The Use of Sodium Metaphosphate for Dispersion of Soils for Mechanical Analysis

"T~HE accuracy of the pipette method for the me-*chanical analysis of soils developed independently by Robinson (is) in Wales, Krauss (10) in Germany, and Jennings, Thomas, and Gardner (8) in the United States is dependent upon the completeness with which the clay fraction is dispersed. The ease with which ultimate dispersion is attained varies considerably with different soils. Experience has shown that no method of mechanical analysis is equally suited for all types of soils. In general, modification of usual routine methods involving the pipette sampling technic is frequently required if satisfactory dispersion of peats, lateritic, calcareous, or gypseous soils is to be secured. Excellent treatment of the factors involved in dispersing soils is to be found in publications by Robinson (14), Joseph and Snow (9), Thomas (19), Bodman (3), Alexander (i) , Olmstead, Alexander, and Middleton (12), and the subcommittee of the Agricultural Education Association of Great Britain (18). The possibility of developing a mechanical analysis method in the future which is applicable to all soil types through improved manipulative technic or chemical treatment prior to dispersion is rather remote. Many of the treatments already proposed are time consuming or too drastic in action. Further standardization of mechanical analysis methods, however, is not impossible if more effective deflocculating agents can be found. In the progress of work by the writer on the fixation of various forms of phosphate, considerable dispersion was observed to occur when the soils were shaken with dilute sodium metaphosphate solutions. This behavior seemed peculiar in so far as sodium metaphosphate is not a basic salt. Chemical analysis of clarified suspensions indicated a negligible absorption of the metaphosphate anion, a very marked absorption of sodium and, furthermore, considerable calcium had come into solution. Recent literature on the chemistry of the metaphosphates (13, 17) revealed that in solution sodium metaphosphate exists in a highly polymerized state probably as (NaPOg)e; hence the commonly used term, sodium hexametaphosphate. Further, sodium metaphosphate reacts with calcium compounds to form soluble highly undissociated sodium calcium metaphosphate anionic complexes which so completely remove calcium ions from solution as to prevent their precipitation by soap or detection by oxalate, phosphate or carbonate anions. It seemed plausible, therefore, that the dispersive action of sodium metaphosphate was due to a more or less complete replacement of cations, particularly calcium, from the exchange complex by sodium, a reverse equilibrium being prevented by the elimination from solution of active calcium as the undissociated sodium calcium metaphosphate complex. It seemed, therefore, that sodium metaphosphate might be very effective for dispersing soils for mechanical analysis. Preliminary data comparing sodium metaphosphate with sodium oxalate indicated the former to be equally as efficient as the latter for the dispersion of gray-brown podzolic soils. Further studies to determine the specific conditions for maximum dispersion of soils by sodium metaphosphate and the efficiency of sodium metaphosphate as a dispersing agent for soils of varied genetic origin were then undertaken. While this work was in progress a paper by Hatch and Rice (7), concerning the surface-active properties of sodium hexametaphosphate as affecting watersoftening processes, appeared. They cite Feldenheimer who in 1922 was granted U. S. Patent 1,438,588 covering the use of sodium hexametaphosphate for the deflocculation of clays. Apparently a knowledge of the existence of this patent never became widespread. It is of interest to note that of the various electrolytes studied by Vinther and Lassen (20) for