synopsisThe development of gel permeation chromatography (GPC) has provided a convenient tool for the rapid determination of molecular weight distribution. The question has arisen as to the suitability of the method for specification purposes. The present work, suggested by the Naval Air Systems Command, represents an attempt to assess the precision of the method through a series of tests carried out by a number of laboratories using identical procedures on the same samples. Ten laboratories agreed to take part. Naval Ordnance Station, Indian Head, worked out standard conditions for operation of the chromatograph, for calibration of the columns, and for analysis of the GPC curves. Two samples of polystyrene were used by the various organizations for calibration of their instruments. Number-average molecular weight, heterogeneity index, and cumulative molecular weight distribution curves were determined on four samples of carboxylterminated polybutadiene (CTPB) and two samples of hydroxyl-terminated polybutadiene (HTPB), all unidentified except by letter code. All laboratories used identical directions for setting up CTPB and HTPB calibration curves which were based on curves determined from vapor-pressure osmometer molecular weights and GPC count numbers of fractionated material. Variation among the different laboratories was 0.15 in heterogeneity index, and a maximum of 1200 in molecular weight provided one aberrant set of values was eliminated. The six samples had heterogeneity indices from 1.15 to 1.54, while molecular weight varied from approximately 3000 to 6000. The average coefficient of variation of the molecular weight values was 6.2 f 0.7%, which is quite acceptable. Variation in heterogeneity index was too great for specification purposes when considered among the different laboratories, but may be sufficiently good when measured by any one laboratory.

No abstract

synopsisWork is reported on the effect of variation of molecular weight distribution on the p r o p erties of binder stocks. This work used three prepolymers (polypropylene glycol 1000, 2000, and 4000 and polypropylene glycol 400, 1000, and 2000) of different molecular weight and known molecular weight distribution. The prepolymer of the middle molecular weight was employed as standard, and stocks were prepared from it. To this prepolymer was added some of the high and low molecular weight prepolymen in proportions such that the crosslink density of stocks made from the mixture was the same as that of stocks made from the middle molecular weight prepolymer alone. Two series were prepared with different crosslinking agents. The proportions of high and low molecular weight prepolymer were successively increased until finally none of the middle molecular weight remained. Heterogeneity indexes of the glycols and their mixtures were measured by gel permeation chromatography. Elongation at break, modulus of elasticity, and extent of solvent swelling were determined on the crosslinked stocks. Properties which vary with crosslink density were found to show decreasing values with increasing heterogeneity index even though the stocks were formulated to a constant crosslink density. Narrow-distribution stocks reach the maximum degree of cure faster than the broader-distribution stocks. Infrared and thermal analysis confirm that monodisperse polymer has a greater extent of reaction than heterodisperse polymer. Results showed that the variation in elongation a t break to be expected bccause of lot-to-lot variations in heterogeneity index is probably not greater than the experimental variation in the elongation test below a heterogeneity index of 1.5. Above 1.5, however, if the effects observed are entirely due to variation in heterogeneity index, lot-to-lot variations in molecular weight distribution cannot be ignored. An explanation is presented based on the varying ability of prepolymer molecules of different size to diffuse through uniform mesh openings resulting in lower final extents of reaction for broad distribution material.