Patterns in Hydrogen Bonding: Functionality and Graph Set Analysis in Crystals

Wiley - Tập 34 Số 15 - Trang 1555-1573 - 1995
Joel Bernstein1,2, R. Davis2, Liat Shimoni1, Nelson Yen-Chung Chang2
1Department of Chemistry Ben‐Gurion University of the Negev Beer Sheva 84105 (Israel) Telefax: Int. code (7)237787
2Department of Chemistry and Biochemistry The University of Texas at Austin Austin, TX 78712‐1167 (USA) Telefax: Int. code + (512)471‐8696

Tóm tắt

AbstractWhereas much of organic chemistry has classically dealt with the preparation and study of the properties of individual molecules, an increasingly significant portion of the activity in chemical research involves understanding and utilizing the nature of the interactions between molecules. Two representative areas of this evolution are supramolecular chemistry and molecular recognition. The interactions between molecules are governed by intermolecular forces whose energetic and geometric properties are much less well understood than those of classical chemical bonds between atoms. Among the strongest of these interactions, however, are hydrogen bonds, whose directional properties are better understood on the local level (that is, for a single hydrogen bond) than many other types of non‐bonded interactions. Nevertheless, the means by which to characterize, understand, and predict the consequences of many hydrogen bonds among molecules, and the resulting formation of molecular aggregates (on the microscopic scale) or crystals (on the macroscopic scale) has remained largely enigmatic. One of the most promising systematic approaches to resolving this enigma was initially developed by the late M. C. Etter, who applied graph theory to recognize, and then utilize, patterns of hydrogen bonding for the understanding and design of molecular crystals. In working with Etter's original ideas the power and potential utility of this approach on one hand, and on the other, the need to develop and extend the initial Etter formalism was generally recognized. It with that latter purpose that we originally undertook the present review.

Từ khóa


Tài liệu tham khảo

From the Introduction to “The Guide of the Perplexed” by the medieval Jewish philosopher/physician Rabbi Moshe ben Maimon also known as the Rambam or Maimonides.

10.1021/ar00172a005

10.1107/S0108768189012929

10.1021/j100165a007

J.Bernstein M. C.Etter J. C.MacDonald J. Chem. Soc. Perkin Trans. 21990 695.

10.1107/S0108768191009345

10.1016/0008-6215(94)00260-M

10.1021/j100004a019

10.1021/cr00032a007

10.1021/ja00103a013

10.1007/BF02252897

10.1107/S0108768193010985

10.1107/S0108270194002180

10.1107/S0108270193014209

10.1107/S0108270194002179

10.1107/S0108270193007012

10.1107/S0108270193007164

10.1107/S0108270193007929

10.1107/S0108270194004439

10.1107/S0108270193006407

10.1021/cm00044a018

10.1021/cm00044a019

10.1021/cm00044a020

10.1021/cm00044a022

10.1021/cm00044a023

10.1021/cm00044a030

10.1021/cm00044a034

10.1021/cm00044a041

10.1021/cm00040a031

10.1021/ja00097a025

10.1021/ja00084a039

C.Glidewell G.Ferguson A. J.Lough C. M.Zakaria J. Chem. Soc. Dalton Trans.1994 1971;

10.1016/0022-328X(94)88028-X

10.1021/j100069a010

10.1021/om00021a029

10.1002/ange.19951072105

Angew. Chem. Int. Ed. Engl.1995 34 in press);

10.1021/cm00044a013

Genkina E. A., 1994, Kristallografiya, 39, 641

10.1107/S0108768193010985

10.1039/CS9932200397

10.1021/ja00065a061

10.1107/S0108768192012278

10.1107/S0108768193000795

10.1107/S0108270192011715

10.1107/S0108270192012927

P.Lightfoot M.Tremayne C.Glidewell K. D. M.Harris P. G.Bruce J. Chem. Soc. Perkin Trans. 21993 1625;

10.1039/JM9930300947

10.1002/jps.2600820908

10.1088/0022-3727/26/8B/010

10.1107/S0108768192007377

10.1021/ja00052a009

C. B.Aakeroy G. S.Bahra P. B.Hitchcock Y.Patell K. R.Seddon J. Chem. Soc. Chem. Commun.1993 152;

10.1007/BF01159140

10.1016/S0960-894X(01)80447-9

N. M.Stanton K. D. M.Harris R. A.Howie J. Chem. Soc. Chem. Commun.1991 1781;

S.Arumgam C.Glidewell K. D. M.Harris J. Chem. Soc. Chem. Commun.1992 724;

C. B.Aakeroy P. B.Hitchcock K. R.Seddon J. Chem. Soc. Chem. Commun.1992 553;

10.1021/jo00029a005

10.1002/poc.610050107

10.1021/cr00012a007

10.1107/S0108270190005078

10.1021/ja00007a037

10.1016/0022-2364(91)90194-X

10.1021/ja00179a028

Note that a hydrogen bond from a cation to an anion or from a molecule to a cocrystallized solvent molecule must be denoted in the first level asD. A‐discrete set that involves more than one hydrogen bond (binary or higher level) would necessarily have more than one donor and/or acceptor and more than two atoms in the pattern so the subscript superscript and parenthetic notation of degree would all be required.

This would correspond to a differentlevelof the graph set assignment (see Section 3).

10.1107/S0567740876002227

10.1107/S0567740880010783

The appearance of these more complex ring notations in some automatic fashion for example in a computerized graph set assignment procedure would be an indication of such an extended ring pattern. The ring system shown here is a pervasive feature in crystal structures of the amino acids.

10.1021/ja00383a012

10.1021/ci00002a004

10.1107/S010876818400210X

10.1021/ja00313a047

10.1007/978-3-642-85135-3

10.1016/0141-8130(82)90048-4

Pauling L., 1960, The Nature of the Chemical Bond, 6

10.1016/0014-5793(86)80730-X

10.1126/science.3892686

10.1016/0022-2836(88)90471-8

10.1007/978-3-642-85135-3

Donohue J., 1968, Structural Chemistry and Molecular Biology

10.1107/S0365110X65003390

10.1107/S0567740868002451

Bernstein J., 1974, Chemistry of the Quinones, 37

10.1107/S0365110X65000439

J.Bernstein E.Harlev unpublished results.

The following discussion is based on the assumption of the availability of rather primitive software for calculating geometry and plotting crystal structures. Surprisingly very few of the currently available molecular modeling software packages have convenient means for displaying and analyzingcrystalstructures as opposed tomolecularstructures.

One of the problems in developing the graph sets stems from the necessity in most cases to generate rather extended descriptions of thecrystal structure including the hydrogen bonding rather than simply themolecularstructure or the “view of the packing” that are sometimes automatically generated. The latter generally does not include sufficient numbers of molecules to portray all the specific intermolecular interactions. In many cases a single view of the crystal structure even in stereo is difficult to interpret. Software for readily generating and rotating crystal structures with a number of unit cells considerably facilitates this task and will make the examination of crystal structures and the assignment of their hydrogen‐bonding graph sets a much more common practice.

Kitaigorodsky A. I., 1955, Organic Chemical Crystallography

L.Shimoni M. Sc. Thesis Ben‐Gurion University of the Negev Beer Sheva Israel 1992.

10.1021/ja01158a079

10.1021/ja00763a016

10.1246/bcsj.53.803

In this admonition the word “all” must be taken in the light of the infinite number of increasingly larger ring systems that appear in a mesh such as that discussed earlier for α‐glycine.

10.1021/ja00259a030

(b) The two hydrogen bonds are crystallographically distinct and should be treated separately as discussed in Section 5.

Martuscelli E., 1967, Ric. Sci., 37, 102

10.1107/S0567740874006820

10.1107/S0567740869005747

10.1021/ma60020a020

10.1107/S0365110X67003640

Ganis P., 1968, J. Phys. Chem., 23, 747

The nomenclature for distinguishing polymorphs has never really been standardized [41]. To avoid even further confusion than already exists in the literature we have chosen in the following discussion to use the names of the polymorphs as they appear in the original literature that is cited here. This has been done at the expense of consistency in style in this paper.

10.1002/jps.2600580802

10.1107/S056774086900505X

10.1007/BF01158908

10.1007/BF01161165

Moore J. C., 1978, Acta Crystallogr. Sect. A, 34, 97

The relative orientation of the two independent molecules in Form 3 suggested to us the possibility of a previously unrecognized element of symmetry (inversion center or twofold axis); a careful check of relationships among the coordinates revealed that no such relationship exists.

10.1021/ja00395a014

10.1098/rspa.1968.0003

10.1107/S0108270185004206

10.1021/ac60032a603

10.1080/15421407208083859

10.1021/j150626a025

10.1021/ja00052a026

10.1107/S0567740877010541

10.1107/S0108270186090182

10.1107/S0365110X65002128

10.1107/S0567740880008606

10.3891/acta.chem.scand.29a-0939

10.1107/S0567740871003583

10.1021/ja01865a029

10.1007/978-3-642-85135-3

10.1107/S0567740878005725

10.1107/S0567740878005737

The original graph set specifications [3] prescribed the degree of aDmotif as “the number of atoms in the entire length of the hydrogen‐bonded set starting with the proton of the first hydrogen bond proceeding along the shortest pathway and ending with the acceptor atom in the last hydrogen bond of the set”. Following those instructions would lead to degree 12 for the binary pattern shown in Fig. 14c; that description would mislead as it does not encompass all the hydrogen bonds that link the discrete cluster. The guideline should be amended so that the degreenincludes counting through all hydrogen bonds that is including all donors and acceptors. The same conclusion was reached by J. C. MacDonald (Ph.D. Dissertation University of Minnesota 1993).

R. E.Davis H.Robb V.Cody unpublished.

H.Nakai M.Takasuka M.Shiro J. Chem. Soc. Perkin Trans. 21984 1459.

10.1107/S0567740882008668

10.1016/0022-2836(69)90488-4

10.1021/bk-1991-0455.ch029

10.1007/BF01131293

J.Emsley M.Arif P. A.Bates M. B.Hursthouse J. Chem. Soc. Chem. Commun.1989 738;

10.1016/S0020-1693(00)83238-7

10.1016/0022-2860(85)80021-1

10.1021/ja00025a028

10.1021/ja00038a092

10.1021/ic00173a027

R. S.Rowland Cambridge Crystallographic Data Centre (UK) personal communication.

10.1107/S0108270184005205

10.1107/S0108270184006570

10.1021/ic50112a043