Colorimetric Determination of Phosphorus in Steels UNO T. HILL Inland Steel Company, East Chicago, Znd.
The colorimetric method of Kitson and Mellon for determining phosphorus in steels is increased in accuracy and extended in scope by the use of a blank which eliminates variables caused by temperature changes and differences in composition of the sample. A more rapid procedure using sodium molybdate and vanadate based on Bogatzlti's method i s applicable when ianadium is not present.
T
H E yellow color formed by ammonium vanadate and molybdate with the phosphate ion was first used by Misson (6) as a basis for estimating phosphorus in steels. Misson's method has been modified by Schroder (8, 9 ) , Getzov (S), Murray and Ashley (7), Bogatzki (f), and Kitson and Mellon (6),and adapted to iron ores by Willard and Center (9, 10) and Hawes (4). As deviations were too great when the methods of Bogatzki ( I ) , Murray and Ashley (r),and Kitson and Mellon (6) were used with photoelectric equipment for determination of phosphorus in steel, this study was undertaken.
Xational Bureau of Standards steel S o . 51 gave a transmlttancy reading of 80% compared with 19d, when dissolved in 30 ml. of 1to 2 nitric acid and diluted to 100 ml. The proposal to use separate curves for each type of steel involves a prior knowledge of the composition of the sample which is not always available to the analyst. It was found possible to compensate for variations in steel composition as well as temperature changes by substituting a blank of the sample for the usual water standard, and cooling both blank and sample t o the same room temperature. By the use of the blank, this colorimetric method is increased in accuracy and is extended in scope to most alloys; a single curve suffices for all types of steels. The Bogatzki (1) method employs a blank consisting of the sample plus ammonium molybdate, and decolorizes both blank and sample with 50 ml. of 3.0% sodium fluoride, although Kitson and Mellon ( 5 ) report that fluoride interferes with the color development. No interference with the color development xas noted in a check of the Bogatzki method; results were erratic, however, when a wide variety of steels was analyzed. Since the 1.5-em. tubes of the Coleman instrument may be manipulated with greater facility than the 5-cm. cell of the Cenco Photelometer or the 4-cm. cell of the Coleman instrument, efforts were directed t o developing a method for this smaller cell thickness. By utilizing sodium vanadate and sodium molybdate in place of the usual ammonium vanadate and ammonium molybdate, and in the absence of all ammonium ions, the type of blank suggested by Bogatzki (1) could be employed without the aid of fluoride, in a concentration of phosphate ions where the yelloxphosphomolybdate precipitated when ammonium ions were present. The ammonium ion prevented development of full background color in some types of steels a t the high concentration of ferric salts, or made development 0.f color exceedingly slow. When potassium nitrite was substituted for hydrogen peroxide or sulfurous acid, the manganese dioxide was reduced more rapidly. Ammonium persulfate oxidation cannot be used, since tbe ammonium ion is introduced. Economies in time and reagents were effected by the use of this smaller cell v5thout.loss in accuracy.
EXPERIMENTAL
Disagreement exists in the literature as to the temperature stability of the molybdovanadophosphoric acid complex. Murray and Ashley (7) found the color stable over a wide range of temperatures, while Center and Willard (8) found it sensitive to temperature and proposed measuring it a t 27" C. Halves (4) reported variations with temperature, while Kitson and Mellon ( 6 )did not report any difficulty and proposed t o measure the color at room temperature. Two 1-gram samples of steel were dissolved in 30 ml. of 1 to 2 nitric acid. The color was developed in one according to the method of Kitson and PIlellon ( 5 ) . The treatment of the other sample was stopped after addition of the persulfate and a 5-minute boiling period. The yellow hue was also developed in a solution containing only potassium hydrogen phosphate in nitric acid. Transmittancies were measured with the Cenco Photelometer, using a 5-em. re11 and a 465 mp filter a t temperatures shown in Table I.
Table I.
Effect of Temperature
Transmittance 465 mg, Filter 6-Cm Cell, TI00 = Water HSOs Phosphorus 21 25 29 32 M g . / f o O ml. C. C. C. C. 30 0 . 1 , color developed 70 68 66 64 30 Color not developed SO i8 76 74 1to 2
Sample
1 gram, Inland
Standard No. 5 Potassium hydrogen phosphate
20
0 . 2 , color developed
44
44
44
44
EQUIPMENT
The findings of Ashley and Murray are confirmed when the color is measured in a solution free of iron salts. In a solution of ferric salts the transmittancy varies with temperature, not because of the varying color of the complex but because of the varying color of the ferric salts. The transmittancy changes in the solution free of iron salts are nil, while the two iron samples show identical changes, irrespective of the addition of ammonium vanadate and ammonium molybdate. These identical changes in transmittancies with temperature are presumed to be caused by changing hydrolysis of the iron salts. Transmittancy also varies with the composition of the steel, even when measured at the same temperature. Thus 1 gram of
The Cenco-Sheard-Sanford filter Photelometer and Coleman PIIodel 11 spectrophotometer were used. SOLUTIOAS REQUIRED
Xitric Acid. Mix 1 liter of concentrated nitric acid, specific gravity 1.42, with 2 liters of water. Ammonium Panadate, 0.25%. Dissolve 2.5 grams of ammonium vanadate in 500 ml. of water with heat. Cool, add 30 ml. of concentrated nitric acid, and dilute to 1 liter. Ammonium hlolybdate, 5%. Dissolve 50 grams of ammonium molybdate in 1 liter of water Sodium Vanadate, 2.5ff'. Dissolve 2.5 grams of sodium vana-
318
319
V O L U M E 1.9, N O . 5, M A Y 1947 Table 11.
Data I'sed for Calibration Curve, Persulfate Method
Steel S o . 12e 12e l2e 12e 12e 12e
Phosphorus
T 465
$0 0 014 0.034 0.054 0.074 0.094 0.104
63 48 37 28 22 19
++ 00 .. 00 24%% PP ++ 00 .. 0068%% PP + O,lO% P
Log T 465 Observed Calculateda
% 1.799 1,681 1,568 1.447 1.342 1,278
1,792 1.680 1,565 1.451 1,336 1,279
Calculated from equation = ~
-
1.874 log T 5.72
Table 111. Analysis of Standard Steels, Persulfate ;\lethod KO, 51% 19d 100
Log
Type
Electric .%.@.€I. ?In rail Sg Bessemer 32c Cr-Xi 111 X-Mo 129.1 Bessemer 106 Cr->lo Inland Hi-steel, Cu 1 . 2 0 , Ki 0 . 6 0
T
1.806 1,677 1,744 1.342 1.819 1.748 1,342 1.760
1,439
Phosphorus Phosphorus Present Found Deviation
%
%
70
0,010
+0.002
0,022 0.094 0.020
0.012 0 034 0.023 0.093 0.010 0,022 0.093 0.020
0.078
0.076
-0,002
0,033 0.023 0.093 0.010
+0.001 0.000 0.000 0.000 0.000 -0.001 0.000
control laboratory using the volumetric method confirmed the accuracy of the method. Samples insoluble in nitric acid may be put in solution with hydrochloric acid, the acid evaporated off, and the sample taken up in 20 ml. of 1 to 2 nitric acid and treated as usual. The limiting amount of chromium up to 14% was found to be the transmittancy of the blank, This was determined with a 1.5-cm. tube on a Coleman Model 11 spectrophotometer. A s the persulfate forms a precipitate m-ith certain alloy steels, permanganate must be employed. Potassium Permanganate. h calibration curve was constructed from Bureau of Standards Steel 13d by adding 0.2-mg. increments of phosphorus as potassium hydrogen phosphate up to 1.0 mg. The data and the equation for the curve, obtained bp the method of least squares, are shown in Table IV. Table V shows the analysis of Bureau of Standards steel samples. This method is applicable to plain carbon steels and alloys where vanadium is absent ( 1 ) . The lorn transmittancy of the blank limit- it othenviqe.
Table IV. Data Used for Calibration Curie, Permanganate Rlethod Sample
Phosphorus
13d B.O.H. 13d 0.02% P 13d 0.04% P 13d 0.06T0 P 13d 0.08% P 13d 0.10% P
0.016 0.036 0.056 0.076 0.096 0.116
T 450
Log T 450 Observed Calculateda
%
date in 50 ml. of nater n-ith heat. When in solution cool, add 20 ml. of concentrated nitric acid, and dilute t o 100 ml. Sodium Molybdate, looc. Dissolve 10 grams of sodium molybdate in 1 liter of water. Potassium Permanganate, 2.5%. Dissolve 2.5 grams of potassium permanganate in 100 ml. of water. Potassium Nitrite. Dissolve 10 grams of potassium nitrite in 100 nil. of water A5.4LTTICAL PROCEDURES
Ammonium Persulfate Method. Weigh two 1-gram samples of the steel into separate 200-ml. Erlenmeyer flasks, one of which is to be the blank. h d d 30 ml. of 1 to 2 nitric acid to each and, after a h o n ceases, place the flask on a hot plat,e. Bring to boil and drive off nitrous oxide fumes. Add 10 ml. of 15% ammonium persulfate solution, made fresh daily, and boil for 5 minutes to destroy excess persulfate. Remove from t,he hot plate and cool blank t'o room temperature. Add to the flask containing the sample 10 ml. of 0.25qc vanadate and mix thoroughly. Add 40 ml. of 57, ammonium molybdate solution, shake t o dissolve precipit.ate, and cool to room temperature. Make up the sample and blank to 100 ml. and measure transmit'tancy of sample against blank, preferably in a 4- or 5-em. cell a t 465 mp. Obtain per cent of phosphorus from a previously prepared transniittancy us. phosphorus curve. Potassium Permanganate Method. Weigh a 1-gram sample of the st,eel into a 200-ml. Erlenmeyer flask, add 25 nil. of 1 to 2 nitric acid, and after action ceases place on a hot plate and boil to expel nitrous oxide fumes. .Idd 1 ml. of 2.57, potassium permanganate and boil for 2 minutes. ,4dd 1 ml. of lOyopotassium nitrite to reduce manganese oxide, and boil 30 seconds to l minute t o expel fumes. Total time on hot plate is about 5 minutes. T o the hot solution add 10 ml. of 10% sodium molybdat,e solution, dilute to 50 ml. in a volumetric flask, and pour back into original flask to mix. Pipet 10 iiil. of the still warm solution to two 1.5cm. cuvettes, one of which is the blank. To the sample add 0.3 ml. of 2.5Tcsodium vanadate, mix well, and cool h0t.h tubes in running water t o room t,emperature. Measure transmitt,ancy of the sample against the blank. From a previously prepared calibration curve, obt,ain the per cent phosphorus. If it is desired to work with cold solutions, the full color is developed in 5 minutes after adding reagents. RESULTS
Ammonium Persulfate. A calibration curve x a s constructed by adding 0.2-mg. increments of phosphorus as potassium hydrogen phosphate to 1-gram samples of Bureau of Standards 12e steel. The data and the equation for the curve, obtained by the method of least squares, are shown in Table 11. Table 111 shorn the analysis of standard steel samples. A daily check nith the
+ +++ +
5
71 57 44 35 27 21
1.851 1.756 1.643 1.544 1.431 1,322
1.862 1.754 1.646 1.538 1,430 1.322
Calculated f r o m equation
% P =
Table V. NO.
8g 16c 19d 65, 100 1069 129X
1.950 - log T 5.41
Analysis of Standard Steels, Permanganate >lethod Type
Bessemer B.O.H. A.O.H. Basic electric Rail Cr->lo Bessemer
Log T 450
1.455 1.788 1.762 1,826 1.836 1.851 1.447
Phosphorus Phosphorus Present Found
Deviation
70
70
n /o
0.093 0 032 0,033 0.023 0.023 0 020 0 094
0.091 0.030 0,035 0.023 0.021 0,018 0.093
-0,002 -0.002 +0.002 0.000
-0,002 -0.002 -0.001
SUAMMARY
By substituting a blank of the sample for the,usual water standard in the colorimetric method of Kitson and Mellon for phosphorus, the deviations caused by temperature changes and by constituents of the steel can be eliminated and the method extended in scope and accuracy. A new method which employs sodium vanadate and molybdate in place of the ammonium salts is economical, rapid, and accurate for carbon steels and low alloys containing no vanadium. LITERATURE CITED
(1) B o g a t e k i , G., Arch. Eissenhuttmw., 12,195 (1938). (2) C e n t e r , E. J . , a n d W i l l a r d , H. H., IND. ENG.CHEM., ANAL.ED., 14, 287 (1942). (3) Getzov, B. B.. Zavodskaya Lab., 4,349 (1935). (4) Han-es, C. C., -km. I n s t . Mining M e t . Engrs., Tech. Pub. 1794 (1945). (5) K i t s o n , R.E . , a n d Mellon, 51.G., ISD.EXG.CHEY.,r l x . ~ED., ~. 16, 378 (1944). (6) ;Misson. G.. Chem.-Zta.. 32.633 (1908).
(7) M u r r a y , IT. M.,Jr.; a n d Ashley, S: E. Q., IND.ENG. CHEM. A N A L . ED.,10, 1 (1938). (8) Schroder, R., Stahl und Eisen, 29,2,1158 (1909). (9) I b i d . , 38,316 (1918). (10) W i l l a r d , H. H., a n d C e n t e r , E. J., IXD.EKG.CHEDI.,ANAL.ED., 13, 81 (1941).