Structure of Poly(acrylic acid) in Electrolyte Solutions Determined from Simulations and Viscosity Measurements Z. Adamczyk,* A. Bratek, B. Jachimska, T. Jasin ´ ski, and P. Warszyn ´ ski Institute of Catalysis and Surface Chemistry Polish Academy of Science, Niezapominajek 8, 30-239 Cracow, Poland ReceiVed: June 26, 2006; In Final Form: September 1, 2006 In this work, the structure of poly(acrylic acid) (PAA) molecules in electrolyte solutions obtained from molecular dynamic simulations was compared with experimental data derived from dynamic light scattering (PCS), dynamic viscosity, and electrophoretic measurements. Simulations and measurements were carried out for polymer having a molecular weight of 12 kD for various ionic strengths of the supporting electrolyte (NaCl). The effect of the ionization degree of the polymer, regulated by the change in the pH of the solution in the range 4-9 units, was also studied systematically. It was predicted from theoretical simulations that, for low electrolyte concentration (10 -3 M) and pH ) 9 (full nominal ionization of PAA), the molecule assumed the shape of a flexible rod having the effective length L ef ) 21 nm, compared to the contour length L ext ) 41 nm predicted for a fully extended polymer chain. For an electrolyte concentration of 0.15 M, it was predicted that L ef ) 10.5 nm. For a lower ionization degree, a significant folding of the molecule was predicted, which assumed the shape of a sphere having the radius of 2 nm. These theoretical predictions were compared with PCS experimental measurements of the diffusion coefficient of the molecule, which allowed one to calculate its hydrodynamic radius R H . It was found that R H varied between 6.6 nm for low ionic strength (pH ) 9) and 5.8 nm for higher ionic strength (pH ) 4). The R H values for pH ) 9 were in a good agreement with theoretical predictions of particle shape, approximated by prolate spheroids, bent to various forms. On the other hand, a significant deviation from the theoretical shape predictions occurring at pH ) 4 was interpreted in terms of the chain hydration effect neglected in simulations. To obtain additional shape information, the dynamic viscosity of polyelectrolyte solutions was measured using a capillary viscometer. It was found that, after considering the correction for hydration, the experimental results were in a good agreement with the Brenner’s viscosity theory for prolate spheroid suspensions. The effective lengths derived from viscosity measurements using this theory were in good agreement with values predicted from the molecular dynamic simulations. I. Introduction Polyelectrolytes or polyions are molecules composed of a large number of covalently linked ionizable subunits. They are abundant in nature and essential for biological systems, including DNA. Polyelectrolytes are often used in pharmaceutical, cos- metic, and food industries, in ternary oil recovery, and for regulating rheological properties of suspensions. Another important field of polyelectrolyte applications is preparing multilayer films on solid substrates of a desired composition and functionality, 1-6 which is often realized by layer-by-layer (LbL) deposition of anionic and cationic poly- electrolytes. The simplicity of this procedure, the feasibility of embedding various molecules, proteins, and colloid particles into the polymeric layer, opens a broad spectrum of possibilities to produce films of targeted architecture. A controlled formation of polymeric film requires a thorough knowledge of the structure of the polyelectolyte molecules in relation to its molecular weight, ionic strength, and pH of the solutions. One of the efficient ways of learning about structural aspects of polyelectolytes are the rheological measurements, which have been performed extensively over the decades. 7-13 Because of the variety of parameters critically influencing the viscosity, especially in the range of low electrolyte concentration, the otherwise valuable results are often misinterpreted. The main source of discrepancies is the assumption of the salt-free solutions. Even in the totally deionized water, the residual concentration of ions must be higher than 2 × 10 -7 M (at neutral pH ) 7). Because of carbon dioxide dissolution, this is normally increased to ∼10 -5 M, if the measurement atmosphere is not controlled. Moreover, adding charged molecules into the solution simultaneously introduces counterions, whose concentration is dependent on the polyelectrolyte ionization (dissociation) degree, which depends in turn on the pH and ion composition. With increasing concentration, many-body effects appear, because of hydrodynamic, electrostatic, and other specific interactions between macromolecule chains. This makes the entire problem nonlinear in respect to the polyelectrolyte concentration, making the interpretation of the viscosity data rather involved. On the other hand, adsorption of polyelectrolytes on the walls of containers (capillaries) creates problems for the range of very low concentrations. This may often lead to the appearance of the maxima on the intrinsic viscosity vs concentration depen- dence. 8,9 Macromolecule aggregates forming in unfiltered solution introduce additional problems because they may break up with increasing shear rate, leading to apparent non-Newtonian behavior, interpreted as the shear-thinning processes. * To whom correspondence should be addressed. 22426 J. Phys. Chem. B 2006, 110, 22426-22435 10.1021/jp063981w CCC: $33.50 © 2006 American Chemical Society Published on Web 10/26/2006