SV\TfIF/rfC RrHHh.R RFSFARCIl 373 



outset ol tlie research program and lor that reason emj)hasis was placed on 

 improving the macro-structure where obvious changes could be effected. 

 We sliall consider this phase of the work next. 



\Vc have already seen how the chain molecules of GR-S vary in size and 

 in composition. They may vary also in over-all shape. Branching and 

 cross-linking leading eventually to net-work formation may result during 

 the chain growth or termination reactions. In this way variously shaped 

 molecules may arise. Obviously the situation may become very complex 

 and in reality we may have to do with mixtures where all types of molecular 

 species are present at once. 



What influence on the properties of the linal compounded and vulcanized 

 rubber do these various branched and netted chain structures have? It was 

 not recognihed at first that the gel part of GR-S was particularly different 

 from the sol in its effect on ultimate properties. This was because no 

 reliable measurements of sol or gel had been made and because sol and gel 

 behaved differently duiing the compounding arid processing steps^-^-. 

 Some workers also did not appreciate that natural rubber and GR-S behave 

 very differently in regard to the effect of processing on their ultimate 

 properties. 



It has since been established that the sol-gel properties are of importance 

 both in the processing and in the final properties of GR-S synthetic rubber. 

 It turns out that the amount of the sol and its molecular weight distribution 

 and the amount of the gel and its swelling volume, which is a measuie of the 

 intensity of netting, enables us to make predictions as to what properties a 

 given sample of rubber will exhibit during processing and in the final 

 product ^^ This does not mean that other features of the sol and gel are 

 unimportant. For example, m.ethods of estimating the degree of branching 

 (by m.eans of concentrated solution viscosity)*^' ^" of the soluble portion have 

 been worked out which undoubtedly will be useful if a more refined control 

 proves desirable. 



It is possible to make GR-S type rubber which is completely soluble. 

 Such a product requires to be characterized only as to molecular weight 

 distribution, composition and perhaps degree of branching. If the distribu- 

 tion of sol is such that there is an excess of low molecular material, the 

 copolymer besides being soft and difficult to handle, provides cured stocks 

 which have low tensile strength, poor tear and abrasion resistance, poor 

 resistance to the growth of cracks and high hysteresis loss. If, on the other 

 hand, an excess of high molecular material is present in the sol the copolymer 

 is very stiff^^ and cannot be handled in the subsequent compounding and 

 j)rocessing procedures. Aside from this ditliculty its ultimate properties 

 seem to be superior the higher the average molecular weight. When all 

 considerations of properties and processing requirements are taken into 



