Dual-mode "co-conformational" switching in catenanes incorporating bipyridinium and dialkylammonium recognition sites.

Three [2]catenanes and three [3]catenanes incorporating one or two pi-electron-rich macrocyclic polyethers and one pi-electron-deficient polycationic cyclophane have been synthesized in yields ranging from 4 to 38%. The pi-electron-rich macrocyclic components possess either two 1,4-dioxybenzene or two 1.5-dioxynaphthalene recognition sites. The pi-electron-deficient cyclophane components incorporate two bipyridinium and either one or two dialkylammonium recognition sites. The template-directed syntheses of these catenanes rely on i) pi...pi stacking interactions between the dioxyarene and bipyridinium recognition sites, ii) C-H...O hydrogen bonds between some of the bipyridinium hydrogen atoms and some of the polyether oxygen atoms, and iii) C-H...pi interactions between some of the dioxyarene hydrogen atoms and the aromatic spacers separating the bipyridinium units. The six catenanes were characterized by mass spectrometry and by both 1H and 13C NMR spectroscopy. The absorption spectra and the electrochemical properties of the catenanes have been investigated and compared with those exhibited by the component macrocycles and by related known catenanes. Broad and weak absorption bands in the visible region, originating from charge-transfer (CT) interactions between electron-donor and electron-acceptor units, have been observed. Such charge-transfer interactions are responsible for the quenching of the potentially fluorescent excited states of the aromatic units of the macrocyclic polyether components. The redox behavior of these novel compounds has been investigated and correlations among the observed redox potentials are illustrated and discussed. The catenanes undergo co-conformational switching upon one-electron reduction of the two bipyridinium units. One of them--in its reduced form--can be also switched by acid/base inputs and exhibits AND logic behavior. The co-conformational rearrangements induced by the redox and acid/base stimulations lend themselves to exploitation in the development of molecular-level machines and logic gates.

Artificial molecular-level machines: which energy to make them work?

The concept of machine can be extended to the molecular level by designing and synthesizing (supra)molecular species capable of performing mechanical movements. The energy needed to make a machine work can be supplied as chemical energy, electrical energy, or light. When a chemical "fuel" is used, waste products are formed, whereas this is not the case when suitable photochemical or electrochemical energy inputs are employed. A number of elementary functions performed by molecular-level machines are illustrated, and more complex ones are foreseen.

Artificial molecular-level machines.

The concept of "machine" can be extended to the molecular level by designing supramolecular species capable of performing mechanical-like movements as a consequence of an appropriate energy supply. Molecular-level machines operate via electronic and nuclear rearrangements, for example, through some kind of chemical reaction. Like macroscopic machines, they are characterized by: (i) the kind of energy input supplied to make them work, (ii) the kind of movement performed by their components, (iii) the way in which their operation can be controlled and monitored, (iv) the possibility to repeat the operation at will and establish a cyclic process, (v) the time scale needed to complete a cycle of operation, and (vi) the function performed. A crucial issue is that concerning energy supply. Artificial machines powered by chemical energy ("fuels") produce waste products whose accumulation compromises the operation of the machine unless they are removed from the system. Photochemical and electrochemical energy inputs, however, can be used to make a machine work without formation of waste products. Examples of chemically, electrochemically, and photochemically powered machines investigated in our laboratory are reviewed, and future directions for the construction of novel machines are illustrated. The two most interesting kinds of applications of molecular-level machines are related to the mechanical aspect, which can be exploited, for example, for molecular-level transportation purposes, and the logic aspect, which can be exploited for information processing at the molecular level and, in the long run, for the construction of molecular level (chemical) computers.

Artificial molecular-level machines. Dethreading-rethreading of a pseudorotaxane powered exclusively by light energy.

The reversible light-driven dethreading-rethreading of a pseudorotaxane is obtained in solution by exploiting the (E)-(Z) photoisomerisation of azobenzene, and monitored through fluorescence signals.

A dendritic antenna for near-infrared emission of Nd3+ ions.

An interior of 18 amide groups and a periphery functionalized with 24 dansyl groups forms a light-harvesting dendrimer which features intense absorption bands in the near-UV spectral region and a strong fluorescence band in the visible region. Upon encapsulation of Nd(3+) ions, the fluorescence of the dansyl groups is quenched and an intense sensitized near-infrared emission of Nd(3+) is observed. The associated energy transfer is shown in the cartoon.

A photochemically driven molecular-level abacus

A molecular-level abacus-like system driven by light inputs has been designed in the form of a [2]rotaxane, comprising the pi-electron-donating macrocyclic polyether bis-p-phenylene-34-crown-10 (BPP34C10) and a dumbbell-shaped component that contains 1) a Ru(II) polypyridine complex as one of its stoppers in the form of a photoactive unit, 2) a p-terphenyl-type ring system as a rigid spacer, 3) a 4,4'-bipyridinium unit and a 3,3'-dimethyl-4,4'-bipyridinium unit as pi-electron-accepting stations, and 4) a tetraarylmethane group as the second stopper. The synthesis of the [2]rotaxane was accomplished in four successive stages. First of all, the dumbbell-shaped component of the [2]rotaxane was constructed by using conventional synthetic methodology to make 1) the so-called "west-side" comprised of the Ru(II) polypyridine complex linked by a bismethylene spacer to the p-terphenyl-type ring system terminated by a benzylic bromomethyl function and 2) the so-called "east-side" comprised of the tetraarylmethane group, attached by a polyether linkage to the bipyridinium unit, itself joined in turn by a trismethylene spacer to an incipient 3,3'-dimethyl-4,4'-bipyridinium unit. Next, 3) the "west-side" and "east-side" were fused together by means of an alkylation to give the dumbbell-shaped compound, which was 4) finally subjected to a thermodynamically driven slippage reaction, with BPP34C10 as the ring, to afford the [2]rotaxane. The structure of this interlocked molecular compound was characterized by mass spectrometry and NMR spectroscopy, which also established, along with cyclic voltammetry, the co-conformational behavior of the molecular shuttle. The stable translational isomer is the one in which the BPP34C10 component encircles the 4,4'-bipyridinium unit, in keeping with the fact that this station is a better pi-electron acceptor than the other station. This observation raises the question- can the BPP34C10 macrocycle be made to shuttle between the two stations by a sequence of photoinduced electron transfer processes? In order to find an answer to this question, the electrochemical, photophysical, and photochemical (under continuous and pulsed excitation) properties of the [2]rotaxane, its dumbbell-shaped component, and some model compounds containing electro- and photoactive units have been investigated. In an attempt to obtain the photoinduced abacus-like movement of the BPP34C10 macrocycle between the two stations, two strategies have been employed-one was based fully on processes that involved only the rotaxane components (intramolecular mechanism), while the other one required the help of external reactants (sacrificial mechanism). Both mechanisms imply a sequence of four steps (destabilization of the stable translational isomer, macrocyclic ring displacement, electronic reset, and nuclear reset) that have to compete with energy-wasteful steps. The results have demonstrated that photochemically driven switching can be performed successfully by the sacrificial mechanism, whereas, in the case of the intramolecular mechanism, it would appear that the electronic reset of the system is faster than the ring displacement.

Artificial Molecular Machines.

The miniaturization of components used in the construction of working devices is being pursued currently by the large-downward (top-down) fabrication. This approach, however, which obliges solid-state physicists and electronic engineers to manipulate progressively smaller and smaller pieces of matter, has its intrinsic limitations. An alternative approach is a small-upward (bottom-up) one, starting from the smallest compositions of matter that have distinct shapes and unique properties-namely molecules. In the context of this particular challenge, chemists have been extending the concept of a macroscopic machine to the molecular level. A molecular-level machine can be defined as an assembly of a distinct number of molecular components that are designed to perform machinelike movements (output) as a result of an appropriate external stimulation (input). In common with their macroscopic counterparts, a molecular machine is characterized by 1) the kind of energy input supplied to make it work, 2) the nature of the movements of its component parts, 3) the way in which its operation can be monitored and controlled, 4) the ability to make it repeat its operation in a cyclic fashion, 5) the timescale needed to complete a full cycle of movements, and 6) the purpose of its operation. Undoubtedly, the best energy inputs to make molecular machines work are photons or electrons. Indeed, with appropriately chosen photochemically and electrochemically driven reactions, it is possible to design and synthesize molecular machines that do work. Moreover, the dramatic increase in our fundamental understanding of self-assembly and self-organizational processes in chemical synthesis has aided and abetted the construction of artificial molecular machines through the development of new methods of noncovalent synthesis and the emergence of supramolecular assistance to covalent synthesis as a uniquely powerful synthetic tool. The aim of this review is to present a unified view of the field of molecular machines by focusing on past achievements, present limitations, and future perspectives. After analyzing a few important examples of natural molecular machines, the most significant developments in the field of artificial molecular machines are highlighted. The systems reviewed include 1) chemical rotors, 2) photochemically and electrochemically induced molecular (conformational) rearrangements, and 3) chemically, photochemically, and electrochemically controllable (co-conformational) motions in interlocked molecules (catenanes and rotaxanes), as well as in coordination and supramolecular complexes, including pseudorotaxanes. Artificial molecular machines based on biomolecules and interfacing artificial molecular machines with surfaces and solid supports are amongst some of the cutting-edge topics featured in this review. The extension of the concept of a machine to the molecular level is of interest not only for the sake of basic research, but also for the growth of nanoscience and the subsequent development of nanotechnology.

Tetrathiafulvalenenaphthalenophanes: planar chirality and cis/trans photoisomerization

A cyclophane incorporating one 1,5-dioxynaphthalene ring system and one tetrathiafulvalene (TTF) unit bridged by [SCH(2)CH(2)O] linkages has been synthesized. In this cyclophane, the TTF unit can adopt either cis or trans configurations. In addition, the 1, 5-dioxynaphthalene ring system imposes one element of planar chirality on this cyclophane. A second element of planar chirality is introduced by the trans form of the TTF unit. Thus, the cyclophane exists in diastereoisomeric forms as three pairs of enantiomers. The enantiomeric pairs associated with the cis form of the TTF unit, as well as one of those associated with the trans form, have been isolated by crystallization, and their structures assigned in the solid state by single-crystal X-ray analyses. In solution, cis/trans isomerization occurs when either the cis or the trans form of the cyclophane is exposed to light. The photoisomerization reaction can be followed by (1)H NMR and UV-vis spectroscopies, as well as by HPLC. The photoisomerization quantum yield has been measured at two different excitation wavelengths (406 and 313 nm). In both cases, the trans --> cis process (Phi = 0.20 at 406 nm) is much more efficient than the reverse cis --> trans process (Phi = 0.030 at 406 nm). Since the absorption spectra of the trans and cis isomers are different and the quantum yield of the trans --> cis photoisomerization reaction depends on the excitation wavelength, the mole fraction of the two diastereoisomers present at the photostationary state depends on the wavelength of the exciting light. No isomerization occurs when the solutions, regardless of the mole fraction of the two diastereoisomers, are stored in the dark.

Switching of pseudorotaxanes and catenanes incorporating a tetrathiafulvalene unit by redox and chemical inputs

An acyclic polyether 1a, incorporating a central tetrathiafulvalene (TTF) electron donor unit and two 4-tert-butylphenoxy groups at its termini, has been synthesized. Two macrocyclic polyethers containing two different electron donors, namely a TTF unit with, in one case, a 1,4-dioxybenzene ring (2a), and, in the other case (2b), a 1,5-dioxynaphthalene ring system, have also been synthesized. These two macrocyclic polyethers have been mechanically interlocked in kinetically controlled template-directed syntheses with cyclobis(paraquat-p-phenylene) cyclophane (3(4+)) to afford the [2]catenanes 2a/3(4+) and 2b/3(4+), respectively. X-ray crystallography reveals that the [2]-catenane 2b/3(4+) has the TTF unit of 2b located inside the cavity of 3(4+). The spectroscopic (UV/vis and 1H NMR) and electrochemical properties of compounds 1a, 2a, 2b, 2a/3(4+), and 2b/3(4+) and of the [2]pseudorotaxane 1a.3(4+) were investigated. The absorption and emission properties of the mono- and dioxidized forms of the TTF unit in these various species have also been studied. The results obtained in acetonitrile solution can be summarized as follows. (a) While TTF2+ exhibits a strong fluorescence, no emission can be observed for the TTF2+ units contained in the polyethers and in their pseudorotaxanes and catenanes. (b) A donor-acceptor absorption band is observed upon two-electron oxidation of the TTF unit in the macrocyclic polyethers 2a and 2b. (c) The spontaneous self-assembly of 1a and 3(4+) to give the [2]pseudorotaxane 1a.3(4+) is strongly favored (Kass. = 5 x 10(5) L mol-1) but slow (at 296 K, k = 11.3 L mol-1 s-1 and delta G++ = 15.9 kcal mol-1) because of the steric hindrance associated with the bulky end groups of 1a. (d) In the pseudorotaxane 1a.3(4+), the reversible displacement of the cyclophane from the TTF unit in the threadlike substrate occurs on oxidation/reduction of its electroactive components. (e) Switching between the two translational isomers of the catenanes 2a/3(4+) and 2b/3(4+) occurs by cyclic oxidation and reduction of the TTF unit contained in 2a and in 2b, respectively. (f) Addition of o-chloroanil to the pseudorotaxane 1a.3(4+) and to the catenanes 2a/3(4+) and 2b/3(4+) causes the displacement of the TTF unit from the cavity of the cyclophane 3(4+) because of the formation of an adduct between the TTF unit and o-chloroanil.

The electrochemically-driven decomplexation/recomplexation of inclusion adducts of ferrocene derivatives with an electron-accepting receptor

The tetracationic cyclophane, cyclobis(paraquat-4,4'-biphenylene), binds 1,1'-disubstituted ferrocene-based polyethers as a result of (i) [pi...pi] stacking between the pi-electron-deficient bipyridinium units and the pi-electron-rich cyclopentadienyl rings and (ii) [C-H...O] hydrogen bonds between the alpha-bipyridinium hydrogen atoms and the polyether oxygen atoms. However, even the presence of a bulky tetraarylmethane group--which is too large to thread through the cavity of the cyclophane host--at the end of each of the two polyether substituents of the ferrocene-containing guest does not discourage adduct formation of the inclusion type. Thus, in these adducts, the ferrocene unit of the guest is located inside the cavity of the host with its two polyether chains protruding outward from the same side of the host. The alternative pseudorotaxane geometry is not observed in solutions of these 1:1 adducts. The host-guest adducts display absorption bands in the visible spectral region, characteristic of charge-transfer interactions. In the case of one of these adducts, reversible decomplexation/recomplexation takes place upon electrochemical oxidation/reduction of the ferrocene-based unit or upon reduction/oxidation of the tetracationic cyclophane.

Eosin molecules hosted into a dendrimer which carries thirty-two dansyl units in the periphery: a photophysical study.

A fourth generation POPAM dendrimer functionalized at the periphery with 32 dansyl groups (circles in the picture) can host up to six eosin dye molecules (black ovals) in two different sites (triangle, square). Efficient energy transfer processes from the dansyl to eosin molecules occur. The picture shows some characteristic wavelengths and lifetimes of these transitions; excitation, emission, and energy transfer processes are represented by dotted, dashed, and full arrows, respectively.

Photoprocesses.

The development of supramolecular chemistry has allowed the construction of structurally organized and functionally integrated chemical systems capable of elaborating the energy and information input of photons to perform complex functions. Model systems capable of mimicking the two fundamental steps of photosynthesis, namely light harvesting and photoinduced charge separation, have recently been developed.

Solar Energy Conversion by Water Photodissociation: Transition metal complexes can provide low-energy cyclic systems for catalytic photodissociation of water.

The basic concepts for direct and catalyzed photodissociation of water have been summarized. Water dissociation in closed-cycle processes based on endothermic photochemical reactions offers a potential solution to the solar energy conversion problem. Transition metal complexes, whose excited state chemistry is extremely rich (23, 24) although mostly unexplored, are, in principle, suitable "catalysts" for cycles of this type. The most interesting cycles are those involving metal hydrido complexes or binuclear complexes in which the two metal atoms are bound into a macrocyclic ligand. Systematic investigations of the photochemistry of transition metal complexes with the aim of designing suitable systems for solar energy conversion have long-range promise and merit further consideration.