REVIEW ARTICLE PUBLISHED ONLINE: 5 FEBRUARY 2015 | DOI: 10.1038/NNANO.2014.337
Supramolecular systems chemistry Elio Mattia and Sijbren Otto* The field of supramolecular chemistry focuses on the non-covalent interactions between molecules that give rise to molecular recognition and self-assembly processes. Since most non-covalent interactions are relatively weak and form and break without significant activation barriers, many supramolecular systems are under thermodynamic control. Hence, traditionally, supramolecular chemistry has focused predominantly on systems at equilibrium. However, more recently, self-assembly processes that are governed by kinetics, where the outcome of the assembly process is dictated by the assembly pathway rather than the free energy of the final assembled state, are becoming topical. Within the kinetic regime it is possible to distinguish between systems that reside in a kinetic trap and systems that are far from equilibrium and require a continuous supply of energy to maintain a stationary state. In particular, the latter systems have vast functional potential, as they allow, in principle, for more elaborate structural and functional diversity of self-assembled systems — indeed, life is a prime example of a farfrom-equilibrium system. In this Review, we compare the different thermodynamic regimes using some selected examples and discuss some of the challenges that need to be addressed when developing new functional supramolecular systems.
S
upramolecular assemblies are essential for biological function and in many aspects the field of supramolecular chemistry has been inspired by biology. Compelling illustrations of complex assemblies carrying out vital cellular tasks abound: bilayer membranes, ribosomes, nucleic acid transcription machinery and so on. In some cases, thermodynamically driven molecular assembly processes are sufficient for biological functions. Examples include the self-assembly of lipids giving rise to a cell membrane or the formation of a stable host–guest complex that triggers a given biomolecular process. However, for many of the more advanced biological machinery, downhill self-assembly processes are insufficient, because the desired function only emerges from processes that are governed by kinetics and that continuously dissipate energy. Continuous freeenergy consumption activates complex biological supramolecular machinery and thereby enables the chemical reactions so that cells can function. Hence, life is far from equilibrium. In biological systems, energy is used in two complementary ways. First, free energy from processes such as ATP hydrolysis is used to perform other chemical reactions that would otherwise be thermodynamically unfavourable, such as phosphorylation enzyme catalysis and transmembrane ion pumping. Second, free energy is used in processes that produce more copies of the molecules carrying out that cellular function, such as during protein synthesis in a ribosome. Far-from-equilibrium behaviour is not only ubiquitous in cell biology but is also at the heart of Darwinian evolution. By having biomolecules (and also entire organisms) that are constantly formed and broken down, the functions carried out by the molecular assemblies (or the organisms) can undergo a process of selection and adaptation. The stability of the resulting far-from-equilibrium supramolecular structures does not derive directly from thermodynamics, but from the degree of adaptation of their function to the environment 1–3. The predominance of far-from-equilibrium thermodynamics in biology contrasts starkly with the mode of operation in most man-made supramolecular systems, for which three different thermodynamic regimes (Fig. 1) may be identified: 1. Equilibrium assemblies. Systems that are likely to persist for a long time due to their thermodynamic stability.
2. Kinetically trapped assemblies. Systems that are transiently durable, trapped in a local minimum of the energy landscape. It would take time, or activation energy, for them to be converted to more stable structures. 3. Far-from-equilibrium assemblies. Systems that require a continuous supply of energy to persist. If the energy supply stops, the system would fall apart and end up in a thermodynamic minimum state (or in a kinetic trap en route). The continuous energy-driven transformation that these structures undergo makes it possible for them to have interesting, and sometimes unpredictable, emergent functions. The field of supramolecular chemistry started with the study of systems under thermodynamic control, but is currently seeing a shift towards kinetically controlled and far-from-equilibrium systems, where the richest functions can be harnessed. For example, the capability of self-sustaining chemical oscillations, formation of concentration gradients, devising unidirectional motion and other phenomena up to the most complex ones such as the functioning of living organisms and their evolution are properties that can be observed only under far-from-equilibrium conditions. Furthermore, the inherent added complexity of a system far from equilibrium, due to the continuous interconversion among multiple structures, adds to the possibility (but is not a guarantee) that entirely new unpredictable functions will emerge. This trend towards increased complexity in supramolecular chemistry (from thermodynamic to kinetic control to far-from-equilibrium systems) resonates with recent activities in the field of systems chemistry, which deals with chemical systems endowed with a high degree of complexity, so as to show emergent properties, that is, properties of a whole system that are not predictable solely from the properties of its constituent parts4–7. Examples of emergent properties of complex chemical systems are oscillatory reactions, molecular self-replication, chiral symmetry breaking and molecular recognition8–10. Supramolecular systems present many possibilities to show emergent behaviour, particularly when under far-from-equilibrium conditions. However, thus far, very few experimental examples of far-from-equilibrium supramolecular systems have appeared in the literature. The structure of this Review will follow the developments in the field of supramolecular chemistry, from thermodynamically
Centre for Systems Chemistry, Stratingh Institute, University of Groningen, Nijenborgh 4, 9747AG Groningen, The Netherlands. *e-mail: [email protected] NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
111
REVIEW ARTICLE a
b
c
Energy
Energy
Energy
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 Input of energy required
Different pathways (not each other’s microscopic reverse) Energy gets dissipated
Thermodynamic equilibrium
Kinetic trap
Far from equilibrium
Figure 1 | Thermodynamic regimes of a chemical system. a, Thermodynamic equilibrium. The final state is determined by the Boltzmann distribution and the pathway through which this final state is reached is irrelevant. b, Kinetic control. The final product distribution depends on the synthetic pathway and the products are kinetically trapped in their final state. c, Far-from-equilibrium systems. Kinetically controlled and dissipative systems in which the final product distribution is a state reflecting the balance between continuous synthesis and degradation pathways.
controlled systems to kinetically controlled systems and finally to farfrom-equilibrium systems. We will highlight a few selected examples from the literature for each of these thermodynamic regimes that are, in our (inevitably rather personal) opinion, representative of the current state of the art in the field.
Equilibrium assemblies
Structures under thermodynamic control are stable systems with properties that result from the molecular structure that they adopt and the corresponding supramolecular interactions within it. Structures of impressive architectural complexity have become accessible thanks to improved understanding of non-covalent interactions. Among them are mechanically interlocked molecular structures that may be accessed by template-directed synthesis. An example is the elegant synthesis of Borromean rings that consist of three macrocycles that cannot be separated without breaking one of the rings11 (Fig. 2a). This structure was templated by transition metals and relied on reversible imine bond formation for its synthesis. Another important class of thermodynamically stable structures are supramolecular polymers, in which the monomer units are held together by non-covalent interactions such as hydrogen bonds12. The properties of the monomers and the effects of the intermolecular interactions that they form are directly translated into mechanical properties that are observable at the macroscopic level, that is, polymer rigidity or flexibility. Supramolecular polymers differ from traditional covalent polymers in that the molecular weight distribution is not fixed during their synthesis, but defined by the experimental conditions including monomer concentration, temperature and the physicochemical properties of the medium. An example of such a structure and the underlying hydrogen bond interactions is shown in Fig. 2b. A recent tool that has proved useful for discovering and studying thermodynamically stable supramolecular systems is dynamic combinatorial chemistry 13, whereby building blocks undergo dynamic covalent exchange leading to the formation of library members whose stability determines their relative abundance within the dynamic combinatorial library that they create. Dynamic combinatorial chemistry has been used for the exploration and discovery of stable host–guest interactions14 and mechanically interlocked structures15, such as a trefoil knot from a dynamic combinatorial library based on a building block containing three naphthalenediimide units (Fig. 2c). This iconic structure forms spontaneously simply by allowing the dithiol groups to oxidize in an aqueous solution; when the system reaches equilibrium, the solution is dominated by the knot structure. Thermodynamically stable structures have also been exploited at the macroscopic level by engineering multistable systems such as muscle-like supramolecular polymers16. The rotaxane shown in 112
Fig. 2d is held together by the crown ether that is bound to the protonated benzylic amines. Deprotonation of these amines causes the crown ether to shift to the triazonium ring, extending the length of the molecule. Several of these rotaxanes were linked together through coordination of the tripyridyl termini with a metal ion. The possibility to use different pH conditions to induce a length change of the polymer by switching between two stable structures is a result of the supramolecular interactions that keep the individual building blocks together within such a polymer. In this example, an emergent property, that is, macroscopic length change, arises not just from one single molecular or supramolecular structure but from the fact that the system can transition between two states (Fig. 2d). Equilibrium supramolecular systems, as we have seen from the examples above, reside in their lowest free-energy state. The only way to make these systems move away from that state is to change the free-energy landscape such that a new energy minimum is created. In the example in Fig. 2d the free-energy landscape was changed by altering the pH allowing the transition to a new state.
Kinetically trapped assemblies
The properties of a well-defined supramolecular chemical system derive from the one single structure that it adopts. Because, in principle, many different kinetically trapped states can exist for a given system, each of these states can give rise to a different set of functional properties that go beyond those found at the equilibrium state. Therefore, the ability to control possible transitions among kinetically trapped states en route to the thermodynamic minimum could endow a supramolecular chemical system with new properties. Most bilayer vesicles formed from phospholipids are examples of supramolecular systems under kinetic control: they form rapidly under certain conditions, but are usually transient products that can convert to the more stable hydrated crystals. Depending on the conditions of formation, the shape and size of the vesicles can be controlled, reflecting different kinetically trapped states. Transitions between states are possible, for example, by allowing vesicles to fuse. However, such transitions are only feasible if the new state has a lower thermodynamic stability than the old state17–19. Many examples exist of kinetically controlled vesicle-formation processes and the topic is of fundamental importance in the study of the origin of life20–24. The kinetics of formation of a supramolecular system might also be characterized by complex pathways along which metastable assemblies are first formed, before they are finally converted into the more thermodynamically stable structure25,26. This scenario was recently described for an unstable supramolecular polymer that undergoes helicity inversion into the final stable product (Fig. 3a). The chiral monomer contains a self-complementary hydrogen
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 a
N
N
H N
O
H2N
O
O
O
N
N
N
H
N
N
N
N
O
O
R
O
N
N
Zn N
H
OR
c R HS
O
O
N
N
O
O
O
O R
N
N
R O
O
pH = 8
O
H N
N N
H N
O N H
O
H n-C13H27
O
H
R
N
RO
NH2
O
N
N
O
H
b
N N
H2N
N
NH2
O
O
N N
O
O
O
N
N
O
O
SH R
R = COOH
H2O
n-C13H27 N
N H N N
N
O
H N
H N
H
O
O
n-C13H27
H
N H
N H
O
N
n-C13H27 N N
O
H N
H N
Knot
N O H n-C13H27
d O N Me N N+ PF6 N
N N
Δd Contracted form
O
O O
O+ N PF6 O H2 O O O
O
O PF6 O H2 N +O N PF6 + O N N Me O O
N N
Decrease pH Increase pH
N
Extended form
Station 1 Station 2 off Station 2 on
Figure 2 | Supramolecular systems under thermodynamic control. a, Molecular Borromean rings. Top: Three different illustrations of the topology and coordination spheres. Bottom: The retrosynthesis showing how amines can condense with aldehydes to form imine macrocycles that are then assembled into the Borromean ring topology though metal–ligand interactions with Zn (right). b, Top: Schematic representation of a supramolecular polymer in which chains are linked by hydrogen bonds (blue). Bottom: Chemical structure of the chains responsible for the macroscopic properties of the polymers. c, Stable trefoil knot discovered by dynamic combinatorial disulfide chemistry made from a tris(naphthalenediimide) building block (top). The equilibrium conditions and final topology of the knot are shown (bottom). d, Muscle-like supramolecular polymers, for which a pH stimulus can switch the system between two relatively stable conformations of different molecular length. On deprotonation of the benzylammonium groups, the crown-ether rings shift to the triazolium rings causing the system to contract. Figures adapted with permission from: a, ref. 11, AAAS; b, ref. 55, Nature Publishing Group; c, ref. 15, AAAS; d, ref. 16, Wiley. NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
113
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
bonding unit that dimerizes. The dimers then stack to form helical aggregates. Controlling the kinetics of formation of kinetically trapped supramolecular structures by carrying out the assembly under different experimental conditions makes it possible to a
access many different metastable structures from the same building blocks27–31. Oligothiophene derivatives, for example, can produce various polymorphic self-assembled structures that are based on hydrogen bonding between the amide groups27 (Fig. 3b).
C12H25O O C12H25O
O NH2 N
C12H25O
N
O N
O
NH HN O
SOPV Nucleation
Elongation
Off-pathway metastable P-SOPV (right-handed) Hydrogenbonded dimer
SOPV
On-pathway stable M-SOPV (left-handed)
Disordered aggregate
Nucleus
Helical aggregates
b
OR O RO
N H
RO
OR
O
H N
S S
S S
O
N H
H N
OR O
OR
R = 2-Ethylhexyl
2 μm
1 μm
1 μm
c O
H N
O
R1
O
O
H N
O O
R2
R1
OMe
N H
O
R2 N H
OH
200 nm
O
Figure 3 | Supramolecular systems under kinetic control. a, Supramolecular polymers composed of the dimer of the S-chiral oligo(p-phenylenevinylene) (SOPV) building block, held together by self-complementary hydrogen bonding; the right-handed helicity is obtained as a transiently stable product that is later transformed into the more stable left-handed assembly. b, Transmission electron microscopy images of different polymorphs of the polythiophene derivative (top) obtained under kinetic control by using different experimental conditions. c, Hydrolytic enzymes convert a methyl ester (top) into a gelating carboxylic acid (bottom). Supramolecular assemblies (blue/yellow tube) are formed from these building blocks in solution. The assemblies constitute the framework for the gel in which enzymes (coloured ribbon structure) are embedded. The enzymes appear as bright spots at the fibre ends in the atomic force microscope image (right). Different enzyme concentrations yield different gel strengths. Figures adapted with permission from: a, ref. 25, Nature Publishing Group; b, ref. 27, American Chemical Society; c, ref. 32, Nature Publishing Group. 114
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 a
b
SH O O
O
H N O
NH3
N H
O
H N O
N H
H N
GTCCA + CAGGT TACGG ATGCC
SH O
NH3
tion ida Ox
Programmable sequence seed n Elo
ion
gat
Growth propagates sequence Fragmentation
Multiple crystals continue cycle
Nucleation
Scission produces daughter crystals
Figure 4 | Supramolecular systems under kinetic control. a, A thiol-functionalized peptide building block forms a dynamic combinatorial library of macrocycles, the larger of which nucleate into fibres that elongate and break thanks to mechanical forces, yielding exponential growth of the replicating fibres. b, DNA origami tiles endowed with sticky ends hybridize selectively with complementary ends on adjacent tiles. In the presence of an initial seed with a predefined tile sequence, sticky end complementarity induces the lateral replication of the initial sequence from a pool of tiles. Further growth of building blocks from solution and breakage of the larger aggregate ensure exclusive kinetically controlled replication of the initial seed sequence. Other possible sequences are not copied competitively. Panel b adapted with permission from ref. 42, NAS.
Catalysis is an exquisite tool for controlling kinetic pathways and thereby accessing kinetically trapped systems. Examples include catalysis performed by external molecules and autocatalysis in which the assembly product accelerates its own synthesis. (Auto)catalysis can be used to control which structure forms at a higher rate and therefore with higher abundance. Using catalysis, transiently stable structures with useful properties can be produced. For example, gel strength can be controlled using enzymatic catalysis: different supramolecular products can be generated by varying enzyme concentration32–34. These products are transiently stable for a sufficiently long time for their properties to be analysed and exploited (Fig. 3c). Autocatalysis can further expand the range of interesting properties that can be attained in kinetically trapped supramolecular systems. One of them is chiral symmetry breaking, which is a major phenomenon of interest in the study of natural homochirality 35. Replication of supramolecular structures, or of molecules, which occurs through supramolecular assembly, is another important phenomenon that directly results from autocatalysis and is of fundamental importance in the studies on the origins of life36–39. Dynamic combinatorial chemistry, usually used in thermodynamically driven processes, can also be used to isolate kinetically trapped supramolecular fibrils that show autocatalytic behaviour through a fibril elongation and breakage mechanism. For example, thiol-functionalized peptides shown in Fig. 4a undergo air oxidation, initially forming a library dominated by interconverting trimeric and tetrameric disulfide macrocycles. A nucleation–growth mechanism subsequently leads to the formation of fibres of larger macrocycles, such as hexamers. These larger macrocycles are only stable when they form aggregates, which only form at a useful rate when the sample is agitated. For this system, mechanically induced fibre breakage is crucial, enabling the autocatalytic, exponential growth of the fibres, and eventually resulting in the total depletion of the smaller macrocycles (Fig. 4a). Experiments yielding different macrocycle distributions under different initial conditions confirmed the kinetic nature of the self-replication process40,41. Autocatalysis is also important in information propagation. In a recently reported system42 based on DNA origami, transient
replicating structures were found to be capable of propagating and maintaining a particular sequence of nucleic acids (Fig. 4b), thereby preventing the sequence from being lost even if the transient nature of the assembly could make some of the sequences undergo degradation. The experimental system involved a pool of different DNA tiles characterized by specific sequences on their sticky ends; the tiles could nucleate larger aggregates that could laterally replicate following a set of rules dictated by sequence complementarity between the available tiles in the pool. In the presence of a seeded tile sequence, growth and breakage ensured that the initial sequence dominated as the kinetic product in the final product distribution.
Far-from-equilibrium assemblies
Unlike the thermodynamically stable and the kinetically trapped systems discussed above, far-from-equilibrium supramolecular systems use energy to maintain their integrity. The interest in thermodynamically stable and kinetically trapped systems derives often from the properties of the state that the system adopts, rather than from the properties of the individual molecules or assemblies. For kinetically trapped systems the selective reduction of energy barriers can allow the one-off move of the system to thermodynamically more stable states on the same energy landscape. For thermodynamically controlled systems a change between states requires changing the free-energy landscape. In far-from-equilibrium systems the continuous supply of energy allows a continuous transition between states. Far-from-equilibrium systems have intrigued experimental scientists but their realization has often been elusive, because devising them is conceptually challenging: a general understanding of the fundamental differences between equilibrium assemblies and far-from-equilibrium systems is still lacking, as well as a complete appreciation of their fundamental importance. A few examples, however, exist and illustrate well the role of far-from-equilibrium conditions in the realization of a supramolecular system43. The experimental realization of a far-from-equilibrium system is in principle straightforward. The simplest paradigm for far-fromequilibrium chemistry is to study a reaction during its approach to equilibrium. A prime example is the well-known Belousov– Zhabotinsky reaction, which features a characteristic new emergent
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
115
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
a
N
trans-MUA HS DDA
b N
Building block
O
–
NHBz S
O
O O– NHBz
S
O
H2N
–
Vis,∆,t trans-MUA/DDA nanoparticles
Fuel (MeI)
O
UV
O O–
O O
Energy dissipation
O
Energy dissipation
cis-MUA/DDA nanoparticles
O–
MeO Energy dissipation
UV
Fuel (MeI)
O Self-assembly
Vis,∆,t
OMe
MeO ‘Activated’ O building block
c Cl
Cl
Cl
N N
Cl
UV
O H
Cl
Cl
Blue
d N N OH
N HO
More acidic
Less acidic
O
UV irradiation 6
6.5
pH 7
7.5
0 −0.5 −1 −1 −0.5 0 0.5 1 x (mm)
Blue irradiation
x
O
HN
0
O
SO3
7-x
NH2
O S O HN
O S
NH
−0.5 −1 −1
O
−0.5
O Si O
0 0.5 x (mm)
O
Stereocentre Double bond isomerization (S) (electronic excitation)
O
O
O
O
O
O O O
Side view C6H13
Helix inversion (vibrational excitation)
H13C6 (R)
(R) (S)
O
N
O
OH
NH
HN
e
C6H13
HO
NH
S
Electrochemical gradient formation and rinsing
0.5 y (mm)
100 µm
y (mm)
1 0.5
8
OH
N
Double-bond isomerization
N (S)
C6H13
H13C6
Helix inversion
Helix inversion
Double-bond isomerization
Figure 5 | Far-from-equilibrium supramolecular systems. a, Nanoparticles functionalized with dodecylamine (DDA) and 4-(11-mercaptoundecanoxy) azobenzene (MAU) are capable of self-assembly in a gel matrix only when continuously irradiated with UV light. Irradiation results in the trans to cis isomerization of the azobenzene units, increasing their polarity and thereby the polarity of the nanoparticle surfaces, which results in nanoparticle assembly. In the absence of UV irradiation, either by visible light irradiation, heat (Δ), or spontaneously over time (t), the azobenzenes relax back to their more stable trans conformation and the self-assembled structures fall apart. b, When a dicarboxylate precursor that exclusively gelates on methylation is exposed to methyl iodide it forms a gel. When the methyl iodide fuel (Mel) is consumed, the methyl esters hydrolyse and the gel collapses. c, A solution containing photoactive phenol molecules exposed to UV irradiation in a specific region locally results in the trans to cis isomerization and an increase in acidity. The cis molecules then diffuse into the region that is exposed to blue light inducing isomerization back to the trans acid. The continuous isomerization and diffusion processes give rise to a pH gradient, as shown in the pH map (right). d, An externally applied electric potential across two electrodes induces a surface concentration gradient of electrochemically active adamantane derivatives, which can be visualized by fluorescence imaging techniques. e, A sequence of excitation pulses from an STM tip induces isomerizations of the alkenes of the ‘nanocar’ allowing it to move in a unidirectional manner on a metal surface. Figures adapted with permission from: a, ref. 56, Wiley; b, ref. 45, Wiley; c, ref. 47, Wiley; d, ref. 48, University of Twente; e, ref. 49, Nature Publishing Group. 116
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 persistent property: in an open system, the concentration of the chemical species involved oscillates in time and space. In a closed system such behaviour can only be transient as the reaction will evolve either towards equilibrium or a kinetic trap. To observe stable features in a far-from-equilibrium system it is therefore necessary to allow it to reach a steady state (which is in some cases intrinsically not possible) for example, by continuously flowing matter into and out of the system using, for example, microfluidic set-ups. Note, however, that a system being in far-from-equilibrium conditions does not guarantee the emergence of qualitatively new properties. Far-from-equilibrium conditions are possibly best understood with a minimalist approach that excludes exchanges of matter and focuses only on the necessary ingredient: a continuous source of energy. Energy flow is the only fundamental element that ensures far-from-equilibrium conditions and can be attained simply by using photons; the area of photochemistry has therefore an essential role in the development of far-from-equilibrium systems chemistry. We will now look at a few examples emphasizing the following key points: 1. What the far-from-equilibrium emergent property is. 2. Whether that property is unique to far-from-equilibrium conditions. 3. What is the continuous energy source that keeps the system far from equilibrium, and without which the system with its peculiar properties would fall apart. 4. What state the system would fall into in the absence of energy consumption. 5. The role of at least two different structures in the realization of far-from-equilibrium conditions. The first example involves nanoparticles functionalized with azobenzene-containing thiol derivatives (Fig. 5a). This system self-assembles only when subjected to UV irradiation, thanks to the increase in polarity that occurs due to the light-induced Z-to-E isomerization of azobenzene44. When the UV irradiation is turned off, the self-assembled structures fall apart spontaneously and the nanoparticles disperse again in solution. To keep the Z-to-E isomerization process steadily active and fast at all times, continuous irradiation with visible light is required. This far-fromequilibrium self-assembly process may be exploited for producing light-sensitive inks; the aggregates appear with a different colour from the dispersed nanoparticles due to plasmon resonance. In this system, the far-from-equilibrium emergent property is the formation of supramolecular self-assemblies. Of course equilibrium or kinetically trapped self-assembled systems also exist, but the farfrom-equilibrium structures are kept together by the utilization of energy in the form of UV light and would fall apart without it. The self-assembled structures exist thanks to the continuous transformation between the two different chemical structures — the E and Z isomers. Without this continuous transformation, the more stable E isomer would prevail and the assembly would fall apart. The second example deals with gelation. A gel can be obtained by continuous methylation of a carboxylate acid precursor (which does not form gels itself due to electrostatic repulsion) by using methyl iodide as the fuel45. The gel falls apart when the fuel is consumed (Fig. 5b). Energy provided by the fuel in this case keeps the supramolecular system in the gel state and the system can be cycled between the solution state and the gel state by subsequent addition of new fuel. In this system, the far-from-equilibrium emergent property is the formation of a gel. Gels can also be produced by catalysis, as explained before, but are in many cases kinetically trapped systems. The gel structure in the present example exists thanks to the utilization of energy in the form of fuel — methyl iodide — and would fall apart without it. The gel structure persists because of the continuous
REVIEW ARTICLE
transformation between the two different chemical structures — the carboxylate and the methylated derivative. Without methyl iodide, this continuous transformation is not possible, the carboxylate would be the most stable species and the gel would disintegrate. An analogous system based on two competing enzymatic reactions has also been reported recently46. Reaction–diffusion systems are also far-from-equilibrium set-ups where supramolecular organization can be achieved by dissipation of energy. For example, it is possible to use UV light to obtain a steady-state system with pH gradients by using a diazobenzene derivative (Fig. 5c); the Z isomer is more acidic than the E isomer due to the breakage of a hydrogen bond47. In this system, the far-from-equilibrium emergent property is long-range supramolecular order in the form of concentration gradients. Stable chemical gradients are an important emergent property, which are ubiquitous in nature and can only be attainable under far-from-equilibrium conditions; without the chemical gradients, homogeneous chemical systems would be obtained. This system is far from equilibrium because it involves continuous isomerization processes fuelled by energy in the form of photons, and this state ceases when energy is no longer supplied. Even though in this case there is no traditional supramolecular assembly based on non-covalent interactions, a long-range supramolecular structure arising from the competitive acid–base reactions and diffusion processes emerges as long as UV light is supplied. Gradients in this chemical system only arise thanks to the continuous transformation between the two diazobenzene isomers. Without energy consumption, the most stable isomer would prevail, cancelling out any gradient. Gradients have also been obtained under electrochemical control48. In this case, the energy is supplied by constant redox processes at two electrodes. Ferrocene carboxylic acid is continuously oxidized and reduced at the two respective electrodes, thereby forming a concentration gradient in the solution between the electrodes. A surface that is functionalized with β-cyclodextrins is placed between the electrodes and brought into contact with the ferrocene solution. The cyclodextrins on the surface are capable of binding a fluorescently modified adamantane guest as well as the reduced (but not the oxidized) ferrocene derivative. Displacement of the adamantane groups by reduced ferrocene leads to the formation of a gradient in the surface concentration of the adamantane derivative opposite to that of the reduced ferrocene derivative (Fig. 5d). In this system, the far-from-equilibrium emergent property is, again, long-range supramolecular order in the form of concentration gradients. The dynamic system is far from equilibrium because it persists only in the presence of continuous redox processes driven by an electrical current; this state ceases to exist when the energy supply is stopped. Without a current, a uniform distribution would prevail and gradients would disappear. Far-from-equilibrium systems can also result in directional motion on the molecular scale. For example, energy for electronic and vibrational excitations supplied through a scanning tunnelling microscope (STM) tip leads to double bond isomerization and stereoisomerization inducing the directional motion of a fourwheeled molecule, a nanocar, across a Cu(111) surface (Fig. 5e)49. In this system, the far-from-equilibrium emergent property is molecular motion. Unidirectional movement is an example of an emergent property unique to far-from-equilibrium conditions. The system is far from equilibrium because it results in directional molecular motion by the utilization of energy in the form of molecular excitations through the STM tip. Movement arises from the continuous interchange between many different electronic and spatial configurations (Fig. 5e). In the absence of an energy supply, the four-wheeled molecule would remain anchored to the copper surface in one of these states. A similar paradigm involving the use of light to attain far-from-equilibrium conditions has been reported recently 50.
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
117
REVIEW ARTICLE Conclusions and outlook
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
There has been a clear trend in supramolecular chemistry to move from thermodynamically to kinetically controlled systems and, more recently, to systems that operate far from equilibrium. Generally, a kinetically controlled system is in a trapped state, one of the many possible states present in the potential energy landscape of the system; the formation of the trapped state requires rather specific conditions, such as the presence of a catalyst, or autocatalysis. Such systems give rise to interesting phenomena such as replication, chiral symmetry breaking and propagation of information. However, the use of kinetic traps is limited by the intrinsically lower thermodynamic stability with respect to other more stable structures. Thermodynamically stable structures, on the other hand, are often sturdy and can be used for structural purposes or for their recognition properties, but are the most limited when it comes to functional potential. Both equilibrium and kinetically trapped systems are limited to the functions that correspond to the properties of the single relatively stable state in which they exist. In contrast, far-from-equilibrium systems owe their function to a continuous transitioning between different structures, rather than to the properties of one particular assembly and, most importantly, to the fact that they use an energy source. Notwithstanding the small number of examples in the literature up to now, far-fromequilibrium supramolecular systems can exhibit unique behaviour emerging from the interplay between the variety of accessible states, a condition unattainable under more traditional thermodynamically or kinetically controlled regimes. Thus, when it comes to systems chemistry, the field of chemistry dealing with complex chemical systems exhibiting unpredictable emergent properties, supramolecular systems that operate far from equilibrium have by far the richest functional potential. The challenge for the coming decades is to unlock this potential. The pathway to further progress involves building far-fromequilibrium conditions into new supramolecular systems, which will allow them to continuously dissipate energy and to eventually use this energy to do work. As illustrated by the examples outlined above, such systems require species that form and degrade continuously, but along pathways that are separate, chemically distinct, hence not each other’s microscopic reverse. The simplest possible scenarios are flow reactors and reaction–diffusion systems, in which out-flow or diffusion, respectively, constitute the destruction pathway. More advanced systems would be based on destruction through a chemical reaction and/or a supramolecular process. At present, the most easily accessible of such systems are those involving photoisomerization-induced processes; these systems relax through a thermal back-reaction that is chemically distinct from the photochemical forward-reaction. Only very few examples have been reported of systems in which the formation and destruction pathways are both chemical (as opposed to photochemical) reactions45,48. Given the wealth of chemical reactions there must be many more processes possible in which formation and destruction occur continuously in a single system, mediated — and therefore fuelled — by separate reagents. The current scarcity of such systems in the literature (at least explicitly) represents a bottleneck for the further development of far-from-equilibrium systems chemistry. In our view, this bottleneck is probably more due to a lack of serious efforts in searching for suitable reaction pairs, rather than as a result of any fundamental experimental difficulty. The organic chemistry community could offer a significant contribution to developing new chemical formation–destruction systems, both by studying already known systems under such conditions and by designing new formation–destruction reaction pairs, as these will be the workhorses for the far-from-equilibrium chemical systems of the future. Interestingly, the general trend that led scientific interest to extend beyond equilibrium systems towards kinetically controlled ones is also found in the historic developments of dynamic combinatorial chemistry. In its early years this field focused exclusively 118
on the discovery of new receptors and guests at equilibrium. More recently, kinetically controlled systems featuring catalysts51,52 and replicators40,53,54 (that is, autocatalysts) have gradually entered the scene. At present, the most compelling illustrations of the functional richness of far-from-equilibrium systems are still found in biology. Yet efforts in supramolecular systems chemistry are increasing as evident from the recent examples discussed in this Review. Probably one of the more appealing objectives in this area would be to attain Darwinian evolution and eventually de novo life via far-from-equilibrium supramolecular chemistry. Several systems of self-replicating molecules have already been developed. Darwinian evolution of such systems should become possible by allowing mutations of the molecular structure of the replicators, while in a far-fromequilibrium regime, where replication and replicator destruction occur simultaneously. Yet, many other functional systems can also be conceived and supramolecular systems chemistry is ultimately only limited by the imagination of the chemist. Received 2 July 2014; accepted 29 December 2014; published online 5 February 2015
References
1. Pross, A. Toward a general theory of evolution: extending Darwinian theory to inanimate matter. J. Syst. Chem. 2, 1 (2011). 2. Pascal, R., Pross, A. & Sutherland, J. D. Towards an evolutionary theory of the origin of life based on kinetics and thermodynamics. Open Biol. 3, 130156 (2013). 3. Pascal, R. & Pross, A. The nature and mathematical basis for material stability in the chemical and biological worlds. J. Syst. Chem. 5, 3 (2014). 4. Whitesides, G. M. & Ismagilov, R. F. Complexity in chemistry. Science 284, 89–92 (1999). 5. Newth, D. & Finnigan, J. Emergence and self-organization in chemistry and biology. Aust. J. Chem. 59, 841–848 (2006). 6. Ludlow, R. F. & Otto, S. Systems chemistry. Chem. Soc. Rev. 37, 101–108 (2008). 7. Peyralans, J. J. P. & Otto, S. Recent highlights in systems chemistry. Curr. Opin. Chem. Biol. 13, 705–713 (2009). 8. Von Kiedrowski, G., Otto, S. & Herdewijn, P. Welcome home systems chemists. J. Syst. Chem. 1, 1 (2010). 9. Nitschke, J. R. Systems chemistry: molecular networks come of age. Nature 462, 736–738 (2009). 10. Gibb, B. C. Teetering towards chaos and complexity. Nature Chem. 1, 17–18 (2009). 11. Chichak, K. S. et al. Molecular Borromean rings. Science 304, 1308–1312 (2004). 12. De Greef, T. F. A. et al. Supramolecular polymerization. Chem. Rev. 109, 5687–5754 (2009). 13. Li, J., Nowak, P. & Otto, S. Dynamic combinatorial libraries: from exploring molecular recognition to systems chemistry. J. Am. Chem. Soc. 135, 9222–9239 (2013). 14. Hamieh, S. et al. A “dial‑a‑receptor” dynamic combinatorial library. Angew. Chem. Int. Ed. 52, 12368–12372 (2013). 15. Ponnuswamy, N., Cougnon, F. B. L., Clough, J. M., Dan Pantoş, G. & Sanders, J. K. M. Discovery of an organic trefoil knot. Science 338, 783–785 (2012). 16. Du, G., Moulin, E., Jouault, N., Buhler, E. & Giuseppone, N. Muscle-like supramolecular polymers: integrated motion from thousands of molecular machines. Angew. Chem. Int. Ed. 51, 12504–12508 (2012). 17. Lasic, D. D. Kinetic and thermodynamic effects on the structure and formation of phosphatidylcholine vesicles. Hepatology 13, 1010–1013 (1991). 18. Guida, V. Thermodynamics and kinetics of vesicles formation processes. Adv. Colloid Interface Sci. 161, 77–88 (2010). 19. Tayebi, L., Vashaee, D. & Parikh, A. N. Stability of uni- and multillamellar spherical vesicles. ChemPhysChem 13, 314–322 (2011). 20. Stano, P. & Luisi, P. L. Achievements and open questions in the selfreproduction of vesicles and synthetic minimal cells. Chem. Commun. 46, 3639–3653 (2010). 21. Hanczyc, H. H. & Szostak, J. W. Replicating vesicles as models of primitive cell growth and division. Curr. Opin. Chem. Biol. 8, 660–664 (2004). 22. Zepik, H. H. & Walde, P. Achievements and challenges in generating protocell models. ChemBioChem 9, 2771–2772 (2008). 23. Loakes, D. & Holliger, P. Darwinian chemistry: towards the synthesis of a simple cell. Mol. BioSyst. 5, 686–694 (2009).
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 24. Takahashi, H. et al. Autocatalytic membrane-amplification on a pre-existing vesicular surface. Chem. Commun. 46, 8791–8793 (2010). 25. Korevaar, P. A. et al. Pathway complexity in supramolecular polymerization. Nature 481, 492–496 (2012). 26. Ogi, S., Sugiyasu, K., Manna, S., Samitsu, S. & Takeuchi, M. Living supramolecular polymerization realized through a biomimetic approach. Nature Chem. 6, 188–195 (2014). 27. Tevis, I. D. et al. Self-assembly and orientation of hydrogen-bonded oligothiophene polymorphs at liquid-membrane-liquid interfaces. J. Am. Chem. Soc. 133, 16486–16494 (2011). 28. Giri, G. et al. Tuning charge transport in solution-sheared organic semiconductors using lattice strain. Nature 480, 504–508 (2011). 29. Hill, J. P. et al. Self-assembled hexa‑peri‑hexabenzocoronene graphitic nanotube. Science 304, 1481–1483 (2004). 30. Würthner, F., Yao, S. & Beginn, U. Highly ordered merocyanine dye assemblies by supramolecular polymerization and hierarchical self-organization. Angew. Chem. Int. Ed. 42, 3247–3250 (2003). 31. Lohr, A., Lysetska, M. & Würthner, F. Supramolecular stereomutation in kinetic and thermodynamic self-assembly of helical merocyanine dye nanorods. Angew. Chem. Int. Ed. 44, 5071–5074 (2005). 32. Hirst, A. R. et al. Biocatalytic induction of supramolecular order. Nature Chem. 2, 1089–1094 (2010). 33. Wang, Q. G. et al. Enzymatic hydrogelation to immobilize an enzyme for high activity and stability. Soft Matter 4, 550–553 (2008). 34. Boekhoven, J. et al. Catalytic control over supramolecular gel formation. Nature Chem. 5, 433–437 (2013). 35. Lohr, A. & Würthner, F. Evolution of homochiral helical dye assemblies: involvement of autocatalysis in the “majority-rules” effect. Angew. Chem. Int. Ed. 47, 1232–1236 (2008). 36. Von Kiedrowski, G. A self-replicating hexadeoxynucleotide. Angew. Chem. Int. Ed. 24, 932–935 (1986). 37. Tjivikua, T., Ballester, P. & Rebek, J. A self-replicating system. J. Am. Chem. Soc. 112, 1249–1250 (1990). 38. Lee, D. H., Granja, J. R., Martinez, J. A., Severin, K. & Ghadiri, M. R. A selfreplicating peptide. Nature 382, 525–528 (1996). 39. Rubinov, B. et al. Transient fibril structures facilitating nonenzymatic selfreplication. ACS Nano 6, 7893–7901 (2012). 40. Carnall, J. M. A. et al. Mechanosensitive self-replication driven by selforganization. Science 327, 1502–1506 (2010). 41. Malakoutikhah, M. et al. Uncovering the selection criteria for the emergence of multi‑building‑block replicators from dynamic combinatorial libraries. J. Am. Chem. Soc. 135, 18406–18417 (2013). 42. Schulman, R., Yulke, B. & Winfree, E. Robust self-replication of combinatorial information via crystal growth and scission. Proc. Natl Acad. Sci. USA 109, 6405–6410 (2012). 43. Fialkowski, M. et al. Principles and implementations of dissipative (dynamic) self-assembly. J. Phys. Chem. B 110, 2482–2496 (2006).
REVIEW ARTICLE
44. Klajn, R., Bishop, K. J. M. & Grzybowski, B. A. Light-controlled selfassembly of reversible and irreversible nanoparticle suprastructures. Proc. Natl Acad. Sci. USA 104, 10305–10309 (2007). 45. Boekhoven, J. et al. Dissipative self-assembly of a molecular gelator by using a chemical fuel. Angew. Chem. Int. Ed. 49, 4825–4828 (2010). 46. Debnath, S., Roy, S. & Ulijn, R. V. Peptide nanofibers with dynamic instability through nonequilibrium biocatalytic assembly. J. Am. Chem. Soc. 135, 16789–16792 (2013). 47. Emond, M. et al. Energy propagation through a protometabolism leading to the local emergence of singular stationary concentration profiles. Chem. Eur. J. 18, 14375–14383 (2012). 48. Krabbenborg, S. O. Surface Gradients Under Electrochemical Control Ch. 7, PhD thesis, Univ. Twente (2014). 49. Kudernac, T. et al. Electrically driven directional motion of a four-wheeled molecule on a metal surface. Nature 479, 208–211 (2011). 50. Ragazzon, G., Baroncini, M., Silvi, S., Venturi, M. & Credi, A. Light-powered autonomous and directional molecular motion of a dissipative self-assembling system. Nature Nanotech. 10, 70–75 (2015). 51. Gasparini, G., Dal Molin, M. & Prins, L. J. Dynamic approaches towards catalyst discovery. Eur. J. Org. Chem. 2010, 2429–2440 (2010). 52. Fanlo-Virgós, H., Roig Alba, A.‑N., Hamieh, S., Colomb-Delsuc, M. & Otto, S. Transient substrate-induced catalyst formation in a dynamic molecular network. Angew. Chem. Int. Ed. 53, 11346–11350 (2014). 53. Nguyen, R., Allouche, L., Buhler, E. & Giuseppone, N. Dynamic combinatorial evolution within self-replicating supramolecular assemblies. Angew. Chem. Int. Ed. 48, 1093–1096 (2009). 54. Del Amo, V. & Philp, D. Integrating replication-based selection strategies in dynamic covalent systems. Chem. Eur. J. 16, 13304–13318 (2010). 55. de Greef, T. F. A. & Meijer, E. W. Materials science: Supramolecular polymers. Nature 453, 171–173 (2008). 56. Klajn, R., Wesson, P. J., Bishop, K. J. M. & Grzybowski, B. A. Writing selferasing images using metastable nanoparticle “inks”. Angew. Chem. Int. Ed. 48, 7035–7039 (2009).
Acknowledgements
This work was supported by the NWO, the ERC, COST CM1304 and the Dutch Ministry of Education, Culture and Science (Gravitation Program 024.001.035). The authors wish to express their gratitude to M. Colomb-Delsuc for helping with the graphics in Fig. 4.
Additional information
Reprints and permissions information is available online at www.nature.com/reprints. Correspondence should be addressed to S.O.
Competing financial interests
The authors declare no competing financial interests.
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
119
Supramolecular systems chemistry Elio Mattia and Sijbren Otto* The field of supramolecular chemistry focuses on the non-covalent interactions between molecules that give rise to molecular recognition and self-assembly processes. Since most non-covalent interactions are relatively weak and form and break without significant activation barriers, many supramolecular systems are under thermodynamic control. Hence, traditionally, supramolecular chemistry has focused predominantly on systems at equilibrium. However, more recently, self-assembly processes that are governed by kinetics, where the outcome of the assembly process is dictated by the assembly pathway rather than the free energy of the final assembled state, are becoming topical. Within the kinetic regime it is possible to distinguish between systems that reside in a kinetic trap and systems that are far from equilibrium and require a continuous supply of energy to maintain a stationary state. In particular, the latter systems have vast functional potential, as they allow, in principle, for more elaborate structural and functional diversity of self-assembled systems — indeed, life is a prime example of a farfrom-equilibrium system. In this Review, we compare the different thermodynamic regimes using some selected examples and discuss some of the challenges that need to be addressed when developing new functional supramolecular systems.
S
upramolecular assemblies are essential for biological function and in many aspects the field of supramolecular chemistry has been inspired by biology. Compelling illustrations of complex assemblies carrying out vital cellular tasks abound: bilayer membranes, ribosomes, nucleic acid transcription machinery and so on. In some cases, thermodynamically driven molecular assembly processes are sufficient for biological functions. Examples include the self-assembly of lipids giving rise to a cell membrane or the formation of a stable host–guest complex that triggers a given biomolecular process. However, for many of the more advanced biological machinery, downhill self-assembly processes are insufficient, because the desired function only emerges from processes that are governed by kinetics and that continuously dissipate energy. Continuous freeenergy consumption activates complex biological supramolecular machinery and thereby enables the chemical reactions so that cells can function. Hence, life is far from equilibrium. In biological systems, energy is used in two complementary ways. First, free energy from processes such as ATP hydrolysis is used to perform other chemical reactions that would otherwise be thermodynamically unfavourable, such as phosphorylation enzyme catalysis and transmembrane ion pumping. Second, free energy is used in processes that produce more copies of the molecules carrying out that cellular function, such as during protein synthesis in a ribosome. Far-from-equilibrium behaviour is not only ubiquitous in cell biology but is also at the heart of Darwinian evolution. By having biomolecules (and also entire organisms) that are constantly formed and broken down, the functions carried out by the molecular assemblies (or the organisms) can undergo a process of selection and adaptation. The stability of the resulting far-from-equilibrium supramolecular structures does not derive directly from thermodynamics, but from the degree of adaptation of their function to the environment 1–3. The predominance of far-from-equilibrium thermodynamics in biology contrasts starkly with the mode of operation in most man-made supramolecular systems, for which three different thermodynamic regimes (Fig. 1) may be identified: 1. Equilibrium assemblies. Systems that are likely to persist for a long time due to their thermodynamic stability.
2. Kinetically trapped assemblies. Systems that are transiently durable, trapped in a local minimum of the energy landscape. It would take time, or activation energy, for them to be converted to more stable structures. 3. Far-from-equilibrium assemblies. Systems that require a continuous supply of energy to persist. If the energy supply stops, the system would fall apart and end up in a thermodynamic minimum state (or in a kinetic trap en route). The continuous energy-driven transformation that these structures undergo makes it possible for them to have interesting, and sometimes unpredictable, emergent functions. The field of supramolecular chemistry started with the study of systems under thermodynamic control, but is currently seeing a shift towards kinetically controlled and far-from-equilibrium systems, where the richest functions can be harnessed. For example, the capability of self-sustaining chemical oscillations, formation of concentration gradients, devising unidirectional motion and other phenomena up to the most complex ones such as the functioning of living organisms and their evolution are properties that can be observed only under far-from-equilibrium conditions. Furthermore, the inherent added complexity of a system far from equilibrium, due to the continuous interconversion among multiple structures, adds to the possibility (but is not a guarantee) that entirely new unpredictable functions will emerge. This trend towards increased complexity in supramolecular chemistry (from thermodynamic to kinetic control to far-from-equilibrium systems) resonates with recent activities in the field of systems chemistry, which deals with chemical systems endowed with a high degree of complexity, so as to show emergent properties, that is, properties of a whole system that are not predictable solely from the properties of its constituent parts4–7. Examples of emergent properties of complex chemical systems are oscillatory reactions, molecular self-replication, chiral symmetry breaking and molecular recognition8–10. Supramolecular systems present many possibilities to show emergent behaviour, particularly when under far-from-equilibrium conditions. However, thus far, very few experimental examples of far-from-equilibrium supramolecular systems have appeared in the literature. The structure of this Review will follow the developments in the field of supramolecular chemistry, from thermodynamically
Centre for Systems Chemistry, Stratingh Institute, University of Groningen, Nijenborgh 4, 9747AG Groningen, The Netherlands. *e-mail: [email protected] NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
111
REVIEW ARTICLE a
b
c
Energy
Energy
Energy
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 Input of energy required
Different pathways (not each other’s microscopic reverse) Energy gets dissipated
Thermodynamic equilibrium
Kinetic trap
Far from equilibrium
Figure 1 | Thermodynamic regimes of a chemical system. a, Thermodynamic equilibrium. The final state is determined by the Boltzmann distribution and the pathway through which this final state is reached is irrelevant. b, Kinetic control. The final product distribution depends on the synthetic pathway and the products are kinetically trapped in their final state. c, Far-from-equilibrium systems. Kinetically controlled and dissipative systems in which the final product distribution is a state reflecting the balance between continuous synthesis and degradation pathways.
controlled systems to kinetically controlled systems and finally to farfrom-equilibrium systems. We will highlight a few selected examples from the literature for each of these thermodynamic regimes that are, in our (inevitably rather personal) opinion, representative of the current state of the art in the field.
Equilibrium assemblies
Structures under thermodynamic control are stable systems with properties that result from the molecular structure that they adopt and the corresponding supramolecular interactions within it. Structures of impressive architectural complexity have become accessible thanks to improved understanding of non-covalent interactions. Among them are mechanically interlocked molecular structures that may be accessed by template-directed synthesis. An example is the elegant synthesis of Borromean rings that consist of three macrocycles that cannot be separated without breaking one of the rings11 (Fig. 2a). This structure was templated by transition metals and relied on reversible imine bond formation for its synthesis. Another important class of thermodynamically stable structures are supramolecular polymers, in which the monomer units are held together by non-covalent interactions such as hydrogen bonds12. The properties of the monomers and the effects of the intermolecular interactions that they form are directly translated into mechanical properties that are observable at the macroscopic level, that is, polymer rigidity or flexibility. Supramolecular polymers differ from traditional covalent polymers in that the molecular weight distribution is not fixed during their synthesis, but defined by the experimental conditions including monomer concentration, temperature and the physicochemical properties of the medium. An example of such a structure and the underlying hydrogen bond interactions is shown in Fig. 2b. A recent tool that has proved useful for discovering and studying thermodynamically stable supramolecular systems is dynamic combinatorial chemistry 13, whereby building blocks undergo dynamic covalent exchange leading to the formation of library members whose stability determines their relative abundance within the dynamic combinatorial library that they create. Dynamic combinatorial chemistry has been used for the exploration and discovery of stable host–guest interactions14 and mechanically interlocked structures15, such as a trefoil knot from a dynamic combinatorial library based on a building block containing three naphthalenediimide units (Fig. 2c). This iconic structure forms spontaneously simply by allowing the dithiol groups to oxidize in an aqueous solution; when the system reaches equilibrium, the solution is dominated by the knot structure. Thermodynamically stable structures have also been exploited at the macroscopic level by engineering multistable systems such as muscle-like supramolecular polymers16. The rotaxane shown in 112
Fig. 2d is held together by the crown ether that is bound to the protonated benzylic amines. Deprotonation of these amines causes the crown ether to shift to the triazonium ring, extending the length of the molecule. Several of these rotaxanes were linked together through coordination of the tripyridyl termini with a metal ion. The possibility to use different pH conditions to induce a length change of the polymer by switching between two stable structures is a result of the supramolecular interactions that keep the individual building blocks together within such a polymer. In this example, an emergent property, that is, macroscopic length change, arises not just from one single molecular or supramolecular structure but from the fact that the system can transition between two states (Fig. 2d). Equilibrium supramolecular systems, as we have seen from the examples above, reside in their lowest free-energy state. The only way to make these systems move away from that state is to change the free-energy landscape such that a new energy minimum is created. In the example in Fig. 2d the free-energy landscape was changed by altering the pH allowing the transition to a new state.
Kinetically trapped assemblies
The properties of a well-defined supramolecular chemical system derive from the one single structure that it adopts. Because, in principle, many different kinetically trapped states can exist for a given system, each of these states can give rise to a different set of functional properties that go beyond those found at the equilibrium state. Therefore, the ability to control possible transitions among kinetically trapped states en route to the thermodynamic minimum could endow a supramolecular chemical system with new properties. Most bilayer vesicles formed from phospholipids are examples of supramolecular systems under kinetic control: they form rapidly under certain conditions, but are usually transient products that can convert to the more stable hydrated crystals. Depending on the conditions of formation, the shape and size of the vesicles can be controlled, reflecting different kinetically trapped states. Transitions between states are possible, for example, by allowing vesicles to fuse. However, such transitions are only feasible if the new state has a lower thermodynamic stability than the old state17–19. Many examples exist of kinetically controlled vesicle-formation processes and the topic is of fundamental importance in the study of the origin of life20–24. The kinetics of formation of a supramolecular system might also be characterized by complex pathways along which metastable assemblies are first formed, before they are finally converted into the more thermodynamically stable structure25,26. This scenario was recently described for an unstable supramolecular polymer that undergoes helicity inversion into the final stable product (Fig. 3a). The chiral monomer contains a self-complementary hydrogen
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 a
N
N
H N
O
H2N
O
O
O
N
N
N
H
N
N
N
N
O
O
R
O
N
N
Zn N
H
OR
c R HS
O
O
N
N
O
O
O
O R
N
N
R O
O
pH = 8
O
H N
N N
H N
O N H
O
H n-C13H27
O
H
R
N
RO
NH2
O
N
N
O
H
b
N N
H2N
N
NH2
O
O
N N
O
O
O
N
N
O
O
SH R
R = COOH
H2O
n-C13H27 N
N H N N
N
O
H N
H N
H
O
O
n-C13H27
H
N H
N H
O
N
n-C13H27 N N
O
H N
H N
Knot
N O H n-C13H27
d O N Me N N+ PF6 N
N N
Δd Contracted form
O
O O
O+ N PF6 O H2 O O O
O
O PF6 O H2 N +O N PF6 + O N N Me O O
N N
Decrease pH Increase pH
N
Extended form
Station 1 Station 2 off Station 2 on
Figure 2 | Supramolecular systems under thermodynamic control. a, Molecular Borromean rings. Top: Three different illustrations of the topology and coordination spheres. Bottom: The retrosynthesis showing how amines can condense with aldehydes to form imine macrocycles that are then assembled into the Borromean ring topology though metal–ligand interactions with Zn (right). b, Top: Schematic representation of a supramolecular polymer in which chains are linked by hydrogen bonds (blue). Bottom: Chemical structure of the chains responsible for the macroscopic properties of the polymers. c, Stable trefoil knot discovered by dynamic combinatorial disulfide chemistry made from a tris(naphthalenediimide) building block (top). The equilibrium conditions and final topology of the knot are shown (bottom). d, Muscle-like supramolecular polymers, for which a pH stimulus can switch the system between two relatively stable conformations of different molecular length. On deprotonation of the benzylammonium groups, the crown-ether rings shift to the triazolium rings causing the system to contract. Figures adapted with permission from: a, ref. 11, AAAS; b, ref. 55, Nature Publishing Group; c, ref. 15, AAAS; d, ref. 16, Wiley. NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
113
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
bonding unit that dimerizes. The dimers then stack to form helical aggregates. Controlling the kinetics of formation of kinetically trapped supramolecular structures by carrying out the assembly under different experimental conditions makes it possible to a
access many different metastable structures from the same building blocks27–31. Oligothiophene derivatives, for example, can produce various polymorphic self-assembled structures that are based on hydrogen bonding between the amide groups27 (Fig. 3b).
C12H25O O C12H25O
O NH2 N
C12H25O
N
O N
O
NH HN O
SOPV Nucleation
Elongation
Off-pathway metastable P-SOPV (right-handed) Hydrogenbonded dimer
SOPV
On-pathway stable M-SOPV (left-handed)
Disordered aggregate
Nucleus
Helical aggregates
b
OR O RO
N H
RO
OR
O
H N
S S
S S
O
N H
H N
OR O
OR
R = 2-Ethylhexyl
2 μm
1 μm
1 μm
c O
H N
O
R1
O
O
H N
O O
R2
R1
OMe
N H
O
R2 N H
OH
200 nm
O
Figure 3 | Supramolecular systems under kinetic control. a, Supramolecular polymers composed of the dimer of the S-chiral oligo(p-phenylenevinylene) (SOPV) building block, held together by self-complementary hydrogen bonding; the right-handed helicity is obtained as a transiently stable product that is later transformed into the more stable left-handed assembly. b, Transmission electron microscopy images of different polymorphs of the polythiophene derivative (top) obtained under kinetic control by using different experimental conditions. c, Hydrolytic enzymes convert a methyl ester (top) into a gelating carboxylic acid (bottom). Supramolecular assemblies (blue/yellow tube) are formed from these building blocks in solution. The assemblies constitute the framework for the gel in which enzymes (coloured ribbon structure) are embedded. The enzymes appear as bright spots at the fibre ends in the atomic force microscope image (right). Different enzyme concentrations yield different gel strengths. Figures adapted with permission from: a, ref. 25, Nature Publishing Group; b, ref. 27, American Chemical Society; c, ref. 32, Nature Publishing Group. 114
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 a
b
SH O O
O
H N O
NH3
N H
O
H N O
N H
H N
GTCCA + CAGGT TACGG ATGCC
SH O
NH3
tion ida Ox
Programmable sequence seed n Elo
ion
gat
Growth propagates sequence Fragmentation
Multiple crystals continue cycle
Nucleation
Scission produces daughter crystals
Figure 4 | Supramolecular systems under kinetic control. a, A thiol-functionalized peptide building block forms a dynamic combinatorial library of macrocycles, the larger of which nucleate into fibres that elongate and break thanks to mechanical forces, yielding exponential growth of the replicating fibres. b, DNA origami tiles endowed with sticky ends hybridize selectively with complementary ends on adjacent tiles. In the presence of an initial seed with a predefined tile sequence, sticky end complementarity induces the lateral replication of the initial sequence from a pool of tiles. Further growth of building blocks from solution and breakage of the larger aggregate ensure exclusive kinetically controlled replication of the initial seed sequence. Other possible sequences are not copied competitively. Panel b adapted with permission from ref. 42, NAS.
Catalysis is an exquisite tool for controlling kinetic pathways and thereby accessing kinetically trapped systems. Examples include catalysis performed by external molecules and autocatalysis in which the assembly product accelerates its own synthesis. (Auto)catalysis can be used to control which structure forms at a higher rate and therefore with higher abundance. Using catalysis, transiently stable structures with useful properties can be produced. For example, gel strength can be controlled using enzymatic catalysis: different supramolecular products can be generated by varying enzyme concentration32–34. These products are transiently stable for a sufficiently long time for their properties to be analysed and exploited (Fig. 3c). Autocatalysis can further expand the range of interesting properties that can be attained in kinetically trapped supramolecular systems. One of them is chiral symmetry breaking, which is a major phenomenon of interest in the study of natural homochirality 35. Replication of supramolecular structures, or of molecules, which occurs through supramolecular assembly, is another important phenomenon that directly results from autocatalysis and is of fundamental importance in the studies on the origins of life36–39. Dynamic combinatorial chemistry, usually used in thermodynamically driven processes, can also be used to isolate kinetically trapped supramolecular fibrils that show autocatalytic behaviour through a fibril elongation and breakage mechanism. For example, thiol-functionalized peptides shown in Fig. 4a undergo air oxidation, initially forming a library dominated by interconverting trimeric and tetrameric disulfide macrocycles. A nucleation–growth mechanism subsequently leads to the formation of fibres of larger macrocycles, such as hexamers. These larger macrocycles are only stable when they form aggregates, which only form at a useful rate when the sample is agitated. For this system, mechanically induced fibre breakage is crucial, enabling the autocatalytic, exponential growth of the fibres, and eventually resulting in the total depletion of the smaller macrocycles (Fig. 4a). Experiments yielding different macrocycle distributions under different initial conditions confirmed the kinetic nature of the self-replication process40,41. Autocatalysis is also important in information propagation. In a recently reported system42 based on DNA origami, transient
replicating structures were found to be capable of propagating and maintaining a particular sequence of nucleic acids (Fig. 4b), thereby preventing the sequence from being lost even if the transient nature of the assembly could make some of the sequences undergo degradation. The experimental system involved a pool of different DNA tiles characterized by specific sequences on their sticky ends; the tiles could nucleate larger aggregates that could laterally replicate following a set of rules dictated by sequence complementarity between the available tiles in the pool. In the presence of a seeded tile sequence, growth and breakage ensured that the initial sequence dominated as the kinetic product in the final product distribution.
Far-from-equilibrium assemblies
Unlike the thermodynamically stable and the kinetically trapped systems discussed above, far-from-equilibrium supramolecular systems use energy to maintain their integrity. The interest in thermodynamically stable and kinetically trapped systems derives often from the properties of the state that the system adopts, rather than from the properties of the individual molecules or assemblies. For kinetically trapped systems the selective reduction of energy barriers can allow the one-off move of the system to thermodynamically more stable states on the same energy landscape. For thermodynamically controlled systems a change between states requires changing the free-energy landscape. In far-from-equilibrium systems the continuous supply of energy allows a continuous transition between states. Far-from-equilibrium systems have intrigued experimental scientists but their realization has often been elusive, because devising them is conceptually challenging: a general understanding of the fundamental differences between equilibrium assemblies and far-from-equilibrium systems is still lacking, as well as a complete appreciation of their fundamental importance. A few examples, however, exist and illustrate well the role of far-from-equilibrium conditions in the realization of a supramolecular system43. The experimental realization of a far-from-equilibrium system is in principle straightforward. The simplest paradigm for far-fromequilibrium chemistry is to study a reaction during its approach to equilibrium. A prime example is the well-known Belousov– Zhabotinsky reaction, which features a characteristic new emergent
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
115
REVIEW ARTICLE
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
a
N
trans-MUA HS DDA
b N
Building block
O
–
NHBz S
O
O O– NHBz
S
O
H2N
–
Vis,∆,t trans-MUA/DDA nanoparticles
Fuel (MeI)
O
UV
O O–
O O
Energy dissipation
O
Energy dissipation
cis-MUA/DDA nanoparticles
O–
MeO Energy dissipation
UV
Fuel (MeI)
O Self-assembly
Vis,∆,t
OMe
MeO ‘Activated’ O building block
c Cl
Cl
Cl
N N
Cl
UV
O H
Cl
Cl
Blue
d N N OH
N HO
More acidic
Less acidic
O
UV irradiation 6
6.5
pH 7
7.5
0 −0.5 −1 −1 −0.5 0 0.5 1 x (mm)
Blue irradiation
x
O
HN
0
O
SO3
7-x
NH2
O S O HN
O S
NH
−0.5 −1 −1
O
−0.5
O Si O
0 0.5 x (mm)
O
Stereocentre Double bond isomerization (S) (electronic excitation)
O
O
O
O
O
O O O
Side view C6H13
Helix inversion (vibrational excitation)
H13C6 (R)
(R) (S)
O
N
O
OH
NH
HN
e
C6H13
HO
NH
S
Electrochemical gradient formation and rinsing
0.5 y (mm)
100 µm
y (mm)
1 0.5
8
OH
N
Double-bond isomerization
N (S)
C6H13
H13C6
Helix inversion
Helix inversion
Double-bond isomerization
Figure 5 | Far-from-equilibrium supramolecular systems. a, Nanoparticles functionalized with dodecylamine (DDA) and 4-(11-mercaptoundecanoxy) azobenzene (MAU) are capable of self-assembly in a gel matrix only when continuously irradiated with UV light. Irradiation results in the trans to cis isomerization of the azobenzene units, increasing their polarity and thereby the polarity of the nanoparticle surfaces, which results in nanoparticle assembly. In the absence of UV irradiation, either by visible light irradiation, heat (Δ), or spontaneously over time (t), the azobenzenes relax back to their more stable trans conformation and the self-assembled structures fall apart. b, When a dicarboxylate precursor that exclusively gelates on methylation is exposed to methyl iodide it forms a gel. When the methyl iodide fuel (Mel) is consumed, the methyl esters hydrolyse and the gel collapses. c, A solution containing photoactive phenol molecules exposed to UV irradiation in a specific region locally results in the trans to cis isomerization and an increase in acidity. The cis molecules then diffuse into the region that is exposed to blue light inducing isomerization back to the trans acid. The continuous isomerization and diffusion processes give rise to a pH gradient, as shown in the pH map (right). d, An externally applied electric potential across two electrodes induces a surface concentration gradient of electrochemically active adamantane derivatives, which can be visualized by fluorescence imaging techniques. e, A sequence of excitation pulses from an STM tip induces isomerizations of the alkenes of the ‘nanocar’ allowing it to move in a unidirectional manner on a metal surface. Figures adapted with permission from: a, ref. 56, Wiley; b, ref. 45, Wiley; c, ref. 47, Wiley; d, ref. 48, University of Twente; e, ref. 49, Nature Publishing Group. 116
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 persistent property: in an open system, the concentration of the chemical species involved oscillates in time and space. In a closed system such behaviour can only be transient as the reaction will evolve either towards equilibrium or a kinetic trap. To observe stable features in a far-from-equilibrium system it is therefore necessary to allow it to reach a steady state (which is in some cases intrinsically not possible) for example, by continuously flowing matter into and out of the system using, for example, microfluidic set-ups. Note, however, that a system being in far-from-equilibrium conditions does not guarantee the emergence of qualitatively new properties. Far-from-equilibrium conditions are possibly best understood with a minimalist approach that excludes exchanges of matter and focuses only on the necessary ingredient: a continuous source of energy. Energy flow is the only fundamental element that ensures far-from-equilibrium conditions and can be attained simply by using photons; the area of photochemistry has therefore an essential role in the development of far-from-equilibrium systems chemistry. We will now look at a few examples emphasizing the following key points: 1. What the far-from-equilibrium emergent property is. 2. Whether that property is unique to far-from-equilibrium conditions. 3. What is the continuous energy source that keeps the system far from equilibrium, and without which the system with its peculiar properties would fall apart. 4. What state the system would fall into in the absence of energy consumption. 5. The role of at least two different structures in the realization of far-from-equilibrium conditions. The first example involves nanoparticles functionalized with azobenzene-containing thiol derivatives (Fig. 5a). This system self-assembles only when subjected to UV irradiation, thanks to the increase in polarity that occurs due to the light-induced Z-to-E isomerization of azobenzene44. When the UV irradiation is turned off, the self-assembled structures fall apart spontaneously and the nanoparticles disperse again in solution. To keep the Z-to-E isomerization process steadily active and fast at all times, continuous irradiation with visible light is required. This far-fromequilibrium self-assembly process may be exploited for producing light-sensitive inks; the aggregates appear with a different colour from the dispersed nanoparticles due to plasmon resonance. In this system, the far-from-equilibrium emergent property is the formation of supramolecular self-assemblies. Of course equilibrium or kinetically trapped self-assembled systems also exist, but the farfrom-equilibrium structures are kept together by the utilization of energy in the form of UV light and would fall apart without it. The self-assembled structures exist thanks to the continuous transformation between the two different chemical structures — the E and Z isomers. Without this continuous transformation, the more stable E isomer would prevail and the assembly would fall apart. The second example deals with gelation. A gel can be obtained by continuous methylation of a carboxylate acid precursor (which does not form gels itself due to electrostatic repulsion) by using methyl iodide as the fuel45. The gel falls apart when the fuel is consumed (Fig. 5b). Energy provided by the fuel in this case keeps the supramolecular system in the gel state and the system can be cycled between the solution state and the gel state by subsequent addition of new fuel. In this system, the far-from-equilibrium emergent property is the formation of a gel. Gels can also be produced by catalysis, as explained before, but are in many cases kinetically trapped systems. The gel structure in the present example exists thanks to the utilization of energy in the form of fuel — methyl iodide — and would fall apart without it. The gel structure persists because of the continuous
REVIEW ARTICLE
transformation between the two different chemical structures — the carboxylate and the methylated derivative. Without methyl iodide, this continuous transformation is not possible, the carboxylate would be the most stable species and the gel would disintegrate. An analogous system based on two competing enzymatic reactions has also been reported recently46. Reaction–diffusion systems are also far-from-equilibrium set-ups where supramolecular organization can be achieved by dissipation of energy. For example, it is possible to use UV light to obtain a steady-state system with pH gradients by using a diazobenzene derivative (Fig. 5c); the Z isomer is more acidic than the E isomer due to the breakage of a hydrogen bond47. In this system, the far-from-equilibrium emergent property is long-range supramolecular order in the form of concentration gradients. Stable chemical gradients are an important emergent property, which are ubiquitous in nature and can only be attainable under far-from-equilibrium conditions; without the chemical gradients, homogeneous chemical systems would be obtained. This system is far from equilibrium because it involves continuous isomerization processes fuelled by energy in the form of photons, and this state ceases when energy is no longer supplied. Even though in this case there is no traditional supramolecular assembly based on non-covalent interactions, a long-range supramolecular structure arising from the competitive acid–base reactions and diffusion processes emerges as long as UV light is supplied. Gradients in this chemical system only arise thanks to the continuous transformation between the two diazobenzene isomers. Without energy consumption, the most stable isomer would prevail, cancelling out any gradient. Gradients have also been obtained under electrochemical control48. In this case, the energy is supplied by constant redox processes at two electrodes. Ferrocene carboxylic acid is continuously oxidized and reduced at the two respective electrodes, thereby forming a concentration gradient in the solution between the electrodes. A surface that is functionalized with β-cyclodextrins is placed between the electrodes and brought into contact with the ferrocene solution. The cyclodextrins on the surface are capable of binding a fluorescently modified adamantane guest as well as the reduced (but not the oxidized) ferrocene derivative. Displacement of the adamantane groups by reduced ferrocene leads to the formation of a gradient in the surface concentration of the adamantane derivative opposite to that of the reduced ferrocene derivative (Fig. 5d). In this system, the far-from-equilibrium emergent property is, again, long-range supramolecular order in the form of concentration gradients. The dynamic system is far from equilibrium because it persists only in the presence of continuous redox processes driven by an electrical current; this state ceases to exist when the energy supply is stopped. Without a current, a uniform distribution would prevail and gradients would disappear. Far-from-equilibrium systems can also result in directional motion on the molecular scale. For example, energy for electronic and vibrational excitations supplied through a scanning tunnelling microscope (STM) tip leads to double bond isomerization and stereoisomerization inducing the directional motion of a fourwheeled molecule, a nanocar, across a Cu(111) surface (Fig. 5e)49. In this system, the far-from-equilibrium emergent property is molecular motion. Unidirectional movement is an example of an emergent property unique to far-from-equilibrium conditions. The system is far from equilibrium because it results in directional molecular motion by the utilization of energy in the form of molecular excitations through the STM tip. Movement arises from the continuous interchange between many different electronic and spatial configurations (Fig. 5e). In the absence of an energy supply, the four-wheeled molecule would remain anchored to the copper surface in one of these states. A similar paradigm involving the use of light to attain far-from-equilibrium conditions has been reported recently 50.
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
117
REVIEW ARTICLE Conclusions and outlook
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337
There has been a clear trend in supramolecular chemistry to move from thermodynamically to kinetically controlled systems and, more recently, to systems that operate far from equilibrium. Generally, a kinetically controlled system is in a trapped state, one of the many possible states present in the potential energy landscape of the system; the formation of the trapped state requires rather specific conditions, such as the presence of a catalyst, or autocatalysis. Such systems give rise to interesting phenomena such as replication, chiral symmetry breaking and propagation of information. However, the use of kinetic traps is limited by the intrinsically lower thermodynamic stability with respect to other more stable structures. Thermodynamically stable structures, on the other hand, are often sturdy and can be used for structural purposes or for their recognition properties, but are the most limited when it comes to functional potential. Both equilibrium and kinetically trapped systems are limited to the functions that correspond to the properties of the single relatively stable state in which they exist. In contrast, far-from-equilibrium systems owe their function to a continuous transitioning between different structures, rather than to the properties of one particular assembly and, most importantly, to the fact that they use an energy source. Notwithstanding the small number of examples in the literature up to now, far-fromequilibrium supramolecular systems can exhibit unique behaviour emerging from the interplay between the variety of accessible states, a condition unattainable under more traditional thermodynamically or kinetically controlled regimes. Thus, when it comes to systems chemistry, the field of chemistry dealing with complex chemical systems exhibiting unpredictable emergent properties, supramolecular systems that operate far from equilibrium have by far the richest functional potential. The challenge for the coming decades is to unlock this potential. The pathway to further progress involves building far-fromequilibrium conditions into new supramolecular systems, which will allow them to continuously dissipate energy and to eventually use this energy to do work. As illustrated by the examples outlined above, such systems require species that form and degrade continuously, but along pathways that are separate, chemically distinct, hence not each other’s microscopic reverse. The simplest possible scenarios are flow reactors and reaction–diffusion systems, in which out-flow or diffusion, respectively, constitute the destruction pathway. More advanced systems would be based on destruction through a chemical reaction and/or a supramolecular process. At present, the most easily accessible of such systems are those involving photoisomerization-induced processes; these systems relax through a thermal back-reaction that is chemically distinct from the photochemical forward-reaction. Only very few examples have been reported of systems in which the formation and destruction pathways are both chemical (as opposed to photochemical) reactions45,48. Given the wealth of chemical reactions there must be many more processes possible in which formation and destruction occur continuously in a single system, mediated — and therefore fuelled — by separate reagents. The current scarcity of such systems in the literature (at least explicitly) represents a bottleneck for the further development of far-from-equilibrium systems chemistry. In our view, this bottleneck is probably more due to a lack of serious efforts in searching for suitable reaction pairs, rather than as a result of any fundamental experimental difficulty. The organic chemistry community could offer a significant contribution to developing new chemical formation–destruction systems, both by studying already known systems under such conditions and by designing new formation–destruction reaction pairs, as these will be the workhorses for the far-from-equilibrium chemical systems of the future. Interestingly, the general trend that led scientific interest to extend beyond equilibrium systems towards kinetically controlled ones is also found in the historic developments of dynamic combinatorial chemistry. In its early years this field focused exclusively 118
on the discovery of new receptors and guests at equilibrium. More recently, kinetically controlled systems featuring catalysts51,52 and replicators40,53,54 (that is, autocatalysts) have gradually entered the scene. At present, the most compelling illustrations of the functional richness of far-from-equilibrium systems are still found in biology. Yet efforts in supramolecular systems chemistry are increasing as evident from the recent examples discussed in this Review. Probably one of the more appealing objectives in this area would be to attain Darwinian evolution and eventually de novo life via far-from-equilibrium supramolecular chemistry. Several systems of self-replicating molecules have already been developed. Darwinian evolution of such systems should become possible by allowing mutations of the molecular structure of the replicators, while in a far-fromequilibrium regime, where replication and replicator destruction occur simultaneously. Yet, many other functional systems can also be conceived and supramolecular systems chemistry is ultimately only limited by the imagination of the chemist. Received 2 July 2014; accepted 29 December 2014; published online 5 February 2015
References
1. Pross, A. Toward a general theory of evolution: extending Darwinian theory to inanimate matter. J. Syst. Chem. 2, 1 (2011). 2. Pascal, R., Pross, A. & Sutherland, J. D. Towards an evolutionary theory of the origin of life based on kinetics and thermodynamics. Open Biol. 3, 130156 (2013). 3. Pascal, R. & Pross, A. The nature and mathematical basis for material stability in the chemical and biological worlds. J. Syst. Chem. 5, 3 (2014). 4. Whitesides, G. M. & Ismagilov, R. F. Complexity in chemistry. Science 284, 89–92 (1999). 5. Newth, D. & Finnigan, J. Emergence and self-organization in chemistry and biology. Aust. J. Chem. 59, 841–848 (2006). 6. Ludlow, R. F. & Otto, S. Systems chemistry. Chem. Soc. Rev. 37, 101–108 (2008). 7. Peyralans, J. J. P. & Otto, S. Recent highlights in systems chemistry. Curr. Opin. Chem. Biol. 13, 705–713 (2009). 8. Von Kiedrowski, G., Otto, S. & Herdewijn, P. Welcome home systems chemists. J. Syst. Chem. 1, 1 (2010). 9. Nitschke, J. R. Systems chemistry: molecular networks come of age. Nature 462, 736–738 (2009). 10. Gibb, B. C. Teetering towards chaos and complexity. Nature Chem. 1, 17–18 (2009). 11. Chichak, K. S. et al. Molecular Borromean rings. Science 304, 1308–1312 (2004). 12. De Greef, T. F. A. et al. Supramolecular polymerization. Chem. Rev. 109, 5687–5754 (2009). 13. Li, J., Nowak, P. & Otto, S. Dynamic combinatorial libraries: from exploring molecular recognition to systems chemistry. J. Am. Chem. Soc. 135, 9222–9239 (2013). 14. Hamieh, S. et al. A “dial‑a‑receptor” dynamic combinatorial library. Angew. Chem. Int. Ed. 52, 12368–12372 (2013). 15. Ponnuswamy, N., Cougnon, F. B. L., Clough, J. M., Dan Pantoş, G. & Sanders, J. K. M. Discovery of an organic trefoil knot. Science 338, 783–785 (2012). 16. Du, G., Moulin, E., Jouault, N., Buhler, E. & Giuseppone, N. Muscle-like supramolecular polymers: integrated motion from thousands of molecular machines. Angew. Chem. Int. Ed. 51, 12504–12508 (2012). 17. Lasic, D. D. Kinetic and thermodynamic effects on the structure and formation of phosphatidylcholine vesicles. Hepatology 13, 1010–1013 (1991). 18. Guida, V. Thermodynamics and kinetics of vesicles formation processes. Adv. Colloid Interface Sci. 161, 77–88 (2010). 19. Tayebi, L., Vashaee, D. & Parikh, A. N. Stability of uni- and multillamellar spherical vesicles. ChemPhysChem 13, 314–322 (2011). 20. Stano, P. & Luisi, P. L. Achievements and open questions in the selfreproduction of vesicles and synthetic minimal cells. Chem. Commun. 46, 3639–3653 (2010). 21. Hanczyc, H. H. & Szostak, J. W. Replicating vesicles as models of primitive cell growth and division. Curr. Opin. Chem. Biol. 8, 660–664 (2004). 22. Zepik, H. H. & Walde, P. Achievements and challenges in generating protocell models. ChemBioChem 9, 2771–2772 (2008). 23. Loakes, D. & Holliger, P. Darwinian chemistry: towards the synthesis of a simple cell. Mol. BioSyst. 5, 686–694 (2009).
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY DOI: 10.1038/NNANO.2014.337 24. Takahashi, H. et al. Autocatalytic membrane-amplification on a pre-existing vesicular surface. Chem. Commun. 46, 8791–8793 (2010). 25. Korevaar, P. A. et al. Pathway complexity in supramolecular polymerization. Nature 481, 492–496 (2012). 26. Ogi, S., Sugiyasu, K., Manna, S., Samitsu, S. & Takeuchi, M. Living supramolecular polymerization realized through a biomimetic approach. Nature Chem. 6, 188–195 (2014). 27. Tevis, I. D. et al. Self-assembly and orientation of hydrogen-bonded oligothiophene polymorphs at liquid-membrane-liquid interfaces. J. Am. Chem. Soc. 133, 16486–16494 (2011). 28. Giri, G. et al. Tuning charge transport in solution-sheared organic semiconductors using lattice strain. Nature 480, 504–508 (2011). 29. Hill, J. P. et al. Self-assembled hexa‑peri‑hexabenzocoronene graphitic nanotube. Science 304, 1481–1483 (2004). 30. Würthner, F., Yao, S. & Beginn, U. Highly ordered merocyanine dye assemblies by supramolecular polymerization and hierarchical self-organization. Angew. Chem. Int. Ed. 42, 3247–3250 (2003). 31. Lohr, A., Lysetska, M. & Würthner, F. Supramolecular stereomutation in kinetic and thermodynamic self-assembly of helical merocyanine dye nanorods. Angew. Chem. Int. Ed. 44, 5071–5074 (2005). 32. Hirst, A. R. et al. Biocatalytic induction of supramolecular order. Nature Chem. 2, 1089–1094 (2010). 33. Wang, Q. G. et al. Enzymatic hydrogelation to immobilize an enzyme for high activity and stability. Soft Matter 4, 550–553 (2008). 34. Boekhoven, J. et al. Catalytic control over supramolecular gel formation. Nature Chem. 5, 433–437 (2013). 35. Lohr, A. & Würthner, F. Evolution of homochiral helical dye assemblies: involvement of autocatalysis in the “majority-rules” effect. Angew. Chem. Int. Ed. 47, 1232–1236 (2008). 36. Von Kiedrowski, G. A self-replicating hexadeoxynucleotide. Angew. Chem. Int. Ed. 24, 932–935 (1986). 37. Tjivikua, T., Ballester, P. & Rebek, J. A self-replicating system. J. Am. Chem. Soc. 112, 1249–1250 (1990). 38. Lee, D. H., Granja, J. R., Martinez, J. A., Severin, K. & Ghadiri, M. R. A selfreplicating peptide. Nature 382, 525–528 (1996). 39. Rubinov, B. et al. Transient fibril structures facilitating nonenzymatic selfreplication. ACS Nano 6, 7893–7901 (2012). 40. Carnall, J. M. A. et al. Mechanosensitive self-replication driven by selforganization. Science 327, 1502–1506 (2010). 41. Malakoutikhah, M. et al. Uncovering the selection criteria for the emergence of multi‑building‑block replicators from dynamic combinatorial libraries. J. Am. Chem. Soc. 135, 18406–18417 (2013). 42. Schulman, R., Yulke, B. & Winfree, E. Robust self-replication of combinatorial information via crystal growth and scission. Proc. Natl Acad. Sci. USA 109, 6405–6410 (2012). 43. Fialkowski, M. et al. Principles and implementations of dissipative (dynamic) self-assembly. J. Phys. Chem. B 110, 2482–2496 (2006).
REVIEW ARTICLE
44. Klajn, R., Bishop, K. J. M. & Grzybowski, B. A. Light-controlled selfassembly of reversible and irreversible nanoparticle suprastructures. Proc. Natl Acad. Sci. USA 104, 10305–10309 (2007). 45. Boekhoven, J. et al. Dissipative self-assembly of a molecular gelator by using a chemical fuel. Angew. Chem. Int. Ed. 49, 4825–4828 (2010). 46. Debnath, S., Roy, S. & Ulijn, R. V. Peptide nanofibers with dynamic instability through nonequilibrium biocatalytic assembly. J. Am. Chem. Soc. 135, 16789–16792 (2013). 47. Emond, M. et al. Energy propagation through a protometabolism leading to the local emergence of singular stationary concentration profiles. Chem. Eur. J. 18, 14375–14383 (2012). 48. Krabbenborg, S. O. Surface Gradients Under Electrochemical Control Ch. 7, PhD thesis, Univ. Twente (2014). 49. Kudernac, T. et al. Electrically driven directional motion of a four-wheeled molecule on a metal surface. Nature 479, 208–211 (2011). 50. Ragazzon, G., Baroncini, M., Silvi, S., Venturi, M. & Credi, A. Light-powered autonomous and directional molecular motion of a dissipative self-assembling system. Nature Nanotech. 10, 70–75 (2015). 51. Gasparini, G., Dal Molin, M. & Prins, L. J. Dynamic approaches towards catalyst discovery. Eur. J. Org. Chem. 2010, 2429–2440 (2010). 52. Fanlo-Virgós, H., Roig Alba, A.‑N., Hamieh, S., Colomb-Delsuc, M. & Otto, S. Transient substrate-induced catalyst formation in a dynamic molecular network. Angew. Chem. Int. Ed. 53, 11346–11350 (2014). 53. Nguyen, R., Allouche, L., Buhler, E. & Giuseppone, N. Dynamic combinatorial evolution within self-replicating supramolecular assemblies. Angew. Chem. Int. Ed. 48, 1093–1096 (2009). 54. Del Amo, V. & Philp, D. Integrating replication-based selection strategies in dynamic covalent systems. Chem. Eur. J. 16, 13304–13318 (2010). 55. de Greef, T. F. A. & Meijer, E. W. Materials science: Supramolecular polymers. Nature 453, 171–173 (2008). 56. Klajn, R., Wesson, P. J., Bishop, K. J. M. & Grzybowski, B. A. Writing selferasing images using metastable nanoparticle “inks”. Angew. Chem. Int. Ed. 48, 7035–7039 (2009).
Acknowledgements
This work was supported by the NWO, the ERC, COST CM1304 and the Dutch Ministry of Education, Culture and Science (Gravitation Program 024.001.035). The authors wish to express their gratitude to M. Colomb-Delsuc for helping with the graphics in Fig. 4.
Additional information
Reprints and permissions information is available online at www.nature.com/reprints. Correspondence should be addressed to S.O.
Competing financial interests
The authors declare no competing financial interests.
NATURE NANOTECHNOLOGY | VOL 10 | FEBRUARY 2015 | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
119
