Biology operates through autonomous chemically fuelled molecular machinery, 1 including rotary motors such as ATP synthase 2 and the bacterial flagellar motor. 3 Chemists have long sought to create analogous molecular structures with chemically powered, directionally rotating, components. 4-17 However, synthetic motor-molecules capable of autonomous 360 directional rotation about a single bond have proved elusive, previous designs lacking either autonomous fuelling 7,10,12 or directionality. 6 Here we show that 1-phenylpyrrole 2,2′-dicarboxylic acid 18,19 (1a) is a catalysis-driven 20,21 motor that can continuously transduce energy from a chemical fuel 9,20-27 to induce repetitive 360 directional rotation of the two aromatic rings around the covalent N-C bond that connects them. Upon treatment of 1a with a carbodiimide, 21,25-27 intramolecular anhydride formation between the rings, and the anhydride's hydrolysis, both occur incessantly. Both reactions are kinetically gated 28-30 causing directional bias. Accordingly, catalysis of carbodiimide hydration by the motor-molecule continuously drives net directional rotation around the N-C bond. The directionality is determined by the handedness of both an additive 31 that accelerates anhydride hydrolysis and that of the fuel, and is easily reversed. >97% of fuel molecules are consumed through the chemical engine cycle 24 with a directional bias of up to 71:29 with a chiralitymatched fuel and additive. In other words, the motor makes a 'mistake' in direction every 3-4 turns. The 26-atom motor-molecule's simplicity augurs well for its structural optimisation and the development of derivatives that can be interfaced with other components for the performance of work and tasks. 32-36In 1999, Feringa reported 13 the first synthetic molecular rotary motor, based on light-driven isomerisation around a double bond. 14 Contemporaneously, 5 Kelly explored a chemically fuelled triptycene derivative, 15 conceptually related to Feynman's ratchet, 16,17 that directionally rotated 120 about a single bond. Our group have described catenanes that undergo pulsed 11 or autonomous 9 chemically fuelled rotation of one macrocycle around another and have recently used 21 carbodiimide chemical fuels to drive a rotaxane-based information 37-41 ratchet. A feature of the latter system's fuelling cycle (Supplementary Information, Section S2.5) is the formation of a transient activated ester. 23 Similar activated ester and anhydride intermediates have been exploited in dissipative assembly [25][26][27] and Hartley has shown that tethering diphenic acids as anhydrides produces a dynamic change in the angle between the biphenyl rings. 42,43 As both carbodiimide-fuelled formation of an anhydride and its subsequent hydrolysis can occur under the same set of conditions, 25,26,42,43 we reasoned that asymmetric catalysis of the formation and/or hydrolysis of the anhydride could potentially directionally bias the rotation of rings joined by

Chemical reaction networks that transform out-of-equilibrium 'fuel' to 'waste' are the engines that power the biomolecular machinery of the cell. Inspired by such systems, autonomous artificial molecular machinery is being developed that functions by catalysing the decomposition of chemical fuels, exploiting kinetic asymmetry to harness energy released from the fuel-to-waste reaction to drive non-equilibrium structures and dynamics. Different aspects of chemical fuels profoundly influence their ability to power molecular machines. Here we consider the structure and properties of the fuels biology has evolved and compare their features to those of the rudimentary synthetic chemical fuels that have been used to date to drive autonomous nonequilibrium molecular-level dynamics. We identify desirable, but context-specific, traits for chemical fuels together with challenges and opportunities for the design and invention of new chemical fuels to power synthetic molecular machinery and other nanoscale processes. MainFuels are consumed to provide the energy that devices and processes require to perform useful work. [1][2][3] The free energy available from a chemical reaction can be harnessed by molecular machines and dissipated to offset work performed, thus preserving the Second Law of Thermodynamics when tasks are carried out through stochastic molecular-level dynamics. 4,5 In this way chemical engines (Fig. 1) 6 transduce energy from chemical fuels and have the potential to power synthetic molecular nanotechnology 7-13 by driving and sustaining processes out-of-equilibrium. [7][8][9]14 While some synthetic molecular machines use light, 15,16 electrochemistry 17 or transmembrane gradients, 18 chemical fuels provide an attractive alternative energy source (Fig. 2). 19 Although the waste generated in chemical fuel-to-waste reactions must be dealt with (either recycled, as happens with ADP in the cell, or removed, as occurs for water and CO2 with aerobic respiration), chemical fuels are unencumbered by many of the issues faced by powering processes with other forms of energy. Photo-and electrochemistry often produce unstable intermediates, reactive radicals or cause photobleaching, all of which can limit the number of cycles that complex molecules survive for. Furthermore, in contrast to photons, chemical fuels have the potential to be stored and transported, enabling flexible and responsive systems that operate autonomously when and where they are needed. 19,20 Indeed, active organisms that require a high-density energy supply (such as animals) tend to rely exclusively on chemical energy sources, whereas photosynthesising organisms cannot harvest enough energy from sunlight to support powerintensive behaviours, such as rapid movement. The same trend is apparent in technology; lightpowered flight, for example, remains a major engineering challenge. For these and other reasons, while the energy input to biology largely originates from light (via photosynthesis), chemical fuels (such as adenosine triphosphate (ATP), Fig. 2A)...

We report a rotaxane-based information ratchet in which the macrocycle distribution is pumped away from equilibrium using a carbodiimide fuel. A carboxylate group on the axle, nonequidistant between two macrocycle binding sites, efficiently catalyzes the hydration of a carbodiimide fuel to the corresponding urea waste, with >80% of the fuel molecules reacting through the machine-catalyzed pathway. The energy of the reaction is harnessed by kinetic differentiation of the mechanical states of the machine driving the macrocycle to the binding site distal to the catalyst. Steric hindrance between the macrocycle and the fuel slows the reaction of the carboxylate group (to form a barrier to macrocycle movement) in the proximal co-conformer, whereas hydrogen bonding between the macrocycle and the barrier accelerates hydrolysis of the activated ester proximal isomer. The two directionally biased processes reinforce each other’s effect, resulting in a doubly kinetically gated ratchet that achieves 1:18 directionality, an exceptional degree of selectivity for a synthetic chemically fueled molecular motor.

Information is a physical quantity, a realisation that transformed the physics of measurement and communication in the latter half of the 20th Century. However, the relationship and flow between information, energy and mechanics in chemical systems and mechanisms remains largely unexplored. Here we analyze a minimalist experimental example of an autonomous artificial chemically-driven molecular motor -a molecular information ratchet -in terms of information thermodynamics, a framework that quantitatively relates information to other thermodynamic parameters. This treatment reveals how directional motion is generated by free energy transfer from the chemical to the mechanical processes involving the motor. We find that the free energy transfer consists of two distinct contributions that can be considered as "energy flow" and "information flow". We identify the efficiency with which the chemical fuel powers the free energy transfer and show that this is a useful quantity with which to compare and evaluate mechanisms of, and guide designs for, molecular machines. The study provides a thermodynamic level of understanding of molecular motors that is general, complements previous analyses based on kinetics, and has practical implications for designing and improving synthetic molecular machines, regardless of the particular type of machine or chemical structure. In particular, the study confirms that, in line with kinetic analysis, power strokes do not affect the directionality of chemically-driven molecular machines. However, we also find that under some conditions power strokes can modulate the molecular motor current (how fast the components rotate), efficiency with respect to how free energy is dissipated, and the number of fuel molecules consumed per cycle. This may help explain the role of such conformational changes in biomolecular machine mechanisms and illustrates the interplay between energy and information in chemical systems.

Chemically fueled autonomous molecular machines are catalysis-driven systems governed by Brownian information ratchet mechanisms. One fundamental principle behind their operation is kinetic asymmetry, which quantifies the directionality of molecular motors. However, it is difficult for synthetic chemists to apply this concept to molecular design because kinetic asymmetry is usually introduced in abstract mathematical terms involving experimentally inaccessible parameters. Furthermore, two seemingly contradictory mechanisms have been proposed for chemically driven autonomous molecular machines: Brownian ratchet and power stroke mechanisms. This Perspective addresses both these issues, providing accessible and experimentally useful design principles for catalysis-driven molecular machinery. We relate kinetic asymmetry to the Curtin−Hammett principle using a synthetic rotary motor and a kinesin walker as illustrative examples. Our approach describes these molecular motors in terms of the Brownian ratchet mechanism but pinpoints both chemical gating and power strokes as tunable design elements that can affect kinetic asymmetry. We explain why this approach to kinetic asymmetry is consistent with previous ones and outline conditions where power strokes can be useful design elements. Finally, we discuss the role of information, a concept used with different meanings in the literature. We hope that this Perspective will be accessible to a broad range of chemists, clarifying the parameters that can be usefully controlled in the design and synthesis of molecular machines and related systems. It may also aid a more comprehensive and interdisciplinary understanding of biomolecular machinery.