Water and alcohol, such as methanol or ethanol, are miscible and, individually, good solvents for poly(N-isopropylacrylamide) (PNIPAm), but this polymer precipitates in water–alcohol mixtures. The intriguing behaviour of solvent mixtures that cannot dissolve a given polymer or a given protein, while the same macromolecule dissolves well in each of the cosolvents, is called cononsolvency. It is a widespread phenomenon, relevant for many formulation steps in the physicochemical and pharmaceutical industry, that is usually explained by invoking specific chemical details of the mixtures: as such, it has so far eluded any generic explanation. Here, by using a combination of simulations and theory, we present a simple and universal treatment that requires only the preferential interaction of one of the cosolvents with the polymer. The results show striking quantitative agreement with experiments and chemically specific simulations, opening a new perspective towards an operational understanding of macromolecular solubility.

Conformational transitions of (bio)macromolecules in aqueous mixtures are intimately linked to local concentration fluctuations of different solvent components. Though computer simulations are ideally suited to investigate such phenomena, in conventional setups the excess of one cosolvent close to the solute leads to depletion elsewhere, requiring very large simulation domains to avoid system size effects. We, here, propose an approach to overcome this depletion effect, which combines the adaptive resolution scheme (AdResS) with a Metropolis particle exchange criterion. In AdResS, a small all-atom region, containing the solute, is coupled to a coarse-grained reservoir, where the particle exchange is performed. The particle exchange would be almost impossible had they been performed in an all-atom setup of a dense molecular liquid. As a first application of the method, we study the concentration driven reentrant collapse and swelling transition of poly(N-isopropylacrylamide) (PNIPAm) in aqueous methanol and demonstrate the role of the delicate interplay of the different intermolecular interactions.

Smart polymers are a modern class of polymeric materials that often exhibit unpredictable behavior in mixtures of solvents. One such phenomenon is co-non-solvency. Co-non-solvency occurs when two (perfectly) miscible and competing good solvents, for a given polymer, are mixed together. As a result, the same polymer collapses into a compact globule within intermediate mixing ratios. More interestingly, polymer collapses when the solvent quality remains good and even gets increasingly better by the addition of the better cosolvent. This is a puzzling phenomenon that is driven by strong local concentration fluctuations. Because of the discrete particle based nature of the interactions, Flory-Huggins type mean field arguments become unsuitable. In this work, we extend the analysis of the co-non-solvency effect presented earlier [D. Mukherji et al., Nat. Commun. 5, 4882 (2014)]. We explain why co-non-solvency is a generic phenomenon, which can only be understood by the thermodynamic treatment of the competitive displacement of (co)solvent components. This competition can result in a polymer collapse upon improvement of the solvent quality. Specific chemical details are not required to understand these complex conformational transitions. Therefore, a broad range of polymers are expected to exhibit similar reentrant coil-globule-coil transitions in competing good solvents.

Poly(N-isopropylacrylamide) (PNIPAm) is a smart polymer that presents a lower critical transition temperature (LCST) of 305 K. Interestingly, this transition point falls within the range of the human body temperature, making PNIPAm a highly suitable candidate for bio-medical applications. However, it is sometimes desirable to have a rather flexible tuning of the LCST of these polymers to further increase their range of applications. In this work, we use all-atom molecular dynamics simulations to study the LCST of PNIPAm-based (co-)polymers. We study different molecular architectures where the polymer sequences are tuned either by modifying its stereochemistry or by the co-polymerization of PNIPAm with acrylamide (Am) units. Our analysis connects global polymer conformations with the microscopic intermolecular interactions. These findings suggest that the collapse of a PNIPAm chain upon heating is dependent on the hydration structure around the monomers, which is strongly dependent on the tacticity and the presence of more hydrophilic acrylamide monomers. Our results are found to be in good agreement with the existing experimental data.

In this review, we summarize recent theoretical and computational developments in the field of smart responsive materials, together with complementary experimental data. A material is referred to as smart responsive when a slight change in external stimulus can drastically alter its structure, function, or stability. Because of this smart responsiveness, these systems are used for the design of advanced functional materials. The most characteristic properties of smart polymers are discussed, especially polymer properties in solvent mixtures. We show how multiscale simulation approaches can shed light on the intriguing experimental observations. Special emphasis is given to two symmetric phenomena: co-non-solvency and co-solvency. The first phenomenon is associated with the collapse of polymers in two miscible good solvents, whereas the latter is associated with the swelling of polymers in poor solvent mixtures. Furthermore, we discuss when the standard Flory–Huggins-type mean-field polymer theory can (or cannot) be applied to understand these complex solution properties. We also sketch a few examples to highlight possible future directions, that is, how smart polymer properties can be used for the design principles of advanced functional materials.