12 BRYANT AND HARDWICK: THE DETERMINATION OF CHROMIUM IN CHROMITE [VOl. 75 The Determination of Chromium in Chromite Part 111. Determination of Chromium in a Standard Sample of U.S. National Bureau of Standards Chrome Refractory No. 103 by Selected Methods BY F. J. BRYANT AND P. J. HARDWICK SYNoPsIs-The two selected methods, both involving fusion with sodium peroxide, were simple, rapid and capable of giving reproducible results. In the standard sample dried to constant weight a t 10.5' to 110OC. they indicated chromium equivalent to 36.96 and 37-03 per cent. of Cr,O, respec- tively. The higher results by the second method were attributed to incomplete elimination of chlorine from the decomposition of traces of manganese dioxide on boiling with dilute hydrochloric acid before the chromate titration.When the boiling was prolonged, the second method gave consistent values of 36-97 per cent. Allowing for the presence of 0.08 per cent. of vanadium in the sample the average results by the two methods were 36.92 and 36.93 compared with the certificate figure of 36.97 per cent. of Cr,O,. h T R o D u c T I o N - o f the four methods previously tested (Part 111) using a synthetic sample containing a known amount of chromium, two (Methods 1 and 2) gave satisfactory results after correction for the small amount of vanadium present and seemed suitable for routine or umpire determinations of chromium in chromite. This Part describes independent analyses of chrome refractory No. 103 of the U.S. National Bureau of Standards by these two methods.In both methods, any vanadium present reacts similarly to chromium, being reduced from Vv to VIV on the addition of ferrous ammonium sulphate and not re-oxidised in the back-titrationJail., 19501 PART 111. DETERMINATION OF CHROMIUM IN A STANDARD SAMPLE 13 with potassium dichromate. On the other hand, titration with potassium permanganate causes the re-oxidation of vanadium to Vv but the reaction is slow, and even under the optimum conditions described in Part I2 there is some sacrifice in the sharpness of the end- point. From information available it seemed that vanadium was unlikely to be present in more than small amounts in most ores. In view of the convenience of the dichromate titration and the satisfactory results obtained with the synthetic sample, it was decided to follow the procedure described in Part I1 and apply a correction for vanadium determined separately, if necessary, by a colorimetric method.EXPERIMENTAL The test sample was a finely ground standard sample of U.S. National Bureau of Standards Chrome Refractory No. 103, previously well mixed (sieving tests showed that all but a very small fraction of the sample passed a 200-mesh B.S. sieve, of aperture 0-0030 inch). Weighed portions of the sample were dried a t 105" to 110°C. to constant weight before test. The certificate value of the percentage of chromium in the sample was unknown to the analysts during the work. The volumetric techniques used by the two independent analysts varied slightly, and are described below. (1) In one case (F. J.B.) the strength of the ferrous ammonium sulphate solution was so adjusted that when 100ml. were added to the chromate solution the excess was of the order of 1 to 5 ml., which was determined by adding the standard dichromate solution from a 10-ml. burette. The standard dichromate solution was accordingly made slightly weaker than the ferrous ammonium sulphate solution and the standardisation performed by taking 100 ml. of the iron solution, adding 100 ml. of the dichromate solution and completing the titration from the 10-ml. burette. The actual concentrations selected, after initial experiments had shown the ore to contain approximately 25 per cent. of chromium, were 0.074 N dichromate and 0.075 N ferrous ammonium sulphate in 5 per cent. sulphuric acid. Potassium dichromate was taken as the primary standard and one calibrated 100-ml.pipette and one 10-ml. burette were used throughout. (2) In the other case (P. J. H.) the graduated apparatus was limited to a 50-ml. pipette and a 50-ml. burette, both previously re-calibrated. Ferrous ammonium sulphate -solution in 5 per cent. by volume sulphuric acid was freshly standardised against 0.1 N potassium dichromate solution both before and after each group of titrations. To minimise burette drainage errors, the ferrous ammonium sulphate solution was made slightly stronger than the dichromate solution and the main bulk of the titrant was added from the pipette. The titration was then ' completed by the addition of dichromate solution from the burette. Usually, on re-standardising the ferrous ammonium sulphate reagent after a period of about an hour, titrations agreed to within 0.02 ml. in a titration of about 50 ml.of 0.1 N dichromate. All reagents were of AnalaR quality and the potassium dichromate was re-crystallised and dried to constant weight a t 105" to 110" C. Spectrographic examination of the crucibles used for sodium peroxide fusions showed that the nickel contained negligible amounts of chromium and vanadium and approximately 0.3 per cent. of manganese. A "blank" with the reagents and a control determination with a known amount of potassium dichromate (added as standard dichromate solution to the alkaline extract of the fused peroxide) were carried out with each group of test determinations. METHODS AND RESULTS METHOD AUXILIARY OXIDATION WITH AMMONIUM PERSULPHATE- This method, which had given satisfactory results in previous work with the synthetic sample, seemed suitable for routine or umpire determinations of chromium in chromite.It is described in detail here for convenience of reference. An accurately weighed amount, approximately 0.5 g., of the well mixed, finely gtuund sample dried to constant weight a t 105" to 110°C. was transferred to a nickel crucible of about 40-ml. capacity and was thoroughly mixed with 5g. of sodium peroxide. A thin layer of peroxide was sprinkled over the surface of the mixture, the crucible was covered with a nickel lid and heated gently over the small flame of a: Bunsen burner for about 5 minutes until the contents were molten. Heating was then continued at dull redness for a further14 BRYANT AND HARDWICK: THE DETERMINATION OF CHROMIUM IN CHROMITE [Vol.75 5 minutes with occasional careful swirling of the contents of the crucible to ensure thorough mixing. On cooling, the crucible and lid were transferred to a 600-ml. beaker having a cover glass, 100 ml. of water were added and the solution was boiled for 10 minutes to decompose most of the peroxide. The crucible and lid were then thoroughly rinsed and removed, the alkaline extract was acidified with 120 ml. of diluted sulphuric acid (1 + 3), warmed to dissolve the precipitate, and transferred to a 1000-ml. conical flask. (In the first four tests the slightly turbid solution was filtered at this stage and the small residue ignited and examined for unattacked ore.The residue was found to consist mainly of manganese dioxide with a trace of chromium (0.0001 g. of Cr,O,) probably not completely washed from the filter paper). Water was added to increase the volume to 500 ml. and followed by 5 ml. of nitric acid (sp.gr. 1-42), 25 ml. of 1 per cent. silver nitrate solution and 5 g. of ammonium persulphate. After addition of one OT two silica chips to prevent bumping, the solution was heated to boiling and boiled vigorously for 15 minutes to destroy excess of persulphate. The solution was cooled slightly, 20 ml. of diluted hydrochloric acid (1 + 3) were added to destroy permanganate and vigorous boiling was continued for a further 16 minutes to expel chlorine. On cooling, a measured volume (100 ml.) of standard ferrous ammonium sulphate solution was added from a pipette and the excess titrated back with potassium dichromate solution after addition of 25 ml.of phosphoric acid (sp.gr. 1-75) and 5 drops of a 0.3 per cent. solution of barium diphenylamine sulphonate in water as indicator. The results obtained by both analysts are set out in detail in Table I. Tests 1 to 12 were carried out a t one laboratory (P. J. H.) and 13 to 20 at the other (F. J. B.). In each group of tests the “blank” was less than 0.00005 g. of Cr,O, (= 0.02 ml. of 0.1 N K,Cr,O,). All weights of Cr203, in this and subsequent tables, are expressed to the nearest 0.05 mg. TABLE I DETERMINATION OF CHROMIUM IN A STANDA:RD SAMPLE OF CHROME ORE BY METHOD 1 Test No. 1 2 3 4 5 6 7 8 Control 9 10 11 12 13 14 15 16 17 18 19 20 Control Control Control Control Control Wt.of air-dried sample g. 0.5238 0.5229 0.5238 0.5213 0.5220 0.5243 0.5232 0.5275 0.5140 0.5102 0.6146 0.5104 Loss on drying at 105” to lloo c. g. 0.0026 0.0026 0.0026 0-0026 0-0025 0-0025 0.0029 0.0029 0.0023 0-0023 0.0025 0.0025 Moisture 0-50 0.50 0.50 0.50 0-48 0.48 0.55 0.55 0.45 0.45 0.49 0-49 % Wft. of dried r-A-------, Total wt. Wt. of Cr,O, found sample taken 0.5205 0-5197 0. 5 20 2 0.5178 0.112730 0.5190 0.5180 0.11 2730 0.5203 0.5246 0. I2730 0.5117 0.5079 0.12740 0.5 12 1 0.5079 0- 112740 0*5000 0*5000 0.5000 0.5000 0*5000 0*5000 0.5000 0-Ei000 0. I. 875 9. in solution g. 0-19200 0.19170 0.1 9260 0-19110 0.12725 0.191 65 0.191 10 0.12735 0.19280 0.1 9390 0.12745 0-18935 0.1 8770 0.12740 0.18910 0.1 8765 0.1 2740 in of Cr,O, residue found g* g.0.00010 0.19210 0.00010 0.19180 0.00010 0-19270 0~00010 0*19120 0.00010 0.12735 0.19165 0.19110 0.12735 0.19280 0.1 9390 0.12745 0.18935 0-18770 0.1 2740 0.18910 0.1 8765 0.12 740 0*18480 0.18500 0.18470 0.1 8520 0.1 8475 0.18460 0.18460 0.18485 0.18760 cr*o, in dried sample yo found 36.91 36.91 37-04 36-93 36-93 36.90 37.06 36.96 37.00 36.96 36.93 36-95 36-96 37.00 36.94 37.04 36.95 36.92 36.92 36.97 Mean .. 36.96 The difference of 0.16 per cent. between t.he extreme values found for the percentage Cr20, in the dried sample was greater than expected in view of the consistent results obtained with the synthetic sample by this method and the sharp titration end-point, which was sensitive to within 0.02 ml. of 0.1 N potassium dichromate in a total titration volume ofJan., 19501 PART 111.DETERMINATION OF CHROMIUM I N A STANDARD SAMPLE 15 about 75 ml. in tests 1 to 12 and 95 ml. in tests 13 to 20. The control determinations, on the other hand, showed an excellent recovery of chromium in line with the precision of the titrations. The slight excess of chromium found in one instance (Control 7, 8) was probably due to traces of chlorine remaining after decomposition of permanganate (derived from the nickel 1 :rucible) with hydrochloric acid although the solution was boiled for 15 minutes before itration and the "blank" was satisfactory. Th: variation in the percentage of Cr,O, found in the test samples was probably not due to differences in moisture content as all samples were dried overnight a t the same temperature (105" to 110" C.) to constant weight.The percentage moisture found in the sample (0.5 &- 0.05 per cent.) varied slightly in tests made on different days. The average percentage of Cr,O, in the dried sample, corrected for the presence of 0.08 per cent. of vanadium" is 36.92 per cent., which is slightly lower than the certificate value of 36.97 per cent. of Cr,O, for chrome refractory No. 103. METHOD 2-hJXILIARY OXIDATION WITH PERMANGANATE I N SMALL EXCESS- Satisfactory results were obtained by this method in previous work (Part 11) with the synthetic sample. With the advantages of being somewhat shorter than Method 1 and requiring less expensive reagents, the method was considered more suitable for routine work. The procedure of fusion with sodium peroxide, subsequent boiling with water and acidi- fication with 120ml.of diluted sulphuric acid (1 + 3) was the same as in Method 1. After addition of 5 ml. of nitric acid (sp.gr. 1.42) and warming to dissolve the precipitate, the almost clear solution was transferred to a 1000-ml. conical flask and diluted to 500 ml. with water. 1 to 2 nil. of 0.1 N potassium permanganate solution were added to give an excess as shown by the deepening amber colour, and the solution was then boiled for 5 minutes. After cooling slightly, 20 ml. of diluted hydrochloric acid (1 + 3) were added and boiling continued for 15 minutes to destroy permanganate or manganese dioxide and remove traces of chlorine. Addition of one or two silica chips helped to maintain even boiling. On cooling, a measured volume (100 ml.) of standard ferrous ammonium sulphate solution was added from a pipette and the excess was titrated back with 0-1 N dichromate solution as described in Method 1.The results obtained are shown in Table 11. TABLE I1 DETERMINATION OF CHROMIUM IN A STANDARD SAMPLE OF CHROME ORE BY METHOD 2 Test No. 21 22 Blank Control 23 24 Blank Control 25 26 Blank Control Wt. of air-dried sample g. 0.5251 0.5229 0-5094 0.5138 0-5 163 0.5086 Loss on drying at 105" to 110" C. 6. 0.0029 0.0026 0.0025 0.0026 0*0025 0.0025 Moisture 0.65 0*60 % 0-49 0.49 0.48 0.49 Wt. of dried sample taken g. 0.5222 0.5203 0.1 2 730 0.5069 0.5113 0.12740 0.5138 0.5061 0.12740 Wt. of found g- 0.1 9335 0.19265 nil 0-12745 0.18780 0.1 8945 0.00030 0.12775 0.19020 0.18735 nil 0.12755 Mean Cr2OfJ Cr20, in dned sample yo found 37.02 37.03 37.05 37.05 37.01 37.02 37-03 The results obtained by this method are more consistent than those obtained by Method 1 and the maximum difference of 0.04 per cent.of Cr,O, is of the order expected from the precision of the titration. The mean percentage of Cr,O, in the sample, however, is 0.07 higher than the average of the twenty determinations by Method 1. Allowing for 0.08 per cent. of vanadium in the sample, the corrected figure of 36.99 per cent. of' Cr,O, is in close agreement with the certificate value of 36.97 per cent. A slight excess of chromium was found in control determinations by this method and in one instance (Blank 23,24) a small "blank" was recorded. These discrepancies were probably due to traces of chlorine remaining in the solution after boiling with dilute hydrochloric acid to destroy permanganate and suggested that the percentage of Cr,O, found in the test samples * 1 ml.of 0.1 N K,Cr,O, = 0.0051 g. of V = 0.00253 g. of Cr,O,.16 BRYANT AND HARDWICK: THE DETERMINATION OF CHROMIUM IN CHROMITE [VOl. 75 was slightly greater than the true value. The presence of chlorine in the final solution after 15 minutes’ boiling may have been due to the slow reaction of hydrochloric acid with the traces of manganese dioxide remaining after treatment of the boiled alkaline solution with dilute sulphuric acid. In Method 1, oxidaiion with ammonium persulphate and silver nitrate in the presence of nitric acid converted all the manganese to permanganate, giving a clear solution, especially noticeable with the “blank.” No difficulty was then experienced in removing traces of chlorine by boiling for 115 minutes with dilute hydrochloric acid in blank or control determinations except in one instance (Control 7, 8).In the present method it appeared necessary to boil for at least 15 minutes after complete reduction of permanganate and the disappearance of the last traces of turbidity due to manganese dioxide. This was confirmed by the results of the following series of tests (Table 111) in which, after addition of dilute hydrochloric acid, boiling was continued for 5 to 10 minutes until the solutions were quite clear and then for a further 15 minutes to expel traces of chlorine. Tests 27 to 30 were carried out at one laboratory (P.J. H.) and 31 to 34 at the other (F. J. B.). It is seen that satisfactory results were obtained in the control determinations, having regard to the limits of experimental error set by the sensitivity of the titration end-point (0.02 ml. of 0.1 N K,Cr,O, = 0.00005 g. of Cr,O,). Consistent results were obtained for the percentage of Cr,O, in the test sample but, as expected, the average was slightly lower than that previously found. After correcting for 0.08 per cent. of vanadium, the average percentage of Cr,O, was 36.93, which agrees with 36.92 per cent. found by Method 1 and compares reasonably well with the certificate value of 36.97 per cent. DETERMINATION Test No. 27 28 Blank Control 29 30 Blank Control 31 32 33 34 Blank Control Wt. of air-dried sample g. 0.5031 0.5022 0.5054 0.504 1 TABLE I11 OF CHROMIUM IN A STANDARD SAMPLE OF CHROME ORE BY MODIFIED METHOD 2 Loss on drying at 106” to 110” C. g. 0.0024 0.0023 0.0024 0.0023 Wt. of dried sample Moisture taken 0.48 0.5007 0.46 0.4999 %I g. 0.12730 0.48 0-5030 0.46 0-5018 0.12730 0.5000 0-5000 0.5000 0.5000 0.1875 Wt. of found g. 0.18500 0.18485 0.00005 0.12740 0.18585 0.18555 nil 0.12735 0-18475 0.18510 0- 18495 0.18485 0~00020 0.1 8750 Mean . . cr*os cr*o, in dried sample yo found 36.95 36.98 36-95 36-98 36.95 37.02 36-99 36.97 36.97 Acknowledgment is made by one of qs (F. J. B.) to Mr. E. Booth for experimental This paper is published by permission of the Chief Scientist, Ministry of Supply, and the assistance. Government Chemist. REFERENCES 1. 2. Hardwick, P. J., Analyst, 1950, 75, 9. Furness, W., Ibid., 1950, 75, 2. CHEMIST INSPECTORATE MINISTRY OF SUPPLY DEPARTMENT OF GOVERNMENT CHEMIST CLEMENT’S INN PASSAGE, LONDON, W.C.2 September, 1949