An experimental investigation of the conditions necessary for the production of compact, single polymer molecules, in a form suitable for direct observation in the electron microscope, is described. Molecules are isolated by dispersing a dilute solution of the polymer as fine droplets on to a suitable substrate: ideally each droplet should contain either one or no polymer molecules. The solution is a mixture of two solvents, a good one and a poor one. Initially the good solvent predominates so that the probability of polymer aggregation is low. Preferential evaporation of the relatively volatile solvent on the substrate itself gives the poor solvent conditions needed for the formation of well-defined molecular spheres. Factors determining the choice of solvent, precipitant, and the composition of the mixture are discussed. There is little difficulty in obtaining single molecules with glassy amorphus polymers; rubbery polymers collapse and spherical molecules are formed only if the entire preparation is carried out at a temperature below that of the glass transition; crystalline polymers are not amenable to this technique. To obtain sufficient contrast the particles have to be shadowed and it is shown that, although certain dimensions are distorted by the metal coating, the shadow length faithfully represents the true particle diameter. Molecular weights, and their distribution, when of the order of a million and above, can readily be accurately determined. Conventional methods are unreliable in this region of high molecular weight.
Intrinsic viscosity and friction coefficient of polymer molecules in solution: Porous sphere model