972
I N D U S T R I A L A N D ENGINEERING CHEMISTRY
Vol. 20. No. 9
Ceric Sulfate as a Volumetric Oxidizing Agent VII-Determination of Vanadium in Presence of Chromium, Tungsten, and Iron1 H. H. Willard and Philena Young UNIVERSITY OF MICHIGAN, ANNA R B O R , MICH.
Vanadyl ion may be determined volumetrically with HE present paper is a or hydrochloric acid solution, ceric sulfate in hot sulfuric, hydrochloric, or percontinuation of prethe reaction a t the end chloric acid solution. The titration of ceric salt with point being more rapid with vious work on the use vanadyl sulfate proceeds quantitatively in hot sulof ceric sulfate as a volumeteither of the last two acids. furic, nitric, or perchloric acid solution. ric oxidizing agent,2 and Correct results may be obThe oxidation of vanadyl ion by ceric sulfate in the deals with the titration of tained a t room temperature, presence of chromic salts, ferric salts, and tungstic vanadyl salts alone and in but the reaction is entirely acid is selective in hot sulfuric acid solution. Also, the presence of iron, chrotoo slow to be practical. the reduction of ceric ion by ferrous sulfate in the presmium, and tungsten. Such a The end-point break averaged ence of vanadic acid is selective at room temperature. study seemed important as 125 to 200 mv. per 0.05 cc. Tungstic acid is kept in solution by dissolving it in permanganate is not entirely of 0.1 N ceric sulfate. Curve sodium hydroxide and pouring it back into the origisatisfactory in this reaction. I, Figure 1, is an example of nal solution. In this soluble form it does not interKelley and Conant3 and a titration in hydrochloric fere. Hamner4 have described the acid solution. Analyses of chrome-vanadium and chrome-vanapermanganate titration of The titration of ceric suldium-tungsten steels show that this method of devanadyl sulfate in the presfate with vanadyl sulfate was termining vanadium is an accurate one. ence of iron and of chroimpossible a t room temperamium-a method which is ture. At 70-80" C. the reacsatisfactorv except that the accurate determination of tion was rapid with 5 to 15 cc. sulfuric acid per 100 cc. of soluthe end ;oilit is* difficult on account of the color, and tion, though the break a t the end point decreased as the sulfuric Kolthoff and Tomiceks have determined the end point acid content was increased. Ten to twenty cubic centielectrometrically, their end point break being extremely meters of perchloric or 5 to 15 cc. of nitric acid per 100 cc. poor. Zintl and ZaimisG have reported that they could of solution were very satisfactory and the break a t the end not obtain accurate results using Kolthoff and Tomicek's point under these conditions averaged 125 to 175 mv. per method, and the authors have had the same experience. Some 0.05 cc. of 0.1 N vanadyl sulfate. Table I1 gives data of chromium was always oxidized. The following work shows typical titrations, showing that this is a satisfactory method that excellent results may be obtained when vanadyl salt of determining cerium after it has been oxidized to ceric salt. is titrated with ceric sulfate.
T
Note-After this nork was submitted (April, 19:s) a paper appeared by Furman, J . A m . Chem. Soc., 60, 1675 (1928), in which the titration of vanadyl salt with ceric su!fate in sslfuric acid so!ut.on is described. He applied the method to ferro-vanadium, but not, hovever, to chromevanadium or chrome-vanadium-tungsten steels.
Experimental
A platinum-0.1 N KC1-AgC1 electrode system was used. The vanadyl sulfate was prepared by reduction of very pure ammonium vanadate in sulfuric acid solution with sulfur dioxide, and removal of the excess by boiling. It was standardized by electrometric titrat,ion in hot sulfuric acid solution with standard potassium permanganate; identical values of 0.1227 gram vanadium were obtained from duplicate analyses of 20-cc. portions of the solution. The ceric sulfate was a part of a supply made for earlier work2 and was 0.5 di in sulfuric acid. It was standardized against sodium oxalate. Measured portions of vanadyl sulfate were taken, and acid and water added to a total volume of 100 cc. The solution was heated to 70-75" C. and titrated electrometrically with ceric sulfate. The results are shown in Table I. Thus, the titration is quantitative in a sulfuric, perchloric, 1 Presented before the Division of Physical and Inorganic Chemistry a t the 75th Meeting of the American Chemical Society, St. Louis, Mo., April 16 t o 19, 1928. From a dissertation submitted by Philena Young to the Graduate School of the University of Michigan in partial fulfilment of the requirements for the degree of doctor of philosophy in chemistry. 2 Willard and Young, J . A m . Chem. Soc., 60, 1322, 1334, 1368, 1372, 1379 (1928). I J. IND. E m . CHEM.,8, 719 (1916). 4 Met. Chem. Eng., 17, 206 (1917). 6 Rcc. trao. chim., 48, 447 (1924). 6 Z.angew. Chem., 40, 1286 (1927).
Table I-Titration of Vanadyl S a l t w i t h Ceric S u l f a t e ACIDPER 100 cc. V V CHARACTER OF END SOLN. TAKEN FOUND POINT cc. Gram Gram 0 0,1227 0.1224 Fairly rapid HzSO4, sp. gr. 1 8 3 5 0.1227 0.1223 Fairly rapid 10 0.1227 Too slow (?) H S o s , sp. gr. 1.42 5 0.1227 0.1217 Fairly rapid HC104 (70;72%) 0,1227 0.1224 Fairly rapid 1: 0.1227 0.1224 Rapid HCI, sp. gr. 1.18 5 0.1227 0.1225 Rapid 0.1227 10 0.1228 Rapid 0,1227 0,1234 20 Rapid 0.0613 0.0612 5 Rapid 0.3067 5 0 3061 Rapid Table 11-Titration ACID PER 100 cc. SOW.
cc.
of Ceric S a l t w i t h Vanadyl S u l f a t e Ce TAKEN Ce FOUND
Gram
Gram
0.3333 0.3333
0.3341 0,3339
0.3333 0.3333
0.3336 0.3328
0.3333 0.3333 0.6666 0.1333
0.3336 0.3333 0.6671 0.1326
EFFECTOF PRESENCE OF CHROMIUM AND IRON-Using measured volumes of vanadyl sulfate, chromic sulfate, ferric sulfate, and acid, and titrating these solutions electrometrically a t 70-75' C. with 0.05 N ceric sulfate, experiments showed that the oxidation of 10 to 60 mg. of vanadium in the presence of as much as 100 mg. of chromium and 5 grams of iron was selective in a sulfuric acid medium. Quantitative results were not obtained when nitric, perchloric, or
*
September, 1928
INDUSTRIAL' B.VD ENGINEERING CHEMISTRY
hydrochloric acid was substituted for sulfuric acid, though a moderate amount of nitric acid along with the sulfuric acid was not harmful. Manganese caused no interference. EFFECTOF PRESEKCE OF TuxGsTm-The only method a t present for determining vanadium in the presence of chromium, iron, and tungsten is one described by the author^.^ I n that paper it was shown that tungsten as sodium tungstate solution may be poured into an acid solution containing considerable iron without any precipitation of tungstic acid. The ferric salt holds the tungstic acid in solution. Such a solution remains clear on boiling if sufficient iron is present. I n the following work, in which the solutions containing tungstic acid were not boiled, but simply heated to 7@75O C.. a ratio of iron to tungsten of 5 : l was satisfactory, such solutions rem:tining clear during the experiment but, becoming cloudy on standing. Experiments were made with solutions containing the amounts of materials indicated in Table 111. The tungsten was added as sodium tungstate and the C-X. o.osYC4So,),.Ct d. cirr, solutions were diluted Figure 1 to 300 cc. Titration Curve 1-20 cc. 0.1 .\' VOSOa titrated with 0.1 X Ce(SO&)n; 5 cc. concd. HCI in initial was made electrometvol. 200 cc. rically a t 70-75" C. Curve 11-Solution from a Cr-V steel. containing excess FeSOI, titrated with 0.05 N Ce- with 0.05 -V ceric ( S p l ) ~ t o F e + + - - + P e + + + and (VO)++-+ sulfate. (\Os) - end point. Table 111-Effect of Presence of T u n g s t i c Arid H?SOa,SP. GR. 1.83, PRESEXT Cr Fe W V TAKENV F o u w cc. M g . Grams Gram Gram Gram 10 50 1 0 2 0 0311 0 0313 10 50 2 0 4 0 0311 0 0312 10 50 3 0 6 0 0311 0 0312 10 100 2 0 4 0 0311 0 0315 20 100 2 0 4 0 0311 0 0312 10 5 cc HaPOh, sp gr 1 3 7 50 3 0 6 0 0311 0 0310
+
The end-point break varied from 55 to 90 mv. per 0.05 CC. of 0.05 ;Ir ceric sulfate in all but the last experiment, in which it amounted to 32 mv. for the same volume of oxidizing agent. Therefore, the use of phosphoric acid as a means of holding the tungstic acid in solution is not to be recommended, not only because it is unnecessary but also because it makes the end point less distinct. Determination of Vanadium in Chrome-Vanadium Steels After the steel was dissolved in sulfuric acid and nitric acid added to oxidize the iron, it was found that further treatment with a suitable oxidizing agent was required to remove all carbonaceous and reducing material from the solution. An excess of permanganate or of ceric salt, follo-i~ed by boiling, vas not satisfactory. Persulfate in the presence of silver nitrate gave better results. However, after the addition of excess ferrous sulfate to this solution, the voltage was found to increase gradually, and if a titration was made a t this stage nith ceric sulfate, the volume of solution required between the Fe++ + Fe++f and (VO)++ + (VOa)end points vias slightly greater than the theoretical valuean indication that there was still reducing material in the solution. If the solution previously oxidized with silver and persulfate and treated with excess ferrous sulfate was allowed t o stand for 10 minutes, treated with a slight excess of dilute permanganate, and then a slight excess of ferrous sulfate. 7 Willard and Young, IXD. EXG.CHEX, 20, 764 (1928).
973
the results for vanadium were quantitative. Therefore, this double oxidizing treatment was always adopted. For a sharp change in potential a t the first end point, a cold solution of fairly high acid concentration was most satisfactory, and a t the second end'point a hot solution of lower acid concentration. Curve 11, Figure 1, was obtained by the titration of such a solution. Results for chrome-vanadium steels are given in Table IV. The recommended procedure is given below. An alternative method for such a steel is to titrate to an end point with ferrous sulfate, after a slight excess of permanganate has been added. Then the titration with ceric sulfate will give the per cent of vanadium. If the ferrous sulfate used in this direct titration is a standard solution, and if the (MnO4)-+ Mn++ end point as well as the (VOa)- + (VO)++ end point is recorded, the volume of ferrous sulfate used between these two end points should indicate directly the amount of vanadium in the solution, provided that no chromic salt was oxidized by the excess of Permanganate. Calculations showed that the percentage of vanadium obtained in this way never differed more than 0.005 per cent from that obtained by the titration of vanadyl sulfate with ceric salt. Figure 2 shows such a titration curve. Other experiments showed that ceric sulfate could be wbstituted for permanganate in the above treatment, and that in the subsequent titration with ferrous sulfate reduction of ceric sulfate in the presence of vanadic acid was select ive . Recommended Procedure
d sample of 4 or 5 grams is convenient when the percentage of vanadium is low (0.15 to 0.25 per cent). Place it in a 600-cc. beaker, add 35 to 40 cc. of water and 10 cc. of H2S04,sp. gr. 1.83. After the steel has been completely decomposed, boil until a considerable quantity of salts separates out, in order to assist in decomposing carbides. Dilute with 20 cc. water and heat until the salts have dir- qa qolved. Add HNO1. bp. gr. 1.42, drop ao by drop, to the hot liquid until the violent oxidation of FeYOi is over (avoid more than 1 cc. excess). Boil the s o h - % tion to destroy oxides of nitrogen. Dilute *y)+o ~ ~ N F ~ s++~005NC1(S&),-cc. - ~ ~ . + to 75-100 cc., add 2 Figure 2 CC..igNOa containing Solution from a Cr-V steel, containing excess KMnOa, titrated with0.05 N FeSOa t o (MnOa) 2.5 grams dg?r'Oi per ---j bin++ and (vo3)--4 (vo) end liter, 2 grams (NE,)?- points: t b - n titrated with 0.03 N Ce(SO4)z to (VO) -+(V03) - end point. S,OR. _ _ and boil for I 5 minutes. Cool to room temperature, add 40 cc. H?S04, sp. gr. 1.5, and then dilute FeSOl (0.05 LV is convenient) until a 4- to j-cc. excess is present as indicated by the voltage. Bfter 10 minutes add a slight excess of dilute permanganate, cool the solution t o 5-10" C. and either add an excess of 0.05 N ferrous sulfate or titrate to the (VOa)- --f (VO)++ end point with FeS04, depending on the method being used. I n the former case the back titration with 0.05 N Ce(SO& to the Fe++ + Fe+++ end point is made in this cold solution. Then the solution is diluted t o 300 cc., heated to 70-75" C., and titrated with 0.05 N Ce(SO& to the (VO)++ + (VOdend point. In the latter case, after the direct titration with 0.05 N FeSOl to the (VOs)- -+ (VO)++ end point, the so++
+ +
974
INDUSTRIAL AND ENGINEERING CHEMISTRY
lution is diluted to 300 cc., heated to 70-75" C., and titrated with 0.05 N Ce(SO&. Determination of Vanadium in Chrome-VanadiumTungsten Steels
After the steel was dissolved in hydrochloric acid, the tungsten oxidized by nitro-hydrochloric acid, and the tungstic acid filtered off, sulfuric acid was added and the solution evaporated to fumes. This material was diluted somewhat, and the tungstic acid added, after it had been converted into sodium tungstate solution either by dissolving in dilute sodium hydroxide or by fusing with sodium carbonate in an open crucible and extracting with hot water. A clear solution was obtained. The use of persulfate or permanganate was unnecessary in this case after the nitro-hydrochloric treatment of the steel. Therefore, ferrous sulfate was added, either in excess, or to an end point, depending upon the method used. Titration was then made with ceric sulfate, two end points being obtained if excess of ferrous sulfate was used, otherwise one end point. The conditions for sharp breaks were the same as for chrome-vanadium steels and the curves entirely similar to those already given. Results for a chrome-vanadium-tungsten steel are shown in Table IV.
Vol. 20, No. 9
the HZWOI, add 5 cc. HCl, 3 cc. HN03, and boil down again. Dilute to 60-70 cc. with hot water, boil a few moments to dissolve any salts, and let settle on the hot plate until the supernatant liquid is clear. Filter the HzWOa, using hot 2 per cent HC1 to transfer the precipitate and to wash it. Add 5 cc. HzSO4, sp. gr. 1.83, to the filtrate, and evaporate on a low-temperature hot plate to fumes. Dilute to 70-80 cc. and heat until all salts have dissolved. Place a 150-cc. beaker under the funnel containing the H2W04, puncture the paper, and wash through most of the H z W O ~with water. Dissolve the remainder of the materia! on the filter with hot 4 per cent NaOH. Add 5 to 10 cc. more to the original beaker to dissolve any precipitate which adheres to the glass, using only about 15 cc. in all. Filter this liquid and pour the filtrate into the main filtrate from the HzW04which has been diluted previously to 70-80 cc. A clear solution will result. Add 40 cc. H2S04, sp. gr. 1.5, cool to 5-10' C., and then add 0.1 N FeS04,either in excess or to the (VO8)- 3 (VO) end point, depending on the method being used. From this point the procedure is the same as that given for chrome-vanadium steels. Summary of Analyses of Steels +
+
Recommended Procedure
I n Table IV, 4- to &gram samples of the chrome-vanadium steels were used, and 1- to 1.3-gram samples of the chromevanadium-tungsten steel.
A 1-gram sample is convenient for a steel containing 1 per cent or more of vanadium, and a 2-gram sample for a smaller per cent. Add 10 cc. of water and 30 cc. HCl, sp. gr. 1.18, to the sample in a 400-cc. beaker. Warm gently until the steel is completely decomposed and the tungsten separates out as a black powder. To the boiling hot solution add 8 cc. HN03, sp. gr. 1.42, at first cautiously, until the violent action is over. Boil, rotate the liquid frequently to expose a fresh surface of the tungsten to the oxidizing agent, and evaporate to 20 cc. If there are any dark particles in
Table IV-Vanadium in Alloy S t e e l s EXCESS FeSOd, FOLLOWEDFeSOd ADDED TO BY TITRATION WITH (VOa)-+ (VO) END Ce(S0dz TO 2 END POINT, FOLLOW~D BY STEEL POINTS TITRATION WITH Ce(SOI)? Per cent Per cent Cr-V, No. 1-(0.230% V, 0.230 0.234 2.35% Cr) 0.227 0.232 0.232 0.211 B. of S.Cr-V, No.30(b) 0.219 0.212 0.216 (0.215% V, 1.03% Cr) 0.213 B. of S. Cr-V-W steel, No. 50 0.764 0.745 0.751 (0.766% V, 3.61% Cr) ++
A Precision Pipet Viscometer' S. W. Ferris THEATLANTICREFININGCOMPANY, PHILADELPHIA, PA.
ANY instruments have been designed for the purpose of determining the viscosity of liquids when the sample at hand is too small to allow the use of a standard instrument such as the Saybolt Universal viscometer. One of the most obvious means of obtaining rough comparative results is to observe the efflux time of the oils from a pipet of small volume. Dean and his co-workers2 have devised and extensively used a pipet which they describe as highly satisfactory for use with small samples, but for which they claim no unusual degree of accuracy. Further reference to their results will be made. The present instrument was also developed primarily as a small-sample accessory to the Saybolt. Its construction is shown in Figure 1. The capacity of the pipet itself is 4 to 5 cc. between the marks. I n constructing the tip, the following method has been found very satisfactory: A glass tube of about 5 mm. inside diameter is drawn in a flame to an inside diameter of about 2 mm. Heating a t the portion having an inside diameter of about 2 mm., the tube is again drawn, reducing the diameter to somewhat less than 0.5
M
1 Presented before the Division of Petroleum Chemistry a t the 74th Meeting of the American Chemical Society, Detroit, Mi&., September 5 to 10,1927. 2 Dean, Hill, Smith, and Jacobs, Bur. Mines, Bull. 307, 45 (1922).
mm. The decrease in diameter should be gradual, as shown in the drawing. The tip is then broken off and polished on a fine emery stone in a plane perpendicular to the long axis of the tip. The tip proper, when so made, is not a true capillary, but the taper is very gradual. By making rough trials of efflux time on some oil of known viscosity, the size of the tip may easily be adjusted to give the approximate efflux time desired. As will be more fully discussed later, there is no attempt whatever to make the efflux times of any two pipets identical. Indeed, they may vary markedly without in any way destroying the usefulness of the instrument. To give times approximating one-half Saybolt, the tip when finished will probably have an orifice of 0.5 to 0.6 mm . It has been found inadvisable to fire-polish the finished orifice, because the walls of the tip are so very thin that unless the work is performed with great care the walls will be thickened at the orifice, so reducing the size of the orifice and very probably giving an orifice not circular in shape. It has not been definitely shown that such an orifice will fail to give the same results as ground orifices; however, it has been found that when the walls are thickened a t the orifice by fire-polishing in such a way that the diameter is considerably smaller than that of the walls just above, small
