Toward Complex Matter: Supramolecular Chemistry and Self-Organization

Introductory Perspective

Toward complex matter: Supramolecular chemistry and self-organization

Jean-Marie Lehn*

Institut de Science et d’Inge´nierie Supramole´culaires, Universite´Louis Pasteur, 67000 Strasbourg, France and Colle`ge de France, 75005 Paris, France s the wind of time blows into the sails and provide a vision. This essay therefore patterns (hydrogen bonding arrays, se- Aof space, the unfolding of the uni- will not be extensively documented (numer- quences of donor and acceptor groups, ion verse nurtures the evolution of matter ous reviews and books are available) but coordination sites, etc.). This venture in- under the pressure of information. From rather outline some conjectures for the fu- volved the design and investigation of divided to condensed and on to organized, ture, mainly based on, illustrated by, and more or less strictly preorganized molec- living, and thinking matter, the path is extrapolated from work performed in the ular receptors of numerous types, capable toward an increase in complexity through author’s laboratories. Looking toward the of binding specific substrates with high self-organization. horizon of supramolecular chemistry, and efficiency and selectivity. Thus emerges the prime question set to more generally of supramolecular science Three overlapping phases may be consid- science, in particular to chemistry, the (1, 3), special attention will be given to ered in the development of supramolecular science of the structure and transforma- exposing the forest(s) rather than to describ- chemistry, each exploring a main theme. tion of matter: how does matter become ing the trees! The first is that of molecular recognition complex? What are the steps and the and its corollaries, supramolecular reactiv- processes that lead from the elementary Supramolecular Chemistry and the ity, catalysis, and transport; it relies on de- particle to the thinking organism, the Information Paradigm sign and preorganization and implements (present!) entity of highest complexity? One of the major lines of development of information storage and processing. And there are two linked questions: an chemical science resides in the ever The second concerns self-assembly ontogenetic one, how has this happened, clearer perception, deeper analysis, and and self-organization, i.e., self-processes in how has matter become complex in the more deliberate application of the infor- general; it relies on design and implements history of the universe leading up to the mation paradigm in the elaboration and programming and programmed systems. evolution of the biological world, and an transformation of matter, thus tracing the The third, emerging phase, introduces PERSPECTIVE epigenetic one, what other and what path from merely condensed matter to adaptation and evolution; it relies on self- INTRODUCTORY higher forms of complex matter can there more and more highly organized matter organization through selection in addition be to evolve, are there to be created? toward systems of increasing complexity. to design, and implements chemical diver- Chemistry provides means to interro- In chemistry, like in other areas, the lan- sity and ‘‘informed’’ dynamics. gate the past, explore the present, and guage of information is extending that of build bridges to the future. constitution, structure, and transforma- From Preorganization Toward Molecular chemistry has created a wide tion as the field develops toward more and Self-Organization and Programmed range of ever more sophisticated mole- more complex architectures and behav- Systems: Design

cules and materials and has developed a iors. It will profoundly influence our per- Supramolecular chemistry has first relied on SPECIAL FEATURE very powerful arsenal of procedures for ception of chemistry, how we think about preorganization for the design of molecular constructing them from atoms linked by it, how we perform it. receptors effecting molecular recognition, covalent bonds. Supramolecular chemistry has paved the catalysis, and transport processes (1, 2). Beyond the molecule, supramolecular way toward apprehending chemistry as an Supramolecular preorganization also chemistry aims at developing highly com- information science through the implemen- has provided new ways and means to plex chemical systems from components tation of the concept of molecular informa- chemical synthesis (1, 7–9). Supramolecu- interacting by noncovalent intermolecular tion with the aim of gaining progressive lar, noncovalent synthesis, i.e., the con- forces (1, 2). It has over the last quarter of control over the spatial (structural) and struction of the supramolecular entities a century grown into a major field and has temporal (dynamic) features of matter and themselves, rests on the making and fueled numerous developments at the in- over its complexification through self- breaking of noncovalent bonds following terfaces with biology and physics, thus organization, the drive to life (4–6). an Aufbau strategy incorporated into the giving rise to the emergence and estab- Supramolecular chemistry has devel- design of the molecular components. On lishment of supramolecular science and oped as the chemistry of the entities gen- the other hand, supramolecular assistance technology, as a broad multidisciplinary erated by intermolecular noncovalent in- to synthesis provides a powerful tool in- and interdisciplinary domain providing a teractions (1, 2). Through the appropriate volving first the noncovalent synthesis of a highly fertile ground for the creativity of manipulation of these interactions, it be- supramolecular architecture, which posi- scientists from all origins. The breadth and came progressively the chemistry of motions the components, followed by post- depth of its scope is evidenced and illus- lecular information, involving the storage assembly modification through covalent trated by the wide selection of many of the of information at the molecular level, in bond formation. Both areas will continue major players in the field gathered in the the structural features, and its retrieval, Special Feature in this issue of PNAS. transfer, and processing at the supramo- Rather than adding another facet to this lecular level, by interactional algorithms Abbreviations: CDC, constitutional dynamic chemistry; already breathtaking panorama, it appeared operating through molecular recognition DCC, dynamic combinatorial chemistry. appropriate here to emphasize perspectives events based on well-defined interaction *E-mail: [email protected].

www.pnas.org͞cgi͞doi͞10.1073͞pnas.072065599 PNAS ͉ April 16, 2002 ͉ vol. 99 ͉ no. 8 ͉ 4763–4768 Downloaded by guest on September 25, 2021 to provide in the future a range of highly tions (such as secondary metal coordina- information and may display kinetic con- sophisticated noncovalent as well as cova- tion, van der Waals stacking, etc.) or trol, generating kinetic products before lent entities. A particularly impressive il- toward modifications of parameters (such reaching the thermodynamic one(s). This lustration of the latter is the synthesis of as concentrations and stoichiometries of is the case in the initial assembly of a triple interlocked compounds (see below). the components, presence of foreign spe- helical complex that evolves toward a cir- Beyond preorganization lies the design cies, etc.). When the assembly occurs only cular helicate (20). Such a process may of systems undergoing self-organization, in a narrow range of conditions, the system either be sequential, if the kinetic product i.e., systems capable of spontaneously gen- is unstable and presents a singularity; it is an intermediate located on the pathway erating well-defined, organized, and func- may also display a bifurcation or a switch- toward the final product, or it may be tional supramolecular architectures by ing point between different assemblies. bifurcated, if this is not so. self-assembly from their components, thus On the other hand, sensitivity to pertur- A sequential process may be either behaving as programmed systems (1, 10). bations, while limiting the operation commutative, if given steps may be inter- Chemical programming requires the in- range, introduces diversity and adaptabil- changed along the overall pathway leading corporation into molecular components ity (3) in the self-organization process. to the final superstructure, or it may of suitable instructions for generation of a Self-selection with self-recognition oc- be noncommutative if its progressive well-defined supramolecular entity. The curs when the structural instructions are build-up occurs through a defined se- program is molecular, the information sufficiently strong, as is the case in the quence of molecular instructions and al- being contained in the covalent structural ‘‘correct’’ pairing of strands of different gorithms, where the generation of a given framework; its operation is supramolecu- lengths in the assembly of helicates, inor- intermediate depends on the previous one lar, making use of recognition algorithms ganic double helices (11). The process and sets the stage for the next one. For based on specific interaction patterns. Un- bears relation to the implementation of example, discotic liquid crystals form by derstanding, inducing, and directing self- combinatorial chemistry in its ‘‘dynamic’’ the assembly of ‘‘sector’’-shaped compo- processes is key to unraveling the progres- version (see below). It reveals a broader nents into disks, which thereafter organize sive emergence of complex matter. Self- perspective with a paradigm shift, from into columns (21), and template bind- organization is the driving force that led ‘‘pure compounds’’ to ‘‘instructed mix- ing by molecular strands induces wrapping up to the evolution of the biological world tures’’, from ‘‘unicity’’ to ‘‘multiplicity ϩ into helical disk-like objects, which from inanimate matter (4–6). information’’ (11). Rather than pursuing then aggregate into large supramolecular Whereas self-assembly may be taken as mere chemical purity of a compound or a assemblies (22). simple collection and aggregation of com- material, one seeks to design instructed The generation of supramolecular ar- ponents into a confined entity, we shall components, which, as mixtures, allow the chitectures and materials through a non- here consider self-organization as the controlled assembly of multiple well- commutative sequence of steps amounts spontaneous but information-directed defined supramolecular species, following to multilevel hierarchical self-organiza- generation of organized functional struc- specific programs and interactional algo- tion, along primary, secondary, tertiary, tures in equilibrium conditions. A relevant rithms. The implementation of this in- etc. structures. Such conditional processes biological example is for instance the for- structed mixture paradigm is crucial for enable the progressive build-up of more mation of a virus particle from its com- the development of complex chemical and more complex systems in a directed, ponents, genomic nucleic acid and coat systems, as witnessed by the build-up of temporally ordered fashion; they also of- proteins. The inclusion of dissipative, non- organized species and the execution of fer the intriguing possibility to intervene equilibrium processes, as present in the highly integrated functions that take place at each step so as either to suppress the living world, constitutes a major goal and side-by-side in the assembly and operation following ones or to reorient the subse- challenge for the future (4–6). of the machinery of the living cell. quent evolution of the system into another A self-organization process may be con- Intense activity has been devoted to the direction, toward another output entity. sidered to involve three main stages: (i) explicit application of molecular recogni- molecular recognition for the selective tion to control the formation, from their Multiple Self-Organization Through ͞ binding of the basic components; (ii) components, of organized supramolecular Multiple Processing Expression of growth through sequential and eventually entities presenting specific physical and Molecular Information hierarchical binding of multiple compo- chemical properties. Three main types of The generation of a given superstructure nents in the correct relative disposition; it investigations have been pursued, based through self-organization results, in its may present cooperativity and nonlinear on the use of hydrogen bonding, donor simplest form, from the operation of a behavior; and (iii) termination of the pro- acceptor, and metal coordination inter- single-code assembly program. A step be- cess, requiring a built-in feature, a stop actions for controlling the processes yond consists in devising systems of higher signal, that specifies the end point and and holding the entities together (12–18). complexity that operate in multimode signifies that the process has reached The latter have been instrumental in the fashion through the implementation of completion. initial introduction of the notions of self- several codes within the same overall Suitable encoding by manipulation of assembly into supramolecular chemistry program, resulting in multiple self- structural subunits and processing (1, 10). The clever exploitation of tem- organization processes (3, 23). through interactional algorithms should plating and self-organization has given Such behavior may take place in the give access to a variety of systems. More or access to a range of molecular and generation of different metallo-architec- less strict programming of the output spe- supramolecular entities of truly impressive tures from the same ligand when using cies may be achieved depending on the structural complexity, which otherwise different sets of metal ions͞coordination robustness of a given directing code (for would have been considered impossible algorithms for reading the binding infor- instance, of hydrogen bonding or metal to construct, such as interlocked entities, mation, as in the generation of two dif- coordination nature), i.e., on the extent to whose components are mechanically ferent helicates from the same strand (3, which it is sensitive to perturbations. In a held together (15, 19) and multicom- 24) and in the assembly of ligands con- robust system the instructions are strong ponent organic or inorganic architectures taining two different subunits coding, re- enough for ensuring the stability of the (1, 12–18). spectively, for the formation of a helicate process, i.e., the self-organization is resis- Because it is a time-dependent process, anda[2ϫ 2] grid-type complex (25). tant, stable toward interfering interac- self-organization also involves temporal Similarly, the differential expression of

4764 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.072065599 Lehn Downloaded by guest on September 25, 2021 hydrogen bonding information contained generated by different modes of process- namic diversity. It allows us to perform in a molecular strand may yield different ing the same information. selection of a given constituent, made up supramolecular structures depending on The combination of different recogni- of a well-defined set of components in the the processing mode (angular or linear) tion͞instruction features in a molecular pro- pool of compounds having all possible defined by the recognition algorithm of gram opens a door to the design of self- constitutions, under the pressure of either the bound effector (22). These consider- organizing systems capable of performing internal [intrinsic stability of the species, ations lead to conjectures that may have molecular computation. Computing by self- as in helicate self-recognition (11)] or far-reaching implications and open novel assembly may yield a powerful alternative to external [interaction with species in the perspectives within the general framework conventional models (28, 29). Recent stud- environment, as in anion binding by cir- of self-organizing, programmed chemical ies described the use of biomolecules and cular helicates (34, 35)] factors. systems (23). DNA-based protocols to solve computa- CDC constitutes thus a general area of The processing of the same ligand in- tional problems (30–32). An approach mak- which one specific expression, when under formation by different interaction algo- ing use of specifically designed non-natural combinatorial conditions, is DCC that has rithms (e.g., through the use of different components could provide higher diversity, actively developed in recent years (27, 36, sets of metal ions or of different H- better resistance to fatigue, and more com- 37). It relies on the dynamic generation of ͞ bonding effectors) allows the controlled pact smaller size. Such potential is latent in molecular and supramolecular diversity self-organization of different output ar- the coordination-controlled assembly through the reversible combination of co- chitectures, resulting in multiple expres- of double helicates (3, 24) and metallo- valently or noncovalently linked building sion of molecular information, through supramolecular architectures (25), as well as blocks (components). Whereas combina- postinformational (postgenomic!) opera- in the differential folding processes induced torial chemistry itself is based on extensive tions. Such a one code͞several outputs by effector H-bonding (22). Numerous libraries of prefabricated molecules, DCC scheme, in addition to the one code͞one types of interactions and recognition units, implements the reversible connection of output (product) mode, also has in prin- be they of inorganic or organic nature, are pools of basic components to give access ciple significant implications in biology. available for exploring these avenues. to virtual combinatorial libraries (VCLs) Multisubroutine self-assembly may dis- The combination of multiple expression whose constituents comprise the full set of play three types of behavior: (i)itmay with reversible diversity generation sug- all possible combinations that may poten- behave as a linear combination of the gests the notion of dynamic computing tially be generated in dynamic equilib- through dynamic information generation rium. The constituents actually expressed subprograms, each running independently ͞ to generate its own encoded output; self- and processing, defining adaptive evolu- among all those accessibles are expected tive programmed systems. Evolutionary to be those presenting the strongest inter- recognition (11) is a related process; (ii)it computation may be envisaged with bi- action with a given target, i.e., the fittest. may present crossover, when the subpro- omolecules (28, 32, 33), but entirely syn- DCC bypasses the need to actually syn- grams operate in a combined fashion; or thetic molecular systems or mixed abio͞ thesize the library constituents and lets (iii) it may also be of dominant͞recessive bio ones should in the future be amenable the target select the optimal partner by type, one of the subprograms imposing its to similar feats, with probably advantages inducing its preferential assembly from its PERSPECTIVE

own output over the other one(s) (26). INTRODUCTORY in design, control, and diversity. components, eventually with a facilitation Multiple processing capacity represents of the connecting reaction. It is thus a a further step in the design of pro- Dynamic Chemistry and Constitutional target͞function-driven self-organization, grammed chemical systems of increasing Diversity: Selection i.e., a self-design process (see below). The complexity capable of producing a variety Supramolecular chemistry is by nature a basic features of the DCC͞VCL approach of more and more complex architectures dynamic chemistry in view of the lability of have been presented together with its im- as outputs. the interactions connecting the molecular plementation in different fields and the Parallel processing, extending eventu- components of a supramolecular entity. perspectives it offers in a variety of areas

ally to massively parallel systems, would The reversibility of the associations allows of science and technology, such as the SPECIAL FEATURE involve the simultaneous operation of a continuous change in constitution, which discovery of biologically active substances, multiple self-organization processes to- may be either internal, by rearrangement new materials, and catalysts, etc. (27). ward the generation of a single functional of the components with modification of CDC introduces a profound change in entity or several different ones. The side- the connectivity between them, or exter- paradigm and opens a range of novel by-side formation of different helicates nal, by exchange, incorporation, or extru- perspectives with respect to constitution- (11) or helicates and inorganic grids (D. sion of components, therefore conferring ally static chemistry (see also refs. 38 and Funeriu and J.-M.L., unpublished work) constitutional plasticity to the system. 39). Whereas the latter relies on design for in a mixture of the corresponding ligands Thus, supramolecular chemistry is a con- the generation of a target molecule or and suitable metal ions may be seen as stitutional dynamic chemistry (CDC). supermolecule, CDC takes advantage of prototypes on a simple level. Dynamic chemistry also can be just dynamic diversity to allow variation and Multiple processing of a single set of morphological, involving reversible implements selection to achieve adapta- instructions allows the generation of di- changes in shape through molecular or tion in a darwinistic fashion. versity, because multiple outputs may ei- supramolecular conformational or config- ther coexist, or be potentially accessible urational modifications, without change in Self-Organization by Design and (virtual diversity) (27). It thus meets dy- (internal or external) constitution and re- Selection: Self-Design namic combinatorial chemistry (DCC) sulting in motional processes. Whereas preorganization relies entirely (see below). CDC may be molecular as well as su- on design, supramolecular self-organiza- Conversely, such chemical systems also pramolecular when the components of the tion introduces in addition the possibility open perspectives for information science, molecular entity are linked by covalent to let the system build up by selection. inasmuch as they raise the question of bonds that may form and break reversibly. Self-organization by design has been pur- going beyond the usual one-to-one corre- This ability to undergo continuous and sued with the goal to achieve full control spondence, established by a given pro- reversible change by reorganization, de- over the output supramolecular entity by gram, between the input information and construction, and reconstruction (alike or means of correctly instructed components, a single output, toward multiple outputs different) generates constitutional dy- specific interaction algorithms, and (as

Lehn PNAS ͉ April 16, 2002 ͉ vol. 99 ͉ no. 8 ͉ 4765 Downloaded by guest on September 25, 2021 much as possible) strict programming. De- processes have been used for developing cial for the development of chemical sys- sign is knowledge-based and has an explicit prototypes of molecular electronic compo- tems of both structural and reactional information content. nents, such as molecular wires, batteries, complexity. It has played a key role in Self-organization by selection requires and rectifiers. The combined operation of biological evolution (4) and presents a dynamic diversity (constitutional and͞or photo- and electro-effects gives rise to major challenge to chemistry. morphological) on which to operate. This photoinduced electron transfer and charge is made possible by the implementation of separation, so that a major thrust has been Functional Supramolecular Materials CDC responding to the pressure of either toward the mimicking of the first steps of The properties of a material depend both internal or external factors. Selection has photosynthesis (45). The search for artificial on the nature of its constituents and the an implicit information content. It is also ion carriers and ion channels provides the interactions between them. Supramolecu- truly a supramolecular process, because it basis for (supra)molecular ionics (41, 42, 46, lar chemistry may thus be expected to occurs in relation to interactions with sur- 47). Controlling the transfer of photons, have a strong impact on materials science roundings (which may be either the me- electrons, and ions sets the stage for semio- by means of the explicit manipulation of dium or a more or less distant part of a chemistry (10), the chemistry of signal gen- the noncovalent forces that hold the con- folded macromolecule). eration and processing, of interest for in- stituents together. These interactions and The introduction of the selection para- stance in devising sensing and logic the recognition processes that they under- digm into (supramolecular) chemistry functions (48, 49). Developments toward lie allow the design of materials and the brings about a fundamental change in supramolecular technology have been ac- control of their build-up from suitable ways, means, and outlook. Of course, the tively pursued, concerning in particular sen- units by self-organization. Thus, supramo- question is not to replace the deliberately sors and other optical or electronic devices, lecular chemistry opens new perspectives planned linear process of design by a possibly of interest for ‘‘molecular comput- in materials science toward an area of multipronged trial-and-error process of ing’’ (42, 50). supramolecular materials, ‘‘smart’’ mate- selection. Design and selection are not Mechano-devices, effecting triggered rials whose features depend on molecular mutually exclusive but are complementary molecular motions, belong to a general information. For instance, liquid-crystal- for reaching systems of higher complexity area of dynamic devices. Molecular mo- line polymers of supramolecular nature through self-organization. The ultimate tions and changes in molecular shape may have been obtained by the self-assembly of goal is to merge design and selection in be produced through external stimuli in complementary subunits (56–58). self-organization to perform self-design, various systems, e.g., cis-trans isomeriza- Supramolecular materials are by nature where function-driven selection among tion in azobenzene derivatives and optical dynamic materials, materials whose constit- suitably instructed dynamic species gener- changes in photochromes are accompa- uents are linked through reversible connec- ates the optimal organized and functional nied by large geometric variations. Inter- tions and which may undergo assembly͞ entity. In fact, self-recognition in helicate locked and intertwined structures have deassembly processes in specific conditions. self-assembly from mixtures of ligand been used for the photochemical or elec- Because of this intrinsic ability to exchange strands (11), the selection of monomers in trochemical induction of relative motion their constituents, they are constitutionally the synthesis of molecular helices under between the mechanically linked compo- dynamic materials and may in principle se- the pressure of folding (40) as well as the nents. They have given access to a range of lect their constituents in response to exter- formation of dynamic helical polymers intriguing processes (e.g., shift registers, nal stimuli or environmental factors. Su- (J.-L. Schmitt and J.-M.L., unpublished circular displacements) related in partic- pramolecular materials thus are instructed, work) may be perceived as such internally ular to the design of ‘‘molecular ma- dynamic, and combinatorial and behave as driven systems. chines’’ and the induction of directional adaptive materials (3). Altogether, harnessing the power of se- motion (51–53). A rich domain emerges from the combi- lection for adaptation and evolution on While the above devices present physi- nation of polymer chemistry with supramo- the molecular scene is ushering in a dar- cal functionality, chemically reactive self- lecular chemistry defining a supramolecular winistic era into chemistry, where the organized entities are formed when the polymer chemistry (56–58). It involves the fittest species survive. assembling brings together components designed manipulation of molecular inter- bearing reactive functional groups. actions (hydrogen bonding, donor-acceptor Self-Organization of Functional Through the appropriate disposition of effects, etc.) and recognition processes to Supramolecular Systems specific subunits in an organized pattern, generate main-chain (or side-chain) su- The self-organization of functional su- they may be amenable to perform efficient pramolecular polymers by the self-assembly pramolecular entities concerns either and selective reactions and catalysis. Such of complementary monomeric components discrete species or, on the other hand, supramolecular processes involve first a (or by binding to lateral groups). In view of extended assemblies in one (e.g., polymo- substrate recognition step followed by a their lability, these associations present fea- lecular chains, fibers), two (e.g., layers, chemical transformation on the bound tures of reversible, ‘‘living’’ polymers capa- membranes), and three (e.g., solids) entity (1), resulting in an activity of arti- ble of growing or shortening, of rearranging dimensions. ficial enzyme type. When the reactions or exchanging components. Supramolecular Functional supramolecular devices are occur within the self-assembled entity, polymers are thus constitutionally dynamic based on the structural organization and they amount to self-transformation and materials, based on dynamic polymer librar- functional integration of active compo- may in particular result in replication and ies whose members possess constitutional nents operating with photons, electrons, self-replication processes (39, 54, 55). diversity determined by the nature and va- and ions or presenting chemical reactivity. They may present autocatalysis, a behav- riety of the different monomers. The com- Photoactive and electroactive devices ior that together with the establishment of ponents effectively incorporated into the performing energy or electron exchange͞ networks of reactions and coupled cata- polyassociations depend in particular on the transfer processes form the core of molec- lytic cycles amounts to self-organization nature of the recognition and core groups, ular and supramolecular photonics and on the chemical reactivity level, present- internal structural compatibility, as well as electronics (41–45). They have, for instance, ing features such as self-regulation, feed- the interactions with the environment. led to labeling and detection procedures of back, and amplification. The controlled These features give access to higher levels of interest for biological studies and medical self-organization of functional systems behavior such as healing, adaptability, re- diagnostics based on energy transfer. Redox displaying reactivity and catalysis is cru- sponse to external stimulants (heat, light,

4766 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.072065599 Lehn Downloaded by guest on September 25, 2021 additives, etc.) by association͞growth͞ cies of nanometric dimensions, defines a Fabricating, manipulating, and imple- dissociation sequences. supramolecular nanochemistry. menting nano-size chemical entities offer The selective, recognition-controlled a wide range of potential applications of incorporation of components presenting Supramolecular Nanochemistry and great value for science and technology specific functional properties (energy Nanomaterials (70). Reducing size to the nano-object and transfer, electron transfer, ion binding, Nanoscience and nanotechnology are re- addressing it are admirable feats that pro- etc.) allows us to envisage applications for ceiving great attention in view of both vide entirely new insight into the proper- such diverse purposes as drug delivery, their basic interest and their potential ties and functioning of chemical as well as gene transfer, mechanical action (e.g., applications (65, 66). Here again, su- biological systems. However, in the long triggered changes in shape or size), vis- pramolecular chemistry and self-organi- run, the goal is complex organization and cosity adjustment, hydrophilicity͞hydro- zation contribute a fundamentally novel collective operation rather than smaller phobicity modulation, optical and elec- outlook of deep impact. Three approaches size and individual addressing. And here tronic effects, etc. to these areas may be considered. the path is traced by self-organization, Supramolecular versions of the various (i) The miniaturization, top-down ‘‘size- covering a full range of self-processes that species and procedures of molecular poly- shrinking’’ approach, mainly pursued up determine the internal build-up and the mer chemistry may be imagined and im- to now, has led to the tremendous achieve- operation of the entity (self-selection, self- plemented, providing a wide field of fu- ments of the microelectronics technology wiring), as well as its interaction with the ture investigation that may offer a wealth and is pushing down the limits of size and environment (self-connection for address- of novel entities and functionalities. Sim- compactness of components and devices. ing and sensing). It follows also a bot- ilar considerations apply to the generation (ii) The nanofabrication and nano- tom-up scale change by growth from the of supramolecular liquid crystals (56–58). manipulation bottom-up approach to mo- nanolevel to the meso level and to the Molecular recognition may be used to lecular nanotechnology, takes advantage macrolevel with internal structural orga- induce and control self-organization in of novel nanolevel materials and methods nization, functional integration, and ex- two and three dimensions for performing (e.g., near-field scanning microscopies) to ternal connection. Indeed, the most com- supramolecular engineering of polymo- generate nanoentities presenting intrigu- plex object around, the brain, builds up by lecular assemblies and materials, layers, ing potential, such as electrical devices self-organization and is self-wired and films, membranes, micelles, gels, me- built on carbon nanotubes (for two recent self-integrated as well as self-connected sophases, and solids as well as on surfaces examples, see refs. 67 and 68) or optical through our senses! or at interfaces. devices like optical sieves (69). Surfaces modified with recognition (iii) The supramolecular self-organiza- Toward Adaptive and Evolutive units may display selective surface binding tion approach, where the goal is not Chemistry on the microscopic level and recognition- smaller size or individual addressing but The combination of the features of su- controlled adhesion on the macroscopic complexity through self-processing, pramolecular systems—information and scale. Intermolecular interactions may be strives for self-fabrication by the con- programmability, dynamics and reversibil- brought to induce the controlled assembly trolled assembly of ordered, fully inte- ity, constitution and diversity—leads toward PERSPECTIVE of macroscopic objects as is the case with grated, and connected operational sys- the emergence of adaptive͞evolutive chem- INTRODUCTORY capillary forces (59). tems by hierarchical growth. istry (3). It is, by essence, of supramolecular Self-organization of polymolecular as- The first two approaches rely on design nature because it is determined by interac- semblies reaches a second level in the and implement physical procedures. Self- tion with an external entity. It may be con- self-organization of objects that are them- organization may take advantage of both stitutional and͞or morphological (as occurs selves self-organized. Vesicles are of spe- design and selection, through its in- in ‘‘induced fit’’ for instance). Adaptive cial interest in this respect, because com- formed, dynamic and adaptive features, chemistry implies selection and growth un- partmentalization must have played a and finds inspiration in the integrated der time reversibility.

major role in the self-organization of com- processes of biological systems. Implementing both design and selec- SPECIAL FEATURE plex matter and thus in the evolution of Indeed, the spontaneous but controlled tion, self-organization offers adjustability living cells and organisms. One may en- generation of well-defined, functional (through self-correction, self-healing un- visage the controlled build-up of architec- supramolecular nanostructures through der internal dynamics); adjustability leads turally organized and functionally inte- self-organization offers a very powerful to adaptation (through reorganization un- grated polyvesicular systems toward the alternative to nanofabrication and nano- der interaction with environmental effec- design of artificial cells and polymolecular manipulation, bypassing the implementa- tors); adaptation becomes evolution, systems of tissue-like character, imple- tion of tedious procedures and providing a when acquired features are conserved and menting specific intravesicular and in- chemical approach to nanoscience and passed on. tervesicular processes (60, 61). The ma- technology. Rather than having to top- Adaptation is illustrated by functionally nipulation of the features of vesicles and down prefabricate or to stepwise construct driven optimization through selection from their behavior is a step in this direction; nanostructures, more and more powerful pools of dynamically interconverting su- thus, liposomes decorated with recogni- methodologies resorting to self-organiza- pramolecular species. Evolutive chemical tion groups, recosomes, present features tion from instructed components will give systems suppose multiple dynamic processes such as selective interaction with molecu- access to highly complex functional archi- with sequential selection͞acquisition͞ lar films, aggregation, and fusion (62, 63). tectures (1). Their dynamic features, al- fixation steps and undergo progressive Molecular recognition interactions also lowing constitutional modification change of internal structure under the pres- provide a powerful entry into solid-state through exchange of components, confer sure of environmental factors. But the world chemistry and crystal engineering. The to them the potential to undergo healing of selection is a brutal world, where only the increasing ability to control the way in and adaptation, processes of great value fittest survives. It is ultimately for thought which molecules associate gives means for for the development of ‘‘smart’’ nanoma- and design to open up a post-Darwinian era the designed generation of supramolecu- terials. Of course, various combinations of by recruiting the forces of information to lar architectures in the solid state (64). self-organization and fabrication proce- override the dictate of selection! The generation of self-organized nano- dures may be envisaged and implemented Beyond programmed systems and in structures, organized and functional spe- at different stages. line with an evolutive chemistry, the next

Lehn PNAS ͉ April 16, 2002 ͉ vol. 99 ͉ no. 8 ͉ 4767 Downloaded by guest on September 25, 2021 step in complexity consists in the design of Complexity implies and results from mul- eling the complexification of matter chemical ‘‘learning’’ systems, systems that tiplicity of components, interaction between through self-organization. The Special are not just instructed but can be trained them and integration i.e., correlation, cou- Feature in this issue of PNAS takes stock (see for instance ref. 71). pling, and feedback (1). The species and and outlines prospects for many of the The incorporation of the arrow of time, properties defining a given level of complex- activities pursued in the field. time irreversibility, leads to self-organiza- ity result from and may be explained on the Together with the corresponding areas in tion in nonequilibrium, dissipative systems basis of the species belonging to the level physics and biology, supramolecular chem- through irreversible processes (4, 5). It below and of their multibody interaction, istry builds up a supramolecular science implies the passage from closed systems to e.g., supramolecular entities in terms of whose already remarkable achievements open and coupled systems that are con- molecules, cells in terms of supramolecular point to the even greater challenges that lie nected spatially and temporally to their entities, tissues in terms of cells, organisms ahead. They lead toward a science of com- surroundings. in terms of tissues and so on, up to the plex matter, of informed, self-organized, Supramolecular Science, the Science of behavior of societies and ecosystems along a evolutive matter. The goal is to progressively Informed, Complex Matter hierarchy of levels defining the architecture discover, understand, and implement the In the long-range perspective, the devel- of complexity. At each level of increasing rules that govern its evolution from inani- opment of chemical science is toward complexity novel features emerge that do mate to animate and beyond, to ultimately complex systems, spanning the broadest not exist at lower levels, which are deducible acquire the ability to create new forms of outlook from divided to condensed matter from but not reducible to those of lower complex matter. The perspectives opened then to organized and adaptive matter, on levels. up become wider and wider as progress is to living matter and thinking matter, up Supramolecular chemistry provides being made and will constitute horizons of the ladder of complexity. ways and means for progressively unrav- the future for quite some time!

1. Lehn, J.-M. (1995) Supramolecular Chemistry: 23. Lehn, J.-M. (2000) Chem. Eur. J. 6, 2097–2102. 48. Irie, M., ed. (2000) Chem. Rev. 100, 1683–1890. Concepts and Perspectives (VCH, New York). 24. Smith, V. & Lehn, J.-M. (1996) Chem. Commun. 49. de Silva, A. P. & McClenaghan, N. D. (2000) 2. Atwood, J. L., Davies, J. E. D., MacNicol, D. D., 2733–2734. J. Am. Chem. Soc. 122, 3965–3966. Vo¨gtle, F. & Lehn, J.-M., eds. (1996) Comprehensive 25. Funeriu, D. P., Lehn, J.-M., Fromm, K. M. & 50. Czarnik, A. W. & Desvergne, J.-P., eds. (1997) Supramolecular Chemistry (Pergamon, Oxford). Fenske, D. (2000) Chem. Eur. J. 6, 2103–2111. Chemosensors for Ion and Molecule Recognition 3. Lehn, J.-M. (1999) in Supramolecular Science: 26. Funeriu, D. P., Rissanen, K. & Lehn, J.-M. (2001) (Kluwer, Dordrecht, The Netherlands). Where It Is and Where It Is Going, eds. Ungaro, R. Proc. Natl. Acad. Sci. USA 98, 10546–10551. 51. Balzani, V., Credi, A., Raymo, F. M. & Stoddart, & Dalcanale, E. (Kluwer, Dordrecht, The Neth- 27. Lehn, J.-M. (1999) Chem. Eur. J. 5, 2455–2463. J. F. (2000) Angew. Chem. Int. Ed. 39, 3348–3391. erlands), pp. 287–304. 28. Conrad, M. (1993) Nanobiology 2, 5–30. 52. Stoddart, J. F., ed. (2001) Acc. Chem. Res. 34, 4. Eigen, M. (1971) Naturwissenschaften 58, 465–523. 29. Rothemund, P. W. K. (2000) Proc. Natl. Acad. Sci. 409–522. 5. Yates, F. E., ed. (1987) Self-Organizing Systems USA 97, 984–989. 53. Sauvage, J.-P. (1998) Acc. Chem. Res. 31, 611–619. (Plenum, New York). 30. Adleman, L. M. (1994) Science 266, 1021–1024. 54. Bag, B. G. & von Kiedrowski, G. (1996) Pure Appl. 6. Brackmann, S. (1997) Biophys. Chem. 66, 133–143. 31. Benenson, Y., Paz-Elizur T., Adar, R., Keinan, E., Chem. 68, 2145–2152. 7. Whitesides, G. M., Simanek, E. E., Mathias, J. P., Livneh, Z. & Shapiro, E. (2001) Nature (London) 55. Robertson, A., Sinclair, A. J. & Philp, D. (2000) Seto, C. T., Chin, D. N., Mammen, M. & Gordon, 414, 430–434. Chem. Soc. Rev. 29, 141–152. D. M. (1995) Acc. Chem. Res. 28, 37–44. 32. Chen, J. & Wood, D. H. (2000) Proc. Natl. Acad. 56. Ciferri, A., ed. (2000) Supramolecular Polymers 8. Fyfe, M. C. T. & Stoddart, J. F. (1997) Acc. Chem. Sci. USA 97, 1328–1330. (Dekker, New York). Res. 30, 393–401. 33. Conrad, M. (1985) Commun. Assoc. Computing 57. Lehn, J.-M. (2000) in Supramolecular Polymers, 9. Prins, L. J., Reinhoudt, D. N. & Timmerman, P. Machinery 28, 464–480. ed. Ciferri, A. (Dekker, New York), pp. 615–641. (2001) Angew. Chem. Int. Ed. 40, 2382–2426. 34. Hasenknopf, B., Lehn, J.-M., Kneisel, B. O., 58. Brunsveld, L., Folmer, B. J. B., Meijer, E. W. & 10. Lehn, J.-M. (1990) Angew. Chem. Int. Ed. Engl. 29, Baum, G. & Fenske, D. (1996) Angew. Chem. Int. Sijbesma, R. P. (2001) Chem. Rev. 101, 4071–4097. 1304–1319. Engl. 35, 1838–1840. 59. Bowden, N. B., Weck, M., Choi, I. S. & Whitesides, 11. Kra¨mer, R., Lehn, J.-M. & Marquis-Rigault, A. 35. Hasenknopf, B., Lehn, J.-M., Boumediene, N., G. M. (2001) Acc. Chem. Res. 34, 231–238. (1993) Proc. Natl. Acad. Sci. USA 90, 5394–5398. Dupont-Gervais, A., Van Dorsselaer, A., Kneisel, 60. Lasicˇ, D. D. (1993) Liposomes: From Physics to 12. Atwood, J. L., Davies, J. E. D., MacNicol, D. D., B. & Fenske, D. (1997) J. Am. Chem. Soc. 119, Applications (Elsevier, Amsterdam). Vo¨gtle, F. & Lehn, J.-M., eds. (1996) Comprehen- 10956–10962. 61. Menger, F. M. & Gabrielson, K. D. (1995) Angew. sive Supramolecular Chemistry (Pergamon, Ox- 36. Cousins, G. R. L., Poulsen, S. A. & Sanders, Chem. Int. Ed. Engl. 34, 2091–2106. ford), Vol. 9. J. K. M. (2000) Curr. Opin. Chem. Biol. 4, 270–279. 62. Marchi-Artzner, V., Gulik-Krzywicki, T., Gue- 13. Whitesides, G. M., Mathias, J. P. & Seto, C. T. 37. Lehn, J.-M. & Eliseev, A. (2001) Science 291, deau-Boudeville, M.-A., Gosse, C., Sanderson, (1991) Science 254, 1312–1319. 2331–2332. J. M., Dedieu, J.-C. & Lehn, J.-M. (2001) Chem. 14. Lawrence, D. S., Jiang, T. & Levett, M. (1995) 38. Quinkert, G., Bang, H. & Reichert, D. (1996) Helv. Phys. Chem. 2, 367–376. Chem. Rev. 95, 2229–2260. Chim. Acta 79, 1260–1278. 63. Paleos, C. M., Sideratou, Z. & Tsiourvas, D. 15. Philp, D. & Stoddart, J. F. (1996) Angew. Chem. 39. Orgel, L. (1995) Acc. Chem. Res. 28, 109–118. (2001) Chem. Biol. Chem. 2, 305–310. Int. Ed. Engl. 35, 1154–1196. 40. Oh, K., Jeong, K.-S. & Moore, J. S. (2001) Nature 64. Desiraju, G. R. (1995) The Crystal as a Supramo- 16. Leinninger, S., Olenyuk, B. & Stang, P. J. (2000) (London) 414, 889–893. lecular Entity, Perspectives in Supramolecular Chem. Rev. 100, 853–908. 41. Lehn, J.-M. (1995) in Supramolecular Chemistry: Chemistry (Wiley, Chichester), Vol. 2. 17. Swiegers, G. F. & Malefetse, T. J. (2000) Chem. Concepts and Perspectives (VCH, New York), pp. 65. Bard, A. J. (1994) Integrated Chemical Systems: A Rev. 100, 3483–3537. 89–138. Chemical Approach to Nanotechnology (Wiley, 18. Lindoy, L. F. & Atkinson, I. M. (2000) Self- 42. Atwood, J. L., Davies, J. E. D., MacNicol, D. D., New York). Assembly in Supramolecular Systems (Royal Soci- Vo¨gtle, F. & Lehn, J.-M., eds. (1996) Comprehen- 66. Chandross, E. A. & Miller, R. D., eds. (1999) ety of Chemistry, Cambridge, U.K.). sive Supramolecular Chemistry (Pergamon, Ox- Chem. Rev. 99, 1641–1990. 19. Sauvage, J.-P. & Dietrich-Buchecker, C., eds. ford), Vol 10. 67. Huang, Y., Duan, X., Cui, Y., Lauhon, L. J., Kim, (1999) Molecular Catenanes, Rotaxanes, and Knots 43. Balzani, V. & Scandola, F. (1991) Supramolecular K.-H. & Lieber, C. M. (2001) Science 294, 1313– (Wiley-VCH, Weinheim). Photochemistry (Ellis Horwood, Chichester). 1317. 20. Hasenknopf, B., Lehn, J.-M., Boumediene, N., 44. Jortner, J. & Ratner, M., eds. (1997) Molecular 68. Bachtold, A., Hadley, P., Nakanishi, T. & Dekker, Leize, E. & Van Dorsselaer, A. (1998) Angew. Electronics (Blackwell, Oxford). C. (2001) Science 294, 1317–1320. Chem. Int. Ed. 37, 3265–3268. 45. Gust, D., Moore, T. A. & Moore, A. L. (2001) Acc. 69. Ebbesen, T. W., Lezec, H., Ghaemi, H. F., Thio, T. 21. Sua´rez, M., Lehn, J.-M., Zimmerman, S. C., Skou- Chem. Res. 34, 40–48. & Wolff, P. A. (1998) Nature (London) 391, lios, A. & Heinrich, B. (1998) J. Am. Chem. Soc. 46. Gokel, G. W. & Mukhopadhyav, A. (2001) Chem. 667–669. 120, 9526–9532. Soc. Rev. 30, 274–286. 70. Service, R. F. (2001) Science 293, 782–785. 22. Berl, V., Krische, M. J., Huc, I., Lehn, J.-M. & 47. Bong, D. T., Clark, T. D., Granja, J. R. & Ghadiri, 71. Hohmann, W., Krauss, M. & Schneider, F. W. Schmutz, M. (2000) Chem. Eur. J. 6, 1938–1946. M. R. (2001) Angew. Chem. Int. Ed. 40, 988–1011. (1998) J. Phys. Chem. A 102, 3103–3111.

4768 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.072065599 Lehn Downloaded by guest on September 25, 2021