A. R. Middleton, H. L. Miller
1916 Journal of the American Chemical Society  
The use of dimethylglyoxime as a reagent for the detection and determination of nickel, discovered by Tschugaevl in 1905 and developed by Brunk,2 has become quite general. In addition to its simplicity of manipulation and freedom from interference by other metallic ions this reagent is unusually delicate. The brilliant scarlet color and extreme insolubility of nickel glyoximine3 permits detection of one part of nickel in at least 350,000 parts of water and a modified method of using the reagent
more » ... f using the reagent will be described in this paper by which we have been able to detect one part of nickel in more than 4,000,000 parts of water. For the detection of traces of nickel in cobalt salts this reagent has not proved entirely ~atisfactory.~ Cobalt combines with dimethylglyoxime to form an extremely soluble compound of brown color. Either because the nickel salt is soluble in this compound or, more probably, because the cobalt appropriates most of the reagent, no nickel is precipitated by ordinary amounts of reagent from cobalt salt solutions even though a considerable amount is present. The object of this investigation was to devise a method by which the cobalt ion should be suppressed, thus permitting the reagent to react with nickel only and avoiding the necessity for large amounts of reagent. Treadwell,6 following a suggestion of Tschugaev, accomplishes this result by transforming the cobalt salt into a cobaltic ammine by concentrated ammonia and hydrogen peroxide before adding dimethylglyoxime. We shall show that this method is unsatisfactory and fails when much cobalt is present. The most striking differences in the chemical behavior of nickel and cobalt are ( I ) the greater susceptibility of the latter to oxidation to the trivalent condition, and (2) the greater stability of its complex ions, both positive and negative. Of the various complex ions formed by cobalt the most stable are the complex cyanides, that of trivalent cobalt being decidedly more stable than that of bivalent cobalt. Nickel forms complex cyanides of a different type, resembling those of bivalent copper, whereas the cobalt cyanides are analogous to the iron cyanides. In the classic method of Liebig6 for detecting nickel in cobalt salts, the inferior stability ,of the nickelocyanide ion together with the ready oxidizability of cobalto-.cyanide to cobalticyanide has long been used to effect a separation. For * Ber., 38, 2 5 2 0 (1905).
doi:10.1021/ja02266a008 fatcat:5mmwztie5vbylnrhcodce5ad6m