122 Table V.
ANALYTICAL CHEMISTRY Some Exanlples, Analysis of Technical Products Substance
Rubber D-0 Rubber D-1 Rubber D-2 Poly(methy1 methacrj-late) Leather 031154 Calves hair Lieht concrete Mortar, weathered
'4mount Taken, Mg.
60.07 66.84 90.40 1081 945.5 194.77 136.90 263.0 69.28 64.40 98.54
% s, Found 0.831 0.785 0.782 0.064 0.063
% s, Expected 0.74
1.61
1.61 3.77 2.12 2.10
1.25
0.79 0.74 0.06
0.06 1.50 1.50 3.66 2.4 2.4
..
studied by Smith ( 1 6 ) , and i n a recciit paper by Iiahane (8) a review is given of its applications in analytical chemistry. In both these papers the risks attending the use of perchloric acid are discussed. In the special case of sulfur determination it seems safe to decompose the sample with a mixture of nitric and perchloric acid, as the nitric acid destroys all the easily oxidizable matter, leaving only a residue of material more resistant to the perchloric acid. When nitric acid is used, the temperature of the acid mixture rises slowly and smoothly, so that the oxidizing power of the perchloric acid is gradually increased, thereby effecting a progressive breakdown of the sample. If a new substance or material is t o be investigated a small sample, 50 mg. or so, should be decomposed by the sernimicromethod and if the decomposition proceeds without charring or excessive foaming the sample weight can be increased. If charring occurs, the amount of nitric acid can be increased or the time for reflux boiling prolonged. The use of catalysts according t o Smith ( 16) may speed up the procedure appreciably. For example, when samples of wood are to he decomposed, 5 ml. of the acid mixture and an additional 5 ml. of nitric acid are used for decomposition. The reartion flask i p heated a little until a reaction starts, giving rise to large amounts of nitrous gases. When this reaction has subsided, the sample is dissolred and the procedure given for semimicro samples is followed. When the nitric acid is almost driven over into the reflux collector, the flame is moderated so that the perchloric acid does not react too rapidly, thereby driving fumes of perchloric acid out of the HI)paratus. These directions for wood samples can be followed with only minor modifications for most types of plant material. The
aniounts of hydrobromic acid and reduction mixture should be chosen so that the proportions remain the same as those in the semimicroprocedure. However, if extra nitric acid is added, there is no need t o increase these quantities, as the amount of perchloric acid left in the reaction flask is unchanged. The method has also been applied to a few samples of technical products of other kinds. Samples of leather, coal, rubber, plastics, hair, light concrete, and mortar were analyzed. Some of the results are listed in Table V. These samples have been analyzed mainly as described for wood samples; in the case of concrete and mortar, the same results were obtained if the wet decomposition was omitted. The method should be suitable for all materials that can be decomposed by wet decomposition. In the case of aqueous solutions it is wise t o add some hydrogen peroxide to the sample, so that neither hydrogen sulfide nor sulfur dioxide may escape before the concentration of nitric acid has reached an Oxidizing level. ACKNOWLEDGMENT
The author thanks Rose llarie Persson for great help in carrying out analyses and John Haslam, Sven Jennerholm, and Gdnther Siessen for providing samples. LITERATURE CITED
Balks, R., and Wehnnann, O., Bodenkunde u. PfEanzeneriiBhr. 5, 48 (1937). 12) Bethge, P. O . , A d . Chim. Acta 9 , 129 (1953). (1)
(3) Ibid., 10, 317 (1954). (4) Grote, W., and Krekeler, H., Angew. Chem. 46, 106 (1933). (5) Johnson, C. M., and Nishita, H., ANAL.CHEM.24, 736 (1952). (6) Jones, J. H., J. Assoc. Ob'ic. Agr. Chemists 26, 182 (1943). (7) Kahane, E., Caoutchouc gutta percha 24, 13549 (1927). (8) Kahane. E., &err. Chem.-Ztg. 55, 209 (1954). (9) Kahane, E., and Kahane, If.,Bull. SOC. chim. 1, 280 (1934). (10) Klingstedt, F. W., 2.a n d . Chem. 112, 101 (1938). (11) Lematte, L., Boinot, G., and Kahane, E., C o m p f . r e m i SOC. bid. 96, 1211 (1927). (12) Lorant,, I. S., 2. physwl. C h e m 185, 246 (1929). ENG.C H E M .A, N A L . E n . 1 5 , 6 0 2 (1943). (13) Luke, C. L., IND. (14) Meyer, F. R., and Ronge, G., Angew. Chem. 52, 637 (1931)). (15) Roth. H., Mikrochemie ~ e r .Mikrochim. Acta 36-37, 379 (1951). (16) Smith, G. F., Anal. Chini. Acta 8 , 397 (1953). ( l i ) Smith, G. F.. and Deem, A. G., IND. ENG.C H E W ,ANAL.ED.4, 227 (1932). (18) Staudinger, H.. and Kiessen, G., Chem. Ber.9, 1223 (1953). (19) Wolesensky, E., I d . Eng. Chem. 20, 1234 (1928). RECEIVEDfor review May 12, 145.5. .4ccepted September 8, 1955.
Tit ration of Gallium with Ferrocyanide Application of Dead-Stop End Point NEIL R. FEllER and D. F. SWINEHART Department o f Chemistry, University o f Oregon, Eugene, Ore.
Dilute solutions of gallium(II1) salts as low as 0.001M may be titrated with potassium ferrocyanide using the dead-stop end point. The optimum conditions for the titration were determined to be 230 mv., pH = 2.0, and 50" C. The presence of ammonium ion, various acids, and pH were studied. The stoichiometry of the reaction corresponds to the formation of gallium ferrocyanide, Gaa[Fe(CN)s],.
A
METHOD for determination of gallium in dilute aqueous solution employs the dead-stop end point apparatus of Foulk and Bawden (1). It consists of titrating gallium(II1) Polutions with potassium ferrocyanide to form gallium ferrocya-
nide, Ga4[Fe(CN)6]3,which is a nhite precipitate. .4s long as any gallium(II1) is present in solution, the galvanometer remains near aero, but as soon as any excess ferrocyanide ion appears in solution, a large deflection occurs. A potentiometric method has been described by Kirschman and Ramsey (9). EXPERIMENTAL
Standard solutions of gallium nitrate and potassium ferrocyanide were prepared from reagent grade materials. The potassium ferrocyanide was recrystallized as the trihydrate and weighed directly t o make standard solutions. The gallium was weighed as the metal, dissolved in nitric acid, and diluted to volume. All other chemicals employed were reagent grade. pilute solutions of gallium nitrate were titrated a t about 50 C. and with a potential difference between the platinum
V O L U M E 2 8 , N O . 1, J A N U A R Y 1 9 5 6 wire elect,rodes of 230 mv. Above 300 mv. the end points became erratic and with the potent,ial difference much below 200 mv. the galvanometer deflect,ions became too small t o yield clearly defined end points. The opt'imum temperature was chosen to be 50" C. because the reaction was too sluggish at room temperature (26' C.) and, a t 75" C., the end points became erratic. However, even at 50" C. the reaction was not instantaneous, a9 immediately after each addition of ferrocyanide there mas a large initial deflection of the galvanometer which decreased to a small steady value after about a minute. The galvanometer employed had a sensitivity OF 1 .T x ampere per millimeter.
3.8
4.0
4.2
44
For 39 titrations with about 0.02111 ammonium sulfate present in the gallium solutions, the average value of R was 1.057 and the average deviation was 6 parts per thousand. I n the case of zinc, the presence of ammonium ion is necessary in order t o obtain sharp end points (3). I n the present case, end points are obtained in either the presence or absence of ammonium ion, but about 5% more potassium ferrocyanide solution is required t o reach the end point when the gallium solution contains about 0.02M ammonium ion than when none is present. Since the theoretical stoichiometric ratio should be unity on the basis of the concentrations of gallium nitrate and potassium ferrocyanide, it appears desirable to titrate in the absence of ammonium ion if possible. The effect of acidity on the titration curve was studied by titrating potassium ferrocyanide into a gallium solution which was 0.75JI wit'h sulfuric acid :tnd into another which was 1.2M with perchloric acid. I n both cases the sharpness of t,he end point was considerably reduced (as indicated in Figure 1). However, the end points were sharp Irhen hydrochloric, sulfuric, or perchloric acid was employed to adjust the pH of the gallium solution to about 2.0, and the stoichiometry was unchanged. T n o experiments giving identical results indicated t,hat it is possible to titrate 0.00750M potassium ferrocyanide with 0.0100.1.1 gallium nitrate, if the time between the addition of the potassium ferrocyanide solution to the acid solution at 50" C. and the commencement of titration with gallium is not more than a few minut,es. Upon waiting for longer times, apparently the ferro'cyanide ion decomposes in warm acid, because the solution and the gallium precipitate become discolored and the end points are erratic. K h e n the gallium nitrate concentration was reduced to 0.001df and titrated with 0.00075.V potassium ferrocyanide, the average deviation for three titrations vias 6% and the average deviation for five titrations of 0.000131 gallium(II1) nitrate with 0.0001 1J1 potassium ferrocyanide was 10%. I n both cases the gallium solut,ions were 0.02.11 in ammonium ion.
4.6
m l o f 0 . 0 0 7 5 0 M K4Fe[CNI6 Figure 1. Titration curves of O.OIOOAVf Ga(NO& with 0.0075OM K,Fe(CN)B 0
123
absent, p H 2.0 with HzSOd 0 NH4+ present, p H 2.0 with H ~ S O I 8 h H 4 + a b s e n t , 0.75.M with HzSO? 0 NI&+ absent. 1.2.11 with HClOa NHd+ absent, p H 2.0 with HClOh 111 all cases 4.00 nil. of 0 01OO.U Ga(NOs)r were titrated with 1GI'e:C'S~~ "I+
Enough acid was added to the gallium solutions to maintain 3 p H of about 2.0 during the titrations. The addition of a little acid seemed t o make the galvanometer deflections at the end point larger and also eliminated the danger of precipitating gallium hydroxide. Acids, with different anions, were employed in some titrations, and, in a feTv cases, higher concentrations were used. Because of the similarity of this titration and that of zinc(I1' with ferrocyanide] experiments were conducted in the presence of ammonium ion to see whether it affected the titration curves markedly as in the case of zinc ( 3 ) . RESULTS
In Table I are typical results of the titration of 0.0100.11 gsllium(II1) nitrate with 0.00750M potassium ferrocyanide in the presence and absence of ammonium ion. Also shown is the stoichiometric ratio, 22, which is the ratio of the volume of ferrocyanide t o that of the gallium solution a t the end point. In all cases the potential difference was 230 mv., the temperature approximately 50" C., and the p H about 2.0. The average value of R for 31 titrations with no ammonium ion was 1.005 and the average deviation was 6 parts per thousand
Table 1. Titration of 0.01OOM Gallium Nitrate with 0.00750M Potassium Ferrocyanide Ga(K0a) 4 , AIl.
n
00
3 00 10 00 10.00 1.5 00 15.00 17 00 1:. 00
a
K4Fe(CN)a. M I . NH; present
absent
NH:
5.28 5.25 IO. 50
__R" S H ; pre.~erit NH;
5 00
5,01
10 02 10 00 15.00 15 03 17. 17.01
10.51 15.70 15 72 17.87 17 90
no
absent
10.5ti 1 055 1 050 1 051 1 047
1 000
1.048 1 051 1.053 1.051
1 002
I . 008
I 002 1,000 I 000
1 000 1 001
Av. I 002 0.002 0 002 A v . deviations R is ratio of volumes of ferrocyanide to gallium a t each end point.
-___~___
..
~~
~
~___
The interference of other cations ivas not studied, but presumably the presence of any other cations which torm insoluble salts with ferrocyanide ion would cause difficulty. I n summary, i t IS possible to titrate gallium(II1) uith ferrocyanide in very dilute acid solutions using the dead-stop end point, but to obtain reproducibility bettei thfiri 1 %, the conditions must be maintained constant. LITER4TI'RE CITED (1) Foulk, C. IT., and R a w d e n , A . T., J . -4rn. Cheirr. SOC.48, 2045 (1926). ( 2 ) Kirschnian, 1%.D., and Knmsey. J . B., Ibid.. 50, 1682 (1928). (3) Sainehart. I). F., . ~ N A I . .CHEM.23, 380 (1951).
RECEIVED f o r review .1;11!. 12. 195.5.
.4ccepted October 17, 19%
