12 crystalline phases showed the
consistent formation of pseudohexagonal channels, demonstrating that the parent porous architecture is stable
both to functional modification of the interior of the channel as well as to enlargement of the pores. Pentoxy-
2,4,6-tris(4-ethynylbenzonitrile)benzene‚AgOTf refined in the monoclinic space group Am. 1,3,5-Tris(4-(4-
ethynylbenzonitrile)phenyl)benzene‚AgOTf was found to be triclinic with space group P1h. These crystals have
pseudohexagonal channels respectively 15 and 25 Å in diameter. Cell constants based on powder data are
compatible with channel diameters ranging from 10 to 30 Å. The latter channel diameters are among the
largest known for organic porous solids. The introduction of the chiral myrtanol unit led to the preparation of
a chiral porous solid. The thermal and chemical stabilities of these phases were investigated. The
pseudohexagonal structure proved stable to complete solvent loss from the channel. It was found in the case
of a host with alcohol functionality that an acid anhydride guest, trifluoroacetic anhydride, reacted with the
host to form an ester with retention of the porous structure type.
Introduction
attractive feature of such systems is that by careful selection of
the constituent molecules one can potentially modify the
inclusion cavities.23 However, one problem with such an
approach is that modification of these organic building blocks
often results in global structural changes such as unwanted
interpenetrations,18,24-32 altered network connectivities,20,33,34
and new metal coordination environments35-38 which can
There has recently been a renewed interest in largely organic
systems whose crystalline structures and porous natures are
reminiscent of traditional inorganic zeolites.1-22 A potentially
* To whom correspondence should be addressed.
(1) MacNicol, D. D.; McKendrick, J. J.; Wilson, D. R. In Inclusion
Compounds; Atwood, J. L., Davies, J. E., MacNicol, D. D., Eds.; Academic
Press: London, 1984; Vol. 2.
(2) Hoskins, B. F.; Robson R. J. Am. Chem. Soc. 1990, 112, 1546.
(3) Abrahams, B. F.; Hoskins, B. F.; Lin, J.; Robson R. J. Am. Chem.
Soc. 1991, 113, 3045.
(4) Simard, M.; Su, D.; Wuest, J. D. J. Am. Chem. Soc. 1991, 113, 4696.
(5) Ermer, O.; Lindenberg, L. HelV. Chim. Acta 1991, 74, 825.
(6) Fujuta, M.; Kwon, Y. J.; Washizu, S.; Ogura, K. J. Am. Chem. Soc.
1994, 116, 1151.
(16) Brunet, P.; Simard, M.; Wuest, J. D. J. Am. Chem. Soc. 1997, 119,
2737.
(17) Yaghi, O. M.; Davis, C. E.; Li, G.; Li, H. J. Am. Chem. Soc. 1996,
119, 2861.
(18) Hoskins, B. F.; Robson, R.; Slizys, D. A. Angew. Chem., Int. Ed.
1997, 36, 2336
(19) MacGillivray, L.; Atwood, J. Nature 1997, 389, 496.
(20) Fujita, M.; Aoyagi, M.; Ogura, K. Bull. Chem. Soc. Jpn. 1998, 71,
1799.
(7) Abrahams, B. F.; Hoskins, B. F.; Michail, D. M.; Robson, R. Nature
1994, 369, 727.
(8) Gardner, G. B.; Venkataraman, D.; Moore, J. S.; Lee, S. Nature 1995,
374, 792.
(9) Venkataraman, D.; Gardner, G. B.; Lee, S.; Moore, J. S. J. Am. Chem.
Soc. 1995, 117, 11600.
(10) Ung, A.; Gizachew, D.; Bishop, R.; Scudder, M.; Dance, I.; Craig,
D. J. Am. Chem. Soc. 1995, 117, 8475.
(11) Endo, K.; Sawaki, T.; Koyanagi, M.; Kobayashi, K.; Masuda, H.;
Aoyama, Y. J. Am. Chem. Soc. 1995, 117, 8341.
(12) Lu, J.; Harrison, W. T. A.; Jacobson, A. J. Angew. Chem. 1995,
107, 2759; Angew. Chem., Int. Ed. Engl. 1995, 34, 2557.
(13) Dolbeq, A.; Boubekeur, K.; Batail, P.; Canadell, E.; Aubansenzier,
P.; Coulon, C.; Lerstrup, K. Mater. Chem. 1995, 5, 1707.
(14) Gardner, G. B.; Kiang, Y.-H.; Lee, S.; Asgaonkar, A.; Venkataraman,
D. J. Am. Chem. Soc. 1996, 118, 6946.
(21) Batten, S. R.; Robson, R. Angew. Chem., Int. Ed. 1998, 37, 1461.
(22) Abrahams, B. F.; Jackson, P. A.; Robson, R. Angew. Chem., Int.
Ed. 1998, 37, 2656.
(23) Russel, V. A.; Evans, C. C.; Li, W.; Ward, M. D. Science
(Washington, D.C.) 1997, 276, 575.
(24) Copp, S. B.; Subramanian, S.; Zawortko, M. J. Am. Chem. Soc.
1992, 114, 8719.
(25) Ermer, O.; Eling, A. Angew. Chem., Int. Ed. Engl. 1988, 27, 829.
(26) Michaelides, A.; Kiritsis, V.; Skoulika, S.; Aubry, A. Angew. Chem.,
Int. Ed. Engl. 1993, 32, 1495.
(27) Reddy, D. S.; Craig, D. C.; Desiraju, G. R. J. Am. Chem. Soc. 1996,
118, 4090.
(28) Ermer, O. J. Am. Chem. Soc. 1988, 110, 3747.
(29) Konoshita, Y.; Matsubara, I.; Higuchi, T.; Saito, Y. Bull. Chem.
Soc. Jpn. 1959, 32, 1221.
(15) Yaghi, O. M.; Li, H.; Groy, T. L. J. Am. Chem. Soc. 1996, 118,
9096.
(30) Ermer, O. AdV. Mater. 1991, 3, 608.
10.1021/ja991100b CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/25/1999

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8205
destroy the desired host properties of the solid.39-42 For example,
one of the most studied host-guest systems is that based on
Dianin’s compound. However, of sixteen derivatives of the
parent,1 only eight exhibit the same inclusion properties as
Dianin’s compound. In another well-investigated system Bishop
and Dance et al. have found a remarkable diol host system for
which stringent variants repeatedly adopt a porous morphology.
However, even here of thirteen studied derivatives of the host
compound, seven form noninclusion solid-state structures.34,43-48
Furthermore, it is often challenging to predict the substituent
effect on the final structure. A good illustration of this is the
Bishop and Dance system for which the methyl derivative is
porous whereas the hydrogen and ethyl derivatives are not.34
This difficulty of predicting the final structure highlights the
need for parent systems which consistently direct the adoption
of a host structure type even in the presence of substantial
building block alterations.
One approach to this packing problem is the selection of
molecular geometries which naturally not only give rise to
porosity but also present sites for modification which are
spatially removed from structurally important intermolecular
network linkages. This modification at a distance, which
perforce requires fairly large organic molecules, potentially
allows both expansion of the cavital dimensions as well as the
ability to define the interior shape. This is especially true as it
is now well established that polytopic ligands49-56 may be used
in the synthesis of extended solids.57-71 We report here that
variants of 1,3,5-tris(4-ethylbenzonitrile)benzene (1) persistently
crystallize with silver(I) trifluoromethanesulfonate (AgOTf)8 into
structure types based on nearly regular hexagonal channels. Such
hexagonal channels are found even in the presence of significant
changes to the organic ligands, and thus this preparative
methodology can be used to tune both the shape and dimensions
of the cavity.
In particular we considered two types of substitution of the
parent molecule 1. In the first set, shown in Figure 1, we placed
different substituents on the centered benzene ring of molecule
1 ranging from hydrocarbons to ether, alcohol, and ester
linkages. Such alterations were used to test the stability of the
porous system to modification of the interior of the hexagonal
channels. We also prepared a second set of compounds, also
shown in Figure 1, in which the initial acetylene bridging units
of molecule 1 are replaced by increasingly long linear rigid units.
Such modifications were used to study the stability of the
pseudohexagonal architecture to greater cavity dimensions. In
both suites of molecules, the crystalline structure of the organic
molecule with AgOTf was found to contain nearly regular
hexagonal channels. The stability of these hexagonal channels
is further proven by an experiment in which a covalent bond
forming reaction occurred between an acid anhydride guest,
trifluoroacetic anhydride, and an alcohol host molecule 6‚AgOTf
without destroying the initial hexagonal pores.
(31) Sinzger, K.; Hu¨nig, S.; Jopp, M.; Bauer, D.; Bietsch, W.; von Shu¨tz,
J. U.; Wolf, H. C.; Kremer, R. K.; Metzenthin, T.; Bau, R.; Kahn, S. I.;
Lindbaum, A.; Lengauer, C. L.; Tillmanns, E. J. Am. Chem. Soc. 1993,
115, 7696.
(32) Hirsch, K. A.; Venkataraman, D.; Wilson, S. R.; Moore, J. S.; Lee,
S. J. Chem. Soc., Chem. Commun. 1995, 2199.
Experimental Section
General Procedure. Unless otherwise indicated, all commercially
available reagents were purchased from Aldrich and used without further
purification. Analytical grade solvents were obtained from commercial
suppliers (Aldrich, Burdick and Jackson, EM Science, Fisher Scientific,
and Mallinckrodt).
(33) Fujita, M.; Kwon, Y.; Sasaki, O.; Yamaguchi, K.; Ogura, K. J. Am.
Chem. Soc. 1995, 117, 7287.
(34) Bishop, R.; Choudhury, S.; Dance, I. J. Chem. Soc., Perkin Trans.
2 1982, 1159.
(35) Kitagawa, S.; Munakata, M.; Tanimura, T. Inorg. Chem. 1992, 31,
1714.
(36) Kitagawa, S.; Katawa, S.; Kondo, M.; Nozaka, Y.; Munakata, M.
Bull. Chem. Soc. Jpn. 1993, 66, 3387
(37) Carlucci, L.; Ciani, G.; Proserpio, D.; Sironi, A. Angew. Chem. 1995,
107, 2037; Angew. Chem., Int. Ed. Engl. 1995, 34, 1895.
(38) Carlucci, L.; Ciani, G.; Proserpio, D.; Sironi, A. J. Chem. Soc.,
Chem. Commun. 1996, 1393.
Powder X-ray diffraction data were recorded on an Enraf-Nonius
Gunier Camera at 40 kV, 13 mA for Cu KR1, λ ) 1.5406 Å, with an
internal standard of RbZn4Ga5S12.72 Lattice constants were fitted and
powder data were indexed by a least-squares method.
Single-crystal X-ray data were collected on a Siemens Smart
diffractometer equipped with a CCD area detector using Mo KR
(56) Su, D.; Wang, X.; Simard, M.; Wuest, J. D. Supramol. Chem. 1995,
6, 171.
(39) Duˇrcˇanska´, E.; Jamnicky´, M.; Koman, M.; Wnek, I.; Glowiak, T.
Acta. Crystallogr., Sect. C 1986, 42, 1157.
(40) Duˇrcˇanska´, E.; Koman, M.; Jamnicky´, M. Acta. Crystallogr., Sect.
C 1991, 47, 1193.
(57) Carlucci, L.; Ciani, G.; Proserpio, D. M.; Sironi, A. Inorg. Chem.
1995, 34, 5698.
(58) Carlucci, L.; Ciani, G.; Proserpio, D. M.; Sironi, A. J. Am. Chem.
Soc. 1995, 117, 4562.
(41) Drew, M.; Gray, N.; Cabral, M.; Cabral, J. Acta. Crystallogr., Sect.
C 1985, 41, 1434.
(42) Nassimbeni, L.; Niven, M.; Taylor, M. Acta. Crystallogr., Sect. B
1990, 46, 354.
(59) Otieno, T.; Rettig, S. J.; Thompson, R. C.; Trotter, J. Inorg. Chem.
1993, 32, 1607.
(60) Otieno, T.; Rettig, S. J.; Thompson, R. C.; Trotter, J. J. Can. Chem.
1989, 67, 1964.
(43) Bishop, R.; Craig, D.; Scudder, M. J. Chem. Soc., Perkin Trans. 2
1989, 1473.
(44) Hawkins, S.; Scudder, M.; Craig, D.; Rae, A. D.; Raof, A.; Bishop,
R.; Dance, I. J. Chem. Soc., Perkin Trans. 2 1990, 855.
(45) Hawkins, S.; Bishop, R.; Craig, D.; Dance, I.; Rae, A. D.; Scudder,
M. J. Chem. Soc., Perkin Trans. 2 1993, 1737.
(46) Bishop, R.; Craig, D.; Dance, I.; Scudder, M.; Marchard, A.; Wang,
Y. J. Chem. Soc., Perkin Trans. 2 1993, 937.
(47) Mallick, M.; Bishop, R.; Craig, D.; Dance, I.; Scudder, M. Aust. J.
Chem. 1991, 44, 343.
(48) Ahn, P.; Bishop, R.; Craig, D. C.; Downing, G. A.; Scudder, M. L.
Aust. J. Chem. 1997, 50, 1053.
(49) Konnert, J.; Britton, D. Inorg. Chem. 1966, 5, 1193.
(50) Hoskins, B. F.; Robson R. J. Am. Chem. Soc. 1989, 111, 5962.
(51) Abrahams, B. F.; Batten, S. R.; Hamit, H.; Hoskins, B. F.; Robson,
R. J. Chem. Soc., Chem. Commun. 1996, 1313.
(52) Venkataraman, D.; Lee, S.; Moore, J. S.; Zhang, P.; Hirsch, K. A.;
Gardner, G. B.; Covey, A. C.; Prentice, C. L. Chem. Mater. 1996, 8, 2030.
(53) Fujita, M.; Ibukuro, F.; Yamaguchi, K.; Ogura, K. J. Am. Chem.
Soc. 1995, 117, 4175.
(54) Yaghi, O. M.; Li, G. Angew. Chem., Int. Ed. Engl. 1995, 34, 207.
(55) Fujita, M.; Aoyagi, M.; Ibukuro, F.; Ogura, K.; Yamaguchi, K. J.
Am. Chem. Soc. 1998, 120, 611.
(61) Otieno, T.; Rettig, S. J.; Thompson, R. C.; Trotter, J. J. Can. Chem.
1990, 68, 1901.
(62) MacGillivray, L. R.; Subramanian, S.; Zaworotko, M. J. J. Chem.
Soc., Chem. Commun. 1994, 11, 1325.
(63) Cromer, D. T.; Larson, A. C. Acta. Crystallogr. 1972, B28, 1052.
(64) Carlucci, L.; Ciani, G.; Proserpio, D. M.; Sironi, A. J. Chem. Soc.,
Chem. Commun. 1994, 2755.
(65) Kuroda-Sowa, T.; Horino, T.; Yamamoto, M.; Ohno, Y.; Maekawa,
M.; Munakata, M. Inorg. Chem. 1997, 36, 6382.
(66) Venkataraman, D.; Gardner, G. B.; Covey, A. C.; Lee, S.; Moore,
J. S. Acta Crystallogr. 1996, 52, 2416.
(67) Lopez, S.; Kahraman, M.; Harmata, M.; Keller, S. W. Inorg. Chem.
1997, 36, 6138.
(68) Lu, J.; Crisci, G.; Niu, T.; Jacobson, A. J. Inorg. Chem. 1997, 36,
5140.
(69) Hirsch, K. A.; Wilson, S. R.; Moore, J. S. Chem. Eur. J. 1997, 3,
765.
(70) Blake, A. J.; Champness, N. R.; Khlobystov, A.; Lemenovskii, D.
A.; Li, W.-S.; Schroder, M. J. Chem. Soc., Chem. Commun. 1997, 2027.
(71) Abrahams, B. F.; Hardie, M. J.; Hoskins, B. F.; Robson, R.;
Sutherland, E. E. J. Chem. Soc., Chem. Commun. 1994, 1049.
(72) Schwer, H.; Keller, E.; Kra¨mer, V. Z. Z. Kristallogr. 1993, 204,
203.

8206 J. Am. Chem. Soc., Vol. 121, No. 36, 1999
Kiang et al.
Figure 1. Parent ligand 1 with modifications 2-10 and 19. Modifications 2-7 and 19 lead to changes in the interior geometry of the pseudohexagonal
channels of their respective silver salts. Modifications 8-10 increase the diameters of the pseudohexagonal channels.
radiation. Diffraction data were collected at 293 K. All structure
solutions were obtained by a direct method and refined using full-matrix
least squares with Shelxl 97. Tables of bond distances, bond angles,
and anisotropic thermal factors appear in the Supporting Information.
1H NMR and 13C NMR were performed on a Bruker AC-200, AM-
300, or A-360 at 25 °C. Apohosts were prepared on a Perkin-Elmer 7
thermal analyzer.
Hz, 1H), 5.62 (dd, J ) 17.7, 1.2 Hz, 1H); 13C NMR (75 MHz, CDCl3)
δ 137.7, 134.6, 134.4, 123.8, 123.1, 121.0.
Pentoxy-2,4,6-tribromobenzene (13). To a flask equipped with a
magnetic stirrer and a reflux condenser were added 2,4,6-tribromophenol
(6.0 g, 18.1 mmol), n-iodopentane (10.8 g, 54.4 mmol), potassium
carbonate (4.0 g, 28.9 mmol), and acetone (40 mL), and the mixture
was heated at reflux overnight. After the solution was allowed to cool
to room temperature, the mixture was extracted with ether (100 mL)
and water (100 mL). The filtrate was washed with 5% sodium hydroxide
solution (100 mL) and water (100 mL) and dried over magnesium
sulfate. The solvent was removed under reduced pressure on a rotary
evaporator to yield 13 as a yellow oil (6.24 g, 86%): 1H NMR (300
MHz, CDCl3) δ 7.64 (s, 2H), 3.97 (t, J ) 6.6 Hz, 2H), 1.84 (m, 2H),
1.44 (m, 4H), 0.94 (t, J ) 7.1 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ
153.2, 135.0 119.2, 117.1, 73.9, 29.9, 28.2, 22.8, 14.3.
[2,4,6-Tribromo-4′-(trans-2-phenylethenyl)biphenyl (11). A heavy-
walled Schlenk flask equipped with a magnetic stirrer was charged with
2,4,6-tribromoiodobenzene (13.0 g, 29.4 mmol), 4-iodostilbene (2.0 g,
6.54 mmol), and powdered copper metal (6.7 g, 0.106 mmol). The flask
was evacuated and sealed with a Teflon screw cap under vacuum. Under
constant stirring, the temperature of the reaction mixture was raised to
200 °C over a period of 6 h. The solid was then suspended in methylene
chloride (100 mL) and filtered through a fritted glass funnel followed
by several washings of the residue solid with methylene chloride (3 ×
100 mL). The combined filtrate was evaporated to dryness under
reduced pressure on a rotary evaporator, and the residue was purified
by column chromatography on silica gel (hexanes) to yield 11 as a
white solid (2.1 g, 66%): 1H NMR (300 MHz, CDCl3) δ 7.81 (s, 2H),
7.61 (d, J ) 7.2 Hz, 2H), 7.38 (t, J ) 7.1 Hz, 2H), 7.29 (d, J ) 7.2
Hz, 1H), 7.17 (m, 4H); 13C NMR (90 MHz, CDCl3) δ 142.0, 139.5,
137.6, 137.4, 134.5, 129.7, 128.9, 128.3, 128.0, 126.8, 126.6, 125.1,
122.1.
(1S,2S,5S)-Myrtanoxy-2,4,6-triiodobenzene (14). A flask equipped
with a magnetic stirrer was charged with sodium (0.4 g, 17.4 mmol)
and anhydrous ethanol (50 mL). After the sodium had dissolved, 2,4,6-
triiodophenol (8.20 g, 17.4 mmol) was added, and the solution was
heated at reflux for 10 min. The solvent was removed in vacuo, and
the residual salt was dissolved in 35 mL of THF-DMSO (1:1). A
solution of myrtanol tosylate (5.3 g, 17.2 mmol) in anhydrous THF
(15 mL) was added via syringe at room temperature and stirred at 60
°C for 3 h. After the solution was allowed to cool to room temperature,
the mixture was extracted with 10% K2CO3 in water (100 mL) and
petroleum ether (100 mL), and the organic layer was dried over sodium
sulfate. The solvent was removed under reduced pressure on a rotary
evaporator and passed through a plug of silica gel (petroleum ether),
and the filtrate was concentrated under reduced pressure to yield 14 as
a white solid (5.2 g, 48%): 1H NMR (360 MHz, CDCl3) δ 8.04 (s,
2H), 3.80-3.69 (m, 2H), 1.84 (m, 2H), 2.71-2.60 (m, 1H) 2.27-2.22
(m, 1H) 2.17-2.10 (m, 1H), 1.94-1.67 (m, 1H) 1.52-1.33 (m, 2H),
1.26 (s, 3H), 0.93 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 158.1, 147.3,
92.3, 89.1, 76.7, 42.6, 41.1, 39.4, 35.9, 27.0, 24.4, 24.2, 20.6, 18.1.
2,4,6-Tribromostyrene (12). A heavy-walled Schlenk flask equipped
with a magnetic stirrer was charged with 2,4,6-tribromoiodobenzene
(2.0 g, 4.54 mmol), tributyl(vinyl)tin (1.74 g, 5.46 mmol), copper(I)
iodide (43 mg, 0.226 mmol), triphenylarsine (0.276 g, 0.902 mmol),
bis(dibenzylideneacetone)palladium(0) (0.12 g, 0.209 mmol), and dry
tetrahydrofuran (30 mL). The flask was capped with a septum,
evacuated, and back-filled with nitrogen three times. The flask was
sealed with a Teflon screw cap and stirred at 65 °C for 20 h. The
mixture was cooled to room temperature and evaporated to dryness
under reduced pressure on a rotary evaporator, and the residue was
purified by column chromatography on silica gel (hexanes) to yield 12
as a white crystalline solid (0.64 g, 42%): 1H NMR (300 MHz, CDCl3)
δ 7.72 (s, 2H), 6.54 (dd, J ) 7.7, 4.5 Hz, 1H), 5.67 (dd, J ) 11.5, 1.2
2,4,6-Tris(trimethylsilylethynyl)toluene (15). A heavy-walled Schlenk
tube equipped with a magnetic stirrer was charged with 2,4,6-
tribromotoluene (2.0 g, 6.08 mmol), copper(I) iodide (0.25 g, 1.32

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8207
mmol), bis(triphenylphosphine)palladium(II) chloride (0.25 g, 0.356
mmol), and triethylamine (50 mL). The flask was capped with a septum,
evacuated, and back-filled with nitrogen. Trimethylsilylacetylene (3.58
g, 36.5 mmol) was added to the flask under nitrogen via syringe. The
flask was sealed with a Teflon screw cap and stirred at 80-90 °C for
12 h. The reaction mixture was then allowed to cool to room
temperature and filtered through a fritted glass funnel. The filtrate was
evaporated to dryness under reduced pressure on a rotary evaporator
on silica gel (3:1 hexanes-chloroform) to yield 15 as a waxy solid
(1.79 g, 77%): 1H NMR (300 MHz, CDCl3) δ 7.49 (s, 2H), 2.52 (s,
3H), 0.25 (s, 18H), 0.22 (s, 9H); 13C NMR (90 MHz, CDCl3) δ 143.5,
135.6, 123.9, 120.8, 99.4, 94.8, 88.2, 19.7, 0.1.
The filtrate was evaporated to dryness under reduced pressure on a
rotary evaporator and transferred to a heavy-walled Schlenk flask. To
this flask were added 4-bromobenzonitrile (0.862 g, 4.74 mmol),
copper(I) iodide (33 mg, 0.174 mmol), triphenylphosphine (0.160 g,
0.611 mmol), bis(triphenylphosphine)palladium(II) chloride (33 mg,
0.047 mmol), and triethylamine (50 mL). The flask was capped with
a septum, evacuated, back-filled with nitrogen, sealed with a Teflon
cap, and stirred at 80-90 °C for 12 h. The reaction mixture was then
allowed to cool to room temperature and filtered through a glass funnel.
The filtrate was evaporated under reduced pressure on a rotary
evaporator and the residue purified by column chromatography on silica
gel (CH2Cl2) to yield 2 as a beige solid (0.2 g, 50%): 1H NMR (300
MHz, CDCl3) δ 7.72 (s, 2H), 7.70 (d, J ) 7.9 Hz, 2H), 7.66 (q, J )
6.7 Hz, 8H), 7.61 (d, J ) 8.6 Hz, 2H), 2.72 (s, 3H); 13C NMR (75
MHz, CDCl3) δ 143.5, 135.6, 132.3, 132.2, 127.7, 123.8, 120.6, 118.4,
112.3, 112.2, 92.9, 91.7, 91.0, 88.8, 19.6; MS (EI, 70 eV) m/z 486
(38), 467 (100), 466 (17); IR (KBr, cm-1) 2924, 2228, 2214, 1602,
1497, 1173, 1103, 836; HRMS (EI, 70 eV) calcd 467.1422, found
467.1415.
2,4,6-Tris(4-ethynylbenzonitrile)styrene (3). A heavy-walled Schlenk
flask equipped with a magnetic stirrer was charged with 12 (50 mg,
0.15 mmol) 4-ethynylbenzonitrile73 (65 mg, 0.51 mmol), copper(I)
iodide (1.1 mg, 5.8 µmol), triphenylphosphine (3.9 mg, 15 µmol), bis-
(triphenylphosphine)palladium(II) chloride (33 mg, 0.047 mmol), and
triethylamine (3 mL). The flask was capped with a septum, evacuated,
back-filled with nitrogen, sealed with a Teflon cap, and stirred at 70
°C for 20 h. The reaction mixture was then allowed to cool to room
temperature and filtered through a glass funnel. The filtrate was
evaporated under reduced pressure on a rotary evaporator and the
residue purified by column chromatography on silica gel (CH2Cl2) to
yield 3 as a yellow solid (38 mg, 53%): 1H NMR (300 MHz, CDCl3)
δ 7.76 (s, 2H), 7.69-7.60 (m, 12H), 7.19 (dd, J ) 17.7, 5.6 Hz, 1H),
6.43 (dd, J ) 17.7, 1.2 Hz, 1H), 5.82 (dd, J ) 5.6, 1.2 Hz, 1H); 13C
NMR (75 MHz, CDCl3) δ 140.9, 136.5, 132.8, 132.1, 132.0, 127.4,
122.2, 121.8, 121.4, 118.2, 112.2, 112.1, 92.9, 91.3, 89.5; MS (EI, 70
eV) m/z 481 (12), 480 (43), 479 (100), 478 (20), 477 (16); IR (KBr,
cm-1) 3057, 2924, 2228, 2214, 1602, 1457, 1173, 1103, 836; HRMS
(EI, 70 eV) calcd 479.1422, found 479.1412.
2,4,6-Tris(4-ethynylbenzonitrile)-4′-(trans)-2-phenylethenyl)bi-
phenyl (4). A Schlenk flask equipped with a magnetic stirrer was
charged with 16 (0.40 g, 0.735 mmol) and tetrahydrofuran (50 mL).
The mixture was capped with a septum, evacuated, back-filled with
nitrogen, and cooled to 5 °C in an ice bath. A 1.0 M solution of
tetrabutylammonium fluoride in 9:1 tetrahydrofuran-water (2.2 mL,
2.2 mmol) was added rapidly Via syringe. After 5 min the solution
was concentrated to near dryness under reduced pressure on a rotary
evaporator, and the residue was eluted through a plug of silica gel with
a 7:3 hexanes-chloroform solution. The filtrate was evaporated to
dryness under reduced pressure on a rotary evaporator and transferred
to a heavy-walled Schlenk flask. To this flask were added 4-bromoben-
zonitrile (0.442, 2.43 mmol), copper(I) iodide (20 mg, 0.105 mmol),
triphenylphosphine (0.104 g, 0.397 mmol), bis(triphenylphosphine)-
palladium(II) chloride (26 mg, 0.037 mmol), and triethylamine (50 mL).
The flask was capped with a septum, evacuated, back-filled with
nitrogen, sealed with a Teflon cap, and stirred at 80-90 °C for 12 h.
The reaction mixture was then allowed to cool to room temperature
and filtered through a glass funnel. The filtrate was evaporated under
reduced pressure on a rotary evaporator and the residue purified by
column chromatography on silica gel (chloroform) to yield 4 as a yellow
solid (0.32 g, 67%): 1H NMR (360 MHz, CDCl3) δ 7.84 (s, 2H), 7.67
(q, J ) 9.0 Hz, 6H), 7.60 (d, J ) 8.2 Hz, 4H), 7.56 (d, J ) 6.7 Hz,
4H), 7.41 (t, J ) 7.1 Hz, 2H), 7.31 (m, 5H), 7.23 (s, 2H); 13C NMR
(90 MHz, CDCl3) δ 146.7, 137.5, 136.9, 135.8, 132.2, 132.0, 131.9,
130.5, 129.6, 128.8, 128.1, 128.0, 127.4, 126.6, 123.1, 122.0, 112.1,
111.9, 92.1, 91.3, 89.4; MS (EI, 70 eV) m/z 633 (16), 632 (56), 631
(100), 630 (19), 554 (13), 529 (21); IR (KBr, cm-1) 2221, 2214, 1602,
1497, 836, 688; HRMS (EI, 70 eV) calcd 631.2048, found 631.2075.
2,4,6-Tris(trimethylsilylethynyl)-4′-(trans-2-phenylethenyl)biphe-
nyl (16). A heavy-walled Schlenk flask was charged with 11 (1.96 g,
3.98 mmol), copper(I) iodide (0.163 g, 0.858 mmol), triphenylphosphine
(0.652 g, 2.49 mmol), bis(triphenylphosphine)palladium(II) chloride
(0.163 g, 0.232 mmol), and triethylamine (50 mL). The flask was
capped with a septum, evacuated, and back-filled with nitrogen.
Trimethylsilylacetylene (2.37 g, 24.2 mmol) was added to the flask
Via syringe. The flask was sealed with a Teflon cap and stirred at 80-
90 °C for 12 h. The reaction mixture was then allowed to cool to room
temperature and filtered through a glass funnel. The filtrate was
evaporated under reduced pressure on a rotary evaporator and the
residue purified by column chromatography on silica gel (3:1 cyclo-
hexane-chloroform) to yield 16 as a yellow solid (1.34 g, 62%): 1H
NMR (300 MHz, CDCl3) δ 7.62 (s, 2H), 7.58-7.48 (m, 6H), 7.38 (t,
J ) 7.2 Hz, 2H), 7.28 (m, 1H), 7.17 (s, 2H), 0.26 (s, 9H), 0.07 (s,
18H); 13C NMR (75 MHz, CDCl3) δ 146.5, 137.4, 136.0, 130.7, 128.9,
127.8, 126.7, 125.4, 123.4, 122.4, 103.3, 99.2, 95.9, 0.18, -0.14.
Pentoxy-2,4,6-tris(trimethylsilylethynyl)benzene (17). A heavy-
walled Schlenk flask equipped with a magnetic stirrer was charged with
13 (1.05 g, 2.62 mmol), copper(I) iodide (0.11 g, 0.589 mmol),
triphenylphosphine (0.44 g, 1.53 mmol), bis(triphenylphosphine)-
palladium(II) chloride (0.11 g, 0.157 mmol), and triethylamine (50 mL).
The flask was capped with a septum, evacuated, and back-filled with
nitrogen. Trimethylsilylacetylene (1.54 g, 15.7 mmol) was added to
the flask Via syringe. The flask was sealed with a Teflon cap and stirred
at 80-90 °C for 12 h. The reaction mixture was then allowed to cool
to room temperature and filtered through a glass funnel. The filtrate
was evaporated under reduced pressure on a rotary evaporator and the
residue purified by column chromatography on silica gel (85:15
hexanes-chloroform) to yield 17 as a yellow oil (0.64 g, 54%): 1H
NMR (300 MHz, CDCl3) δ 7.64 (s, 2H), 3.97 (t, J ) 6.6 Hz, 2H),
1.84 (m, 2H), 1.44 (m, 4H), 0.94 (t, J ) 7.1 Hz, 3H); 13C NMR (75
MHz, CDCl3) δ 153.2, 135.0, 119.2, 117.1, 73.9, 29.9, 28.2, 22.8, 14.3.
(2,4,6-Tris(trimethylsilylethynyl)phenoxy)ethanol (18). A heavy-
walled Schlenk flask equipped with a magnetic stirrer was charged with
(2,4,6-tribromophenoxy)ethanol (3.00 g, 8.00 mmol), copper(I) iodide
(0.168 g, 0.882 mmol), triphenylphosphine (0.64 g, 2.57 mmol), bis-
(triphenylphosphine)palladium(II) chloride (0.168 g, 0.239 mmol), and
triethylamine (50 mL). The flask was capped with a septum, evacuated,
and back-filled with nitrogen. Trimethylsilylacetylene (3.92 g, 40.0
mmol) was added to the flask Via syringe. The flask was sealed with
a Teflon cap and stirred at 80-90 °C for 12 h. The reaction mixture
was then allowed to cool to room temperature and filtered through a
glass funnel. The filtrate was evaporated under reduced pressure on a
rotary evaporator and the residue purified by column chromatography
on silica gel (65:35 hexanes-chloroform) to yield 18 as a clear oil
(3.08 g, 91%): 1H NMR (360 MHz, CDCl3) δ 7.50 (s, 2H), 4.39 (t, J
) 4.2 Hz, 2H), 3.82 (t, J ) 4.4 Hz, 2H), 0.23 (s, 18H), 0.22 (s, 9H);
13C NMR (75 MHz, CDCl3) δ 161.2, 139.2, 119.1, 117.7, 102.8, 100.8,
100.0, 19.1, 16.2, 1.26.
2,4,6-Tris(4-ethynylbenzonitrile)toluene (2). A Schlenk flask
equipped with a magnetic stirrer was charged with 15 (0.30 g, 0.789
mmol) and tetrahydrofuran (50 mL). The mixture was capped with a
septum, evacuated, back-filled with nitrogen, and cooled to 5 °C in an
ice bath. A 1.0 M solution of tetrabutylammonium fluoride in 9:1
tetrahydrofuran-water (2.7 mL, 2.7 mmol) was added rapidly Via
syringe. After 5 min the solution was concentrated to near dryness under
reduced pressure on a rotary evaporator, and the residue was eluted
through a plug of silica gel with a 9:1 hexanes-chloroform solution.
(73) Takahashi, S.; Kuroyama, Y.; Sonogashi, K.; Hagihara, N. Synthesis
1980, 8, 627.
(74) Lavastre, O.; Cabioch, S.; Dixneuf, P. H. Tetrahedron 1997, 53,
7595.

8208 J. Am. Chem. Soc., Vol. 121, No. 36, 1999
Kiang et al.
Pentoxy-2,4,6-tris(4-ethynylbenzonitrile)benzene (5). A Schlenk
flask equipped with a magnetic stirrer was charged with 17 (0.45 g,
1.00 mmol) and tetrahydrofuran (50 mL). The mixture was capped with
a septum, evacuated, back-filled with nitrogen, and cooled to 5 °C in
an ice bath. A 1.0 M solution of tetrabutylammonium fluoride in 9:1
tetrahydrofuran-water (3.2 mL, 3.2 mmol) was added rapidly Via
syringe. After 5 min the solution was concentrated to near dryness under
reduced pressure on a rotary evaporator, and the residue was eluted
through a plug of silica gel with a 7:3 hexanes-chloroform solution.
The filtrate was evaporated to dryness under reduced pressure on a
rotary evaporator and transferred to a heavy-walled Schlenk flask. To
this flask were added 4-bromobenzonitrile (1.09, 5.99 mmol), copper-
(I) iodide (42 mg, 0.221 mmol), triphenylphosphine (0.210 g, 0.802
mmol), bis(triphenylphosphine)palladium(II) chloride (42 mg, 0.060
mmol), and triethylamine (50 mL). The flask was capped with a septum,
evacuated, back-filled with nitrogen, sealed with a Teflon cap, and
stirred at 80-90 °C for 12 h. The reaction mixture was then allowed
to cool to room temperature and filtered through a glass funnel. The
filtrate was evaporated under reduced pressure on a rotary evaporator
and the residue purified by column chromatography on silica gel
(chloroform) to yield 5 as a white solid (0.34 g, 63%): 1H NMR (360
MHz, CDCl3) δ 7.70 (s, 2H), 7.68 (d, J ) 8.6 Hz, 4H), 7.66 (d, J )
8.5 Hz, 2H), 7.61 (d, J ) 8.6 Hz, 4H), 7.59 (d, J ) 8.7 Hz, 2H), 4.38
(t, J ) 6.3 Hz, 2H), 1.84 (m, 3H), 1.36 (m, 4H), 0.87 (t, J ) 7.3 Hz,
3H); 13C NMR (75 MHz, CDCl3) δ 161.8, 137.4, 132.2, 132.1, 132.0,
127.7, 118.4, 117.9, 117.4, 112.2, 112.0, 92.9, 91.4, 88.8, 88.5, 75.2,
30.4, 28.5, 22.7, 14.2; MS (EI, 70 eV) m/z 540 (10), 470 (42), 469
(100), 440 (10); IR (KBr, cm-1) 2931, 2227, 1602, 1499, 1442, 1236,
1175, 1113, 837; HRMS (EI, 70 eV) calcd 539.1998, found 539.1990.
2.04-1.98 (m, 1H), 1.90-1.84 (m, 1H), 1.82-1.68 (m, 2H), 1.16 (s,
3H), 0.87 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 161.9, 137.5, 127.2,
118.5, 118.4, 117.9, 117.3, 112.2, 112.0, 93.0, 91.5, 88.8, 88.6, 78.5,
42.8, 41.1, 39.5, 36.3, 26.9, 24.3, 23.9, 20.5, 18.4; IR (KBr, cm-1
)
2923, 2806, 2227, 2213, 1603, 1499, 1440, 1405, 1374, 1234, 1175,
1116, 1103, 989, 836, 756, 552.
1,3,5-Tris(4-butadiynylbenzonitrile)benzene (8). A Schlenk flask
equipped with a magnetic stirrer was charged with 4-bromoethynyl-
benzonitrile (0.77 g, 3.73 mmol) and ethanol (40 mL). To this rapidly
stirring mixture were dropwise added a solution of NH2OH‚HCl (62
mg, 0.90 mmol), n-butylamine (1.68 g, 23.0 mmol), copper(I) chloride
(1.0 g, 0.01 mmol), and 1,3,5-triethynylbenzene (186 mg, 1.24 mmol)
in ethanol (5 mL) at room temperature. This flask was then evacuated,
back-filled with nitrogen, capped with a Teflon cap, and stirred at room
temperature. After 2 h, the precipitate was filtered and chromatographed
on silica gel (CH2Cl2) to yield 8 as a yellow solid (128 mg, 20%): 1H
NMR (300 MHz, CDCl3) δ 7.67 (s, 3H), 7.64 (m, 12H); 13C NMR (90
MHz, CDCl3) δ 136.9, 133.2, 132.3, 126.4, 123.1, 118.3, 113.1, 81.0,
80.8, 75.5; IR (KBr, cm-1) 3057, 2924, 2228, 2214, 1602, 1457, 1173,
1103, 836; HRMS (EI, 70 eV) calcd 525.1266, found 525.1260.
1,3,5-Tris(4-(4-ethynylbenzonitrile)phenyl)benzene (9). A heavy-
walled Schlenk flask equipped with a magnetic stirrer was charged with
1,3,5-tris(4-bromophenyl)benzene (700 mg, 1.3 mmol) 4-ethynylben-
zonitrile (10.6 g, 4.7 mmol), copper(I) iodide (2.5 mg, 0.15 mmol),
triphenylphosphine (34 mg, 0.13 mmol), bis(dibenzylideneacetone)-
palladium(0) (14 mg, 26 µmol), and triethylamine (30 mL). The flask
was capped with a septum, evacuated, back-filled with nitrogen, sealed
with a Teflon cap, and stirred at 70 °C for 16 h. The reaction mixture
was then allowed to cool to room temperature and filtered through a
glass funnel. The filtrate was evaporated under reduced pressure on a
rotary evaporator and the residue purified by column chromatography
on silica gel (CH2Cl2) to yield 9 as a yellow solid (0.33 g, 37%): 1H
NMR (300 MHz, CDCl3) δ 7.85 (s, 3H), 7.76-7.74 (m, 24H); 13C
NMR (90 MHz, CDCl3) δ 141.7, 141.3, 132.1, 127.4, 125.4, 121.7,
118.5, 111.6, 93.5; IR (KBr, cm-1) 3039, 2224, 2217, 1599, 1515, 1017,
832; HRMS (EI, 70 eV) calcd 681.2205, found 681.2177.
1,3,5-Tris(4-(4-ethynylbenzonitrile)benzoethylnyl)benzene (10). A
heavy-walled Schlenk flask equipped with a magnetic stirrer was
charged with 1-trimethylsilylethynyl-4-((4-cyanophenyl)-ethynyl)ben-
zene (500 mg, 1.7 mmol), potassium carbonate (0.12 g, 0.80 mmol),
and a methanol-dichloromethane (1:3) solution (20 mL) and then the
mixture was stirred under nitrogen for 3 h. This mixture was transferred
to another heavy-walled Schlenk flask equipped with a magnetic stirrer
which was then charged with 1,3,5-tribromobenzene (153 mg, 0.48
mmol), copper(I) iodide (1.0 mg, 5.3 µmol), triphenylphosphine (12.6
mg, 48 µmol), bis(dibenzylideneacetone)palladium(0) (5.2 mg, 9.7
µmol), and triethylamine (10 mL). The flask was capped with a septum,
evacuated, back-filled with nitrogen, sealed with a Teflon cap, and
stirred at 70 °C for 16 h. The reaction mixture was then allowed to
cool to room temperature and filtered through a glass funnel. The filtrate
was evaporated under reduced pressure on a rotary evaporator and the
residue purified by column chromatography on silica gel (CH2Cl2) to
yield 10 as a yellow solid (148 mg, 41%): 1H NMR (300 MHz, CDCl3)
δ 7.66 (s, 3H), 7.62-7.56 (m, 12H), 7.53 (s, 12H); 13C NMR (90 MHz,
CDCl3) δ 134.2, 132.0, 131.7, 131.6, 127.8, 123.8, 123.3, 122.4, 118.3,
111.7, 93.1, 90.2, 89.9, 89.6; IR (KBr, cm-1) 2226, 2217, 1658, 1600,
1579, 1103, 835; HRMS (EI, 70 eV) calcd 753.2205, found 753.2184.
Preparation of Microcrystalline Powders 1‚, 2‚, 3‚, 4‚, 5‚, 7‚, 8‚,
9‚, and 10‚AgOTf. The organic ligand (40 mg, 0.1 mmol) was
dissolved in benzene (10 mL) and added to a solution of silver(I)
trifluoromethanesulfonate (25 mg, 0.1 mmol) in benzene (10 mL) in a
vial sealed with a Teflon-lined screw cap. The vial was heated in a
programmable furnace equipped with a Eurotherm temperature control-
ler from 1 to 100 °C, held at 100 °C for 60 min, and then cooled to
room temperature at 0.02 °C/min. The resultant crystalline powder was
composed of particles typically 25 µm × 10 µm × 10 µm as determined
by optical microscopy.
(2,4,6-Tris(4-ethynylbenzonitrile)phenoxy)ethanol (6). A Schlenk
flask equipped with a magnetic stirrer was charged with 18 (1.00 g,
2.35 mmol) and tetrahydrofuran (50 mL). The mixture was capped with
a septum, evacuated, back-filled with nitrogen, and cooled to 5 °C in
an ice bath. A 1.0 M solution of tetrabutylammonium fluoride in 9:1
tetrahydrofuran-water (8.3 mL, 8.3 mmol) was added rapidly Via
syringe. After 5 min the solution was concentrated to near dryness under
reduced pressure on a rotary evaporator, and the residue was eluted
through a plug of silica gel with chloroform. The filtrate was evaporated
to dryness under reduced pressure on a rotary evaporator and transferred
to a heavy-walled Schlenk flask. To this flask were added 4-bromoben-
zonitrile (1.71, 9.40 mmol), copper(I) iodide (49 mg, 0.257 mmol),
triphenylphosphine (0.200 g, 0.763 mmol), bis(triphenylphosphine)-
palladium(II) chloride (49 mg, 0.070 mmol), and triethylamine (50 mL).
The flask was capped with a septum, evacuated, back-filled with
nitrogen, sealed with a Teflon cap, and stirred at 80-90 °C for 12 h.
The reaction mixture was then allowed to cool to room temperature
and filtered through a glass funnel. The filtrate was evaporated under
reduced pressure on a rotary evaporator and the residue purified by
column chromatography on silica gel (chloroform) to yield 5 as a white
solid (0.30 g, 25%): 1H NMR (360 MHz, CDCl3) δ 7.73 (s, 2H), 7.71-
1
7.61 (m, 12H), 4.50 (t, J ) 4.2 Hz, 2H), 3.99 (t, J ) 4.2 Hz, 2H); H
NMR (300 MHz, acetone-d6) δ 7.89-7.77 (m, 14H), 4.55 (t, J ) 4.8
Hz, 2H), 4.02 (t, J ) 4.8 Hz, 2H); 13C NMR (90 MHz, CDCl3) δ 160.9,
137.4, 132.4, 132.3, 132.2, 127.5, 127.2, 118.8, 118.5, 118.4, 117.7,
112.5, 112.2, 93.6, 91.1, 88.9, 88.1, 76.4, 62.1; IR (KBr, cm-1) 3428,
3066, 2228, 1603, 1501, 1460, 1443, 1406, 1242, 1117, 1106, 1071,
883, 837; HRMS (EI, 70 eV) calcd 513.1477, found 513.1460.
(1S,2S,5S)-Myrtanoxy-2,4,6-tris(4-ethynylbenzonitrile)benzene (7).
A Schlenk flask equipped with a magnetic stirrer was charged with 14
(1.00 g, 2.35 mmol), 4-ethynylbenzonitrile (1.3 g, 10.20 mmol), copper-
(I) iodide (30 mg, 0.15 mmol), triphenylphosphine (170 mg, 0.65
mmol), bis(dibenzylideneacetone)palladium(0) (90 mg, 0.17 mmol), and
triethylamine (30 mL). The flask was capped with a septum, evacuated,
back-filled with nitrogen, sealed with a Teflon cap, and stirred at 50
°C for 12 h. The reaction mixture was then allowed to cool to room
temperature and filtered through a glass funnel. The filtrate was
evaporated under reduced pressure on a rotary evaporator and the
residue purified by column chromatography on silica gel (chloroform)
to yield 7 as a yellow solid (1.2 g, 53%): [R]23D ) -7.22° (c ) 0.024
in CHCl3); 1H NMR (360 MHz, CDCl3) δ 7.71 (s, 2H), 7.69-7.69 (m,
12H), 4.21-4.12 (m, 2H), 2.59 (m, 1H), 2.19 (t, J ) 5.2 Hz, 1H),
Preparation of Microcrystalline Powders 6‚ and 19‚AgOTf. The
organic ligand (40 mg, 0.1 mmol) was dissolved in benzene (10 mL)
and acetone (1 mL), and then the solution was added to a solution of
silver(I) trifluoromethanesulfonate (25 mg, 0.1 mmol) in benzene (10

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8209
mL). The vial was left undisturbed and exposed to the atmosphere to
allow escape of solvent. Crystalline precipitate typically formed after
24 h.
Preparation of X-ray-Quality Single Crystals of 5‚AgOTf. A
solution of silver(I) trifluoromethanesulfonate (0.9 mg, 25 µmol) in
benzene (1.5 mL) was added to a solution of 5 (2 mg, 3.5 µmol) in
benzene (1.5 mL) in a clean vial. A white precipitate formed
immediately, which was then dissolved by adding acetonitrile (0.3 mL).
Slow evaporation of the solvents yielded six single crystals suitable
for X-ray data collection after seven days.
Preparation of X-ray-Quality Single Crystals of 9‚AgOTf. A
solution of silver(I) trifluoromethanesulfonate (0.05 mg, 0.2 µmol) in
benzene (1.5 mL) was added to a solution of 9 (0.13 mg, 0.2 µmol) in
benzene (1.5 mL) in a clean vial. Slow evaporation of the solvent
yielded three single crystals suitable for X-ray data collection after 10
days.
Preparation and Vapor Exchange of Apohosts 4‚, 5‚, and
7‚AgOTf. The crystals grown in benzene were first filtered until the
supporting filter paper was slightly damp. The sample was then
immediately placed in a thermogravimetric analyzer (TGA) and heated
from room temperature to 200 °C at 10 °C/min under nitrogen to void
the solid of initially included solvent (4‚AgOTf was heated only to
120 °C). The apohost was then suspended above benzene, taking care
to maintain spatial separation between host and guest, and then sealed
in a vial. The vial was heated to 40 °C for 20 h. Powder patterns of
these materials were obtained from special glass capillary tubes.
Solid-State Esterfication of 6‚AgOTf. 6‚AgOTf crystals grown in
benzene were gravity filtered until the supporting filter paper was
slightly damp. A Schlenk tube was prepared with trifluoroacetic
anhydride at its bottom and a loose glass wool plug in its neck. The
sample was immediately transferred onto this glass wool. The tube was
then exposed to vacuum for 10 s before being sealed. After 1 h the
tube was pumped down to an ultimate pressure of 15 mTorr to remove
excess anhydride. Following this same procedure, the sample was
further allowed to react, this time though in a Schlenk tube charged
with benzyl alcohol. In this latter procedure the tube was initially
exposed to vacuum for 15 min rather than 10 s. The final sample was
analyzed by XRD, IR, and 1H NMR: 1H NMR (300 MHz, acetone-d6)
δ 7.89-7.77 (m, 14H), 4.93 (s, 4H); IR (KBr, cm-1) 3444, 2232, 1788,
1604, 1501, 1409, 1222, 1165, 1037.
Figure 2. View down the a axis of (a) 1‚AgOTf and (b) 5‚AgOTf
showing the pseudohexagonal channels. Solvent molecules and triflate
ions have been removed for clarity. (c) A single net of 1‚AgOTf oriented
to show its three-dimensional character. (d) Two eclipsed nets of 5‚
AgOTf.
Scheme 1
Trifluoroacetic ((2,4,6-Tris(4-ethynylbenzonitrile)phenoxy))ethyl
Ester (19). 6 (5.8 mg, 0.01 mmol) in CH2Cl2 (3 mL) was added to
trifluoroacetic anhydride (1.05 g, 5 mmol). This solution was stirred
at room temperature for 10 min and then evacuated to dryness to yield
19 as a white solid (100% by thin layer chromatography): 1H NMR
(300 MHz, CDCl3) δ 7.73 (s, 2H), 7.70-7.60 (m, 12H), 4.76 (dd, J )
3.3, 2.2 Hz, 2H), 4.69 (dd, J ) 3.3, 2.2 Hz, 2H); 1H NMR (300 MHz,
acetone-d6) δ 7.89-7.75 (m, 14H), 4.90 (s, 4H); IR (KBr, cm-1) 3431,
2228, 1788, 1603, 1501, 1402, 1221, 1155, 837; MS (FAB) m/z (M+)
609 (14).
Results and Discussion
Functionalized Structures 2-7‚ and 19‚AgOTf. In Scheme
1 we show the general synthetic route to molecules 1-7.75
Preparation of powder crystalline samples of 1-7‚ and 19‚
AgOTf and single crystals of 5‚AgOTf and 9‚AgOTf is further
described in the previous section. We begin our discussion of
these systems with the crystal structure of 5‚AgOTf, illustrated
in Figure 2b. A summary of parameters for the X-ray data
collection and subsequent refinement is given in Table 1. We
compare this structure to the previously reported14 structure of
1‚AgOTf illustrated in Figure 2a. In both structures each silver
ion is coordinated to three nitrogen atoms in a trigonal planar
environment while each organic molecule is in turn coordinated
to three silver ions again in a trigonal planar manner. Silver-
nitrogen distances range from 2.26 to 2.12 Å for 1‚AgOTf and
from 2.24 to 2.21 Å for 5‚AgOTf. The triflate ions remain near
the silver cation. In the compound 1‚AgOTf, oxygen atoms of
one of the triflate anions form relatively weak bonds with a
Ag-O bond distance of 2.4 Å while the other triflate anion
forms no oxygen-silver bonds. In 5‚AgOTf, due to crystalline
disorder, it proved impossible to locate the oxygen atoms of
the triflate anion. However, the more heavy sulfur ion could be
located. One sulfur atom lies 3.32 Å from a Ag ion while the
other sulfur atoms are 5.20 Å away from silver ions. By
(75) For 7, the starting material was the triiodide (1S,2S,5S)-Myrtanoxy-
2,4,6-triiodobenzene, and for 3 and 7, the trihalides were directly reacted
with the 4-ethynylbenzonitrile to make the desired final products. See the
Experimental Section for specific reactions.

8210 J. Am. Chem. Soc., Vol. 121, No. 36, 1999
Kiang et al.
Table 1. Crystal Data and Structure Refinement for 5‚AgOTf
formula
formula weight
T, K
wavelength, Å
crystal system
space group
a, Å
C39H25Ag1F3N3O4S1
748.55
173(2)
0.71073
monoclinic
Am
3.7528(8)
22.259(5)
32.295(7)
90
92.65(3)
2694.8(10)
2
b, Å
c, Å
R ) γ, deg
â, deg
V, Å3
Z
F
calcd, g/cm3
0.923
0.445
absorption coefficient, mm-1
θ range for data collection, deg
limiting indices
1.83-20.30
-4 e h e 4, -9 e k e 24,
-20 e l e 33
2575/11/152
2873/2757
no. of data/restraints/parameters
no. of reflections collected/unique
reflections
Figure 3. (a) Three-dimensional silicon network in ThSi2 oriented to
show its pseudohexagonal channels. (b) Two-dimensional boron net
in AlB2.
absorption correction
goodness-of-fit on F2
R1a
SADABS
1.765
0.117 (I > 2σ(I))
0.2887 (I > 2σ(I))
wR2a
In both crystal structures shown in Figure 2, the hexagonal
channels do not have perfect hexagonal symmetry. This can be
seen by examining the ratio of the c and b axes. In the case of
a perfect hexagonal channel placed into an orthorhombic cell
c/b ) x3 ) 1.71.76 In 1‚AgOTf this ratio is elongated while in
5‚AgOTf it is shortened to c/b values of, respectively, 2.029
and 1.491. The structural cause for these deviations may be
understood by examining Figure 2a,b. In Figure 2a one of the
three arms of molecule 1 lies directly along the c cell axis while
the other two arms are rotated in such a way as to shorten the
b axis from its maximal possible value, hence shifting the c/b
ratio to larger values. Similarly in Figure 2b the honeycomb
sheets are rotated out of the bc plane. The axis of rotation can
be taken to be the a axis itself. Hence, the b axis length
corresponds to the ideal hexagonal cell axis length while the c
axis is shortened with respect to the ideal value. In 5‚AgOTf
the c/b ratio therefore is smaller than the ideal c/b ratio.
Of particular importance are the c axis length of the ThSi2
type structure and the b axis length of the AlB2 structure. This
is so these lengths can be directly compared to distances
calculated from known bond distances. Assuming standard
tabulated bond lengths77 and a Ag to N distance of 2.21 Å, one
finds a distance of 13.02 Å between the center of the molecules
1-7 and the adjacent silver ions. For the ThSi2 structure, we
find that the c axis is invariant and should be 3 times this
distance or 39.06 Å while for AlB2 the b axis is also invariant
and should be x3 times this distance or 22.55 Å. These
theoretical values compare well with the single-crystal lattice
parameters of, respectively, 38.75 and 22.26 Å for 1‚AgOTf
and 5‚AgOTf. Such comparisons will be used in this paper for
systems for which there are only powder data and for which
one wishes to distinguish between the AlB2 and ThSi2 structural
alternatives.
a R1 ) ∑||Fc| - |Fo||/∑|Fo|, wR2 ) [∑[w(Fo2 - Fc2)]2/∑[w(Fo )2]]1/2
.
2
comparison, in 1‚AgOTf‚benzene Ag-S distances are 3.55 and
5.00 Å. The similarity of the shorter of the two sulfur-silver
distances suggests that one of the oxygen atoms on some triflate
ions is coordinated to a silver atom.
The probable proximate cause for the crystalline disorder in
5‚AgOTf is that the molecule 5 itself has lower point group
symmetry than the D3h symmetry of molecule 1. In particular
the pentoxy group is covalently attached to only one of the three
possible sites on the central benzene ring. In the solved single-
crystal structure model this pentoxy group is found to be
disordered with an equal distribution of three possible orienta-
tions of this molecule. Due to this disorder only the oxygen
atom of the pentoxy group could be located as may be seen by
inspection of Figure 2a.
The overall crystal structures shown in Figure 2a,b are related
to one another. Passing through the center of the unit cell in
both figures are slightly irregular hexagonal channels. In both
structures, there are two such hexagonal channels per unit cell
running along the a axis. Despite the similar coordination
environments of the two structures and the similarity of their
pseudohexagonal channels, the structures represented in Figure
2a,b are quite different. This is most clearly seen in Figure 2c,d.
In these two illustrations we have emphasized the overall
topology of the network connectivity in 1‚AgOTf and 5‚AgOTf.
In Figure 2d one sees that the trigonal planar coordination leads
to planar honeycomb sheets which stack one on top of each
other in a nearly eclipsing fashion along the a axis. By contrast,
the crystal structure of 1‚AgOTf is three-dimensional with half
the trigonal planes oriented one way and half another way.
Both network types illustrated in Figure 2c,d correspond to
well-known inorganic crystal structure types. In Figure 3 we
illustrate the known network formed by the boron atoms in AlB2
and by the silicon atoms in ThSi2. It may be seen that, just as
in 5‚AgOTf, the boron atoms of AlB2 form into planar
honeycomb sheets which lie eclipsed to each other. By contrast
the silicon atoms of ThSi2 form a three-dimensional array like
the one found in 1‚AgOTf. Following standard practice where
network types are named after original members of the structural
families, it seems appropriate to term the structures of 1‚AgOTf
and 5‚AgOTf as, respectively, the ThSi2 and AlB2 types.
We now turn to the other substituted versions of 1‚AgOTf‚
benzene, to 2-7‚ and 19‚AgOTf. As Figure 1 shows this
includes a range of substituents ranging from both saturated
and unsaturated hydrocarbon moieties to ether, alcohol, and ester
groups. Of particular interest is 7, a chiral substituent, which
(76) In this paper the a cell axis is pointed consistently in the direction
of the channels. Were the structure truly hexagonal, the a axis would then
be (unconventionally) the unique axis.
(77) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley: New York,
1992; p 21.

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8211
will perforce lead to a chiral crystal structure. As we were unable
to prepare crystals suitable for single-crystal X-ray diffraction,
we consider instead X-ray powder data. We find in these systems
that the powder data give strong insight into the actual crystal
structures. An illustration of these powder data is given in Figure
4 for 1‚AgOTf‚toluene, 1‚AgOTf‚benzyl alcohol, 5‚AgOTf, and
6‚AgOTf. Also shown in Figure 4 is the calculated powder
pattern based on the single-crystal data for 1‚AgOTf‚benzene.
As may be seen, not only do the various powder patterns all
look rather similar to one another, there is also excellent
agreement between the observed powder patterns and those
calculated from the single-crystal models.
The refined cell constants for 2-7‚ and 19‚AgOTf are given
in Table 2. In Table 2, we report only values for c, b, and R.
This is due to the high Debye-Waller factors in these systems.
Due to these high values the decay in intensity as a function of
θ is significant, as may be seen in Figure 4. Recalling that for
5‚AgOTf the a cell axis is only 3.75 Å, this decay means that
while we generally have a reasonable number of reflections
(10-20 reflections) to index the larger c and b axes, there are
only two to three reflections useful for determining a, â, and
γ. Hence, only values for the former lattice parameters are
reliable.
It may be seen that the cell axes in Table 2 separate into
three somewhat distinct families. In the first family b and c
axes are, respectively, 19-20 and 38-39 Å while for the second
and third family b is 22.2-22.7 Å. The distinction between the
second and third families is in their c axes. In the former
compounds c axes are 33-34 Å while for the latter systems
they are 35-36 Å. Thus, while the division between the type
one compounds and the others is quite sharp, that between types
two and three is somewhat diffuse.
Given the agreement of both the intensities of the observed
reflections and the near equivalence of the cell parameters, it is
reasonable to expect that all type one compounds are isotypic
with the ThSi2 structure reported for 1‚AgOTf‚benzene. Simi-
larly, agreement in both reflection intensities and cell parameters
suggests that all type two systems are isotypic with the AlB2
structure reported for 5‚AgOTf. For type three compounds it
would appear the structure is also of AlB2 type, but where the
rotation of the honeycomb sheets out of the bc plane is reduced
with respect to those found for type two.
One curious point is the strong similarity between the powder
patterns of the AlB2 and ThSi2 structures. For example, there
is a marked similarity between the ThSi2 pattern shown in Figure
4d and the AlB2 patterns shown in Figure 4a-c. The source of
this similarity may be traced to the similarities in their actual
crystal structures. In particular, the data observed in the various
powder patterns are 0kl reflections. The 0kl data set corresponds
in real space to the projected electronic density onto the bc plane.
As may be seen in Figure 2a,b, it is exactly under this projection
that the two crystal types most resemble one another. The
observed powder pattern can be taken to be the characteristic
pattern for pseudohexagonal channels.
Figure 4. X-ray powder patterns for (a) 1‚AgOTf‚benzyl alcohol, (b)
6‚AgOTf‚benzene, (c) 5‚AgOTf‚benzene, (d) 1‚AgOTf‚toluene, (e) 8‚
AgOTf‚benzene, (f) 9‚AgOTf‚benzene, and (g) 10‚AgOTf‚benzene and
(h) the calculated powder pattern for the 0kl reflections of 1‚AgOTf‚
benzene based on the single-crystal model.
It may also be observed that in one respect the AlB2 and
ThSi2 structures differ from one another. Both are distortions
of the ideal hexagonal cell axes. Hence, reflections which are
degenerate in θ value for a hexagonal cell split into two distinct
powder pattern peaks. As noted above, the distortion is an
elongation of c/b for ThSi2 but a diminution for AlB2. However,
as both this elongation and this diminution are roughly equal
displacements from the ideal value, the pairs of reflections
appear roughly equally spaced but in reversed order in the two
different structural families. For example, the pair of reflections
observed at 2θ ) 5° in Figure 4 correspond to the 002 and 011
reflections. In the ThSi2 system the 002 reflection is at the lower
2θ value while for the AlB2 type it is the higher 2θ reflection.
One final point of interest is that one can prepare 1‚AgOTf
in both the ThSi2 and AlB2 structure types. If crystals of 1‚

8212 J. Am. Chem. Soc., Vol. 121, No. 36, 1999
Kiang et al.
Table 2. Cell Constants for Phenylacetylene‚AgOTf Crystalsa
compound
b (Å)
c (Å)
â (deg)
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
85.75(2)
86.87(2)
90
90.0(3)
90
c/b
structure type
ref
1‚AgOTfb‚benzene
1‚AgOTf‚toluene
1‚AgOTf‚m-xylene
7‚AgOTf‚apohost
1‚AgOTf‚benzyl alcohol
1‚AgOTfc‚benzene
1‚AgOTf‚phenol
2‚AgOTf‚benzene
3‚AgOTf‚benzene
5‚AgOTf‚benzene
5‚AgOTfb‚benzene
6‚AgOTf‚benzene
7‚AgOTf‚benzene
19‚AgOTf‚benzene
19‚AgOTfd‚benzene
19‚AgOTfe‚benzene
1‚AgOTf‚apohost
4‚AgOTf‚benzene
5‚AgOTf‚apohost
8‚AgOTff‚benzene
8‚AgOTff‚benzene
9‚AgOTf‚benzene
9‚AgOTfg‚benzene
10‚AgOTf‚benzene
19.110(7)
19.29(2)
19.58(3)
20.08(2)
22.43(1)
22.385(2)
22.48(2)
22.58(3)
22.70(2)
22.64(2)
22.259(5)
22.63(1)
22.23(2)
22.57(2)
22.37(1)
22.47(2)
22.77(2)
22.17(6)
22.98(8)
24.128(5)
26.94(5)
29.106(3)
30.327(6)
34.103(4)
38.856(15)
39.91(15)
38.50(5)
37.9(4)
2.03
2.07
1.97
1.89
1.51
1.51
1.52
1.48
1.47
1.47
1.45
1.46
1.52
1.49
1.51
1.51
1.60
1.58
1.54
1.92
1.54
1.67
1.52
1.55
ThSi2
ThSi2
ThSi2
14
14
14
h
14
h
14
h
h
h
h
h
h
h
h
h
h
h
h
h
h
33.98(2)
33.828(7)
34.08(6)
33.42(5)
33.37(3)
33.43(2)
32.295(7)
33.23(3)
33.82(6)
33.58(3)
33.91(3)
33.95(3)
36.39(4)
35.22(8)
35.42(5)
46.441(1)
41.53(8)
48.515(7)
46.214(7)
52.751(6)
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
AlB2
h
h
h
a Unless otherwise specified, cell constants are from powder data. b Single-crystal data. c From the solid-state exchange of benzene for benzyl
alcohol by placing 1‚AgOTf‚benzyl alcohol in liquid benzene. d From the solid-state reaction of crystalline 6‚AgOTf‚benzene with trifluoroacetic
anhydride vapor. e From the solid-state reaction of crystalline 6‚AgOTf‚benzene with trifluoroacetic anhydride vapor, followed by introduction of
benzyl alcohol vapor. f Two different A-centered pseudorectangular cells corresponding to the same oblique Bravais lattice (b ) 25.362(4) Å, c )
24.128(5) Å, R ) 114.07(2)°). g Pseudorectangular supercell of the reported P1h Bravais lattice (from single-crystal data). h This work.
Scheme 2
AgOTf-ThSi2 type are placed at room temperature in a benzyl
alcohol solution, the crystal cell parameters evolve into the AlB2
type. Further, if these then altered benzyl alcohol crystals are
now placed at room temperature back into benzene solution,
the unit cell parameters remain in the honeycomb type.
AgOTf. The above data demonstrate the stability of the
hexagonal channel structure subject to modifications which
change the functional groups in the interior of these channels.
We have also studied a series of molecules in which the original
3-fold symmetry of the molecule 1 has been maintained but
where the distance between the centered benzene ring and the
peripheral nitrile groups has increased from 13.02 to 19.96 Å.
The molecules prepared include 8-10. The overall synthetic
route to these compounds is shown in Scheme 2. In each case
we could prepare microcrystalline samples of the organic
molecule together with silver triflate. For 9 it proved possible
to find a crystal suitable for a single-crystal X-ray data
collection. Parameters relevant to the single-crystal data col-
lection and structure refinement are given in Table 3.
1
However, H NMR studies show that there is a quantitative
replacement of the benzyl alcohol molecules by benzene. Thus,
one can prepare 1‚AgOTf‚benzene in two different near
hexagonal channel structure types. The observed unit cells are
given in the first and sixth lines of Table 2. As the cell constants
differ by 3-5 Å, it is clear that we have prepared two different
forms of the same compound. These results indicate that Ag-
nitrile bonds both break and re-form at room temperature in
the presence of solvent. It is interesting that optical microscope
pictures of crystals of 1‚AgOTf show no change in crystalline
morphology, other than a loss of crystal transparency.14
Structures with Enlarged Channel Diameters, 8-10‚
As in the earlier discussed crystal structures, the trigonal
organic molecule 9 is coordinated to three silver ions through
linear nitrile to silver bonds. There are two crystallographically

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8213
Table 3. Crystal Data and Structure Refinement for
9‚AgOTf‚4.5benzene
formula
formula weight
T, K
wavelength, Å
crystal system
space group
a, Å
C79H54Ag1F3N3O3S1
1290.18
173(2)
0.71073
triclinic
P1h
14.3765(11)
27.628(2)
27.648(2)
66.549(2)
84.674(3)
79.620(3)
9906.9(13)
4
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
F
calcd, g/cm3
0.865
0.264
absorption coefficient, mm-1
θ range for data collection, deg
limiting indices
0.8-17.23
-11 e h e 11, -23 e k e 16,
-23 e l e 23
11525/12/724
no. of data/restraints/parameters
no. of reflections collected/unique 21609/11525
reflections
absorption correction
SADABS
1.212
0.157 (I > 2σ(I))
0.4005 (I > 2σ(I))
goodness-of-fit on F2
R1a
wR2a
2
a R1 ) ∑||Fc| - |Fo||/∑|Fo|, wR2 ) [∑[w(Fo2 - Fc2)]2/∑[w(Fo )2]]1/2
.
inequivalent silver atoms in the structure. Both silver atoms are
bonded to three nitrogen atoms; in both cases Ag-N bond
distances range from 2.08 to 2.23 Å with Ag-N coordination
environments that are nearly trigonal planar. The N-Ag-N
bond angles for Ag1 are 105.1°, 113.8°, and 129.1° while for
Ag2 they are 112.9°, 120.1°, and 124.8°. The net sum of each
set of three angles are 348° for Ag1 and 358° for Ag2. These
net sums approach the 360° angles expected for perfectly planar
coordination. Also as in the earlier structures the silver ions
form a single weak bond to an oxygen atom of a triflate anion.
For Ag1 this bond is 2.39 Å long while for Ag2 it is 2.46 Å
long.
An illustration of the crystal structure is given in Figure 5.
As may be seen we are able to locate the first layer of solvent
atoms which are directly adjacent to the rigid phenylacetylene
host. More interior to the channel, benzene molecules proved
too solvent-like to locate by diffraction methods.
As may be inferred from Figure 5b the overall topology of
this crystal is of the AlB2 type. As in the previous AlB2
structures the planar honeycomb sheets do not lie exactly in
the bc planes but instead are rotated out of this plane using the
b axis as the axis of rotation. For this crystal the c/b ratio is
1.52, similar to those for the other previously discussed AlB2
structures. There is however one notable difference between the
AlB2 type structures reported for 5‚AgOTf and 9‚AgOTf. While
both these two structures are comprised of stacks of tilted
honeycomb sheets, in 5‚AgOTf the silver atoms of one layer
lie nearly directly above the silver atoms of the adjacent layer;
by contrast in 9‚AgOTf the silver atoms lie half the time above
silver atoms and half the time above the central benzene ring
of molecule 9. One therefore needs four layers to form a
translational repeat unit for 9‚AgOTf but only a single layer
for 5‚AgOTf. This is reflected in the a axis lengths for the two
crystals of, respectively, 14.38 and 3.75 Å.
Figure 5. Crystal structure of 9‚AgOTf. (a) Molecule 9 is shown in
green, silver in blue, and triflate ions and solvent molecules in yellow.
Channel diameters are roughly 25 Å. (b) A view down the a axis
showing the nearly eclipsed hexagonal layers. The layer on top is red,
the middle layer yellow, and the bottom layer green. Silver ions are
represented as small spheres.
Å distance between the Ag atom and the center of molecule
9’s central benzene ring. For an AlB2 structure the b axis length
would be roughly 30.0 Å (x3 × 17.3) while for the ThSi2
structure we would expect a c axis length of 51.9 Å (3 × 17.3).
The actual cell axes are b ) 30.3 Å and c ) 46.2 Å. The cell
constants clearly suggest that 9‚AgOTf forms in the AlB2
structure type. Furthermore, the c/b ratio of 1.52 suggests a tilt
angle of the honeycomb layers very similar to the tilt angle
found for 5‚AgOTf. An examination of the full single-crystal
structure shows that both surmises are correct. However, we
were not able to deduce the different stacking sequences actually
observed in 9‚AgOTf.
We now turn to powder samples of 8-10. Their powder
patterns are shown in Figure 4. In each case the resultant powder
pattern can be indexed to a cell similar in shape to those found
for 1-7. The refined cell constants are reported in Table 2. Of
the three powder patterns the one for 10‚AgOTf is most similar
to the previous powder patterns. Indeed, as Figure 4 shows,
except for the scale change in the θ angles caused by the larger
size of molecule 10, the similarity between this powder pattern
and the ones for 1-7‚AgOTf is striking. The cell axis lengths
for 10‚AgOTf confirm this similarity. If we use standard
carbon-carbon and carbon-nitrogen bond distances77 and
silver-nitrogen bond distances of 2.21 Å, we can calculate a
Ag to centered benzene ring distance of 19.9 Å. In the AlB2
We may use the solved crystal structure of 9‚AgOTf to assess
the prediction of structure type from cell constants. As before,
using literature reference values for organic bond distances and
a Ag-N bond distance of 2.21 Å, we find for 9‚AgOTf a 17.3

8214 J. Am. Chem. Soc., Vol. 121, No. 36, 1999
Kiang et al.
structure type this corresponds to a b axis length of 34.4 Å
compared to the observed distance of 34.1 Å. The c/b ratio is
1.55, well within the range of c/b of 1.47-1.59 found for the
other AlB2 structures. We therefore have good evidence
suggesting that 10‚AgOTf forms in the AlB2 structure type.
Channel dimensions have however increased from 10-15 Å
for 1‚AgOTf to 25-30 Å for 10‚AgOTf.
For 9‚AgOTf we find two unit cells, one from the powder
sample and the other from a single-crystal data set where the
crystal was grown from slow evaporation from benzene. The
cell parameters are considerably different from one another. In
the single-crystal sample, the b axis is 1.2 Å longer than that
found by the powder pattern while the c axis is 2.3 Å shorter
than that of the powder pattern. These differences are signifi-
cantly larger than those typically expected between powder and
single-crystal X-ray data. As in our earlier example where 1‚
AgOTf‚benzene formed in two different structures depending
on the route of synthetic growth, it is quite plausible that the
differences in solvent and crystal growth conditions change the
exact unit cell dimensions while leaving the fundamental powder
patterns quite similar.
Figure 6. X-ray powder patterns for 6‚AgOTf‚benzene (a) before and
(b) after exposure to trifluoroacetic anhydride and benzyl alcohol vapor.
these data and the other powder pattern data reported in this
paper suggests that their crystal structures are still of the same
overall family. As may be seen in Table 2, apohost 5‚AgOTf
and apohost 7‚AgOTf have b cell axes of 22-23 Å and c axes
of 35-36 Å. The fairly close agreement of the b axes with the
b axes of 1-7‚AgOTf‚benzene suggests, just as in those crystals
and just as in the single-crystal structure of 5‚AgOTf‚benzene,
that hexagonal sheets have formed which lie in a plane which
includes the b axis. The c axis is however roughly 2-3 Å longer
than what is found for 5‚AgOTf‚benzene. This suggests that
the structures in these systems are of the AlB2 type but with a
smaller rotation of the honeycomb sheets out of the bc plane
than that found in the single-crystal structure of 5‚AgOTf‚
benzene. For one of the unit cells, that of 7, cell parameters are
closer to those of the ThSi2 type.
Among all the powder patterns shown in Figure 4, the powder
patterns for 8‚ and 9‚AgOTf are the most different. For 9‚AgOTf
the different appearance in the powder profile can be attributed
to a difference in the c/b ratio. The observed c/b ratio is 1.67.
This is very near the ideal hexagonal c/b ratio of 1.71, and
therefore the resolution of the densitometer trace of our Gunier
film data; many nearly equivalent reflections overlap.
The powder pattern for 8‚AgOTf is the most different from
the other reported powder patterns. Whereas in the other powder
patterns reflections appear in doublets, reflections for 8‚AgOTf
are in triplets. This can be understood in the following way. In
a two-dimensional hexagonal cell the three reflections 10, 01,
and -11 have exactly equal d spacing. In rectangular distortions
of a two-dimensional hexagonal cell 10 and 01 can have exactly
equal d spacing, but -11 has a different d spacing. Hence, a
rectangular cell leads to a doubling of the original hexagonal
reflections. If we further lower the symmetry to an oblique one,
10, 01, and -11 each have different d spacings. We therefore
tried to index the 8‚AgOTf pattern to such an oblique cell. This
proves possible with concomitant good standard deviations in
the cell parameters. Two equivalent cells are presented in Table
2. By the previously described method of estimating cell
constants for an AlB2 structure, we find a putative b axis of 27
Å while for the ThSi2 structure we find a potential c axis of
46.7 Å. In Table 2 we see that in one oblique cell the b axis is
26.95 Å and in another oblique cell the c axis is 46.4 Å. It is
therefore difficult to ascribe to which structure type the 8‚AgOTf
corresponds. Nevertheless, the high agreement between dcalcd
and dobsd, as well as the overall powder pattern intensities,
suggests that some sort of pseudohexagonal channels are present
in this structure.
We and others have shown that such apohosts can reincor-
porate a number of guests.11,78-80 Rather than just repeat earlier
experiments, we investigated more stringent conditions on the
porosity of the channel structures.81 We chose to study the
reactivity of the channel structure itself. In particular 6‚AgOTf
contains an alcohol unit inside the channel. We therefore
combined a sample of 6‚AgOTf with a strong electrophile,
trifluoroacetic anhydride. While the reaction in solution between
an alcohol and an acid anhydride is straightforward and leads
to the ester 19, we need to demonstrate that such a reaction can
also take place in the solid state in good yield and with retention
of the initial crystal structure.
The design of our experiment is as follows. We first placed
a sample of crystalline 6‚AgOTf in the vapor of trifluoroacetic
anhydride (bp 39.5-40 °C). After some time we quenched the
reactions with benzyl alcohol vapor to remove excess trifluo-
roacetic anhydride. X-ray powder patterns were taken before
reaction, before addition of the benzyl alcohol vapor, and after
addition of the benzyl alcohol vapor. The cell constants after
each of these steps are shown in Table 2. As may be seen there
is little evolution of cell parameters, indicating little change in
the overall crystal structure. In Figure 6 we show the actual
powder patterns before and after reaction. Again, little change
in the diffraction pattern can be seen.
Chemical Reactions between Host and Guest. The previous
results show that channels of 10-30 Å are present in compounds
such as 1‚AgOTf‚benzene through 10‚AgOTf‚benzene. We now
turn to a few illustrative examples of the chemical transforma-
tions possible in this general family of compounds. We consider
two types of transformations. The first is that of preparing an
apohost, i.e., a material in which all the solvent has been
removed. As described in the Experimental Section, if we take
5‚AgOTf‚benzene or 7‚AgOTf‚benzene and quickly heat the
samples to 200 °C, we can show by 1H NMR that all the solvent
has been removed. The X-ray diffraction data for these systems
are given in the Supporting Information. The similarity between
Solid-state IR studies reveal however clear changes. The
original crystalline sample shows no bands between 2100 and
1650 cm-1. After addition of the anhydride but before the
(78) Kondo, M.; Okubo, T.; Asami, A.; Noro, S.; Yoshitomi, T.;
Kitagawa, S.; Ishii, T.; Matsuzaka, H.; Seki, K. Angew. Chem., Int. Ed.
1999, 38, 140.
(79) Li, H.; Davis, C. E.; Groy, T. L.; Kelly, D. G.; Yaghi, O. M. J. Am.
Chem. Soc. 1998, 120, 2186.
(80) Li, H.; Eddaoudi, M.; Groy, T. L.; Yaghi, O. M. J. Am. Chem. Soc.
1998, 120, 8571.
(81) Chui, S.; Lo, S.; Charmant, J.; Orpen, A. G.; Williams, I. D. Science
(Washington, D.C.) 1999, 283, 1148.

Porous Phenylacetylene SilVer Salts
J. Am. Chem. Soc., Vol. 121, No. 36, 1999 8215
These results suggest that the porous crystalline framework is
itself intact throughout these transformations.
Conclusion
We have examined 20 crystals in this paper. All of the 20
compounds have similar powder patterns, indicating a common
crystalline motif. The three solved single-crystal structures
clearly demonstrate that this motif is pseudohexagonal channels
which are regularly packed and parallel to one another. The
channel structure type exhibits a somewhat unexpected stability.
It is stable to modification of the chemical functional groups,
to alteration of the size of the channel, to removal of all guest
molecules within the channel, and to a covalent bond forming
reaction between guest and host. Modifications allowable
include the inclusion of ether, alcohol, unsaturated hydrocarbons,
and esters. With inclusion of chiral guests, chiral porous hosts
could be prepared with diameters ranging from 10 to 30 Å.
Figure 7. Portion of the 300 MHz 1H NMR spectra of 6‚AgOTf‚
benzene in acetone-d6 (a) before exposure to trifluoroacetic anhydride
vapor, (b) after exposure to trifluoroacetic anhydride vapor, and (c)
after exposure to trifluoroacetic anhydride and benzyl alcohol vapor.
Resonances correspond to the methylene protons of the pendant group.
In (c) the resonance marked with an asterisk is due to the benzyl alcohol
molecule.
We may compare this family of structures to the very
important inorganic porous solid MCM-41. In this latter system
one is able to prepare rather stable hexagonal channel crystal
structures with variable pore sizes ranging up to 100 Å82 in
which the channel walls are entirely composed of strong
covalent bonds. The chemical stability of the MCM-41 structure
is far superior to the stability of the structures reported in this
paper. However, the MCM-41 structure is nearly amorphous at
atomic length scales. In this paper we have seen that if we
sacrifice the strength of covalent bonds and replace them with
coordination bonds, we can prepare porous solids of some
chemical stability in a well-crystallized form which allows
structural resolution at atomic distances. A clear goal for the
future is the preparation of covalently bonded structures which
retain the structural resolution of the compounds reported in
this paper but also the chemical stability of materials such as
MCM-41.
addition of the benzyl alcohol, a new peak appears at 1788 cm-1
.
Such a peak is compatible with the ester 19. To verify this, we
prepared the ester 19 by regular solution methods. The carbonyl
peak for this sample was also at 1788 cm-1, in good agreement
with our solid-state results.
Further confirmation of the solid-state reaction was found
1
by a solution H NMR study. Here again, we took solid-state
samples before reaction, after the addition of the anhydride but
before addition of benzyl alcohol, and after addition of benzyl
alcohol. We then dissolved these solid-state samples in acetone-
d6. The NMR data show that addition of the trifluoroacetic
anhydride results in the near quantitative conversion of 6 into
1
Acknowledgment. This work was supported by the National
Science Foundation (Grant DMR-9812351). We thank Professor
D. Venkataraman for the synthetic procedure of 9.
a single new compound. The H NMR pattern of this latter
compound is in excellent agreement with that of molecule 19
prepared by solution reaction methods. The NMR shifts are quite
marked for the ethylene region near the reacting alcohol moiety
in molecule 6. We show these data in Figure 7. As a final check
that the reaction has indeed taken place, we took molecule 19
and crystallized it with AgOTf. The cell parameters for this
compound are reported in Table 2 and are in agreement with
the cell parameters of the reacted solid.
We conclude that it is possible to chemically modify the 6‚
AgOTf system with a covalent bond to an introduced guest
molecule without destroying the overall crystalline topology.
Furthermore, the solid-state transformation can be made by
introduction of a vapor at room temperature into the solid crystal.
Supporting Information Available: Tables of crystal
refinement data, bond distances, bond angles, anisotropic thermal
factors for 5‚ and 9‚AgOTf, and powder diffraction data for
compounds in Table 2, and IR spectra for 19 and 19‚AgOTf
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA991100B
(82) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck,
J. S. Nature 1992, 359, 710.