J. MATER. CHEM., 1993,3(1 l), 1129-1 135
`Hydrogen-bonded Layers of Hydrogentartrate Anions:
`Two-dimensional Building Blocks for Crystal Engineering
`
`1129
`
`Christer B. Aakeroy,*"t and Peter B. Hitchcockb
`a Department of Chemistry, University of Minnesota, 207 Pleasant St. S.E., Minneapolis, MN 55455,
`USA
`School of Chemistry and Molecular Sciences, University of Sussex, FaImer, Brighton, UK BN1 9QJ
`
`An investigation of hydrogen-bond patterns in a family of ionic crystals has established the use of hydrogen
`bonding as a potential design strategy for crystal engineering of ionic materials. In 11 out of 12 crystal structures
`of anhydrous hydrogentartrate salts (data extracted from the Cambridge Structural Database), the anions were
`found to generate infinite two-dimensional sheets created by relatively short O-H...O
`hydrogen bonds. The
`regular occurrence of the infinite layer, even in the presence of widely differing cations, demonstrates the
`selectivity, strength and directionality of hydrogen bonds between ions in the solid state. The two-dimensional
`sheet was subsequently employed as a building block ('scaffolding') in the synthesis of a material with a three-
`dimensional hydrogen-bond network, 1H-imidazolium hydrogen L-tartrate (the X-ray single-crystal structure is
`reported), where the cation provides the anticipated cross-l ink between adjacent anionic layers. Each anionic
`network was also described using a notation based upon graph theory which simplifies the process of recognizing
`and communicating complex structural information.
`
`Keywords: Hydrogen bonding ; Crystal engineering ; Graph set; Hydrogentartrate; Crystal structure
`
`Ever since Pasteur carried out the first reported separation of
`crystalline enantiomers,' tartaric acid has held a firm place in
`the archives of ground-breaking chemical discoveries. 145
`years later, it is possible that hydrogentartrate salts may play
`a key role in inspiring further development in one of the most
`exciting and challenging of current scientific disciplines: crystal
`engineering.2
`Many elegant studies have illustrated that hydrogen bond-
`ing can be used as a versatile tool, with which molecular
`subunits can be joined together in a relatively predetermined
`However, hydrogen bonding is also known to be
`energetically strong enough to have a profound effect on the
`spatial arrangement of ions in the solid state,7 and it would
`therefore seem reasonable to use the strength and direc-
`tionality of the hydrogen bond as an interionic synthetic route
`towards specific ionic aggregates and material^.^,^
`Earlier work on some organic hydrogentartrates (the mate-
`rials were prepared and investigated for their non-linear
`optical properties)" showed that their crystal structures incor-
`porated a highly ordered, infinite layer of hydrogentartrate
`anions linked together by relatively short O-H-..O
`inter-
`actions. This paper presents an extensive analysis of 12 crystal
`structures of hydrogentartrate salts with the purpose of estab-
`lishing whether an infinite 2D anionic layer is a regular feature
`of these salts. Such aggregates could be used as structural
`building blocks (molecular 'scaffolding'), provided that they
`can be shown to possess structural and spatial consistency in
`the presence of chemically different counterions. This survey
`focuses on the anionic networks, since they form a point of
`reference throughout these materials and they are also
`expected to generate the most distinctive patterns, due to the
`availability of several hydrogen-bond donors and acceptors
`on each anion.
`Potential applications for the design of novel materials with
`predictable structural features and properties can be found in
`almost every commercial and academic area concerned with
`macroscopic structure-property considerations, e.g. non-
`
`t Permanent address: University of Sussex, School of Chemistry and
`Molecular Sciences, Falmer, Brighton, UK BN 1 9QJ.
`
`linear optics, shape-selective catalysis and pharmaceutical
`drug design.
`
`Encoding Hydrogen-bond Patterns
`An important aspect of crystal engineering pertains to the
`manner in which hydrogen-bond patterns are described.
`Etter et al. have developed a methodology''
`for classifying
`hydrogen-bond networks in, primarily, molecular solids.
`However, since the existing graph-set language is not always
`adequate for examining multi-dimensional aggregates with a
`high hydrogen-bond density (a consequence of the inherent
`difficulties with translating 3D information into scalar quanti-
`ties), an additional descriptor,$ based upon the terminology
`of Etter et al., has been utilized as a means of specifically
`representing 2D hydrogen-bonded networks.
`The original encoding" comprises a classification of pat-
`terns generated by specific hydrogen-bond types and by
`combinations of hydrogen bonds. Most patterns in organic
`molecular solids can be adequately described in terms of four
`principal motifs; chains (C), dimers (D), rings ( R ) and intramol-
`ecular hydrogen bonds (S).V Although this notation may not
`
`$ The 2D descriptor utilized in this paper is employed as a means of
`facilitating a comparison and analysis of specific aggregate-types. The
`descriptor may have general applicability to a range of 2D structural
`types and this is currently being investigated. In the course of this
`investigation, certain modifications to the 2D descriptor may
`become necessary.
`lr A graph set is specified using the pattern designator (G), its degree
`(n) and the number of donors (d) and acceptors (a); G; (n). G, the
`descriptor referring to the pattern of hydrogen bonding, can be either
`S (intramolecular bond), C (infinite chain), R (intermolecular ring) or
`D (acyclic dimers and other finite structures) and the parameter n
`refers to the number of atoms in a ring, or the repeat unit of a chain.
`The set of ions/molecules to be analysed is called an array. Graph
`sets are assigned initially to motifs (patterns constructed by only one
`type of hydrogen bond), and then to the first-level network (a
`sequential listing of every motif). For example, if the array contains
`four different hydrogen-bond types, then the first-level graph set is a
`combination of four motifs. Higher-level networks are assigned by
`listing patterns generated by combinations of different hydrogen-
`bond types.
`
`
`
`View Article Online
`
` / Journal Homepage
`
` / Table of Contents for this issue
`
`Apotex Exhibit 1024.001
`
`

`

`1130
`
`J. MATER. CHEM., 1993, VOL. 3
`
`Plate 1 Anionic network of caesium hydrogen-L-tartrate; N,, = L{R:( ll)R:( 17)}.19 Covalent bonds, red; hydrogen bonds, yellow, green and
`white.
`
`Table 1 Aggregate types and a graph-set description for the anionic networks in hydrogentartrate salts”
`r(0 * * .O)/Ab
`
`aggregate type
`
`graph set
`
`L(O-H-..O)/degrees
`
`salt
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`
`double chain
`buckled layer
`buckled layer
`buckled layer
`buckled layer
`buckled layer
`flat layer
`buckled layer
`buckled layer
`flat layer
`buckled layer
`buckled layer
`flat layer
`
`2.49
`2.559
`2.55
`2.51
`2.50
`2.49
`2.438
`2.57
`2.54
`2.498
`2.48
`2.558
`2.50
`
`177
`178
`161
`174
`167
`179
`163
`165
`168
`168
`174
`173
`167
`
`“The search yielded structural data for (a) hydrogen-L-tartrates with the following cations; (-)-3-benzoyl-3-ethyl-l -methylpiperidinium ( 1),13
`rubidium (2),14 S-(+)-N-methylamphetammonium (3),’ 3-methylene-4-(prop-2-enyl)pyrrolidinium (4),16 (-)-adrenalinium (6),” 3-hydroxypyridi-
`nium (7),” caesium (8),19 R-(-)-N-methylamphetammonium (9),” (-)- 1-phenylethylammonium
`piperazinium(2 +) (11),9 (b) ammonium
`hydrogen-D-tartrate ( 12),,l and (c) (-)-1-phenylethylammonium hydrogen-meso-tartrate 13.,, bThe hydrogen-bond interaction responsible for
`the head-to-tail linking of adjacent anions.
`
`provide an unambiguous assignment of every structural
`arrangement, it is flexible enough to facilitate a systematic
`description of a wide range of hydrogen-bonded solids.”
`The presence of hydrogen-bonded, infinite, 2D aggregates
`(sheets or layers) in crystal structures is not a unique phenom-
`enon. Although a hydrogen-bonded sheet can be generated
`by the interconnection of very different (chemically and geo-
`metrically) building blocks, certain structural aspects of a
`
`sheet are amenable to a compact, yet informative, notation.
`One way of describing the topology of a hydrogen-bonded
`sheet is to identify the size, and the number, of unique
`hydrogen-bonded rings within the sheet. This paper will
`employ the following notation for characterizing such aggre-
`gates:
`
`N z D = L{R:(n”R:(n)”R:(n). . .}
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.002
`
`

`

`J. MATER. CHEM., 1993, VOL. 3
`
`where NzD indicates a 2D network, L specifies it to be of a
`layered, or sheet-like, type, and the remaining variables are
`the same as those defined previously.
`Plate 1 shows an example of how this notation is applied.
`Note that NZD does not include intramolecular hydrogen
`-
`-
`bonds since such bonds are not instrumental in connecting
`adjacent molecules/ions into aggregates. Furthermore, only
`
`
`
`1131
`
`rings that do not contain smaller hydrogen-bonded rings as
`part of their structure are listed in this 2D descriptor.
`
`Results and Discussion
`Table 1 summarizes the observed aggregate types and a graph-
`set description for each structure, and it also contains infor-
`
`Plate2 Anionic networks of (a) 2,14 (b) 3,15 and (c) 4.16 The relevant graph sets are listed in Table 1. Covalent bonds, red; hydrogen bonds
`yellow, green and white.
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.003
`
`

`

`1132
`
`mation about the geometry of the O-H-..O hydrogen bond
`(coloured yellow in Plate 1) which is responsible for the head-
`to-tail linking of adjacent anions into infinite chains. This
`interaction is present in every hydrogentartrate structure
`included in this survey.
`The most commonly occurring hydrogen-bonded aggregate
`in these salts is an infinite 2D anionic sheet; a structural
`feature exhibited by 11 of the 12 structures in this family. The
`sheet is generated by crosslinking of neighbouring chains and
`each chain is created by a O-H...O hydrogen bond linking
`anions in a 'head-to-tail' fashion. Adjacent chains are then
`cross-linked either by one O-H-..O hydrogen bond, creating
`graph set, or by two 0-
`a network with a N 2 D = L { & ( n ) }
`H..-O hydrogen bonds, generating a network with a N 2 D =
`L(Ri(n)'Rzn)} graph set. The average O...O distance for the
`'head-to-tail' hydrogen bond is 2.51(4) A with an average 0-
`H..-O bond angle of 170(6)", indicating the presence of a
`strong, almost linear, hydrogen bond.
`The anionic sheet is absent in only one crystal structure,
`(-)-3-benzoyl- 3-ethyl- 1 -methylpiperidinium hydrogen-L-tar-
`trate ( l ) . I 3 There does not seem to be a simple explanation
`for the observed structure of this compound; salts containing
`similar cations are currently being prepared with a view to
`establishing whether this is a structurally unique material.
`The remarkable consistency of the infinite hydrogen-bonded
`sheet is illustrated by the fact that hydrogen-L-tartrate salts
`with very different (both chemically and geometrically) cations
`e.g. rubidium, S-(+)-N-methylamphetaminium and 3-methyl-
`315 and 416) display
`ene-4-(prop-2-enyl)pyrrolidinium, (in
`very similar anionic configurations, Plate 2. The cations of 3
`and 4 are shown below. Consequently, hydrogentartrate
`anions may be employed as a means of incorporating a
`specific structural 2D feature into a crystalline material whilst,
`at the same time, imposing severe restrictions on the positional
`freedom of the cation.
`
`I
`
`I
`
`cation of 3
`
`cation of 4
`
`Design of Novel Hydrogen-bonded Structures
`Hypothesis
`Having established that most of the materials in this diverse
`series are dominated by rigid, structurally consistent, anionic
`networks, a new hydrogentartrate salt 1 H-imidazolium hydro-
`gen-L-tartrate (5) was synthesized. The cation, 1 H-imidazol-
`ium, was chosen in a deliberate attempt to provide a specific
`cross-link between adjacent sheets. The combination of hydro-
`gen-bond donors at opposite ends of a small, rigid spacer,
`and a lack of hydrogen-bond acceptors (which could disrupt
`the anionic network), suggested that this cation would provide
`a cross-link for the anionic scaffolding, thereby creating a
`three-dimensional network. This hypothesis was tested by the
`subsequent preparation and crystallographic characterization
`of 1 H-imidazolium hydrogen-L-tartrate.
`
`Crystal Structure of 1H-Imidazolium Hydrogen-L-tartrate?
`The cation and anion do not exhibit any unusual features
`(Fig. I), with the anion displaying the expected zig-zag con-
`t Fractional coordinates are listed in Table 2 and selected bond
`lengths and angles in Table 3.
`
`J. MATER. CHEM., 1993, VOL. 3
`
`Table 2 Final fractional coordinates (x lo4 for C, N, 0; x lo3 for H)
`and equivalent isotropic thermal factors for 5 with e.s.d.s in
`parentheses
`
`X
`
`Y
`
`2
`
`8234(2)
`5900(2)
`6469(2)
`6592(2)
`9624(2)
`7439(2)
`4076( 3)
`1798(2)
`7 1 55(2)
`761 6(3)
`7904(3)
`8378(3)
`3 198(3)
`3198(3)
`1775(3)
`803(5)
`533(5)
`861(4)
`94(4)
`5W5)
`876(4)
`648(4)
`362(5)
`357(5)
`90(5)
`~ U,, is defined as one third of the trace of the orthogonalised Uij tensor.
`
`3609(2)
`2680(2)
`- 487(2)
`- 482(2)
`- 1990(2)
`- 3466(2)
`5 582( 3)
`5986(3)
`2435(3)
`648(3)
`- 265(3)
`- 2054( 3)
`6739(3)
`4057(3)
`4307(4)
`501(5)
`607( 5)
`90(5)
`65 l(5)
`2 1(5)
`57(5)
`- 146(5)
`804(5)
`3 18(5)
`358(5)
`
`871(0)
`- 290(4)
`- 1436(4)
`2684(3)
`305(4)
`1544(3)
`58 l(4)
`639(4)
`68(4)
`- 370(4)
`1509(4)
`1083(4)
`193(5)
`130 l(5)
`1338(5)
`1 18(8)
`5703)
`-1 1q8)
`5703)
`-162(8)
`223(9)
`298(7)
`- 48(8)
`176(7)
`173(8)
`
`Table3 Selected bond lengths (/A) and angles (/degrees) for 5 with
`e.s.d.s in parentheses
`
`bond lengths/A
`
`1.303(2)
`1.206( 2)
`1.86(4)
`0.176(4)
`1.264(3)
`1.357(3)
`1.31 l(3)
`0.94(4)
`1.525(4)
`1.528(3)
`1.09(4)
`0.86(4)
`
`C( 1)-O( 1)-H(O 1)
`C( 3)-0(4)-H(04)
`C(5)-N( 1)-H(N1)
`C( 5)-N(2)-C(7)
`C(7)-N(2)-H(N2)
`O( I)-C( 1)-C(2)
`0(3)-C(2)-C( 1)
`O( 3)-C(2)-H(2)
`C( 1)-C(2)-H(2)
`O(4)-C( 3)-C( 2)
`0(4)-C(3)-H(3)
`C( 2)-C( 3)-H( 3)
`0(5)-C(4)-0(6)
`O( 6)-C(4)-C( 3)
`N( l)-C(5)-H(5)
`N( l)-C(6)-C( 7)
`C(7)-C(6)-H(6)
`N( 2)-C( 7)- H( 7)
`
`bond angles/degrees
`
`119(2)
`103( 3)
`1 19(2)
`109.0(2)
`124(2)
`1 13.2(2)
`IlOS(2)
`109(3)
`109(2)
`I 09.5( 2)
`107(3)
`108(3)
`126.3(2)
`116.2(2)
`123(2)
`107.4(2)
`130(3)
`122(3)
`
`C(2)-O( 3)-H( 0 3 )
`C(5)-N( 1)-C(6)
`C(6)-N( 1)-H(N 1)
`C( 2)-C( 3)-C(4)
`C( 5)-N( 2)- H( N2)
`O( I)-C( 1)-O(2)
`O(2)-C( 1)-C(2)
`0(3)-C(2)-C( 3)
`C( l)-c(2)-c(3)
`C(3)-C(2)-H(2)
`0(4)-C(3)-C(4)
`C(4)-C( 3)-H(3)
`0(5)-C(4)-C( 3)
`N( 1)-C(5)-N(2)
`N(2)-C( 5)-H( 5)
`N( 1)-C(6)-H(6)
`N(2)-C( 7)-C( 6)
`C( 6)-C( 7)-H( 7)
`
`1.13(4)
`1.4 13(3)
`1.4 1 7( 3)
`1.236(3)
`1.319(4)
`1.1 2(4)
`1.357(4)
`1.522(3)
`1.03(4)
`1.02(4)
`1.331(4)
`0.91(4)
`
`lOl(3)
`108.5(2)
`131(2)
`109.8(2)
`127(2)
`125.1(2)
`12 1.7(2)
`110.1(2)
`10942)
`109( 3)
`112.5(2)
`1 lO(2)
`117.5(2)
`108.2(2)
`1 29( 3)
`123(3)
`106.9(2)
`131(3)
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.004
`
`

`

`J. MATER. CHEM., 1993, VOL. 3
`
`1133
`
`Plate 3 Infinite 2D sheet of anions, parallel to the a-b plane, in 5. Covalent bonds, red; hydrogen bonds, yellow and green.
`
`Plate4 Cross-linking of adjacent anionic sheets by the IH-imidazolium cation in 5. Cationlanion layer parallel to the a-c plane. Covalent
`bonds, red; hydrogen bonds, green, pink and black.
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.005
`
`

`

`1134
`
`Fig. 1 ORTEP drawing (20%) showing the geometry and numbering
`scheme of the cation and anion of 5
`
`torsion angle of almost 180".
`figuration with a C-C-C-C
`The anions form infinite chains through a head-to-tail arrange-
`ment, in which anions from neighbouring unit cells are linked
`along the a axis via a short hydrogen bond r[O(1)...0(6)']
`2.503(2) A, Table 4. These chains are then cross-linked in the
`a-b plane by ions related by the 2, screw axis parallel to b,
`via one hydrogen bond r[0(3)...0(4)"] 3.049(3) A, to form an
`infinite 2D sheet (Plate 3) with the graph set NZD = L(R:(22)),
`Table 1. In addition, there is an intramolecular hydrogen
`bond within the sheet, r[0(4)...0(6)] 2.645(3) A, but as it does
`not provide a link between anions, it does not appear in the
`2D descriptor (although it would be listed in a first-level
`graph set).
`The imidazolium ion does, as intended, provide a cross-
`link between the anionic 'scaffolding' since each cation is
`engaged in two hydrogen bonds, rCN(2). ..0(5)] 2.724(3) and
`r[N( 1)-0(6)'] 3.040(3) A, parallel to c, between adjacent
`anionic layers (Plate 4). This also results in an interlinked
`anion/cation layer in the a-c plane with layers located at
`y z 0 and yz0.5.
`Although it is impossible to predict, a priori, the exact
`arrangement of the cations within this salt, their spatial
`freedom' has been severely restricted by the presence and
`geometry of the anionic sheets, thereby locking them into the
`structural framework in the anticipated fashion.
`
`Summary
`This study has demonstrated. that hydrogentartrate anions
`can be employed as ionic building blocks in crystal engineering
`by virtue of their tendency to create rigid, 2D aggregates
`which restrict and direct the possible orientation of the cation
`in the solid state. This observation also highlights the strength
`and directionality of hydrogen bonds between ions. A novel
`salt, 1 H-imidazolium hydrogen-L-tartrate, was then designed
`based upon the 2D anionic sheet. The cation provided the
`expected cross-link between adjacent sheets, thereby creating
`a 3D hydrogen-bonded network throughout the material. A
`language based upon graph theory" (with an additional 2D
`descriptor) has facilitated the classification and analysis of the
`complex hydrogen-bond patterns present in this series of
`differing hydrogentartrate salts.
`The availability of structurally consistent hydrogen-bonded
`
`J. MATER. CHEM., 1993, VOL. 3
`
`ionic aggregates, in combination with an increased awareness
`of how to restrict the spatial freedom of the counterion, will
`undoubtedly play a crucial role in the future engineering of
`ionic crystalline materials.
`
`Experiment a1
`Analysis of Literature Data
`A search was made of the Cambridge Crystallographic
`Database (CSD, January 1992 Release) for organic and inor-
`ganic hydrogentartrate salts. The search was limited to struc-
`tures of anhydrous hydrogentartrates reported after 1970, and
`only structures without disorder and where hydrogen atoms
`had been located were included. Salts containing transition
`metals were also excluded.
`The relevant crystallographic data were the transferred to
`a Macintosh IIfx and a Silicon Graphics Workstation where
`software from CAChe Tektronix, Molecular Simulation Inc.
`(Cerius 3.2 and Quanta 3.3) and Biosym (Insight/Discover
`2.10) was employed in the analysis of the patterns generated
`by the hydrogen bonds within these structures.
`
`Preparation of 1H-Imidazolium Hydrogen-L-tartrate
`An aqueous solution (20 cm3) of 1H-imidazole (1 .O g; 15 mmol)
`was mixed with an aqueous solution (15 cm3) of L-tartaric
`acid (2.2 g; 15 mmol). The solvent was evaporated by warming
`it on a hot plate until a white precipitate formed. The product
`was collected by filtration and recrystallized from water to
`produce clear, colourless crystals.$ Found: C, 38.4; H, 4.7;
`N, 12.7%. Calculated for C,H,,N206: C, 38.54; H, 4.62;
`N, 12.84%. m.p. 179-181 "C.
`
`X-Ray Structure Determination
`Crystal Data
`C7H10N206, A4 = 218.2, monoclinic, space group P2,, a =
`0.7569(1) nm, b=0.6953(1) nm, c=0.8993(2) nm, p= 101.55(1)",
`U=0.4637 nm3, Z=2, D,= 1.56 g ~ m - ~ ,
` F(000)=228.
`IE = 0.07 1069 nm, ,,I =
`Monochromatic Mo-Ka radiation,
`1.3 cm-', T= 173 K.
`
`Data Collection
`Data were collected on a crystal (0.4 mm x 0.2 mm x 0.2 mm),
`using an Enraf-Nonius CAD4 diffractometer fitted with a cold
`dinitrogen stream low-temperature attachment, in the 8-28
`mode, with A8={0.8+0.35 tan(0))' and a maximum scan
`time of 1 min.
`
`$ In order to verify that the chosen single crystal was representative
`of the bulk material, the X-ray powder diffraction pattern was
`simulated from the single-crystal data (using CERIUS 3.2 from
`Molecular Simulation Inc.) and compared with the experimental
`X-ray powder pattern, recorded on a bulk sample. The match between
`simulated and experimental pattern demonstrated that only one
`structural form of the salt was present.
`
`Table 4 Geometry of the hydrogen bonds in 5 with e.s.d.s in parentheses"
`
`0(1)-H(01)-..0(6)'
`O( 3)- H(03). . .0(4)"
`0(4)-H(04)*..0(6)
`N( 1)- H(N 1). . O(6)'
`N(2)- H(N2)-. -O( 5)"'
`
`1.1 3(4)
`0.86(4)
`0.76(4)
`I. I2(4)
`0.94(4)
`
`1.38(4)
`2.29(4)
`2.14(4)
`1.98(4)
`1.80(4)
`
`2.503(2)
`3.049(3)
`2.645(3)
`3.04q3)
`2.724(3)
`
`167(4)
`148(3)
`125(4)
`158(4)
`166(3)
`
`"Symmetry code: (') x + l , y, z; (") -x, y-4, 1-z; ("') x + l , y , z-1.
`
`~
`
`~~
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.006
`
`

`

`J. MATER. CHEM., 1993, VOL. 3
`
`Solution and ReJlnement of Structure
`A total of 890 reflections were measured for 2<B/degrees
`<25 and +h+k&/, and 841 unique reflections with
`IF2J>30(FZ) were used in the refinement, where u(F2)=
`{aZ(I) + (0.041)2)'12/Lp. There was no crystal decay and no
`correction was made for absorption.
`Non-hydrogen atoms were located by direct methods using
`SHELXS-86 and refined with anisotropic thermal parameters
`by full-matrix least-squares analysis. Hydrogen atoms were
`located on a difference map and their positions refined but
`thermal parameters were fixed at Biso = 0.04 nm2, with a
`weighting scheme of w = l/u2(F). The refinement converged
`to give R=0.041, R'=0.046, S = 1.8 with 165 variables.
`Programs from the Enraf-Nonius SDP-Plus package were
`run on a Micro VaxII computer.
`the Cambridge
`from
`Additional material available
`Crystallographic Data Centre comprises thermal parameters.
`
`Many thanks go to Prof. J. Bernstein (Ben-Gurion University
`of the Negev), Dr. J. C. MacDonald (Harvard University) and
`Prof. K. R. Seddon (The Queen's University of Belfast) for
`helpful suggestions and invaluable discussions. Generous fin-
`ancial support (C.B.A.) from The Royal Swedish Academy of
`Sciences and DRA (Fort Halstead) is gratefully acknowledged.
`
`1
`2
`
`References
`L. Pasteur, Ann. Chim. Phys., 1848,24,442.
`G. R. Desiraju, Crystal Engineering, The Design of Organic Solids,
`Elsevier, Amsterdam, 1989; J. M. Lehn, Angew. Chem. Znt. Edn.
`Engl., 1988, 27, 89; J. M. Lehn, Angew. Chem. Znt. Edn. Engl.,
`1990, 29, 1304; G. M. Whitesides, J. P. Mathias and C. T. Seto,
`Science, 199 1, 254, 13 12; V. Vogtle, Supramolecular Chemistry,
`Wiley, Chichester, England, 1991; M. C. Etter, J. Phys. Chem.,
`1991,95,4601.
`F. Garcia-Tellado, S. J. Geib, S . Goswami and A. D. Hamilton,
`J. .4m. Chem. SOC., 1991,113,9265.
`
`3
`
`1135
`4 X. Zhao, Y-L. Chang, F. W. Fowler and J. W. Lauher, J. Am.
`Chem. SOC., 1990,112,6627.
`5 D. Simard, D. Su and J. D. Wuest, J . Am. Chem. Soc., 1991,
`113,4696.
`6 G. T. Set0 and G. M. Whitesides, J. Am. Chem. SOC., 1990, 112,
`6409.
`7 C. B. Aakeroy, M. Leslie and K. R. Seddon, Struct. Chem., 1992,
`3,63.
`8 C. B. Aakeroy, P. B. Hitchcock and K. R. Seddon, J. Chem. SOC.,
`Chem. Commun., 1989, 1856; C. B. Aakeroy, G. S . Bahra,
`P. B. Hitchcock, Y. Patell and K. R. Seddon, J. Chem. SOC. Chem.
`Commun., 1993, 152; C. B. Aakeroy, D. Phil. Thesis, University of
`Sussex, 1990.
`9 R. Masse and J. Zyss, J . Mol. Eng., 1991,1, 141; J. Zyss, J. Pecaut,
`J. P. Levy and R. Masse, Acta Crystallogr., Sect. B, 1993,49, 334.
`10 C. B. Aakeroy, P. B. Hitchcock and K. R. Seddon, J. Chem. SOC.,
`Chem. Commun., 1992,553.
`11 M. C. Etter, J. C. MacDonald and J. Bernstein, Acta Crystallogr.,
`Sect. B, 1990,46,256; M. C. Etter, Acc. Chem. Res., 1990,23, 120.
`12 T. Steiner and W. Saenger, J. Am. Chem. SOC., 1992,114, 10146;
`R. G. Ball and M. W. Baum, J. Org. Chem., 1992, 57, 801;
`S. Larsen and H. L. Dediego, Acta Crystallogr., Sect. C, 1993,
`49, 303.
`13 J. R. Ruble, G. Hite and J. R. Soares, Acta Crystallogr., Sect. B,
`1976,32, 136.
`14 L. K. Templeton and D. H. Templeton, Acta Crystallogr., Sect. C ,
`1989,45, 675.
`15 E. Fogassy, M. Acs, F. Faigl, K. Simon, J. Rohonczy and
`Z. Ecsery, J . Chem. SOC., Chem. Commun., 1986, 1881.
`16 W. Oppolzer, T. N. Birkinshaw and G. Bernardinelli, Tetrahedron
`Lett., 1990,31,6995.
`17 D. Carlstrom, Acta Crystallogr., Sect. B, 1973,29, 161.
`18 V. A. Tafenko, B. P. Bespalov and V. R. Gakel, Zh. Strukt. Khim.,
`1990,31, 165.
`19 L. K. Templeton and D. H. Templeton, Acta Crystallogr., Sect. A,
`1978,34,368.
`20 E. Molins, C. Miravitlles, F. Lopez-Calahorra, J. Castells and
`J. Raventos, Acta Crystallogr., Sect. C, 1989,45, 104.
`21 A. L. Spek, Acta Crystallogr., Sect. C, 1987,43, 1633.
`22 J. Kroon, A. J. M. Duisenberg and A. F.. Peerdeman, Acta
`Crystallogr., Sect. C , 1984,40, 645.
`
`Paper 3/03632D; Received 25th June, 1993
`
`
`
`View Article Online
`
`Apotex Exhibit 1024.007
`
`