Temperature-Dependent Solvent Disruption of Guanidinium-1,5-Naphthalenedisulfonate Networks Yields a One-Dimensional Pore Structure Jason N. Voogt and Harvey W. Blanch*
CRYSTAL GROWTH & DESIGN 2005 VOL. 5, NO. 3 1135-1144
Chemical Engineering Department, University of California, Berkeley, California 94720 Received October 26, 2004
ABSTRACT: Crystallization of mixtures of the host compounds 1,5-naphthalenedisulfonate and guanidinium results in a new three-component network structure in the presence of linear fatty acid esters. The formation of this new topology is temperature-dependent and arises from the disruption of the basic guanidinium-sulfonate sheet structure through the incorporation of the solvent 2-methoxyethanol within the host network. Examination of the hightemperature crystal phase, the fatty acid ester-free 2-methoxyethanol solvate, reveals that the GS sheet topography of this network is directed by the hydrogen-bond interactions of the solvent and host. In the fatty acid ester inclusion compound, which crystallizes at lower temperatures, the host-guest interaction is much more explicit, with each 2-methoxyethanol molecule participating in three hydrogen bonds, becoming the third component of the host structure. The result of this new assembly is the creation of hydrophobically lined pores with an available cross-section area of 4.1 × 4.7 Å2 within which the linear fatty acid ester guests are contained. Hydrophobic interactions and tight host-guest packing provide an explanation for formation of the entropically unfavorable three-component host network. Introduction Nanoporous organic crystalline networks have been widely investigated in recent years for their use in optoelectronic applications, guest recognition, catalysis, and in adsorption and separation processes. Crystal engineering attempts to synthesize materials through atomistic control of molecular positions, orientations, and assembly to yield functional supramolecular networks for such applications. Most synthesis approaches for organic crystalline networks have employed hydrogen bonds to drive network formation as they provide directionality, predictable donor-acceptor pairings,1 and relatively strong interaction energies. There are now a number of systems based primarily on hydrogen bonding,2 some of which provide for network assembly in a predictable manner.3 Unfortunately, these approaches are often frustrated by ubiquitous nondirectional interactions such as van der Waals forces; these competing packing forces often result in unpredictable changes in network structure. One of the most versatile examples of a tunable hydrogen-bond network is the guanidinium organodisulfonate system developed by Ward and co-workers.4-7 These networks are characterized by pervasive hydrogenbonded sheets that form between guanidinium (G) ions and sulfonate (S) moieties. The organic “pillars” of the disulfonate ions connect adjacent sheets and provide inclusion galleries within which guests are bound, typically stabilized only through packing forces. The tunability of the system resides primarily in the choice of the organodisulfonate ion, although guest templating of alternate structures with the same pillar has also been observed.4 The large number of reported crystal structures with various organodisulfonates and wide* Corresponding author; 201 Gilman Hall, Department of Chemical Engineering University of California, Berkeley, CA 94720-1462; tel.: + 1-510-642-1387; fax: + 1-510-643-1228; e-mail:
[email protected].
ranging guest size is due, in part, to the inherent flexibility of the basic GS sheet structure. The hydrogenbonded sheets are able to fold, or pucker, to alter the cavity size and shape and optimize host-guest packing. The robustness of the GS system is evidenced by the report of more than 300 structures that conserve the GS hydrogen-bonding connectivity while varying widely in the size and configuration of the inclusion pores. While guest inclusion with the basic GS network structure alone has proven to be extremely versatile, several extensions of the system have also been reported. Ward and co-workers have employed the GS sheet structure in the absence of guest inclusion to crystallize ordered hydrogen-bonded helices through the use of a sterically bulky pillar,8 and more recently used organomonosulfonates to form crystalline inclusion compounds at the air-water interface, providing a route to the synthesis of functional monolayers that contain nonamphiphilic molecules.9 Other studies have investigated applications of GS systems, including the exploration of hydrogen-bonding competition between sulfonic and carboxylic groups,10 generation of liquid crystalline solids alkylbenzenesulfonates,11,12 as well as the incorporation of metals within the sulfonate network, for which guest-free sulfonated phosphane networks13 and an inorganic analogue of the G-organodisulfonate network using second-sphere interactions14 have been reported. Herein we present a new development of the GS network structure in which the temperature-dependent solvent disruption of the hydrogen-bonded sheet topology results in a system for the inclusion of long, linear guests. Interruption of the two-dimensional sheet connectivity has not previously been observed for inclusion networks of guanidinium disulfonates, although it has been reported with crystalline networks based on monosulfonates. Guanidinium para-substituted benzene-
10.1021/cg049636i CCC: $30.25 © 2005 American Chemical Society Published on Web 02/11/2005
1136
Crystal Growth & Design, Vol. 5, No. 3, 2005
sulfonates have been shown to disrupt the two-dimensional sheet motif in two instances: when the para substituent contains hydrogen-bonding character that competes for the hydrogen-bonding capacity of the G and S moieties,15 and with long-chain alkylbenzenesulfonates where interchain dispersive forces become significant as the alkyl chain length is increased, leading to a discontinuous GS sheet.11 The structures reported here result from similar interactions, but instead disruption results from interactions of both the solvent and the guest molecules, rather than the organic moiety of the sulfonate. Crystallization from solutions of long-chain fatty acid esters with the pillar 1,5-naphthalenedisulfonate (NDS) in 2-methoxyethanol forms a new host structure in which the solvent is incorporated within the network host structure. This disruption of the two-dimensional sheet motif yields a tubular hydrogen-bonded connectivity that exhibits narrow, hydrophobically lined pores within which long fatty acid ester guests are included, stabilized by dispersion forces. Interestingly, this new three-component host structure forms only during crystallization at low temperatures, whereas when crystallized at higher temperature, identical mixtures yield the conventional GS sheet structure of the 2-methoxyethanol solvate. The formation of alternate structures of varying composition from the same host-guest-solvent mixture is of interest in its relationship to polymorphism, and several recent studies of such behavior have been reported.16 Formation of this new network topology is explained in part due to the strong hydrogen-bonding character of 2-methoxyethanol. In the 2-methoxyethanol solvate, the solvent molecule is shown to direct formation of the fully shifted ribbon motif of the GS sheet structure in lieu of the more intrinsically stable quasihexagonal motif through the formation of discrete host-guest hydrogen bonding. 2-Methoxyethanol forms three hydrogen bonds in the new fatty acid ester clathrate structure, and this, along with the segregation of hydrophobic and hydrophilic moieties in the final structure, contributes to the formation of the three-component host topography. This connectivity no longer exhibits the flexibility of the GS sheet structure that enables it to accommodate guests of varying size. The rigid, narrow pores of this three-component host may provide new opportunities for exploring guest structure and dynamics,17 intermolecular interactions,18 and separations.19 Experimental Section General Considerations. Single crystals of G-NDS-2methoxyethanol inclusion complexes, 1, were grown at constant temperature upon mixing of solutions of guanidine hydrochloride (Aldrich, 99%) and 1,5-naphthalenedisulfonic acid tetrahydrate (Aldrich, 97%) in 2-methoxyethanol (Aldrich, 99%) at a 2:1 G:NDS ratio. Single crystals of the inclusion complexes of stearic and oleic acid 2-methoxyethyl esters, 2 and 3, were grown in the same manner, but with the fatty acid guest combined with NDS solution before addition of G. 1 H NMR spectra were recorded at 300 MHz on a Bruker Avance spectrometer. TGA was performed on a TA instruments 2950 system using ∼5 mg of crystal and heating from 25 to 900 °C at a heating rate of 2 °C/min. Ester Synthesis. Fatty acid 2-methoxylethyl esters were synthesized by the acid-catalyzed transesterification of stearic acid methyl ester (Sigma, 99%) and oleic acid methyl ester
Voogt and Blanch Table 1. Crystallographic Data for 1-3 cryst. system space group a (Å) b (Å) c (Å) R (deg) β (deg) γ (deg) V (Å3) Z temp (K) R Rw GOF
1
2
3
orthorhombic Pbam 7.8217(8) 19.228(2) 7.4380(5) 90 90 90 1118.7(5) 2 123 0.070 0.107 4.49
triclinic P1 h 7.5842(8) 7.7710(8) 13.812(1) 102.670(2) 95.096(2) 94.973(2) 786.3(1) 2 124 0.054 0.063 2.12
triclinic P1 h 7.6103(8) 7.7693(8) 13.808(1) 103.478(1) 95.362(1) 92.951(1) 788.2(1) 2 127 0.056 0.074 2.24
(Sigma, 99%). Fatty acid methyl esters were dissolved in 5% acetyl chloride (Aldrich, 98%) in 2-methoxyethanol and stirred in a nitrogen atmosphere at 90 °C for 1 h. 2-Methoxyethanol esters were extracted with hexane, which was then removed by evaporation under a nitrogen stream. Gas chromatography and 1H NMR analysis confirmed 98% conversion to the 2-methoxyethyl ester. X-ray Crystallography. Single-crystal X-ray diffraction data were collected at 124 K on a Bruker SMART CCD diffractometer with graphite monochromated Mo-KR radiation. The structures were solved by direct methods using SIR9720 and expanded using Fourier techniques with the program DIRDIF94.21 An empirical absorption correction was applied to the data using SADABS.22 Some non-hydrogen atoms were refined anisotropically, while the rest were refined isotropically. Hydrogen atoms were included but not refined. The measurement conditions and structural details are provided in Table 1. Molecular Graphics. Molecular graphics and volume calculations were performed using MSI Cerius2 v.4.9. Host volumes (Vhost) were calculated by subtracting the “available volume” (probe radius ) 0.5 Å) of the unit cell (after removal of the guest) from the unit cell volume, Vcell. Guest volumes (Vguest) were calculated using a Connolly surface with a probe radius of zero.
Synthesis and Characterization of GS Sheet Structure Networks 2-Methoxyethanol Solvate. Crystallization of G and NDS from 2-methoxyethanol solutions at 20 °C yields colorless, bladelike crystals suitable for single-crystal X-ray structure analysis. The mixture crystallizes in the primitive orthorhombic space group Pbam, and crystallographic data are presented in Table 1. The approximate composition of the inclusion compound is G2NDS‚1.1(2-methoxyethanol) (1), and the structure is shown in Figure 1. The host composition G2NDS is exact, as confirmed by NMR and X-ray crystallography, and is consistent with the stoichiometry reported previously for guanidinium organodisulfonate networks.7 However, unlike most previously reported structures, here the guest solvate, 2-methoxyethanol, is disordered and inclusion is incommensurate. NMR and TGA of single crystals of the inclusion compound as well as X-ray diffraction suggest the approximate composition of 1.1 2-methoxyethanol for each G2NDS unit. As with other GS networks, the structure of 1 is characterized by major hydrogen-bonded ribbons that form between G protons and S oxygens. These major GS ribbons are interconnected through additional G to S hydrogen bonds to form two-dimensional sheets. Different motifs of the GS sheet structure are possible,
Guanidinium-1,5-Naphthalenedisulfonate Networks
Crystal Growth & Design, Vol. 5, No. 3, 2005 1137
Figure 1. Packing diagram of 1 viewed (a) along and (b) perpendicular to the pore. The disordered 2-methoxyethanol guest was omitted for clarity, except for the hydrogen-bonded guest oxygen shown in green in the upper channel. Host carbon, nitrogen, oxygen, and hydrogen atoms are shown in gray, blue, red, and white, respectively. Table 2. Hydrogen Bonding Detailsa for Two-Dimensional Sheet Structures7
H-bond motif
1-hex nitrile
1
MeOH
triglyme
fully shifted 2.92 177.5 2.94 168.9
fully shifted 2.92 171.0 2.92 165.2
quasi-hex
major ribbon (Dmr1 ) Dmr3) (Dmr2 ) Dmr4)
Dmr1 θmr1 Dmr2 θmr2
fully shifted 2.91 179.9 2.90 164.0
inter-ribbon (Dir1 ) Dir2)
Dir1 θir1
3.16 108.5
3.22 127.9
3.08 119.0
3.03 163.3
host-guest (Dhg1 ) Dhg2)
Dhg1 θhg1
3.05 143.6
3.09 144.3
3.04 138.8
-
34° 0.55 0.70
42° 0.55 0.68
34° 0.56 0.68
61° 0.42 0.60
θIR Pf,b guest-free Pf, with guest
2.91 177.2 2.95 167.5
aHydrogen-bond lengths are denoted in Figure 2. The methanol, triglyme, and 1-hexanenitrile structures are from Russell et al. D ) (donor)-(acceptor) distance, Å. θ ) (donor)-H‚‚‚(acceptor) angle, deg. b Pf ) packing fraction. Calculated as Pf (guest-free) ) (Vhost/ Vcell) and Pf (with guest) ) (Vhost + Vguest)/Vcell.
Figure 2. Diagram of the two-dimensional GS sheet architecture in (a) the quasihexagonal motif,7 and (b) the fully shifted ribbon motif, including the guest oxygen atom shown in green for (b). Dotted lines indicate major ribbon, interribbon, and host-guest hydrogen bonding.
resulting from translation of adjacent ribbons with respect to one another. In the most commonly occurring arrangement, the quasihexagonal motif shown in Figure 2, the hydrogen-bonding capacity of the host G and S moieties is fulfilled by providing each S oxygen with two strong (O‚‚‚H ∼ 2 Å) hydrogen bonds from the G protons. Guests are contained between the organic “spacers” of the sulfonates that project to either side of the sheet, and host-guest van der Waals interactions are enhanced by the ability of the sheet to pucker along the interconnection seam.4 1 maintains this general structure, with the naphthalene “spacer” projecting to opposite sides of the sheet on successive major GS ribbons. These ribbons pucker with θIR ) 34°, and this creates 1D channels in which the 2-methoxyethanol guest is contained, as shown in Figure 1. However, the sheet connectivity in 1 is not
the quasihexagonal motif, but is a variation of the second type of ribbon connectivity observed, the shifted ribbon motif, which is also shown in Figure 2. The shifted ribbon motif is typically characterized by a shift in adjacent ribbons that results in one strong and one weak (O‚‚‚H ∼ 2.5 Å) inter-ribbon hydrogen bond per G. However, in 1 adjacent ribbons are fully translated (half of their repeat distance), resulting in the alignment of adjacent G and S moieties. This weakens both interribbon hydrogen bonds by significantly reducing their linearity as shown in Table 2. These two weak (O‚‚‚H ∼ 2.5 Å) hydrogen bonds per G decreases the intrinsic stability of the two-dimensional hydrogen-bonded sheet structure in comparison to the quasihexagonal motif. Crystallization in the fully shifted ribbon motif is directed by host G-guest 2-methoxyethanol hydrogen bonding. Despite the disorder of the guest 2-methoxyethanol, a contact is made between the two G protons participating in the weak inter-ribbon hydrogen bonds and a partial occupancy oxygen of the 2-methoxyethanol, as shown in Figure 1, with hydrogen bonding details presented in Table 2. This interaction is possible due to the projection of alternating G N-H bond vectors toward the center of the channel in the fully shifted ribbon motif, shown in Figure 3. This results in the formation bifurcated hydrogen bonds in which the channel-projecting G protons donate to both adjacent S
1138
Crystal Growth & Design, Vol. 5, No. 3, 2005
Figure 3. Structure of NDS pores with guests removed for clarity in (a) the fully shifted ribbon motif (1), and (b) the quasihexagonal motif. Alternating G ions have been removed from the top row of (a) to show the guest-projecting N-H bond more clearly.
and guest 2-methoxyethanol oxygen atoms. Host-guest interactions of this type are not possible in the quasihexagonal motif where all G N-H bond vectors project toward adjacent S oxygen atoms and not toward the channel or guest. Figure 3 compares the structures of a single pore in the fully shifted ribbon motif (1) with the previously reported crystal structure for the inclusion of 1-hexanenitrile with NDS, which is in the quasihexagonal motif.7 In both cases, the pores have identical channel components, with aromatic naphthalene walls and rows of G ions separating adjacent pores. The difference in the pore structure is the orientation of the G molecules that results in projection of the N-H bond vectors either toward the guest as in 1, or toward the S oxygen atoms as seen in the quasihexagonal motif. 1-Hexanenitrile, with only one potential hydrogen-bond acceptor per unit of host, results in the formation of the intrinsically more-stable quasihexagonal motif, which interacts with the host only through van der Waals contacts. Guest functionality directs the formation of either of these connectivities, with the increased density of hydrogen-bond accepting atoms in 2-methoxyethanol resulting in the fully shifted ribbon motif. When formed, these host-guest hydrogen bonds provide two G protons with a second weak hydrogen bond, resulting in four strong and four weak hydrogen bonds per G. However, due to the conformationally restricted positions of the two oxygen atoms of an individual 2-methoxyethanol molecule, host-guest hydrogen bonds cannot form with every available G proton. The result is incommensurate inclusion of the guest,
Voogt and Blanch
limiting the stability imparted by the host-guest interaction. Comparison with the Previously Reported Methanol Solvate. It is instructive to compare the crystal structure of 1 with the previously reported structure of the methanol solvate of NDS,7 which also exhibits the fully shifted ribbon motif resulting from discrete hostguest hydrogen bonding. Unlike 2-methoxyethanol, however, the methanol solvate exhibits commensurate inclusion, with two methanol molecules per G2NDS, and the positions of the methanol guests are well resolved. The fixed position of the guests more clearly illustrates the existence of hydrogen bonding between the G channel-projecting protons and the guest oxygen atoms, and the hydrogen-bonding details for this solvate are also presented in Table 2. The hydrogen-bond distances and angles reveal the high degree of similarity between the two solvate structures. The most significant difference in the two solvate structures is their puckering angles, which relates to the observed difference in commensurism. The puckering angle of the methanol solvate, θIR ) 42°, allows host-guest hydrogen bonding between all of the methanol oxygen atoms and the channel projecting G protons, resulting in the observed commensurism. While a single 2-methoxyethanol molecule occupies a volume (74 Å3) comparable to two methanol molecules (68 Å3), and both have two hydrogen-bond accepting oxygen atoms, the 2-methoxyethanol solvate is unable to attain equivalent commensurate inclusion. Puckering of the hydrogenbonded sheet to smaller angles brings successive G molecules closer together, as observed in the 2-methoxyethanol solvate (θIR ) 34°), increasing hydrogen-bond formation between host and guests with narrowly spaced hydrogen-bond acceptors. However, as the sheets pucker, the inter-ribbon hydrogen bonds become weaker as they become less linear. This is reflected in the interribbon hydrogen-bond angles for the two solvates: 108.5° for the tightly puckered 2-methoxyethanol solvate, and 127.9° for the methanol solvate. With the narrowly spaced, conformationally restricted guest oxygen atoms of 2-methoxyethanol the sheets are unable to fold to provide each G with host-guest hydrogen bonding. It is the conformational freedom of the methanol molecules that allows the structure to form a commensurate inclusion compound. This decreases the lattice energy of the methanol solvate by providing all of the channel-projecting G protons with hydrogen-bond accepting guest oxygens, unlike the partial formation of host-guest hydrogen bonds found with 2-methoxyethanol. The increased stability of the methanol solvate is also supported by the higher observed temperature for crystallization of the solvate from identical host concentrations. The solvate of methanol forms at 50 °C, compared to the 2-methoxyethanol solvate which forms at 20 °C at the same host concentration. This temperature difference is more significant considering the increased entropic penalty of fixing the positions of the all of the guests in the methanol solvate compared to the enhanced conformational freedom of the 2-methoxyethanol guest molecules, some of which remain unbound to the host. Despite this slight increase in the puckering angle, the guest-free packing fractions of the two solvates are
Guanidinium-1,5-Naphthalenedisulfonate Networks
both 0.55 (the value of 0.47 cited by Russell et al.7 appears to be in error). Including the guests in the packing fraction yields a slightly larger packing fraction of 0.70 for the 2-methoxyethanol solvate compared to 0.68 for methanol. This increased packing fraction for 2-methoxyethanol provides additional van der Waals contacts to offset the decreased stability resulting from incomplete host-guest hydrogen bonding. Linear Guest Inclusion. Inclusion of short-chain linear guests (eight carbons and less) with NDS from methanolic solutions has been reported.7 These structures are characterized by the typical GS sheet structure, larger puckering angles, commensurate inclusion, and guests that are individually constrained within successive cavities along a one-dimensional pore. The only previously reported structure of NDS that exhibits incommensurate inclusion is the crystallization of the intermediate length guest triglyme from methanol, with each triglyme molecule spanning nearly two unit cells of the host. The structure exhibits the fully shifted ribbon motif of the GS sheet architecture, and, as with the methanol and 2-methoxyethanol solvates, hostguest hydrogen bonding occurs between the channel projecting G protons and partial occupancy triglyme oxygen atoms. The structure is essentially identical to the 2-methoxyethanol solvate, including the same puckering angle, θIR ) 34°. This is reflective of the structural similarity of triglyme and two 2-methoxyethanol molecules, both with two ethylene-interrupted oxygen atoms per unit of inclusion host. As with 2-methoxyethanol, however, the structural constraints of these interconnected oxygen atoms prevents complete fulfillment of the potential host-guest hydrogen bonding. Extension of the G-NDS system for the inclusion of longer chain guests with decreased host-guest interaction potential was investigated. Stearic acid methyl esters (18:0-ME) were crystallized from G-NDS solutions in methanol, and the crystals were analyzed with both NMR and TGA. For all crystals examined, the stoichiometry of the host is consistent, G2NDS as shown by NMR. The guest composition, however, varies considerably. Often, especially at higher temperatures or low fatty acid ester concentrations, only the methanol solvate is observed, G2NDS‚2(methanol). As the initial fatty acid ester concentration is increased, crystals begin to include a small quantity of the linear fatty acid ester guest. However, crystals could not be obtained where the inclusion of linear guests completely displaced the co-included methanol. The stoichiometry of the crystals that maximized 18:0-ME inclusion was G2NDS‚1.8(methanol)‚0.02(18:0-ME), obtained from solution at 20 °C, as determined by NMR of single crystals of the inclusion compound as well as by TGA. The inability to form complete inclusion compounds of the long chain guests relates to the stability of the observed structures. Mixtures of G and NDS in methanol readily form the commensurate solvate G2NDS‚ 2(methanol). There are moderate strength, pervasive hydrogen bonds between all channel-projecting host G protons and guest oxygen atoms that bind the methanol, leading to the observed commensurism and decreased lattice energy. Triglyme maintains this ability to form a crystalline inclusion compound with NDS despite
Crystal Growth & Design, Vol. 5, No. 3, 2005 1139
incomplete formation of host-guest hydrogen bonding compared to methanol. Similar crystalline networks are not observed in the presence of long chain fatty acid esters, for which only limited co-inclusion was observed. Two factors are the primary cause of the inability to form pure clathrates of the fatty acid ester guests: their lack of hydrogenbond acceptors, and their length, which precludes commensurate inclusion. These guests are more than twice as long as triglyme, but contain only two potential hydrogen-bond acceptors per molecule compared to four spatially distributed oxygen atoms in triglyme. In addition, the two oxygens of the ester moiety are closely spaced, leaving the majority of the guest molecule requiring dispersive stabilization to be included within the network pores. Short alkanenitriles have previously been shown to form inclusion compounds from methanol by dispersive stabilization alone. However, unlike longer guests, short alkanenitriles are able to be commensurately included in the quasihexagonal motif, with larger θIR angles providing for strong, linear inter-ribbon hydrogen bonds. The ability of the network to “shrinkwrap” around these shorter guests allows it to optimize packing interactions while maintaining the inherently more stable quasihexagonal motif. With longer guests, such as triglyme or fatty acid esters, commensurate inclusion in successive cavities is not possible due to the length of the guests. Puckering angles increase in order to increase the packing coefficient of the host, but this weakens the inter-ribbon hydrogen bonding. The long fatty acid ester guests, unlike triglyme, are unable to offset this loss of host GS sheet hydrogen bonding through host-guest hydrogen bonding. This, along with the inability to provide for commensurate inclusion in the quasihexagonal motif, makes the long fatty acid ester guests unable to effectively compete with the stability imparted by the strongly hydrogen bound inclusion of methanol. The low inclusion specificity for the linear guests over the competing methanol solvent would limit the possible application of this system. It was anticipated that the solvent 2-methoxyethanol, with its reduced host-guest interaction in the solvate, would provide a route to exclusive noninteracting linear guest inclusion. Synthesis and Characterization of Three-Component Host Networks Crystallization of mixtures of G, NDS, and fatty acid ester guests in 2-methoxyethanol reveal the same preference for solvate formation with increasing temperature that is observed in methanol. In the presence of stearic acid 2-methoxyethyl ester, mixtures of equivalent host concentration crystallized as the solvate 1 at temperatures in excess of 60 °C. No fatty acid esters were found in these high-temperature crystals. However, at temperatures below 25 °C, a new pure fatty acid ester clathrate crystal phase is formed from identical initial mixtures. Analysis of the low-temperature fatty acid clathrate reveals a new host architecture in which the solvent has disrupted the GS sheet architecture and has become a third component of the host structure. The effect of the solvent-host hydrogen-bonding interaction that leads to the crystallization of the fully shifted ribbon motif in the solvate here results in the complete
1140
Crystal Growth & Design, Vol. 5, No. 3, 2005
incorporation of the solvent molecules within the hydrogen-bonded network host structure. Crystallization of G and NDS from 2-methoxyethanol solutions containing stearic acid 2-methoxyethyl ester (18:0-2M) yields colorless, platy inclusion compounds of approximate composition G2(2-methoxyethanol)2NDS‚ 0.25(18:0-2M) (2). Single crystals were obtained from solution at 20 °C, and its structure was determined by X-ray crystallography. Crystallographic data are presented in Table 1, and views of the structure are shown in Figures 4 and 5. The host composition G2(2-methoxyethanol)2NDS is exact, as confirmed by both NMR and X-ray crystallography. The approximate guest occupancy of 0.25 per unit of host is suggested by NMR of single crystals, TGA measurements, and from the crystal structure of the host and disordered fatty acid guest from X-ray diffraction. The network structure of 1 and the methanol solvate is an indication of the propensity of solvents with hydrogen-bond acceptors to interact with the host more specifically than through van der Waals contacts alone. In the structure of the NDS solvates, this contributes to a change in the ribbon connectivity, forming the fully shifted ribbon motif in preference to the more commonly observed, and more intrinsically stable, quasihexagonal motif. In 2, the G-2-methoxyethanol interaction is much more explicit as each 2-methoxyethanol participates in three strong hydrogen bonds. This results in the transformation of 2-methoxyethanol from a weakly bound guest solvate in 1 to a third component of the host structure in 2. The crystals of 2 (triclinic, P1 h ) yield a new host network in which narrow 1D pores contain long, linear guests. The major GS ribbon observed in 1 and all previously reported guanidinium organodisulfonate networks is conserved, shown in Figure 4a. However, these ribbons are no longer interconnected through weak hydrogen bonds from S oxygen atoms to G protons as in 1. Instead, the G protons not participating in the major GS ribbon form strong hydrogen bonds with the oxygen atoms of 2-methoxyethanol, with hydrogenbonding distances for O‚‚‚N of 2.89 and 2.81 Å, shown in Table 3. This fulfills the hydrogen-bonding capacity of all six of the G protons, four in the major GS ribbon and two with 2-methoxyethanol. The result is the formation of a new type of GS-2-methoxyethanol ribbon, which can be described as the GS-2-methoxyethanol tape. Two adjacent GS-2-methoxyethanol tapes, related by an inversion center, are interconnected through the third 2-methoxyethanol hydrogen bond that forms between each alcohol proton and an S oxygen atom of the adjacent tape. In this configuration, each S moiety participates in five strong hydrogen bonds, a deficit of one compared to the four strong and two weak hydrogen bonds observed in the fully shifted ribbon motif. Interconnection of adjacent tapes forms GS-2-methoxyethanol tubes as shown in Figure 4b. This tube structure is similar to the box bilayer connectivity observed with alkylbenzenesulfonates,11 although here it arises with disulfonates and through the incorporation of the solvent molecule. The GS-2-methoxyethanol tubes, connected through the NDS pillars, stack to form the inclusion network
Voogt and Blanch
Figure 4. Hydrogen-bonding motif of 2. (a) The GS-2methoxyethanol tape, (b) view of GS-2-methoxyethanol tube perpendicular to the plane of the page, showing back-to-back connectivity of two GS-2-methoxyethanol tapes through 2-methoxyethanol alcohol proton-S oxygen hydrogen bonding, and (c) network connectivity resulting from NDS linkage through successive GS-2-methoxyethanol tubes. Table 3. Hydrogen Bonding Details for Three-Component Host Structuresa major ribbon
G-2-methoxyethanol
S-2-methoxyethanol NDS-NDS (DSS1 ) DSS2) Pf, no guest Pf, with guest
Dmr1 θmr1 Dmr2 θmr2 Dmr3 θmr3 Dmr4 θmr4 DGm1 θGm1 DGm2 θGm2 DSm θSm DSS1 θSS1
2
3
2.97 167.2 2.86 163.9 2.99 162.8 2.92 158.7 2.81 146.0 2.89 169.6 2.82 173.8 3.53 167.0 0.57 0.69
2.95 165.8 2.87 163.6 3.00 162.5 2.92 159.4 2.83 147.0 2.88 169.6 2.83 164.3 3.52 174.5 0.57 0.69
a Hydrogen-bond lengths are denoted in Figures 4 and 5. D ) (donor)-(acceptor) distance, Å. θ ) (donor)-H‚‚‚(acceptor) angle, deg.
shown in Figure 5, supported by weak C-H‚‚‚O hydrogen bonds between aromatic protons and sulfonate oxygen atoms of successive NDS molecules. This stacking forms bilayers of alternating G and 2-methoxyethanol molecules from the GS-2-methoxyethanol tapes, with the NDS pillars extending between adjacent bilayers. A one-dimensional channel is formed between the NDS “walls” and the G-2-methoxyethanol bilayers, in which the disordered 18:0-2M guest is contained. The alternating G and 2-methoxyethanol molecules form a nearly continuous van der Waals surface due to their close
Guanidinium-1,5-Naphthalenedisulfonate Networks
Crystal Growth & Design, Vol. 5, No. 3, 2005 1141
Figure 5. Packing diagram of 2. Disordered guests have been removed for clarity. (a) Side-view perpendicular to channel, with primary hydrogen-bonding connectivity outlined, excluding weak C-H‚‚‚O bonds, which are shown for the bottom pore walls only. (b) View down channel, with host van der Waals radii depicted. (c) Top-view illustrating rotation of NDS planes from channel axis and an expanded view of the NDS molecule illustrating O-H‚‚‚S (green) and C-H‚‚‚O (blue) hydrogen bonds.
packing, and provide a consistent available pore height of 4.7 Å. As with previous guanidinium disulfonate networks, the width of the pores is determined by the repeat distance of the GS ribbons. In 2, this is the value of the a-axis, and accounting for the van der Waals dimension of the NDS “walls” of the channel, the available pore width is 4.1 Å. Down the channel, successive NDS molecules are staggered as the aromatic planes of the NDS pillars are rotated out of parallel with the channel axis. This is due to the increased linearity of the 2-methoxyethanol-S hydrogen bond that results from this rotation, which extends the NDS oxygen atoms into the channel and toward the 2-methoxyethanol alcohol protons. In 2, the aromatic planes are rotated 5.1° out of parallel with the channel axis. This is shown in Figure 5, and the resulting O-H‚‚‚O hydrogen bond angle is 173.8°. Additionally, this twisting enhances the formation of weak C-H‚‚‚O hydrogen bonds (H‚‚‚O ) 2.6 Å) between successive NDS molecules by increasing their linearity to a hydrogen-bond angle of 167.0°.23 The result of this shifting of NDS planes are pores with consistent crosssectional area of 19.3 Å2, but slightly staggered every 7.77 Å. The approximate occupancy of 0.25 by the disordered 18:0-2M guest was determined by NMR and TGA measurements. This is consistent with occupancy calculated by assuming completely filled pores by an extended, all-trans acyl chain,24 which is suggested by the disordered guest. Although the approximate stoichiometry of the clathrate is 4(G2(2-methoxyethanol)2NDS)‚(18:0-2M), guest inclusion is incommensurate. Incommensurism is confirmed by the lack of long-range ordering observed in the guest electron density from X-ray diffraction, as well as from an analysis of chain length specificity. When complexation mixtures contain
an excess of guests of varying lengths (16-22 carbon chains), no preference is exhibited toward a particular guest, suggesting that none of these guests form commensurate clathrates. A significant difference between the three-component network structure and the GS sheet structure is the rigidity of the pores. The flexibility of the hydrogenbonded sheet puckering and pillar twisting of the GS sheet networks results in a wide range of inclusion networks formed with different pillars. This is most dramatically exhibited by the conformationally flexible pillar 4,4′-biphenyldisulfonate for which structures have been reported with over 20 different guests of varying size.5-7,25 GS-2-methoxyethanol networks are able to coinclude guests of varying chain length nonspecifically in pores identical to those in 2. However, the rigidity of the system is evidenced by the inclusion of bulkier linear guests, introduced through cis unsaturation of the acyl chain. Crystallization of G and NDS from 2-methoxyethanol solutions containing oleic acid 2-methoxyethyl ester (18:1-2M) yields colorless, platy inclusion compounds of approximate composition G2(2-methoxyethanol)2NDS‚ 0.25(18:1-2M) (3). Single crystals were obtained from solution at 0 °C, and its structure was determined by X-ray crystallography. Crystallographic data are presented in Table 1. Again, the host composition G2(2methoxyethanol)2NDS is exact, as confirmed by both NMR and X-ray crystallography. The guest occupancy is also the same as in 2, 0.25 per unit of host, determined by NMR and TGA, as well as from the crystal structure of the host and disordered fatty acid ester guest from X-ray diffraction. As evidenced by the similar unit cell dimensions in Table 1, the structure of 3 is nearly identical to 2, again resulting in one-dimensional pores with a cross-sectional
1142
Crystal Growth & Design, Vol. 5, No. 3, 2005
Voogt and Blanch
Figure 6. Dark field optical microscopy views of (a) needlelike 2-methoxyethanol solvate crystal 1 and (b) blocky stearic acid 2-methoxyethyl ester inclusion compound 2, both at 5× magnification.
area of 19.3 Å2. The hydrogen-bond geometries and packing fractions are essentially equivalent, as shown in Table 3. The most significant difference in the host structure is the increase in the staggering of the channel along the length of the pores. This results from an increase in the rotation of the aromatic planes out of parallel with the channel axis in 3 to 8.1°. The other significant difference between the two structures is the disordered guest. In 2, the guest is essentially linear, with deviations of only 0.2 Å from the channel centerline. In 3 however, the disordered guest electron density follows the staggered channel more closely, with deviations from linearity of 0.7 Å. The increase in the rotation of the NDS pillars observed in 3 accommodates the sterically rigid cis bond in the oleic acid ester guest, which the electron density suggests occur most often at the channel stagger. However, despite this slight increase in NDS twisting and channel stagger, the network maintains the same structure as 2, evidence of the rigidity of the pores. As mentioned in the discussion of the crystal structure of 2, formation of either the 2-methoxyethanol solvate crystal or fatty acid ester clathrate crystal phase is temperature dependent. At temperatures in excess of 60 °C, the crystal phase formed is exclusively the solvate crystal 1, whereas at temperatures below 25 °C, formation of the fatty acid ester crystal 2 is the only phase observed. Interestingly, in the intervening temperature range, concomitant crystallization of both phases occurs. In all cases, rapid identification of the crystal phase is simplified due to the morphological differences in the observed crystal habit, illustrated in Figure 6. The solvate crystals display a high aspect ratio, needlelike morphology due to the two-dimensional GS structure reinforcing growth along the channel axis. With the new network connectivity of the fatty acid ester clathrates illustrated in Figure 5a, the hydrogen-bond connectivity links adjacent pores, no longer reinforcing growth along the channel axis and leading to a blocky morphology with much lower aspect ratios. The same temperature-dependent behavior is observed in the presence of the monounsaturated oleic acid ester guest, albeit at lower temperatures. For identical host mixtures, exclusive formation of 3 occurs at a temperature of 0 °C or lower, whereas temperatures in excess of 25 °C leads to the formation of the solvate
Figure 7. Cross-section view of a single pore with a guest fatty acid schematically shown in an all-trans, vertical orientation with van der Waals radii. Aromatic pillar of NDS, ethyl backbone of 2-methoxyethanol, and sideways-projecting N-H vectors of G form a primarily hydrophobic-lined pore. Arrows indicate N-H hydrogen-bond projection.
crystal, again with concomitant formation of both phases in the intervening temperature range. The solvate and clathrate structures likely have similar lattice energies, suggested by their concomitant formation, similar hydrogen-bonding density, and packing fractions (Tables 2 and 3). The temperature-dependent behavior of both fatty acid ester clathrate phases is indicative the entropic penalty associated with linearization of the fatty acid esters, as well as with formation of a three-component host structure. Comparison of the inclusion of the two fatty acid esters also suggests the importance of the entropic penalty of fatty acid ester guest inclusion. Because the two clathrates have essentially the same structures, packing fractions, and thus host lattice energies, the decrease in temperature required for fatty acid ester clathrate formation can be attributed primarily to the increased entropic penalty of constraining the unsaturated cis bond of oleic acid within the narrow pores, although this is also due in part to the decrease in host-guest stabilization due to tight packing of the cis bond. An additional contribution to the observed temperature-dependent crystal phase formation and synthesis of the entropically unfavorable three-component host system is the segregation of hydrophobic and hydrophilic moieties within the crystal structure. In the structures of 2 and 3, hydrophilic moieties of the host structure are segregated within the interpore region where they participate in hydrogen bonding. Figure 7 provides a cross-section view of a single pore, demonstrating the hydrophobic environment that surrounds the fatty acid guest. The walls of the pores are framed by the aromatic rings of NDS, and the ethyl backbone of the 2-methoxyethanol molecule is rotated toward the pore and guest. The orientation of the G ion minimizes its interaction with the guest by projecting its protons
Guanidinium-1,5-Naphthalenedisulfonate Networks
toward adjacent S oxygens, identical to the motif observed for the inclusion of a noninteracting guest molecule in the two-component sheet structure, shown in Figure 3a. The contribution of the segregation of hydrophobic and hydrophilic moieties to network structure formation is similar to that observed with cyclodextrins,26 and other engineered crystalline systems.6,27 In addition to their hydrophobic character, the pore dimensions and configuration are also are well-suited to the stabilization of linear guests. Their available width of 4.1 Å, determined by the repeat distance of the major hydrogen-bonded ribbon, is consistent with the structure of 1 and all previously reported GS structures. Their 4.7 Å height is also indicated by an examination of the previous GS sheet structures of NDS, and results primarily from the intramolecular S-to-S distance in NDS. Figure 1b illustrates a cross-section view of the GS sheet structure of 1, and it shows that the available pore height varies due to the GS sheet puckering, but achieves a minimum value of approximately 5.0 Å at its narrowest point. In 2 and 3, this dimension has decreased slightly to 4.7 Å, and this results in a rectangular pore, uniform along its axis. These rectangular pores are similar in both size and shape to the values observed for closely packed crystalline alkyl chains, which have ca. 4.3 × 4.8 Å unit cell dimensions.28 However, unlike alkyl chain packing, the networks in 2 and 3 provide the potential for CH‚‚‚π interactions with the aromatic walls of the pore. It was recently demonstrated through 2D solid state NMR that these interactions stabilize the inclusion of aliphatic polymers in the slightly larger 5 Å pores of tris-(ophenylenedioxy) spirocyclotriphosphazene,29 and it is likely that these same interactions stabilize the linear guests within the narrower 4.1 × 4.7 Å2 pores of 2 and 3. Taken together, the hydrophobic environment, small pore size, and CH‚‚‚π interaction potential likely provide the additional stability for fatty acid inclusion that results in the formation of the complex three-component network observed with decreasing temperature. Summary We have expanded the potential application of GS hydrogen-bonded networks by incorporating a third component in the host structure, yielding a new narrow, one-dimensional pore architecture. Hydrogen-bond inclusion networks with the pillar NDS were shown to direct different GS sheet architectures based on the guest type. Formation of the shifted ribbon motif in preference to the quasihexagonal motif occurs when the guest contains a high density of hydrogen-bond acceptors, as with methanol, triglyme, and 2-methoxyethanol, forming explicit host-guest interactions through the formation of hydrogen bonds. This host-guest interaction stabilizes the formation of the fully shifted ribbon motif, which is intrinsically less stable than the full complement of strong hydrogen bonds formed in the quasihexagonal motif. A new hydrogen-bonding architecture forms upon the addition of long chain fatty acid esters to mixtures of G and NDS in 2-methoxyethanol. At high temperatures, this system continues to crystallize in the fully shifted ribbon motif 2-methoxyethanol solvate. However, decreasing the crystallization temperature results in the
Crystal Growth & Design, Vol. 5, No. 3, 2005 1143
formation of a new crystal phase consisting of a threecomponent host architecture in which the disordered fatty acid esters are contained within narrow, onedimensional pores. This architecture results from the interaction of 2-methoxyethanol with G and NDS, which leads to the fully shifted ribbon motif in the solvate 1, but in 2 and 3 fixes the position of solvent molecule as a third component in the host structure through the formation of three hydrogen bonds per 2-methoxyethanol molecule. Segregation of hydrophobic and hydrophilic moieties in the crystal structure also contributes to the observed formation of this three-component host architecture. The pores surround the fatty acid ester guest with the aromatic pillars of the NDS and the ethyl backbone of 2-methoxyethanol, and the cross-sectional area matches the value observed for closely packed alkyl chains.28 The resulting packing interactions in these narrow pores, including the potential for CH‚‚‚π interactions along with the hydrogen bonding in the interpore region, provide the network with sufficient lattice energy to permit its formation as the temperature and entropic penalty of the three-component network is decreased. This network structure provides one of the narrowest pore sizes observed that is still capable of including linear alkyl guests as well as slightly bulkier cis unsaturated chains. Additional investigation into the confinement of cis unsaturated guests in these narrow pores through X-ray diffraction or solid-state NMR would provide new insight into their structure and dynamics.30 Other potential applications of this new host system include investigations into intermolecular interactions through terminally functionalized linear guests, polymerization reactions, and separations of linear compounds such as unsaturated fatty acid esters. Acknowledgment. This work was supported by the Engineering Research Centers program of the National Science Foundation under award number EEC-9731725. We thank Dr. Frederick Hollander and Dr. Allen Oliver of the College of Chemistry X-ray Crystallographic Facility for the their assistance with the structure determinations. Supporting Information Available: Results of the X-ray crystallographic studies are available in the form of a crystallographic information file (CIF). This material is accessible free of charge via the Internet at http://pubs.acs.org.
References (1) Etter, M. C. Acc. Chem. Res. 1990, 23, 120-126. (2) Maspoch, D.; Domingo, N.; Ruiz-Molina, D.; Wurst, K.; Vaughan, G.; Tejada, J.; Rovira, C.; Veciana, J. Angew. Chem., Int. Ed. 2004, 43, 1828-1832; Sada, K.; Inoue, K.; Tanaka, T.; Tanaka, A.; Epergyes, A.; Nagahama, S.; Matsumoto, A.; Miyata, M. J. Am. Chem. Soc. 2004, 126, 1764-1771; Sozzani, P.; Comotti, A.; Bracco, S.; Simonutti, R. Angew. Chem., Int. Ed. 2004, 43, 2792-2797; Fournier, J.-H.; Maris, T.; Wuest, J. D.; Guo, W.; Galoppini, E. J. Am. Chem. Soc. 2003, 125, 1002-1006; Saied, O.; Maris, T.; Wuest, J. D. J. Am. Chem. Soc. 2003, 125, 14956-14957; Stanley, N.; Sethuraman, V.; Muthiah, P. T.; Luger, P.; Weber, M. Cryst. Growth Des. 2002, 2, 631-635; Carrasco, H.; Foces-Foces, C.; Perez, C.; Rodriguez, M. L.; Martin, J. D. J. Am. Chem. Soc. 2001, 123, 11970-11981; Dapporto, P.; Paoli, P.; Roelens, S. J. Org. Chem. 2001, 66, 4930-4933; Kawasaki, T.; Tokuhiro, M.; Kimizuka, N.; Kunitake, T. J. Am. Chem. Soc. 2001, 123, 6792-6800; Mak, T. C. W.; Xue, F. J. Am. Chem. Soc. 2000, 122, 9860-9861.
1144
Crystal Growth & Design, Vol. 5, No. 3, 2005
(3) Holman, K. T.; Pivovar, A. M.; Ward, M. D. Science 2001, 294, 1907-1911. (4) Holman, K. T.; Martin, S. M.; Parker, D. P.; Ward, M. D. J. Am. Chem. Soc. 2001, 123, 4421-4431. (5) Holman, K. T.; Pivovar, A. M.; Swift, J. A.; Ward, M. D. Acc. Chem. Res. 2001, 34, 107-118; Holman, K. T.; Ward, M. D. Angew. Chem., Int. Ed. 2000, 39, 1653-1656; Plaut, D. J.; Holman, K. T.; Pivovar, A. M.; Ward, M. D. J. Phys. Org. Chem. 2000, 13, 858-869; Swift, J. A.; Pivovar, A. M.; Reynolds, A. M.; Ward, M. D. J. Am. Chem. Soc. 1998, 120, 5887-5894; Swift, J. A.; Russel, V. A.; Ward, M. D. Adv. Mater. 1997, 9, 1183-1186. (6) Horner, M. J.; Holman, K. T.; Ward, M. D. Angew. Chem., Int. Ed. 2001, 40, 4045-4048. (7) Russell, V. A.; Evans, C. C.; Li, W.; Ward, M. D. Science 1997, 276, 575-579. (8) Custelcean, R.; Ward, M. D. Angew. Chem., Int. Ed. 2002, 41, 1724-1728. (9) Plaut, D. J.; Martin, S. M.; Kjaer, K.; Weygand, M. J.; Lahav, M.; Leiserowitz, L.; Weissbuch, I.; Ward, M. D. J. Am. Chem. Soc. 2003, 125, 15922-15934. (10) Videnova-Adrabinska, V.; Turowska-Tyrk, I.; Borowiak, T.; Dutkiewicz, G. New J. Chem. 2001, 25, 1403-1409. (11) Martin, S. M.; Yonezawa, J.; Horner, M. J.; Macosko, C. W.; Ward, M. D. Chem. Mater. 2004, 16, 3045-3055. (12) Mathevet, F.; Masson, P.; Nicoud, J.-F.; Skoulios, A. Chem. Eur. J. 2002, 8, 2248-2254. (13) Burrows, A. D.; Harrington, R. W.; Mahon, M. F.; Teat, S. J. Eur. J. Inorg. Chem. 2003, 1433-1439. (14) Reddy, D. S.; Duncan, S.; Shimizu, G. K. H. Angew. Chem., Int. Ed. 2003, 42, 1360-1364. (15) Russell, V. A.; Etter, M. C.; Ward, M. D. Chem. Mater. 1994, 6, 1206-1217. (16) Mondal, R.; Howard, J. A. K.; Banerjee, R.; Desiraju, G. R. Chem. Commun. 2004, 644-645; Kobayashi, K.; Sato, A.; Sakamoto, S.; Yamaguchi, K. J. Am. Chem. Soc. 2003, 125, 3035-3045; Pedireddi, V. R.; PrakashaReddy, J. Tetrahedron Lett. 2003, 44, 6679-6681; Raj, S. B.; Muthiah, P. T.; Rychlewska, U.; Warzajtis, B. CrystEngCommun 2003, 5, 48-53; Udachin, K. A.; Enright, G. D.; Brown, P. O.; Ripmeester, J. A. Chem. Commun. 2002, 2162-2163; Thaimattam, R.; Xue, F.; Sarma, J. A. R. P.; Mak, T. C. W.; Desiraju, G. R. J. Am. Chem. Soc. 2001, 123, 4432-4445. (17) Toudic, B.; Le Lann, H.; Guillaume, F.; Lechner, R. E.; Ollivier, J.; Bourges, P. Chem. Phys. 2003, 292, 191-199; Sidhu, P. S.; Enright, G. D.; Ripmeester, J. A.; Penner, G. H. J. Phys. Chem. B 2002, 106, 8569-8581; Comotti, A.; Simonutti, R.; Catel, G.; Sozzani, P. Chem. Mater. 1999, 11,
Voogt and Blanch
(18)
(19)
(20) (21)
(22) (23) (24) (25) (26) (27) (28) (29) (30)
1476-1483; Sozzani, P.; Bovey, F. A.; Schilling, F. C. Macromolecules 1991, 24, 6764-6768. Chao, M.-H.; Harris, K. D. M.; Kariuki, B. M.; Bauer, C. L.; Foxman, B. M. J. Phys. Chem. B 2002, 106, 4032-4035; Lee, S.-O.; Harris, K. D. M.; Jupp, P. E.; Yeo, L. J. Am. Chem. Soc. 2001, 123, 12913-12914; Hollingsworth, M. D. J. Am. Chem. Soc. 1993, 115, 5881-5882. Nakano, K.; Mochizuki, E.; Yasui, N.; Morioka, K.; Yamauchi, Y.; Kanehisa, N.; Kai, Y.; Yoswathananont, N.; Tohnai, N.; Sada, K.; Miyata, M. Eur. J. Org. Chem. 2003, 24282436; Allcock, H. R.; Sunderland, N. J. Macromolecules 2001, 34, 3069-3076; Pivovar, A. M.; Holman, K. T.; Ward, M. D. Chem. Mater. 2001, 13, 3018-3031; Caira, M. R.; Nassimbeni, L. R.; Toda, F.; Vujovic, D. J. Am. Chem. Soc. 2000, 122, 9367-9372; Caira, M. R.; Nassimbeni, L. R.; Vujovic, D.; Toda, F. J. Phys. Org. Chem. 2000, 13, 7579. Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. J. Appl. Crystallogr. 1999, 32, 115-119. Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; de Gelder, R.; Israel, R.; Smits, J. M. M. The DRDIF-94 program system; University of Nijmegen: Nijmegen, The Netherlands, 1994. Sheldrick, G., Siemens Area Detector ABSorption correction programs, 1996. The Weak Hydrogen Bond: Applications to Structural Chemistry and Biology; Desiraju, G.; Steiner, T.; Oxford University Press: Oxford, 1999. Jorgensen, W. L.; Madura, J. D.; Swenson, C. J. J. Am. Chem. Soc. 1984, 106, 6638-46. Swift, J. A.; Reynolds, A. M.; Ward, M. D. Chem. Mater. 1998, 10, 4159-4168. Connors, K. A. Chem. Rev. 1997, 97, 1325-1357. Tabellion, F. M.; Seidel, S. R.; Arif, A. M.; Stang, P. J. J. Am. Chem. Soc. 2001, 123, 11982-11990. Gerson, A. R.; Roberts, K. J.; Sherwood, J. N. Acta Crystallogr., Sect. B 1991, B47, 280-284. Sozzani, P.; Comotti, A.; Bracco, S.; Simonutti, R. Chem. Commun. 2004, 768-769. Akita, C.; Kawaguchi, T.; Kaneko, F.; Yamamoto, H.; Suzuki, M. J. Phys. Chem. B 2004, 108, 4862-4868; Eldho, N. V.; Feller, S. E.; Tristram-Nagle, S.; Polozov, I. V.; Gawrisch, K. J. Am. Chem. Soc. 2003, 125, 6409-6421; Feller, S. E.; Gawrisch, K.; MacKerell, A. D., Jr. J. Am. Chem. Soc. 2002, 124, 318-326; Binder, H.; Gawrisch, K. J. Phys. Chem. B 2001, 105, 12378-12390; Everts, S.; Davis, J. H. Biophys. J. 2000, 79, 885-897.
CG049636I