COLORIMETRIC METHOD FOR THE DETERMINATlON OF IRON. 93X.-A Coloiimetric Method foy the Determination ofSmall Percentages of I?;.on in Coppe,. Alloys.By ARNOLD WILLIAM GREGORY, B.Sc. (Lond.).IT is often n matter of considerable importance to be able to determineaccurately the amount of iron in copper alloys, on account of theeffect of this element on their physical properties. The gravimetricmethods are such that only by working with very large quantities ofmaterial can accurate results be obtained. Moreover, there is alwaysa danger OF iron being introduced into the solution of the alloy by theaddition of large quantities of reagents which may contain a trace o94 COLORIMETRIC METHOD FOR THE DETERMINATION OF IRON.that element, and from external sources during the lengthy method ofprocedure.The following method has been found to give extremely accurateresults ; it is simple in operation, and very rapid in execution.It isbased upon the colour reaction given by salicylic acid and ferricchloride. The violet coloration produced when salicylic acid is addedto ferric chloride, although affording a delicate test for iron underproperly chosen conditions, cannot be relied on for the quantitativedetermination of that element, since the colour is destroyed in thepresence of mineral acids and also by excess of alkalis.If, however, an excess of ti solution of sodium acetate be added to aferric salt, and then a solution of salicylic acid in acetic acid, a deepred colour is produced. Under these conditions, the depth of colouris proportional to the amount of iron present, and this method may beused for the estimation of small quantities of iron.I n the case of copper alloys, the blue or green colout.a€ the solutionentirely masks the red colour produced by the iron, This difllculty isovercome by the addition of a weak solution of potassium cyanide inquantity suficient for the formation of the colourless, complex cyanideof copper and potassium. The red colour is unchanged by this treat-ment. The exact method of procedure is as follows : 0.2 gram of thealloy is dissolved in a minimum quantity of strong nitric acid. If aprecipitate is produced, due to tin or antimony, the liquid is dilutedslightly and filtered,To this solution, 20 C.C. of a concentrated solution of sodium acetateare added, and 10 C.C.of a 2 per cent. solution of salicylic acid inglacial acetic acid. A 3 per cent. solution of potassium cyanide is nowadded gradually until the green colour of the solution has disappeared,and the precipitate of copper cyacide is re-dissolved. The solution,which is red if iron is present, is now made up t o a definite volume(depending on the intensity of the colour), and a measured amount istransferred to a Nessler comparison tube.Into a similar tube, 20 C.C. of sodium acetate solution and 10 C.C. ofthe salicylic acid solution are placed, and diluted to approximately thesame volume as the solution of the alloy. A standard solution offerric chloride is added drop by drop, with constant stirring, until thecolour produced is similar in intensity in the two tubes.From theamount of the standard solution used, the percentage of iron may becalculated.By this method, it is possible to detect as little as 0*00002 gram ofiron in the presence of 0.2 gram of copper.A strong solution of potassium cyanide must on no account be used,as this gives a coloured solution with pure copper salts, especially onwarming.Lead, if present, must be removed as sulphateDERIVATIVES OF TETRAMETHYL GLUCOSE. 95This test cannot be satisfactorily employed in the case of alloys con-taining considerable percentages of bismuth. Zinc and antimony,however, may be present without any appreciable error beingintroduced.Ex PE R I M ENT A L.A solution of pure copper sulphate was made, such that 1 C.C.wasequivalent to 0.02 gram of copper. Ten C.C. of this solution and3 C.C. of concentrated nitric acid were placed in each of seven beakers,and to all but the first of these varying amounts of a standardsolution of ferric chloride were added. The tests were then carriedout as described above, the solutions in each case being made up to100 C.C. Thesecond column shows the number of C.C. of ferric chloride solutionadded to the copper sulphate; the third gives the percentage of ironthat each quantity represents; the fourth gives the number of C.C.of ferric chloride required to produce the same colour as that obtainedin each of the test experiments, and the fifth shows the percentageof iron calculated from the volume of ferric chloride added :The results obtained are given in the following table.Number.1.2.3.4.5.6.7.C.C. ofFeC1, used.nil125101520Per cent, Fe.nil0.010 *020 -050.100.150.20C . C . ofFeCI, required.nil1'12.15.210.415.721 *oPer cent. Fenil0.0110.0210'0520.1040.1570'210APPLEBY IRON WORKS,FRODINGHAM