Brush-Like Polymers: New Design Platforms for Soft, Dry Materials with Unique Property Relations Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 22, 2019
  • Daniel, William
    • Affiliation: College of Arts and Sciences, Department of Chemistry
  • Elastomers represent a unique class of engineering materials due to their light weight, low cost, and desirable combination of softness (105-107 Pa) and large extensibilities (up to 1000%). Despite these advantages, there exist applications that require many times softer modulus, greater extensibility, and stronger strain hardening for the purpose of mimicking the mechanical properties of systems such as biological tissues. Until recently, only liquid-filled gels were suitable materials for such applications, including soft robotics and implants. A considerable amount of work has been done to create gels with superior properties, but despite unique strengths they also suffer from unique weaknesses. This class of material displays fundamental limitations in the form of heterogeneous structures, solvent loss and phase transitions at extreme temperatures, and loss of liquid fraction upon high deformations. In gels the solvent fraction also introduces a large solvent/polymer interaction parameter which must be carefully considered when designing the final mechanical properties. These energetic considerations further exaggerate the capacity for inconstant mechanical properties caused by fluctuations of the solvent fraction. In order to overcome these weaknesses, a new platform for single component materials with low modulus (<105 Pa) must be developed. Single component systems do not suffer from compositional changes over time and display more stable performance in a wider variety of temperatures and humidity conditions. A solvent-free system also has the potential to be homogeneous which replaces the large energetic interactions with comparatively small architectural interaction parameters. If a solvent-free alternative to liquid-filled gels is to be created, we must first consider the fundamental barrier to softer elastomers, i.e. entanglements - intrinsic topological restrains which define a lower limit of modulus (~105 Pa). These entanglements are determined by chemistry specific parameters (repeat unit volume and Kuhn segment size) in the polymer liquid (melt) prior to crosslinking. Previous solvent free replacements for gels include elastomers end-linked in semidilute conditions. These materials are generated through crosslinking telechelic polymer chains in semidilute solutions at the onset of chain overlap. At such low polymer concentrations entanglements are greatly diluted and once the resulting gel is dried it creates a supersoft and super-elastic network. Although such methods have successfully generated materials with moduli below the 105 Pa limit and high extensibilities (~1000%) they present their own limitations. Firstly, the semidilute crosslinking methods uses an impractically large volume of solvent which is unattractive in industry. Second, producing and crosslinking large monodisperse telechelic chains is a nontrivial process leading to large uncertainties in the final network architecture and properties. Specifically, telechelics have a distribution of end-to-end distances and in semidilute solutions with extremely low fraction of chain ends the crosslink reaction is diffusion limited, very slow, and imprecise. In order to achieve a superior solvent-free platform, we propose alteration of mechanical properties through the architectural disentanglement of brush-like polymer structures. In recent year there has been an increase in the synthetic conditions and crosslinking schemes available for producing brush-like structures. This makes brush-like materials an attractive alternative to more restrictive methods such as end-linking. Standard networks have one major control factor outside of chemistry, the network stand length. Brush-like architectures are created from long strands with regularly grafted side chains creating three characteristic length scales which may be independently manipulated. In collaboration with M. Rubinstein, we have utilized bottlebrush polymer architectures (a densely grafted brush-like polymer) to experimentally verify theoretical predictions of disentangled bottlebrush melts. By attaching well-defined side chains onto long polymer backbones, individual polymer strands are separated in space (similar to dilution with solvent) accompanied by a comparatively small increase in the rigidity of the strands. The end result is an architectural disentangled melt with an entanglement plateau modulus as much as three orders of magnitude lower than typical linear polymers and a broadly expanded potential for extensibility once crosslinked. Disentangled brush-like molecules are well suited to create biologically soft materials. However such architectures possess multiple characteristic length scales, making quantitative analysis of their structures challenging. Because of their large molecular dimensions, densely grafted bottlebrush structures can be readily measured by molecular imaging techniques. We have combined AFM with other methods to characterize the important molecular parameters of n_bb(degree of polymerization (DP) of the backbone) and n_sc⁄n_g (number of side chains monomers per backbone monomer). This technique allows for fast and accurate analysis of molecules for comparison with theory and computer simulation. Taking the analysis one step further, we have also explored the behavior of bottlebrush molecules with bimodal side chain distributions on 2D surfaces. We have shown that side chain adsorption is a statistically random process leading characterizing the adsorbed side chain population as the weight average of all side chains (adsorbed and unadsorbed). This discovery allowed for extraction of the side chain dispersity data using molecular widths taken from Langmuir-Blodgett film AFM analysis. With the architectural and entanglement properties of the brush-like architectures characterized we turned to designing supersoft elastomers (modulus <105 Pa). There are two approaches to consider when forming elastomers: (1) chemical crosslinking which offers ease of synthesis and superior environmental stability, and (2) physical crosslinking which allows for reversible self-assembly, molding of complex shapes, and greater control over mechanical performance via microphase separation effects. We first turned to chemical crosslinking in order to explore the potential of simple bottlebrush elastomers. An elastomer series was synthesized with changing network strand degree of polymerization (DP) (n_x) and side chain DP (n_sc) using a radical based polymerize-through method. This synthetic technique allowed for bulk polymerization of prefabricated mono-functional side chains (macromonomers) and bifunctional crosslinkers. The resulting networks possessed uniform structures in terms of mesh size and grafting density and could be molded into experimentally useful shapes. The samples were subjected to mechanical testing and analyzed using a universal stress-strain relation developed by A. Dobrynin. It was discovered that the stress evolution of bottlebrush networks could be completely described by two key parameter: the network modulus and the network pre-extension factor (β). The materials’ network modulus displayed a linear dependence on network strand DP and side chain DP allowing for easy tuning of modulus from 102 to 104 Pa. It was discovered that bottlebrush elastomers possess very high values of β (10 to 100 times higher than typical linear architectures). This high pre-extension factor created strong strain hardening behavior (rapid increase in modulus with respect to deformation) while maintaining extensibility and low modulus. The strain hardening behavior was found to scale inversely with respect to modulus as the side chain molecular weight was altered. This correlation allowed for a simultaneous decrease in initial modulus combined with faster onset of strain hardening which is unseen in single component linear polymer systems. Transitioning from chemically to physically crosslinked elastomer allows us to create supersoft materials which can be molded into complex shapes and recycled for numerous uses. To facilitate such materials we have generated ABA type triblock copolymers with bottlebrush middle blocks and crystallizable linear A blocks. We have shown the potential for these systems to spontaneously microphase separate into well-defined, super-soft polymer networks which can be easily melted and reformed. The cross-link density was effectively controlled by the DP of the side chains with respect to the DP of the linear tails. Shorter side chains allowed for crystallization of the linear tails of neighboring bottlebrushes forming a soft network without forming a continuous crystalline phase. However, steric repulsion between longer side chains hindered phase separation and crystallization, thus preventing network formation altogether. Initial stress-strain analysis of these networks display even higher values of β (2 to 4 times higher despite possessing longer network strand lengths than those employed by our chemically crosslinked networks). The steric repulsion between the micro phase-separated domains created high pre-extension of the bottlebrush backbones leading to a combination of supersoft modulus with strong strain hardening. Such properties are rarely seen outside of biological tissues and points to ABA type polymers as a forerunner in future research on the topic of mimicking tissue mechanical response. Our previous work has dealt mostly with highly grafted brush-like materials (dense bottlebrushes). Recently we have developed a reliable method to synthesize materials with well controlled n_g (number of backbone monomers per side chain). These materials displayed a series of mechanical property correlations as functions of the changing molecular parameter triplet (n_x,n_sc,n_g). These trends include: (1) Inverse relations between modulus and extensibility (typical in polymer materials), (2) concurrent increase of modulus and extensibility (unseen in linear systems), (3) and (4) holding one property constant while altering the other (previously impossible without altering chemical makeup). The comb-like materials (high n_g) allow for vast increases in the extensibility of branched networks, beyond the practical limitation set by linear polymer entanglements (~5 time extension). Control of the completed architectural triplet effectively expanded the possible combinations of mechanical properties from a strain line (trend 1) to a rich 2D space of possibilities (trends 1-4 combined). The implications of these new trends are discussed at the end of the dissertation, as well as how fine control of all three brush-like parameters (n_x,n_sc,n_g) represents the future of studies in this field.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Craig, Stephen
  • Ashby, Valerie
  • You, Wei
  • Rubinstein, Michael
  • Sheiko, Sergei
  • Doctor of Philosophy
Degree granting institution
  • University of North Carolina at Chapel Hill
Graduation year
  • 2017

This work has no parents.