346
K.
s.
1)t-I.IS.
the permanganate imparts a pink ccilor to the solutioti. X few trials will enable the operator to ascertain this point Lvitli accuracy and rapidity. This work was conducted more for t h e purpose oi ascertaining the values of the methods when working \vitli the rapitlity whicli is required in actual practice, than for :tl)solute :irciirac! , to 0.01 or 0.001 per cent. I have no tioulit that \-olliartl',q proc be made to check closer than 0.64 or 0.461)er ceiit. r.espwti\.el!., of t h e metallic coiitent of a i l ore, but i i i pr:ictirnl ii!c.talluq,iral anal>.ses, it is not alwa\.s practicx1)le to esci-c%iwtlic iirc preca:itioiis. The method is cn1):il)le of \vitle :i11l)liration, aiid where rapidity is esseiitial, is huflicieiitly accurate iii most, if iiot all cases. .
.~
. .
.
THE WET A S S A Y FOR COPPER. I l r R S. X)I-I~IS.
I'
seeins difficult for the iiletallurgical chemists of this country to settle up011 a uriiforiii nietliod for the rapid determination of copper. We should have a staiitlartl method, applicable for all coinniercial work, which woul~lbe fairly accurate under a s many possible varying coiitlitioiis. so that results obtained froin the s a n e ore, by diffcreiit chemists, should be substantially uniforiii. For about nine niontlis past I have been engaged upon an esteiided series of esperiiiieiits. 1iaI.itig for their object a determination of the chief causes for \.ai-iation aiid error i i i tlie methods inost generally e~~iplo!.edaiitl for the purpose of finding a niodification of coiiinioii methods ivliicli noulcl be an iniproveinelit upon those now used. At the same time I have iiiade myself c o n v e r s n t with iiiuch of the current literature upon the subject aiid the observations hereiii offered, while based iipoii my individual experiments. are corroborated, i l l most part, by the published results of others. There are at present tliree nrell-recognized methods employed i n tlie United States for the deterinination of copper. Each iiiethod has its own advocates, and it is perfectly fair to s a y , that either method in the Iiaiids of a skilled chemist. thoroughly u 11derst a n d i ng the re actions of the met 11o d s , work i 11g with all due care, xi11 J-ield substailtially the same results. T h e nietliods referred to are :
THE WET ASSAY FOR C O P P E R .
347
I . T k cyanide method. 2 . T h e iodide. 3 . T h e electrolytic. Acolorimetric method is also employed for the deterniination of copper, when the percentages fall below two per cent. I have made no particular study of this method, and as it is only employed in special cases, I shall make no further reference to it. T h e cyanide method depends upon the fact, that when a solution of potassium cyanide is run into an ammoniacal solution of copper the blue color is discharged. T h e reaction is as follows : (”,)z(”,),,CuO(NO,)zf 4KCN f 3HzO= K,Cu(CN), f 2 K N O , + 4NII,OH. This method is fully described in Furinan’s Manual of Practical Assaying. T h e following precautions should always be carefully observed. I . T h e bulk of the liquid titrated should always be uniform. By inattention to this an error of from two tothree per cent. is possible. 2 . T h e solution should always be cooled to the temperature of the laboratory before titrating, otherwise an error of about three per cent. is possible. 3. T h e amount of ammonia added should be nearly constant, otherwise the possible error niay amount to as much as five per cent., or even more. These precautions are general and must always be carefully observed, in every modification which may be made in the method. In the ordinary modification of the method, ferric hydroxide is aliiiost certain to be precipitated, upon the addition of the animoniuni hydro’xide. If the amount be small, no error is apt to occur, but it should always be filtered off. If the amount be large, it is not easy to wash out all of the copper salt, thereby causing lower results, unless large quantities of washwater are employed, thus increasing and varying the bulk. T h e error arising from increased hulk may be obviated, by taking, after mixing; an aliquot part of the solution ; if the solution be not thereby made very dilute, the error is so slight that it niay be neglected. If salts of manganese are present, the end reaction cannot be determined, owing to the liquid first turning green, finally black. T h e presence of large quantities of calcium, I found, confused the end reaction, causing error of importance. Experiments made showed that magnesia did not interfere, and the presence of antimony and arsenic was found to cause no sensible variations.
348
R. S. DULIN.
Zinc, which is almost certain to be present in varyingamount, is a possible source of great error. T h e following resclts were obtained from a large number of carefully conducted experiments. Only the averages are given and they are substantially the same as the extremes. In these experiments the bulk of the liquid varied from twenty-five to as much as fifty cc., thereby causing a slight error, for which no correction has been made. This error would not substantiall>r change the results Careful attention was paid to the precautions previously enumerated. Weight of copper. Weight of ziuc.
Cyanide u3ed.
Iiicreast ,
0.05
0.00
0.05
0.01
10.4 cc. 11.0 I '
0.05
0.02
11.7
1.3
0.05
0.03
12.3
1.9 ,'
0.0j
0.04 0.05 0.06
12.0
0.05 0.05 0.05 0.Oj
0.07 0.08
0.05
0.09
14.0 21.6 24.j
/L
j '
2.5
)'
3.6 ''
16.1
18.o
0.6 m.
( (
5.7 8.5
''
11.2
I '
13.9
',
'' I'
These results show that there is a gradual increase of about six-tenths cc. in the amount of the cyanide solution required, until the amount of zinc present nearly equals the amount of copper, when the increase becomes variable, until the amount of zinc becomes greater than the amount of copper to the extent of twenty per cent., when the increase, though larger, about two and seven-tenths cc., again becomes regular. T h e effects of cadmium are similar, as shownby the following results, which are also the averages of a large number of experiments, in which the extremes are farther removed from the mean than was found with zinc. As in the preceding experiments the precautions previously enumerated were very carefully observed, except in the case of bulk, in which the variations were identically the sanie a s with the experiments with zinc. Weight of copper.
Weight of cadmium
0.05
0.00
0.05
0.01
0.05 0.05 0.05
Cyanide used.
Increase.
0.02
10.4 cc. 10.6 10.8 "
0.2 cc. 0.4 ' '
0.03
11.1
0.7
' I
0.04
11.6 I C
1.2
I'
0.Oj
0.05
12.0
1.6
' I
0.05
0.06
12.4
0.05 0.05
( '
"
2.0
0.07
13.5 ''
0.08
14.5
3.1 4.1
"
)' ' I
'
