Routine Determination of Nickel in Cobalt-Base Alloys Ferricyanide Oxidation of Cobalt LOUIS SILVERMAN' AND HERRlAN K. LEMBERSKY Westinghouse Electric Corporation, East Pittsburgh, Pa. In ammoniacal solution, cobaltous salts are oxidized to the trivalent stage by ferricyanide and nickel is subsequently precipitated by dimethylglyoxime in acetate buffered solution. One part of nickel may be detected in the presence of 200 parts of cobalt. The usual components of high temperature alloys-tungsten molybdenum, chromium, iron, manganese, columbium, and titanium-do not interfere and no special separations are needed.
T
acids, or a perchloric acid solutioii, must be dilut,ed and t m i l t d until fret. of chlorine.
Hb: principle of the ferricy:iiiidc method ( 2 , 8) for the determination of cobalt may be applied to the gravimetric determination of nickel in the presence of very large amounts of cohalt. In ammoniacal citrate solution, olie atomof trivalent cobalt forms a Kernerlike complex with two molecules of dimet'hylglyoximc ( 1 ) ; and this complex forms no precipitate with ferric iron. Thus, the proposed procedure does not require the preliminary renioval of iron. Chromium (300 mg.), columbium (20 nig.), and tungsten or molybdenuni (60 my.) do not interfere. Tantalum, \\.hose reactions are similar t,o those of columbium, should not' int erf't,re. 3Iaterials insoluhle in aqua regia (silica, columbium carbide, Ptc.) are removed, after perchloric aci? dehydration, hy filtration. The precautions observed iii the dcteriiiination of nickel iii steel, such as regulation of acidity, volume of alcohol, amount of reagent dimethylglyoxime, tiinf, of st:mding, and temperature, of filtration are followed.
RESULTS
In Table I tire listed results obtained by this niethotf its conipared to those obtained by other acceptttble methods. Sample 1 is Bureau of Standards S o . 153 in which t h r cobalt content is more than 50 times that of the nickel and the iron content is more than 500 times that of nickel. Molybdenum, vanadium, and tungsten are also present. One-, 2-, arid 5-gram samples were analyzed by the proposed method; thc 5-gram samples yielded the more accurate results. Samples 2 to 7 are Westinghouse alloys lyith approximate analyses as listed. Samples 2 and 3 Kere checked by t,he cyanide procedure of Feigl and Kapulitzas (3, 6). Samples 4,5, 6, and 7 show check results with t,he method of Kirtchik ( 5 ) . Sam le 6 is t,he same as Sample 5 with columbium added in the form of ferrocolunibium and titanium added as ferrotitanium. Spectrographic analysis showed only traces of columbium and titanium in the final nickel precipitate.
The minimum amount, of dimethylgl~-oximereagent required is 10 ml. of a 1% solution for each 10 mg. of nickel expected t o bcpresent, plus 10 ml. for each 6 mg. of cobalt expected, plus an vxcess of 10% of reagent to obtain complete precipitation. Hoaever, it is simpler to add 1 gram of solid dimethplglyoxime for each 200 mg. of cobalt expected, and later reprecipitate the nickel under more closely controlled conditions. An acetate solutioii with a pH of 8 is preferred. The time of standing after precipitation is 1 hour for nickel in amounts of 1 to 4%, 2 hours for nickel from 0.5 to I % , and 4 or more hours for less than 0.5%. The volume of alcohol should be less than 257, ( 4 ) and the nickel precipitate should be filtered at room temperature. One important precaution must be taken. In hypochlorite solution (?), nickel is quickly oxidized to a higher valence, in which form the glyoxime is soluble; hence an acid mixture of hydrochloric acid and nitric I
Table I1 shims check results by clifferrnt,itntdysts of some highcoh:ilt :illoys. SOLUTIOKS
Cobalt Buffer Solution. Dissolve 500 grams of citric acid in 675 ml. of ammonium hydroxide and 1000 ml. of water. Alkaline Dimethylglyoxime Solution. Dissolve 5 grams of potassium hydroxide in 100 ml. of rrater, add 10 grams of dimethylglyoxime, and stir until dissolved. Dilute to 250 ml. with water. Alcoholic Dimethylglyoxime Solution. Dissolve 10 grams of dimethylglyoxime in 1 liter of alcohol. Filter before using. Tartaric Acid Solution, 50%. Dissolve 500 grams of tartaric acid in 1 liter of water. Filter before using. Potassium Ferricyanide Solution, 10%. Dissolve 10 grams of potassium ferricyanide in I00 ml. of water. One milliliter is approximately equivalent to 0.02 gram of cobalt or 0.02 gram of manganesr. Discard after 30 days.
Present address, 827 South Cramerry Place. 1.0s Angeles ;, Calif
Table I . Nickel Determination in Cobalt iIIn?s Sauiple Sample So. Size 1
(B.S. (B.S.1.53) 1.53) 2
1
Co 8.43
2 5
1
(420'mg.) 60 60 60 60
4.14
1.58
8.36
... 20
(70 mg.) 6
(-1lO'Ag.)
... ...
20 6 ZD b 26 6 0.2 26 6 U .2 1 0.2 6 0.2 6~1 1 26 6 7 26 6 1 0.1 Value listed on Bureau of Standards certificate. b Cyanide-peroxide procedure of Kirtohik (6). C Cyanide method of Feigl and Kapulitzas ( 6 ) . d Same as sample B with columbium and titanium added. 3
-1 .i
1 1
60 60 60 60 60 60
Cb
hlo
Ti
N1 Proposed Method
SI
PROCEDURE
Other Method. 0.107a 107" llb O.llb
The sample should not contain more than 50 mg. of nickel. Weigh a 1.000-gram sample and transfer to a 400-ml. beaker. Digest the alloy with 30 mi. of hydrochloric acid until most of the sample has been dissolved. Add 10 ml. of nitric acid and 15 ml. of perchloric acid (70%); then evaporate to fumes of perchloric acid. Boil 3 to 5 minutes longer, cool, add 100 ml. of water, and boil for 5 minutes to remove free chlonnv. Pour this solution into
2.03
. . . . . .
0.10
...
. . . . . .
0.10 0.11
, , ,
... ... ... ... ... ...
. . . . . . . . . . . . . . . . . .
. . . . . . . . .. . . .. . . 2
22
. . . . . .
983
0.54
0.59 1.98 2.01 2.02 1.97
... ...
0 , ' j52 i ' cc
0.59c 59 c 2 . O00 Ob
2.036 2.03b 2.02b 96 b 1.96b
984
ANALYTICAL CHEMISTRY Table 11. Reproducibility of Kickel Results Sample so.
Sample Size Grams
Kickel Analyst A
rlnalyst B
7%
%
acid solution to the filtrate. Xeutralize to litmus paper with ammonium hydroxide and add 1 ml. in excess. Dilute the solution to 250 ml. with water, heat to 60’ C., and add a 1% alcoholic dimethylglyoxime solution (10 ml. for each 10 mg. of nickel expected). Stir and allow to warm for 0.5 hour (longer for low nickels). Cool, filter through a weighed Gooch crucible, and wash 15 times with warm (50” C.) water. Dry the crucible a t 110” C. for a t least 2 hours and weigh. The increasr in weight is nickel dimethylglyoxime.
yo I i i a 600-ml. beaker containing 100 ml. of cobalt buffer solution and 60 ml. of ammonium hydroxide. Wash the 400-ml. beaker with dilute ammonium hydroxide to remove any tungstic oxide, and
stir until a clear solution is obtained. Add sufficient lOy0potassium ferricyanide solution to oxidize the cobalt and manganesr plus a 10% excess (6 ml. for each 0.1 gram of cobalt or manganese present), and mix well. The red color of the cobaltic ion appears. Add 50 ml. of alcohol and 100 ml. of alkaline dimethyl lyoxime solution, and stir well. Allow to stand for 10 minutes. (!arefuly adjust the solution with glacial acetic acid to pH 8, using a pH meter. Allow to stand for the prescribed time and if necessary, readjust the acidity with acetic acid or ammonium hydroxide. A white precipitate of excess dimethvlglyovime along with the nickel glyoxime forms. Filter through a 12.5-ern. No. 40 Whatman paper, and wash 6 times with marm (50” C.) Tvater. Discard the filtrate. Return the paper to the 600-ml. beaker and add 25 ml. of nitric acid and then 10 ml. of perchloric acid (70%). Break the paper with a stirring rod, heat gently until the paper decomposes, and then evaporate to fumes of perchloric acid. Boil 3 to 5 minutes longer, cool, add 100 ml. of water, and boil for 5 minutes to volatilize free chlorine. Filter through an 11-em. S o . 40 Whatman paper into a 400-ml. beaker and wash 6 times with hot water. Discard the paper (silica). .4dd 10 ml. of a 50% tartaric
=
weight of nickel glyoxime X 20.32 weight of sample ACKNOWLEDGMENT
The authors wish to express their thanks to Walter Mot2 who performed some of the check analyses and to George Wiener for the spectrographic work. LITERATURE CITED
(1) Ablov, A., Bull. soc. chim., 7, 151 (1940). (2) Dickens, P., and Maassen, G., Mitt. Kaiser-Wilhelm Inst., Eisenjorschung, DBsseldorf, 17, 191 (1935). (3) Diehl, H., “Applications of the Dioximes to Analytical Cheniis(4) (5)
(G)
(7) (8)
try,” Chap. 11, Part E, Columbus, Ohio, G . Frederick Smith Chemical Co., 1940. Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” p. 315, New York, John Wiley & Sons, 1929. Kirtchik, H., ANAL. CHEM.,19, 95 (1947). Lundell, G. E. F., Hoffman, J. I., and Bright, H. A., “Chemical Analysis of Iron and Steel,” p. 279, New York, John Wiley & Sons, 1931. McAlpine, R. K., J . Chem. Education, 23, 301 (1946). Steele, G. J., and Phelan, J. J.. Gen. Elec. Rev., 42, 218 (1939).
RECEIVED May
5 , 1947.
Determination of Potash in Fertilizers C. H. PERRIN Canada Packers Limited, Toronto, Canada .A modification of the official A.O.A.C. method for the determination of potash in fertilizers introduces a very rapid wet combustion procedure which operates simultaneously with the precipitation of potassium chloroplatinate. A s a result the method offers a saving in time and a simplification of apparatus.
A
RAPID and accurate method for determining potash
is badly needed by laboratories controlling fertilizer manufacture. Thornton (IO) emphasizes the importance of prompt analysis of fertilizer samples by control laboratories. This paper describes a modification of the official A.O.A.C. method, wherein the elimination of ammonium salts is effected by a new wet combustion technique. Use of this proposed method makes it possible to run a potash analysis in less than 2 hours and thus improve laboratory service and efficiency. The wet combustion method for the destruction of ammonium salts and organic matter, suggested by DeRoode in 1895 (2) and studied by Keitt and Shiver ( 7 ) , was modified in an effort to make it applicable to a variety of fertilizer mixtures (3, 8). However, collaborative trials ( 4 , 5 ) of the improved procedure gave disappointing results and interest has decreased considerably. Joy (6) reviews some disadvantages associated with the official ( 1 ) A.O.A.C. method: possible losses of potassium due to spattering or volatilization during ignition and the formation of water-insoluble residues. Shuey (9) points out advantages of avoiding the ignition step. Some advantages of n e t combustion, as practiced in the proposed method, include:
A new and extremely rapid method of destroying interfering substances, combined with precipitation of the weighing form, results in a substantial saving in time and simplification of procedure. Former methods making use of wet combustion of ammonia offered little or no saving in time. There is little danger of loss of potassium from spattering or decrepitation, whereas in the official method these losses frequently occur. I t is unnecessary to wash the precipitate with the regular ammonium chloride solution (saturated with respect to potassium chloroplatinate). I n the official method this wash is frequently necessary to remove visible impurities. The precipitate is completely soluble in water, whereas in the official method it is frequently necessary to make allowance for insoluble residues. At no stage in the new method is it necessary to evaporate solutions to dryness (as in former wet combustion methods) or ignite residues (as in the official method). Thus the danger of forming insoluble or difficultly soluble impurities is reduced and phosphates cause no interference. The precipitate does not adhere to the containing vessel and can be transferred to the crucible without the aid of a rubber policeman. In the official method this transfer is occasionally troublesome. The precipitate obtained is more uniform in particle size and appearance than in the official method. The precipitate formed is purer than that secured in the official A.O.A.C. method before the ammonium chloride wash.
