Systematic Structural Study of Asymmetric Supramolecular Assembly by a Series of Bile Acid Derivatives with Different Side-Chain Lengths Kazuaki Kato,* Michihiro Sugahara, Norimitsu Tohnai, Kazuki Sada,† and Mikiji Miyata*
CRYSTAL GROWTH & DESIGN 2004 VOL. 4, NO. 2 263-272
Department of Material and Life Science, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan Received August 1, 2003;
Revised Manuscript Received November 6, 2003
ABSTRACT: A systematic structural study of asymmetric supramolecular assembly in the crystalline state has been performed by one-by-one insertions of methylene spacers into the side-chain of steroidal bile acids. Five derivatives of bile acid with different side-chain lengths, bisnorcholic acid (0), norcholic acid (1), cholic acid (2), homocholic acid (3), and bishomocholic acid (4), were recrystallized from many solvents. X-ray diffraction studies of the resulting crystals revealed that these derivatives form various bilayer and layerlike structures in which two steroidal planes are diversely aggregated by using the hydrophilic faces inside. Such asymmetric bimolecular aggregation modes are logically classified into 10 categories on the basis of the relative positions between two steroidal planes. Actually, four modes among them were observed by means of one-by-one methylene insertions. This is attributable to the fact that four hydrogen-bonding groups, three hydroxyl groups on the skeletons and one carboxyl group at the side-chain, work cooperatively to form various hydrogen-bonding networks with the corresponding host frameworks. In this way, the insertion of methylene spacers plays a decisive role to diversify the asymmetric bimolecular aggregation modes. Introduction The prediction and control of molecular assemblies in the crystalline state have generated much interest in recent years as huge amounts of crystal structural data are accumulated.1 Organic molecules keep affording challenging subjects of research because of their diverse shape and flexibility.2-4 Thirty years ago, Kitaigorodskii took a historical approach to the prediction of two-dimensional assembly modes based on the closepacking principle.5 Systematic structural studies were undertaken to define several representative close-packing assemblies on a series of aromatic hydrocarbons.6 Current computational techniques can generate several crystal structures that approximately satisfy the principle.7 In addition, robust supramolecular synthons have become a successful strategy in handling molecular assemblies.8 Such synthons are composed of various hydrogen bonds, and are maintained against chemical modifications.9-14 According to the tunable pillar strategy, guanidinium sulfonate gave precisely controlled molecular arrangements with a two-dimensional hydrogen-bonding network, resulting in adjustable porosities.8,15 Another strategy is based on a control of multiple hydrogen bonds and molecular symmetry, making it possible to design various molecular arrangements.16-21 Subsequently, molecules with cooperative hydrogenbonding groups were employed.22-28 Such groups form an intermolecular network independently. For example, the assembly modes of a series of diol compounds have * To whom correspondence should be addressed. E-mail: katochan@ molrec.mls.eng.osaka-u.ac.jp, miyata@ molrec.mls.eng.osaka-u.ac.jp. † Present Address, Department of Chemistry and Biochemistry, Graduate School of Engineering, Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan. e-mail:
[email protected]. kyushu-u.ac.jp.
been investigated in detail.29-31 The usage of three or four hydroxyl groups rather than two would enable us to observe the diversity of molecular assembly due to a large number of combinations among the cooperative hydrogen-bonding groups. Moreover, we can adopt another strategy to use alicyclic rather than aromatic compounds, and asymmetric rather than symmetric molecules. Plural cooperative hydroxyl groups are seen in asymmetric bile acid molecules.32 Cholic acid, which is one of the well-known bile acids, has three hydroxyl groups on its alicyclic skeleton and a carboxyl group at its sidechain. The three hydroxyl groups at the C3, C7, and C12 positions (hereafter abbreviated as OH(C3), OH(C7), and OH(C12)) are directed to the same face (R-face) of the skeleton to give a hydrophilic side, while two methyl groups at C10 and C13 are directed to the opposite face (β-face) to give a lipophilic side (Figure 1a). Therefore, cholic acid has facial amphiphilicity to form bilayer structures.33 Another feature of cholic acid is that the three hydroxyl groups do not yield intramolecular hydrogen bonds, because each distance between the groups is in a range of about 5-6 Å. Therefore, we expect that these groups function as cooperative hydrogen-bonding groups to form various intermolecular networks among the neighboring molecules. A notable difference between the alicyclic and aromatic molecules lies in the symmetry of their molecular skeletons. Namely, the former asymmetric molecules result in much more diverse assembly modes than the latter symmetric molecules. Cholic acid and its derivatives have alicyclic skeletons, and serve as a typical example of the former. Logically, they produce a great number of asymmetric assemblies, but the number would be limited to some degree by the facial amphiphilicity mentioned above. This is because the hydrophilic
10.1021/cg0341499 CCC: $27.50 © 2004 American Chemical Society Published on Web 01/20/2004
264
Crystal Growth & Design, Vol. 4, No. 2, 2004
Kato et al.
Figure 1. (a) Facial amphiphilic molecule and (b) the bimolecular aggregate that is the minimum constitution unit of (c) the corresponding bilayer structure of bile acid. (d) Ten representations of the bimolecular aggregation mode of the steroidal molecules.
side of one molecule faces the same side of the other molecule inside, as shown in Figure 1b. Figure 1d illustrates 10 representatives of asymmetric bimolecular aggregation modes of the steroidal molecules based on the geometrical possibility of the asymmetric molecules within the restriction of the bilayer structure. Rotation right above the molecule of facing would lead to four distinct bimolecular aggregations: P, P#, R-, and R+. Alternatively, sliding along the long axis and the minor axis of the molecule of facing would lead to six distinct aggregations: Sh, Sh#, Sv, Sv#, Shv, and Shv#. It is noteworthy that these bimolecular aggregates are the minimum constitutional units of the corresponding bilayer structures (Figure 1c). It can be seen that each aggregation mode accompanies the corresponding positional relation among OH(C3), OH(C7), and OH(C12). The dotted lines between the hydroxyl groups mean that each group is within the distance of possible hydrogen bonds. In these figures, the carboxyl group at the sidechain is omitted for clarity. We prepared bisnorcholic acid (0), norcholic acid (1), homocholic acid (3), and bishomocholic acid (4), starting from cholic acid (2). Each number in parentheses is based on the number of methylene spacers at the sidechain. In previous papers, we reported some typical structures of 134, 2,35-42 and 4.43 Although the hydrogen-
bonding mode of 4 was the same as that of 2, the mode of 1 was distinct from that of 2. These results indicated that various hydrogen-bonding networks and the corresponding bimolecular aggregation modes might be observed by the modifications of the side-chain length without changing the hydrogen-bonding groups. Further studies have made it clear that these five derivatives form various layer structures. In this paper, we focus on the asymmetric bimolecular aggregation mode in each layer structure, and demonstrate that several modes in Figure 1d are induced by one-by-one insertions of methylene spacers into the side-chains. To our knowledge, this is the first example of systematic structural studies on crystalline supramolecular assemblies of facially amphiphilic and asymmetric molecules.
Asymmetric Supramolecular Assembly
Crystal Growth & Design, Vol. 4, No. 2, 2004 265
Table 1. Inclusion Abilities and Crystal Structures of Five Steroidal Derivatives with Different Side-Chain Length guest
0
methanol ethanol 1-propanol 2-propanol 1-butanol 2-butanone 2,4-pentadione acetophenone 1,4-dioxane tetrahydrofuran ethyl acetate methyl propionate acetonitrile propionitrile benzene toluene o-xylene 1-methylnaphthalene 1-tetralone
GFa GF GF nc nc GF nc nc GF GF GF nc GF nc GF GF nc nc nc nc
1 b
Ba Ba Ba nc nc Ba nc nc Ba Ba Ba nc Ba nc Ba nc nc nc nc
ncc nc 2:1 2:1:1f 2:1 2:1:1 2:1 2:3 2:1 2:1:2 nc 2:1 2:1:1 2:1:2 nc nc nc nc 2:4
2 nc nc Bb Bb Bb Bb Bb Bc Bb Bb nc Bb Bb Bb nc nc nc nc Bd
1:1d
crossinge
1:1 1:1 1:1 GF 1:1 1:1 1:1 1:1 nc 1:1 1:1 1:1 nc 1:1 1:1 2:1 2:1 1:1
crossing crossing crossing crossing Bf Bg Be Be nc Bf Be Bf nc Be Be Be Be Be
3 nc 1:2 nc 1:2 GF GF GF GF GF GF GF GF GF GF GF GF GF GF GF
nc crossing nc crossing T T T T T T T T T T T T T T T
4 1:1 1:1 1:1 1:1 1:1 GF GF 1:1 GF GF GF GF 1:1 GF GF GF 1:1 1:1 1:1
crossing crossing crossing crossing crossing W W crossing W W W W crossing W W W Be Be Be
a Guest-free crystal. b The symbols of molecular assembly modes, corresponding to the structures shown in Figure 2. c Crystals are not generated yet. d Host-guest ratio. e Crossing type structures are not treated in this paper. f Host-guest-water ratio.
Figure 2. Molecular packing diagrams of layer structures. These layer structures are classified into seven bilayer types (Ba, Bb, Bc, Bd, Be, Bf, and Bg) and two layerlike types (T and W).
Experimental Section General Methods. Compound 2 was commercially available. The other four derivatives, 0, 1, 3, and 4, were prepared by the previous method starting from 2.44,45 All chemicals and solvents were commercially available and used without any further purification. Infrared spectra were recorded on a HORIBA FT-IR spectrometer. Thermal gravity (TG) analyses were performed on a Rigaku TAS100 system; ca. 5 mg from 40 to 250 °C at a heating rate of 5 °C min-1. X-ray powder diffraction (XRD) patterns were measured by a Rigaku RINT1100 at room temperature. Preparation of Inclusion Crystals. Each derivative was dissolved with warming in the liquid guest (usually 1-3 mL), and the resulting solution was allowed to stand at room temperature. Collected crystals were dried on the filter papers, and then characterized by IR, TG, and XRD. Part of the results are summarized in Table 1. Crystal Structure Determinations. X-ray diffraction data were collected on a Rigaku RAXIS-IV diffractometer or a Rigaku RAPID diffractometer with 2D area detector with graphite-monochromatized Mo-KR or Cu-KR radiation. Lattice parameters were obtained by least-squares analysis from three oscillation images in the 2D area detector. The structures were solved by direct methods (SHELEX8646 or SIR9247). All calculations were performed using the program TEXSAN48
crystallographic software packages of the Molecular Structure Corporation. Non-hydrogen atoms were refined with anisotropic displacement parameters. Hydrogen atoms were placed in idealized positions, and no further refinement was applied. The measurement conditions and structural details are listed in Table 2.
Results and Discussion Inclusion Behaviors and Crystal Structures of the Compounds 0-4. The steroidal derivatives 0-4 with different side-chain lengths yield diverse inclusion abilities and crystal structures, as summarized in Table 1. Compound 0 gives only guest-free crystals, while 1-4 give diverse inclusion crystals with crossing49 and/or layer structures. As schematically shown in Figure 2, this paper deals with only the latter layer structures, which can be classified into seven bilayer types (Ba, Bb, Bc, Bd, Be, Bf, and Bg) and two layerlike types (T and W). Compound 0 has only a bilayer structure Ba. 1 and 2 have dominant bilayer structures (Bb - Bg), while 3 and 4 have dominant layerlike structures (T and W). Their crystallographic data are summarized in Table 2.
266
Crystal Growth & Design, Vol. 4, No. 2, 2004
Kato et al.
Table 2. Crystallographic Data for the Crystals Shown in Figure 2
compound formula formula wt crystal system space group a (Å) b (Å) c (Å) R (deg) β (deg) γ (deg) V (Å3) Z Dcalc (g/cm3) R1 wR2 G.O.F. reference
compound formula formula wt crystal system space group a (Å) b (Å) c (Å) R (deg) β (deg) γ (deg) V (Å3) Z Dcalc (g/cm3) R1 wR2 G.O.F. ref
Ba
Bb
Bc
Bd
Be
0 C22H36O5 380.52 orthorhombic P21212 16.898(2) 11.011(1) 11.022(3) 90 90 90 2050.8(5) 4 1.232 0.068 0.099 1.21 this work
(1)2‚(THF) C50H84O13 893.21 triclinic P1 10.834(3) 15.354(6) 7.745(4) 93.77(2) 79.99(2) 109.00(3) 1199.5(9) 1 1.236 0.092 0.226 1.21 this work
(1)2‚(acetophenone)3 C70H100O13 1149.55 triclinic P1 11.711(1) 15.745(8) 10.514(1) 101.75(2) 113.468(7) 105.05(2) 1611.1(9) 1 1.185 0.092 0.265 1.25 this work
(1)2‚(tetralone)4 C70H116O14 1373.85 triclinic P1 12.355(8) 15.0734(9) 10.2995(8) 98.553(2) 99.128(2) 88.287(3) 3247.3(8) 1 1.281 0.065 0.169 2.25 this work
2‚(acetophenone) C32H48O6 528.73 monoclinic P21 13.83(1) 8.26(2) 14.28(2) 90 114.4(1) 90 1485(4) 2 1.182 0.071 0.162 1.91 33
Bf
Bf
Bf
Bf
Bf
2‚(2,4-pentadione) C29H48O7 508.69 triclinic P1 12.264(2) 14.452(7) 8.178(2) 90.48(3) 92.88(1) 105.19(2) 1396.8(1) 1 1.209 0.052 0.123 1.77 this work
(2)2‚(o-xylene) C56H90O10 923.32 monoclinic P21 7.515(10) 25.612(6) 13.827(9) 90 90.99(8) 90 2661.0(2) 2 1.152 0.054 0.080 3.55 50
3 C100H168O20 1690.42 tetragonal P41 11.5429(5) 11.5429(5) 69.338(4) 90 90 90 9236.9(8) 4 1.215 0.171 0.418 4.83 this work
4‚(1-tetralone) C36H52O6 556.78 monoclinic P21 14.718(3) 26.610(5) 8.294(1) 90 114.144(4) 90 3247.3(8) 4 1.139 0.085 0.229 1.22 43
4 C26H44O5 436.63 orthorhombic P212121 9.886(1) 37.500(4) 6.584(7) 90 90 90 2441.0(4) 4 1.188 0.067 0.115 1.55 43
These results indicate the fascinating fact that oneby-one methylene insertions into the side-chain induce diverse changes of the layer structures. Such changes accompany the changes of the host frameworks, which may be attributable to those of various hydrogenbonding networks among the steroidal host molecules. Moreover, the networks can be reduced to various relative positions between neighboring molecules, which are represented by bimolecular aggregation modes (Figure 1d). Hereafter, we describe relationships among the host frameworks, the hydrogen-bonding networks, the relative positions, and the bimolecular aggregation modes. Compound 0; Host Framework with a Robust Hydrogen-Bonding Network. Compound 0 forms a definite bilayer structure Ba (Table 1). Figure 3a,b shows its host framework and hydrogen-bonding network, respectively. It can be seen that arched hydrogenbonding networks spread out horizontally to form the bilayer structure. Figure 3c illustrates a relative position between two steroidal skeletons with the network among OH(C7), OH(C12), and COOH(C22). Figure 3d shows the bimolecular aggregate, which corresponds to mode R- in Figure 1d. Hereafter, this mode of the derivative 0 is named (R-, 0). Compound 1; Rearrangement of the HydrogenBonding Network and Aggregation Mode. Compound 1 has three kinds of bilayer structures, Bb, Bc,
and Bd (Table 1). Their host frameworks and their corresponding hydrogen-bonding networks are shown in Figure 4a,b, respectively. It can be seen that these have a common bilayer, which is retained by cyclic hydrogen bonds among two OH(C3) and two COOH (C23) groups. The other OH(C7) and OH(C12) are free or hydrogen-bonded with guest compounds. For the active hydrogen bonds between host and guest, the host frameworks are governed by the size of the guest compounds. Small guests can be enclosed within the host cavity of the Bb framework. On the other hand, larger guests spread out the host cavity to form distinct host frameworks Bc or Bd. Figure 4c shows a relative position between two steroidal molecules. As shown in Figure 4d, this common bimolecular aggregate corresponds to mode P# in Figure 1d, named (P#, 1). Interception of the Robust Hydrogen-Bonding Network. We notice that an insertion of one methylene spacer brings about a great change of the host frameworks, the hydrogen-bonding networks, and the bimolecular aggregation modes. Another notable change is seen in the distances between two OH(C12) groups. That is, the distances between two OH(C12) groups in (R-, 0) and (P#, 1) are 2.9 Å and over 3.3 Å, respectively. The former hydrogen bond is dominant in the mode (R-, 0). The latter weak one is not dominant in mode (P#, 1), but the other strong hydrogen bond between OH(C3) and COOH(C23) (2.6-2.9 Å) is dominant.
Asymmetric Supramolecular Assembly
Crystal Growth & Design, Vol. 4, No. 2, 2004 267
Figure 3. A schematic overview of the structure Ba in compound 0. (a) Host framework, (b) hydrogen-bonding network, (c) the relative position between two steroidal molecules, and (d) the bimolecular aggregation mode. The aggregation mode corresponds to R- in Figure 1d, named (R-, 0).
Figure 4. Schematic overviews of the structures Bb, Bc, and Bd, in compound 1. (a) Host frameworks and (b) the corresponding hydrogen-bonding networks. (c) A relative position between two steroidal molecules and (d) a common bimolecular aggregation mode among the three structures. The aggregation mode corresponds to P# in Figure 1d, named (P#, 1).
Figure 5 shows an easy simulation for the change by the insertion of a methylene spacer at the side-chain. In the mode (R-, 0), COOH(C22) links OH(C3) and OH(C7). However, the insertion leads to cleavage of the hydrogen bond between OH(C7) and COOH(C23), indicating an impossible mode (R-, 1) This is because one methylene spacer is long enough to keep the COOH(C23) away from OH(C7). In other words, any conformations make it impossible to locate the COOH(C23) within the effective distances of hydrogen bonds with both OH(C3) and OH(C7). This simulation demonstrates that R- is a unique mode to 0, meaning that (R-, 1), (R-, 2), (R-, 3), and (R-, 4) are not observed. Compound 2; Diverse Hydrogen-Bonding Networks by the Carboxyl Group. 2 has been known to form various bilayer structures depending on the included guests. Table 1 contains three bilayer structures, Be, Bf, and Bg, which have different host frameworks and hydrogen-bonding networks, as shown in Figure 6a,b, respectively. It can be seen that Be has a cyclic hydrogen-bonding network composed of OH(C3), OH(C7), OH(C12), and COOH(C24). Bf has both a cyclic and a linear network, while Bg has a branched one. Because
Figure 5. Easy simulation for the change by the insertion of a methylene spacer. Interception of the robust hydrogenbonding network of Ba indicates the impossible mode (R-, 1).
268
Crystal Growth & Design, Vol. 4, No. 2, 2004
Kato et al.
Figure 6. Schematic overviews of the structures Be, Bf, and Bg, in compound 2. (a) Host frameworks and (b) the corresponding hydrogen-bonding networks. The shadowed partial hydrogen-bonding networks, OH(C7)- - -OH(C3)- - -OH(C12), are common with Be and Bf. (c) Relative positions among four neighboring steroidal molecules in each hydrogen-bonding network. The arrows represent the linking modes of the carboxyl groups between the partial hydrogen-bonding networks in Be and Bf. (d) Bimolecular aggregation modes. The bimolecular aggregate in the structure Be and Bf corresponds to a common mode Shv# in Figure 1d, named (Shv#, 2), while the structure Bg corresponds to mode Shv, named (Shv, 2).
guest compounds are not hydrogen-bonded to the host, the host frameworks would be determined by the size and shape of the guest compounds.51 Figure 6c illustrates each relative position among four neighboring steroidal molecules. As shown in Figure 6d, Be and Bf are composed of a common antiparallel bimolecular aggregate with the mode corresponding to Shv# in Figure 1d, named (Shv#, 2). On the other hand, Bg has the parallel mode corresponding to Shv, named (Shv, 2). It is reasonable that the mode Shv# produces a partial hydrogen-bonding network, OH(C7)- - -OH(C3)- - -OH(12), in common with Be and Bf (Figure 6c). When COOH(C24) participates in the network as an internal or external linker, the cyclic or linear network in the bilayer structure Be or Bf is produced, as represented by the arrow in Figure 6c. On the other hand, the mode Shv forms a different hydrogen-bonding network from Shv#, COOH(C24)- - -OH(C3)- - -OH(12), indicating no hydrogen bonding between OH(C3) and OH(C7). This is ascribed to the asymmetric positions of the four hydrogen-bonding groups. Fundamental Difference in the Direction of the Carboxyl Group. The aggregation modes, Shv# and Shv, are observed in 2 but not in 1, whereas P# is observed in 1 but not in 2. This drastic change due to the insertion of a methylene spacer is attributable to a difference in the direction of the carboxyl group, as schematically shown in Figure 7. This change follows the so-called odd-even rule. Each carboxyl group plays a different role in hydrogen bonds with the neighboring molecule. In case of 1 in Figure 7a,b, one oxygen atom of COOH(C23) is directed
to the lipophilic face (β-face) to link OH(C3) of the neighboring molecule in the same layer, while the other one is directed to the hydrophilic face (R-face) to link OH(C3) in the opposite layer. These directions result in the mode (P#, 1) in Figure 7c. In contrast, in the case of 2 in Figure 7a,b, both oxygen atoms of COOH(C24) are directed to the hydrophilic face (R-face). In the structure Be in 2, COOH(C24) links OH(C7) and OH(C12) of different molecules in the opposite layer. These directions result in the vertically slide mode (Shv#, 2) in Figure 7c. The other slide mode (Shv, 2) can be understood similarly. In this way, the insertion of one methylene spacer into the side-chain induces such fundamentally different aggregation modes due to the different directions of COOH(C23) and COOH(C24). Compound 3; Inherited Hydrogen-bonding Network and Bimolecular Aggregate. As shown in Figure 8a, compound 3 does not form a bilayer structure but a layerlike structure T, being different from compounds 0-2. Although the host framework of T in 3 appears to be greatly different from those of Be, Bf, and Bg in 2 (Figures 6a and 8a), the cyclic hydrogen-bonding network of T in 3 is the same as that of Be in 2 (Figures 6b and 8b). Figure 8c shows the relative position of four neighboring molecules, which is the same as Be in 2 (Figure 6c). Therefore, the bimolecular aggregate of 3 (Figure 8d) corresponds to the same mode Shv# as that of 2 (Figure 6d), named (Shv#, 3). In this way, 3 inherits the hydrogen-bonding network and bimolecular aggregation from 2.
Asymmetric Supramolecular Assembly
Crystal Growth & Design, Vol. 4, No. 2, 2004 269
Figure 7. Effect of the different directions of the carboxyl group between 1 and 2 on the bimolecular aggregation mode. (a) The directions of the two oxygen atoms of the carboxyl group represented by small arrows and (b) the role of the carboxyl group in linking two neighboring molecules. (c) Resulting bimolecular aggregation modes.
Figure 8. A schematic overview of the structure T in compound 3. (a) Host framework, (b) hydrogen-bonding network, (c) the relative position between neighboring four steroidal molecules, and (d) the bimolecular aggregation mode. The aggregation mode corresponds to Shv# in Figure 1d, named (Shv#, 3).
Twist of the Side-Chain. We consider a relationship between 2 and 3. First, we notice that they belong to different space groups. The structure Be in 2 belongs to space group P21 and has a 21 axis in the direction designated in Figure 9a. Steroidal skeletons arrange along this axis to yield a 21 helical assembly (block with an arrow in Figure 9b). On the other hand, the structure T in 3 belongs to space group P41. Although this space group does not designate any 21 axes, the structure T is composed of a quasi-21 helical assembly (Figure 9d). The reason is that both Be and T have a common relative position (Figure 9c,f). As shown in Figure 9d, the quasi-21 axis is orthogonal to the 41 axis, indicating that the space group of T can be approximately regarded as P41212 (Figure 9e). Such consideration indicates that T is just the result of a “twist” of Be. Figure 9g schematically shows that the twisted side-chains link the neighboring 21 helical assemblies. Figure 9h depicts that a plane with the cyclic hydrogen-bonding network along the 41 axis is orthogonal to the neighboring one. It can be seen that the central molecule (X in Figure 9h) links the two orthogonal planes. Figure 9i reveals that the sidechain of 3 is twisted 90 degrees from that of 2. This is the reason 3 inherits the hydrogen-bonding network from 2, and why 3 does not inherit the mode (P#) from 1. These results are contrary to the odd-even rule.
Compound 4; Inherited and Unique Layer Structures. Compound 4 has a bilayer structure Be and a layerlike W. As shown in Figures 6a-d and 10a-d, the structure Be in 4 has the same host framework, hydrogen-bonding network, and bimolecular aggregation mode Shv# as Be in 2, named (Shv#, 4). That is, Be in 4 is the bilayer structure inherited from 2. This is because the insertion of two methylene spacers can bring about the same direction of COOH(C26) as that of COOH(C24) in agreement with the odd-even rule. On the other hand, the structure W is unique to 4 in the following way. W has the same antiparallel aggregation moder as Be (Figure 10c), indicating the same bimolecular aggregation mode Shv# (Figure 10d). However, W has a different host framework and branched hydrogen-bonding network (Figure 10a,b) from Be, suggesting that such changes do not follow the odd-even rule. Flexible Side-Chain beyond the Odd-Even Rule. Further description of the unique structure W is given as follows. Despite different space groups between Be (P21) and W (P212121), both structures are composed of 21 helical assemblies. Figure 11a shows two orthogonal 21 axes of W. Steroidal molecules form 21 helical assemblies along both axes. As shown in Figure 11b, the vertical assembly may be represented with hydrogenbonded blocks, as in the case of Figure 9b,e. It can be seen from the figure that the vertical assemblies twist 180 degrees. It is noteworthy that the structures Be and
270
Crystal Growth & Design, Vol. 4, No. 2, 2004
Kato et al.
Figure 9. (a) 21 helical assembly formed by partial hydrogen-bonding networks of the three hydroxyl groups and (b) pattern diagram of the structure Be based on the 21 helical assembly, which is represented by block with an arrow. (c) Position of the 21 axis in the bimolecular aggregation mode (Shv#, 2). (d) Quasi-21 helical assembly orthogonal to a 41 axis and (e) pattern diagram of the structure T. (f) Position of the quasi-21 axis in the bimolecular aggregation mode (Shv#, 3). (g) Pattern diagram of twisted side-chains linking the neighboring 21 helical assemblies in the structure T, and (h) the highlight of a twisted side-chain. The small arrows represent the orthogonal directions between OH(C3) and COOH(C25). (i) Comparison of direction of the carboxyl group toward the steroidal skeleton between Be in 2 and T in 3.
Figure 10. Schematic overviews of the structures Be and W in compound 4. (a) Host frameworks, (b) the corresponding hydrogenbonding networks, (c) the relative position between neighboring four steroidal molecules, and (d) the bimolecular aggregation mode. Both aggregation modes correspond to Shv# in Figure 1d, named (Shv#, 4).
W have the same bimolecular aggregation mode Shv# (Figures 9c and 11c), although the structures have different host frameworks and hydrogen-bonding networks. Figure 11d shows the warped side-chain along the 21 axis. A plane with the branched hydrogen-bonding network turns over from the neighboring one along the axis, and the central molecule (X in Figure 11d) links the two reversed planes. Figure 11e reveals that the side-chain of Be is warped with twist of 180 degrees as compared with that of W. However, the carboxyl group of Be is on the plane of the cyclic hydrogen-bonding network, while the one of W is under the branched
network. These results indicate that the side-chain is no longer affected by the odd-even rule. Prospects of the Aggregation Modes by Further Insertion of Methylene Spacers. As mentioned above, drastic diversification of the bimolecular aggregation mode was observed in the case of the derivatives having shorter methylene spacers than 3. On the other hand, for the derivatives 3 and 4 having longer side-chains, a common bimolecular aggregation was observed in their different host frameworks because of the flexibility of the side-chains. This means that the mode is subject to hydrogen bonds of three hydroxyl groups on the steroidal skeleton rather than the carboxyl group at the side-
Asymmetric Supramolecular Assembly
Crystal Growth & Design, Vol. 4, No. 2, 2004 271
Figure 11. (a) 21 helical assembly orthogonal to another 21 axis and (b) pattern diagram of the structure W based on the 21 helical assembly, which is represented by block with an arrow. (c) Position of the 21 axis in the bimolecular aggregation mode (Shv#, 4). (d) Highlight of the warped side-chain. The small arrows represent the orthogonal directions between OH(C3) and COOH(C26). (e) Comparison of direction of the carboxyl group toward the steroidal skeleton between Be and W in 4.
Figure 12. Summary of bimolecular aggregation modes and the cooperative hydrogen-bonding networks in each side-chain length.
chain. These results may indicate that the longer methylene spacers would result in the constant mode Shv#. Accordingly, it is considered that such short methylene spacers attached to the skeletons are highly effective for diversification of the aggregation modes and hydrogen-bonding networks. Conclusion We demonstrated a systematic structural study on asymmetric supramolecular assemblies by using a series of bile acid derivatives with different side-chain lengths. One-by-one methylene insertions into the side-chains caused the diversity of host frameworks with the characteristic hydrogen-bonding networks. On the basis of the relative positions between two steroidal planes, bimolecular aggregation modes were classified into 10 categories, and four of them were observed. The hydrogen-bonding groups of steroidal bile acids form cooperative hydrogen bonds and bimolecular aggregates, as summarized in Figure 12. This strategy based on the one-by-one methylene insertion into the side-chain enables us to observe the diverse assembly modes of the asymmetric molecule. From the viewpoint of crystal engineering, this strategy may give us an idea of controlling crystal structures by changing methylene spacers at side-chains, without changing the main molecular skeleton.
Acknowledgment. This work was supported by a Grant-in-Aid Scientific Research from the Ministry of Education, Science, Sports, and Culture, Japan. Supporting Information Available: X-ray crystallographic information files (CIF) are available. This material is available free of charge via the Internet at http://pubs.acs.org.
References (1) For a depository of more than 250 000 small molecule crystal structures, see Cambridge Structural Database (http:// www.ccdc.cam.ac.uk/prods/csd/csd.html). (2) (a) Gavezzotti, A.; Filippini, G. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1993, 235, 225-230. (b) Gavezzotti, A.; Filippini, G. J. Am. Chem. Soc. 1996, 118, 7153-7157. (3) Chaka, A. M.; Zaniewski, R.; Youngs, W.; Tessier, C.; Klopman, G. Acta Crystallogr. 1996, B52, 165-183. (4) (a) Lommerse, J. P. M.; Motherwell, W. D.; Ammon, H. L.; Dunitz, J. D.; Gavezzotti, A.; Hofmann, D. W. M.; Leusen, F. J. J.; Mooij, W. T. M.; Price, S. L.; Schweizer, B.; Schmidt, M. U.; Eijck, B. P.; Verwer, P.; Williams, D. E. Acta Crystallogr. 2000, B56, 697-714. (b) Motherwell, W. D.; Ammon, H. L.; Dunitz, J. D.; Dzyabchenko, A.; Erk, P.; Gavezzotti, A.; Hofmann, D. W. M.; Leusen, F. J. J.; Lommerse, J. P. M.; Mooij, W. T. M.; Price, S. L.; Scheraga; H.; Schweizer, B.; Schmidt, M. U.; Eijck, B. P.; Verwer, P.; Williams, D. E. Acta Crystallogr., Sect. B 2002, 58, 647661. (5) Kitaigorodskii, A. I. Molecular Crystals and Molecules; Academic Press: New York, 1973; pp 1-37. (6) (a) Gavezzotti, A.; Desiraju, G. R. Acta Crystallogr., Sect. B 1988, 44, 427-433. (b) Desiraju, G. R.; Gavezzotti, A. J. Chem. Soc., Chem. Commun. 1989, 621-623. (c) Desiraju, G. R.; Gavezzotti, A. Acta Crystallogr., Sect. B 1989, 45, 473-482.
272
Crystal Growth & Design, Vol. 4, No. 2, 2004
(7) Gavezzotti, A. Synlett 2002, 201-214. (8) Russell, V. A.; Evans, C. C.; Li, W.; Ward, M. D. Science 1997, 276, 575-579. (9) (a) Desiraju, G. R. Angew. Chem., Int. Ed. Engl. 1995, 34, 2311-2327. (b) Vishweshwar, P.; Thaimattam, R.; Jasko´lski, M.; Desiraju, G. R. Chem. Commun. 2002, 1830-1831. (10) Guth, H.; Heger, G.; Dru¨ck, U.; Jenneskens, L. W.; Z. Kristallogr. 1982, 159, 185-190. (11) Herbstein, F. H. Top. Curr. Chem. 1987, 140, 107-139. (12) Takemoto, K.; Sonoda, N. In Inclusion Compounds; Atwood, J. L.; Davies, J. E. D.; MacNicol, D. D., Eds.; Academic Press: London, 1987; Vol. 2, pp 44-67. (13) Keegstra, E. M. D.; Spek, A. L.; Zwikker, J. W.; Jenneskens, L. W.; J. Chem. Soc. Chem. Commun. 1994, 1633-1634. (14) Russell, V. A.; Etter, M. C.; Ward, M. D. J. Am. Chem. Soc. 1994, 116, 1941-1952. (15) (a) Evans, C. C.; Sukarto, L.; Ward, M. D. J. Am. Chem. Soc. 1999, 121, 320-325. (b) Holman, K. T.; Pivovar, A. D.; Ward, M. D. Science 2001, 294, 1907-579. (16) MacDonald, J. C.; Whitesides, G. M. Chem. Rev. 1994, 94, 2383-2420. (17) Aoyama, Y.; Endo, K.; Anzai, T.; Yamaguchi, Y.; Sawaki, T.; Kobayashi, K.; Kanehisa, N.; Hashimoto, H.; Kai, Y.; Masuda, H. J. Am. Chem. Soc. 1996, 118, 5562-5571. (18) Bhyrappa, P.; Wilson, S. R.; Suslick, K. S. J. Am. Chem. Soc. 1997, 119, 8492-8502. (19) Kolotuchin, S. V.; Thiessen, P. A.; Fenlon, E. E.; Wilson, S. R.; Loweth, C. J.; Zimmerman, S. C. Chem. Eur. J. 1999, 5, 2537-2547. (20) Moulton, B.; Zaworotko, M. J. Chem. Rev. 2001, 101, 16291658. (21) Kobayashi, K.; Sato, A.; Sakamoto, S.; Yamaguchi, K. J. Am. Chem. Soc. 2003, 125, 3035-3045. (22) Del Bene, J.; Pople, J. J. Chem. Phys. 1970, 52, 4858-4866. (23) Ceccarelli, C.; Jeffrey, G. A.; Taylor, R. J. Mol. Struct. 1981, 70, 255-271. (24) Newton, N. D. J. Acta Crystallogr., Sect. B 1983, 39, 104113. (25) Biradha, K.; Dennis, D.; MacKinnon, V. A.; Sharma, C. V. K.; Zaworotko, M. J. J. Am. Chem. Soc. 1998, 120, 1189411903. (26) Prins, L. J.; Reinhoudt, D. N.; Timmerman, P. Angew. Chem., Int. Ed. 2001, 40, 2382-2426. (27) Szcze¸ sna, B.; Urban´czyk-Lipkowska, Z. New J. Chem. 2002, 26, 243-249. (28) Farrell, D. M. M.; Ferguson, G.; Lough, A. J.; Glidewell, C. Acta Crystallogr., Sect. B 2002, 58, 530-544. (29) Weber, E.; Csoregh, I.; Stensland, B.; Czugler, M. J. Am. Chem. Soc. 1984, 106, 3297-3306. (30) Bishop, R. In Comprehensive Supramolecular Chemistry, Solid-State Supramolecular Chemistry: Crystal Engineering; MacNicol, D. D., Toda, F., Bishop, R., Eds.; Pergamon: New York, 1996; Vol. 6, pp 85-115. (31) Toda, F., In Comprehensive Supramolecular Chemistry, Solid-State Supramolecular Chemistry: Crystal Engineering; MacNicol, D. D., Toda, F., Bishop, R., Eds.; Pergamon: New York, 1996; Vol. 6, pp 456-516. (32) Miyata, M.; Sada, K. In Comprehensive Supramolecular Chemistry, Solid-State Supramolecular Chemistry: Crystal Engineering; MacNicol, D. D., Toda, F., Bishop, R., Eds.; Pergamon: New York, 1996; Vol. 6, pp 147-176. (33) Miki, K.; Masui, A.; Kasai, N.; Miyata, M.; Shibakami, M.; Takemoto, K. J. Am. Chem. Soc. 1988, 110, 6594-6596. (34) Sugahara, M.; Sada, K.; Miyata, M. J. Chem. Soc., Chem. Commun. 1999, 293-294.
Kato et al. (35) (a) Johnson, P. L.; Schaefer, J. P. Acta Crystallogr., Sect. B 1972, 28, 3038-3088. (b) Jones, E. L.; Nassimbeni, L. R. Acta Crystallogr., Sect. B 1990, 45, 399-405. (36) (a) Shibakami, M.; Sekiya, A. J. Inclusion Phenom. 1994, 18, 397-412. (b) Shibakami, M.; Sekiya, A. J. Chem. Soc. Chem. Commun. 1994, 429-430. (c) Shibakami, M.; Sekiya, A. J. Inclusion Phenom. 1995, 22, 155-168. (d) Shibakami, M.; Tamura, M.; Sekiya, A. J. Inclusion Phenom. 1995, 22, 299-311. (e) Shibakami, M.; Tamura, M.; Sekiya, A. J. Am. Chem. Soc. 1995, 117, 4499-4505. (37) (a) Caira, M. R.; Nassimbeni, L. R.; Scott, J. L. J. Chem. Soc. Chem. Commun. 1993, 612-614. (b) Caira, M. R.; Nassimbeni, L. R.; Scott, J. L. J. Chem. Soc., Perkin Trans. 2 1994, 623-628. (c) Caira, M. R.; Nassimbeni, L. R.; Scott, J. L. J. Chem. Soc., Perkin Trans. 2 1994, 1403-1405. (38) (a) Scott, J. L. J. Chem. Soc., Perkin Trans. 2 1995, 495502. (b) Scott, J. L. Supramol. Chem. 1996, 7, 201-207. (c) Scott, J. L. Supramol. Chem. 1997, 8, 231-239. (d) Scott, J. L. Supramol. Chem. 1997, 8, 241-251. (39) (a) Lessinger, L. Cryst. Struct. Comm. 1982, 11, 1787-1792. (b) Lessinger, L.; Low, B. W. J. Crystallogr. Spectrosc. Res. 1993, 23, 85-99. (40) (a) Gdaniec, M.; Polonski, T. J. Am. Chem. Soc. 1998, 120, 7353-7354. (b) Gdaniec, M.; Milewska, M. J.; Polonski, T. Angew. Chem. 1999, 111, 405. (c) Gdaniec, M.; Milewska, M. J.; Polonski, T. Angew. Chem., Int. Ed. 1999, 38, 392395. (d) Nowak, E.; Gdaniec, M.; Polonski, T. J. Inclusion Phenom. 2000, 37, 155-169. (41) (a) Imashiro, F.; Kuwahara, D.; Terao, T. J. Chem. Soc., Perkin Trans 2 1993, 1759-1763. (b) Dastidar, P. CrystEngComm 2000, 8. (42) (a) Miki, K.; Kasai, N.; Shibakami, M.; Chirachanchai, S.; Takemoto, K.; Miyata, M. Acta Crystallogr. 1990, 24422445. (b) Miki, K.; Masui, A.; Kasai, M.; Goonewardena, W.; Shibakami, M.; Takemoto, K.; Miyata, M. Acta Crystallogr. 1992, 503-507. (c) Nakano, K.; Sada, K.; Miyata, M. Chem. Commun. 1996, 989-990. (d) Nakano, K.; Katsuta, M.; Sada, K.; Miyata, M. CrystEngComm 2001, 11. (e) Yoswathananont, N.; Chirachanchai, S.; Tashiro, K.; Nakano, K.; Sada, K.; Miyata, M. CrystEngComm 2001, 19. (f) Sada, K.; Nakano, K.; Hirayama, K.; Miyata, M.; Sasaki, S.; Takemoto, K.; Kasai, N.; Kato, K.; Shigesato, M.; Miki, K. Supramol. Chem. 2001, 13, 35-44. (43) Sada, K.; Sugahara, M.; Kato, K.; Miyata, M. J. Am. Chem. Soc. 2001, 123, 4386-4392. (44) Kobayashi, N.; Hagiwara, C.; Morisaki, M.; Yuri, M.; Oya, I.; Fujimoto, Y. Chem. Pharm. Bull. 1994, 42, 1028-1035. (45) Schteingart, C. D.; Hofmann, A. F. J. Lipid Res. 1988, 29, 1387-1395. (46) Sheldrick, G. M. Comput. Crystallogr. 1978, 34-42. (47) Burla, M. C.; Camalli, M.; Cascarano, G.; Giacovazzo, C.; Pilidori, G.; Spagna, R.; Viterbo, D. J. Appl. Crystallogr. 1989, 22, 389-393. (48) TEXSAN, X-ray structure analysis package; Molecular Structure Corporation: The Woodlands, TX, 1985. (49) Johnson, P. L.; Schaefer J. P. Acta Crystallogr., Sect. B 1972, 28, 3083-3089. (50) Nakano, K.; Sada, K.; Miyata, M. J. Chem. Soc. Chem. Commun. 1995, 953-954. (51) Nakano, K.; Sada, K.; Kurozumi, Y.; Miyata, M. Chem. Eur. J. 2001, 7, 209-220.
CG0341499