1626
ANALYTICAL CHEMISTRY
LA series of heteropoly tungstocobaltates ( 3 ) )containing both (2) Baker, L. C. W., Loev, B., and McCutcheon, T. P.. [bid.. 72, bivalent and tervalent cobalt, was analyzed by these methods. 2374 (1950). I n some cmes the proportion of tungstate was several times that (3) Baker, L. C. W.,and McCutcheon, T. P., Abstracts of Papers cited above, but the end point was always sharp. The substances Presented at . Fourth Meeting-in-Miniature. Philadelphia are not good primary standards, but the results showed excellerit Section, AMERICAX CHEMICAL SOCIETY, p. 81, 1951. consistency. Examples: percentages calculated for CS6H(4) Baker, L. C. W., and AfcCutcheon, T. P.. ANAL.CHEM.,22, 944 [ C O + ~ C O + ~ W.13H20: ~ ~ O ~ Zcesium, ] 19.8; tungsten, 54.8; water, (1950). 6.04; cobalt (tervalent), 1.47. Found: cesium, 19.9; tungsten, ( 5 ) Biltz, H., and Biltr, W., “Laboratory Methods of Inorganic 54.9; water, 6.00; cobalt (tervalent), 1.47. (Total cobalt is Chemistry;” 2nd ed. (adapted from the German by W Hall always a little high in compounds of this anion because the method and A. Blanchard), pp. 173-4, Wiley, New York, 1928. of preparation leaves them impure in that respect.) Percentages (6) Friedheim and Keller, Ber., 394, 4301 (1907). s +3W12042 1.18H20 : potassium, 4.57 : calculated for K ~ H[ Co (7) Horan, H. A., and Eppig. H. J., J . -4m.Chem. SOC.,71, 581 tungsten, 64.5; water, 10.8; cobalt (tervalent), 1.72. Found: (1949). potassium, 4.60; tungsten, 64.5; water, 10.5; cobalt (tervalent), (8) Kolthoff, I. M.. and Sandell, E. R., ”Textbook of Quant,itative 1.79. The method is applicable in the presence of molybdates Inorganic Analysis,” rev. ed., p. 493, >lacmillan, Xew York, also. Percentages calculated for (SH~)3H5[Co(OH)(MoOa)6] 1946. 3 H 2 0 (2,6): cobalt. 5.97; nitrogen, 4.28; molybdenum, 18.6. (9) Sarver, L. A., IXD.EXG.C m x . , .&SAL. ED., 5, 276 (1933). Found: cobalt, 5.96; nitrogen, 4.42; molybdenum, 48.3. Percentages calculated for (CH3”3)6[(Co06~lo60,5)~],18&0 (1,2, (10) Sarver, L. A., and Kolthoff. I. AI., J . Am. Chem. Soc., 53, 2902 15): cobalt, 4.81; nitrogen, 3.42: molybdenum, 47.0. Found: (1931). cobalt, 4.80; nitrogen, 3.53; molybdenum, 46.7. A sample of (11) Treadwell, F. P., and Hall, W. T., “ilnalytical Chemistry,” vol. the ammonium salt of t,he last-mentioned anion (1,2,6,15) gave 11, 8th ed., p. 145, Wiley, New York, 1935. 4.74% cobalt by the 1-nitroso-2-naphthol method and 4.74% (12) Willard, H. H., and Diehl, H., “Xdvanced Quantitative h a l y cobalt by the method here described. sis,” pp. 233, 236, 243-4, 249-50, Van Nostrand, New York, 1943. ACKNOWLEDGMENT (13) Willard, H. H.. and Young, P., Ind. Eng. Chem., 2 0 , 764 (1928). Helpful advice was received from Donald Cooke, Arno Heyn, (14) Willard, H. H., and Young, P., TND. ENG.CHEM.,ANAL.ED.,4, and Wallace hlcNabb. 187 (1932); 5, 154 (1933). (15) Wolfe, C. W., Block, AI., and Baker, L. C. W,, J . A n i . C‘hem. REFERENCES Soc., 77, 2200 (1955). ( 1 ) Baker, L. C. W., Foster, G., Tan, K., Scholnick. F., and RlcR E C E I V Efor D review January 28, 1955. Arcepted M a y 3, 1955 Cutcheori, T. P., .I. A n ! . (‘hen!. Soc., 77, 2136 (1955).
Spectrophotometric Determination of Phosphorus as Molybdovanadophosphoric Acid KENNETH P. QUINLAN
and
MICHAEL A. DESESA
Raw M a t e r i a l s D e v e l o p m e n t Laboratory, N a t i o n a l L e a d Co., Inc., Winchester, Mass.
The molybdovanadophosphoric acid method for the spectrophotometric determination of phosphorus has been extensively reviewed. The optimum concentrations of acid, vanadium(V), and molybdenum(V1) were determined by factorial experiment. The optimum color development occurs in solutions which are 0.4M in acid, 0.02 to 0.06M molybdenum(VI), and 1.0 to 4.0mM vanadiumw). The optimum range is 3 to 20 p.p.m. of phosphorus pentoxide for 1-em. cells. Dichromate is the only serious interference, but can be eliminated by volatilization of the chromium as chromy1 chloride.
A
N EFFORT 4s currently being made to review critically the
analytical methods employed in this laboratory, R ith the intention of replacing any which may have become outdated, by more rapid or accurate procedures. The tlvo requirements for routine control analyses are speed and a reasonable degree of accburacy. With these needs in mind, the literature was reviewed for methods of analysis for phosphorus. Various spectrophotometric procedures have been proposed and employed for the determination of phosphorus. Because spectrophotometric provedures offer the speed and accurary mentioned as requisites for a routine analysis, these procedures were examined in greater detail. I n common with arsenic, germanium, and silicon, phosphorus forms a yellow heteropoly acid with excess molybdate. These heteropoly acids may be utilized directly in spectrophotometric procedures or they may be reduced to molybdenum blue with an increase in sengtivity. While either of these methods can be used
for phosphorus determination, they are subject to serious interference if arsenic, silicon, or germanium is also present. .Is t,he 9amples treated in this laboratory are ores and ore products, all of these interferences may be encountered. Therefore, the heteropoly methods, as such, \+-erenot considered further. Various workers have attempted to make the molybdophosphorir arid method more selective by extracting the complex into an organic solvent. The literature on this subject has been recently reviewed by Wadeljn and Mellon ( 1 7 ) . However, the reagents which are most selective are difficult t o separate from the aqueous phase, and the other proposea reagents are not selective enough to justify the extra time involved in performing the extraction. Another popular spectrophotometric method for phosphorus is based on the formation of the yellow molybdovanadophosphoric acid as originally proposed by Misson ( 2 3 ) . This method seems to be the most sperific spectrophotometric method proposed for phosphorus, and has been used for t’he determination of phosphorus in steels and iron ores (3, 5 , 7-10! I S , 14, 16, 1 8 ) ; in uranium metal, uranium oxides, and uranium phosphate ( 1 , 1 2 ) ; and in phosphate rock ( 2 , 6 ) , limestone ( 4 , 16), and biological materials ( 1 1 ) . It was decided to investigate the possible ap plication of this method to the routine determination of phosphorus in uraniferous ores and ore products. A review of t h r literature reveals a surprising lack of conformity in the procedures for developing the color of molybdovanadophosphoric acid. Therefore, it was necessary to make a critical study of the optimum conditions for color development before the method could be employed. The results of this study and a modified procedure are reported in this paper.
V O L U M E 2 7 , NO. 10, O C T O B E R 1 9 5 5 Table I. V,
InJf
0.4 0.4 0.4 0.4
1.0 1.0 1.0 1.0 2.0 2.0 2.0 2.0
0 . 2 M HClOb &Io, .M Duplicates Av. 0 . 0 1 0.06.5, 0 , 0 6 2 0.0635 0 . 0 2 0.075,O. 075 0.0750 0.04 0 . 0 8 7 , 0 . 0 8 1 0.0840 0 . 0 6 0 , 0 9 5 , O . 095 0.0960 0 . 0 1 0.084,O. 087 0.0855 0.02 0 . 1 0 4 , 0 . 1 0 0 0.1020 0 . 0 4 0.117,0.117 0.1170 0.00 0.130,O. 131 0,1305 0 . 0 1 0.093,O. 087 0.0900 0 . 0 2 0.146,O. 141 0.1435 0 . 0 4 0.170,O. 165 0 . 167.5 0.06 0.190, 0.181 0 18.55
1627
Absorbance of Blanks
0.4M HClOi Duplicates Av.0 . 0 0 7 , 0 . 0 1 8 0.0125 0 . 0 3 1 , 0 . 0 2 7 0.0290 0.081,O. 074 0.0778 0,084.0.082 0 0830 0.017,O. 021 0.0190 0.042,O. 042 0.0420 0.103,O. 100 0.1015 0 . 1 1 1 , 0 . 1 0 8 0.1095 0.054,O. 054 0.0540 0.074,O. 075 0.0745 0.148,O. 133 0.1405 0 . l.X, 0 . 145 0.1500
0.6.M HClOI Duplicates Av. 0.020.0.012 0.0160 0 . 0 1 4 , 0 . 0 2 0 0.0170 0 . 0 3 8 , 0 . 0 5 2 0.04.50 0 . 0 6 2 , 0 . 0 7 4 0.0680 0 . 0 2 7 , 0 . 0 3 5 0.0310 0.026,O. 037 0.0315 0.054, 0.053 0.0538 0 . 0 7 8 , 0 . 0 7 8 0.0765 0 , 0 3 9 , O .028 0.0335 0.065,O.062 0.0655 0 . 1 1 8 , 0 . 0 9 8 0.1080 0,128,O. 118 0.1230
0.8.M HClOi Duplicates Av. 0.014,O.Oll 0 0125 0 . 0 1 2 , 0 . 0 2 0 0 0160 0 . 0 2 9 , 0 , 0 3 4 0 0315 0.046,O. 040 0 0430 0.027,O. 030 0 0283 0 , 0 3 1 , O .026 0 0285 0 . 0 5 % 0.047 0 0530 0,068,O. 067 0 0675 0 . 0 2 7 , 0 . 0 2 7 0 0270 0 . 0 4 3 , 0 . 0 6 3 0 0530 0 . 0 8 7 , 0 . 0 8 7 0 0870 0,072,O. 101 0 0805
Table 11. Absorbance of Samples 0.2M HCIO, V, m M h i o , M Duplicates Av. 0.4 0 01 0.775,O. 798 0,7865 0.4 0.02 0 . 8 2 8 , O .842 0 , 8 3 5 0 0.4 0 . 0 4 0.836, 0 . 8 4 2 0.8388 0.4 0.06 0.845,0 845 0.8450 1.0 0.01 0 . 7 9 5 , 0 . 8 1 0 0.8025 1 .o 0 . 0 2 0 . 8 5 6 , 0 . 8 7 7 0.8665 1 .o 0 . 0 4 0.867,O. 876 0.8715 1.0 0.06 0 . 8 6 8 , 0 , 8 6 6 0.8670 2.0 0 . 0 1 0.810,O. 841 0.8255 a 0 0 . 0 2 0.910,0.526 0.9180 0 04 0.930,O. 920 0 . 9 2 5 0 2.0 2 ,0 0 ofi 0.948.0.955 0,9500
0.4.M HClOi Duplicates Av. 0 . 7 4 4 , 0 . 7 6 1 0 7525 0.788,O. 788 0 7880 0 830,0.830 0 8300 0 . 8 4 5 , 0 , 8 4 3 0 8440 0.775, 0.812 0 7935 0 . 7 9 0 , O .812 0 8010 0 860.0 857 0 8585 0.8R8,O. 880 0 8740 0 794,0.834 0 8140 0 . 8 2 3 , O . 875 0 8300 0.910,O. 896 0 9030 0.927,O. 923 0 9250
0.6.W HCIOb Duplicates Av. 0.602, 0 . 6 5 6 0.6290 0.751,O. 770 0.7605 0.760,O. 807 0.7835 0.794,O. 821 0.8075 0 713,0.758 0.7340 0 . 7 7 3 , O .788 0,7805 0.784,O 818 0.8010 0.820,0.847 0,8335 0.773,O. 785 0.7790 0.840,O. 835 0,8375 0 . 8 7 5 , O .873 0.8740 0 870, 0 . 8 9 1 0.8805
0.8.M HClOa Duplicates A\. 0 . 3 7 3 , 0 . 4 2 2 0 3975 0 . 7 2 4 , O .741 0 732.5 0.770,O. 786 0 7780 0.784,O. 800 0 7920 0 . 5 2 4 , o .575 0 ,5495 0 . 7 8 8 , 0 . 7 9 0 0 7890 0 . 8 2 3 , 0 . 7 9 7 0 8100 0 . 8 3 0 , 0 . 8 1 2 0 8210 0.607,O. 606 0.6065 0.782,O. 798 0 7900 0 . 8 3 0 , 0 . 8 7 5 0 8525 0.814,O. 874 0 8440
rinse down the sides of the beaker or platinum dish, bring to fumes again, and add another gram of sodium chloride. Repeat this procedure until fumes of chromyl chloride are no longer observed. Dilute, and transfer the sample t o a 1OO-ni1. volumetric flask. To the sample in the 100-ml. volumetric flask add exactll- 10 ml. of the vanadium stock solution, mix, and add 20 ml. of the molybdenum stock solution. Dilute to the mark and mix well. Sfter 15 minutes, measure the absorbance a t 400 mp (0.055-mni. slit) against a reagent blank. EXPERIMENTAL
Color-Developing Agents. llnt’il 1948, t8he vanadate and molybdate Tvere added as separate reagents (IO), but t8hen13arton ( 8 ) proposed the use of a mixed reagent containing both the color developing Table 111. Absorbance of Samples - Blanks agents. Since then most other 0.2M HC104 0.4M HCIO4-0.6.M HCIOd 0.8N H C ~ O I --workers h a w recommended V, m.%f &Io, M Duplicates Av. Duplicates Av. Duplicates Av. Duplicates Av. the mixed reagent ( 1 , 6, 1 2 ) . 0.737,O.743 0.7400 0 . 5 8 2 , 0 . 5 7 4 0.5780 0 . 3 5 9 , 0 . 4 2 1 0.3900 0 1 0 01 0 . 7 1 0 , 0.736 0.7230 Gee and Dietz (6) found the, 0.737,O. 746 0.7115 0 . 7 1 2 , 0 . 7 1 5 0.7155 0 4 0 02 0.753, 0.766 0,7595 0 757,0.760 0.7585 0 . 7 4 1 , 0 . 7 5 2 0.7465 0 . ~ 4 9 , 0 . 7 5 6 0 , 7 5 2 5 0 . 7 2 2 , 0 . 7 5 5 0.738: 0 4 0 04 0.748, 0.761 0.7545 reagent stable for only a n.ec.i; 0 . 7 3 4 . 0 . 7 4 7 0.7465 0 . 7 3 8 , 0 . 7 3 9 0.7383 0 , 6 1 , 0 . 7 6 7 0,7640 0 4 0 06 0 , 7 5 0 , 0.756 0.7530 0.496,0.545 0,5205 0 086, 0 . 720 0.7030 1 0 0 01 0.711, 0 . 7 2 3 0.7170 0 . 7 5 8 , 0 . 7 9 1 0.7745 in warn1 we:Lther, but hnder0.757,O. 764 0.76Od 0.747, 0.751 0 7490 0 748.0.770 0.7S90 1 0 0 02 0.752, 0 . 7 7 7 0,7645 Ron and Wright ( 1 ) claim t h a t 0.764,O. 750 0.7670 0.730, 0 766 0 7475 0.737, 0.7.57 0.7670 0 04 0.750, 0.759 0.7545 1 0 0.762, 0.745, 0.7533 0 . 7 4 3 , 0 . 7 6 9 0 7570 1 0 0 06 0.738, 0 . 7 3 5 0.7305 0 7.57, 0 . 7 7 2 0 , 7 f i 4 5 thereagent is stable for a “long 0 . 5 5 0 , 0 . 5 7 9 0.5635 0 . 7 2 4 , O .757 0 7405 0 . 7 4 0 , O .780 0,7600 2 0 0 01 0.717. 0.744 0 7305 0 , 7 3 9 , O .722 0.7305 0 . 7 6 1 , O .800 0 . 7 7 % 0 02 0.764, 0 . 7 8 5 0.7745 0 771.0.773 0 7720 2 0 period.” During developnient 0.743,O. 788 0,705.i 0 . 7 5 7 . 0 . 7 7 5 0 7660 2 0 0 04 0.760, 0 . 7 5 5 0.7575 0 . 7 6 2 , O .763 0 7623 work on the method separate 0 . 7 4 2 . 0 . 7 7 3 0.7575 0 772,0.778 0.7780 0 7 4 2 . 0 . 7 7 8 0 7575 2 0 0 06 0 755, 0.774 0 7045 reagents were e m p l o y e d , Once a final procedure \viis developed, the possibility of APPARATUS AND REAGElrrTS using H, mixed reagent WLS examined. However, the mixture Absorbance measurements were made with a Beckman Llodel of vanadate and molybdate ip soluble only a t high acid concenD U quartz spectrophotometer using matched 1.00-em. Coi e \ trations. Since excess acid \vas found t o result in poor color decells. velopment, it was considered preferable to continue t o use t h e 411 chemicals were of analytical reagent purity. A standard solution of phosphate was prepared by dissolving separate reagents. 1.917 grams of dried potassium dihydrogen phosphate (KH,PO,) Concentration of Acid, Vanadate, and Molybdate. The v x i in deionized water and diluting t o 1 liter. This solution was ous final molar concentrations of acid which have been reromstandardized gravimetrically, in triplicate, and was found to contain 1.01 mg. of phosphorus pentoxide per ml. as soluble mended are 0.20 ( 6 ) , 0.53 ( I O ) , 0.64 ( I ) , and 0.80 ( 2 , 12). If the phosphate. acid concentration is too low, a yellow vanadium molyhdenum A stock solution of 0.20M molybdenum(V1) was prepared 1)) complex (6, 10) or a precipitate ( 1 4 , 18) forms. If the acid condissolving 35.3 grams of ammonium molybdate tetrahydratp centration is too high, the color developnient is slow or incomplete in deionized water and diluting t o 1 liter. A stock solution of 0.02051 vanadium(V) in 0.4X perchloric ( I , IO, 18). Nitric acid has been most widely used, but ICitsoii acid was prepared by dissolving 1.17 grams of ammonium metaand Mellon ( I O ) claim that sulfuric, hytlrorhloric, or perchloric, vanadate in 400 ml. of deionized 11-ater, acidifying with 25 ml. acid is also satisfactory. I n t,he present work perchloric acid \vas of 8JP perchloric acid, and diluting t o 500 ml. A stock solution of 8M perchloric acid was prepared by dilutchosen, as nitric acid niay contain dissolved colored oxitlw of ing 3-15 ml. of 70 to 72% perchloric acid to 500 ml. nitrogen and sulfuric or hydrochloric acid may give rise to colored complexes with cations. T h e optimum final ooncentration of RECOMLMENDED PROCEDURE vanadate has generally ( 2 , 10, 12, 15) been specified as 2.0 to Solid Samples. JT’eigh out a sample which contains 0.3 to 2.2m31, although 1.7mX ( 1 ) and 0.4 t o 1.2m.V (6) have been 2.0 mg. of phosphorus entoxide into a platinum dish. Cove: employed. While some excess of vanadate is required for full the sample with hydroluoric acid, and add 5 or 10 ml. more. Add 5 ml. 70% perchloric acid, and evaporate to dense fumes of color development, too large an excess simply adds t o t h e abperchloric acid. Dilute, and transfer the solution to a 100-ml. sorbance of the blank. There is no unanimity as t o the optimum volumetric flask, filtering off any residue. molybdate concentration. T h e following final molar concentraLiquid Samples. Pipet an aliquot which contains 0.3 to 2.0 mg. of phosphorus pentoxide into a 250-ml. beaker. Neutralize tions have been recommended: 0.006 t o 0.018 (6),0.028 ( 1 0 1 , any excess acid or base t o the phenolphthalein end point with 0.038 (151, 0.045 (I), and 0.057 ( 2 , 1 2 ) . Gee and Dietz (fi) atconcentrated sodium hydroxide or perchloric acid, respectively. tempted to evaluate the nature of the complex acid, and conAdd 5 ml. of 70y0perchloric acid, and evaporate to dense fume? of perchloric acid. Dilute, and transfer the sample to a 100-ml. cluded t h a t the ratio of vanadium t o phosphorus is 1 to 1 and volumetric flask. that the ratio of molybdenum t o phosphorus is a t least 14 to 1. Samples Suspected to Contain Chromium. To the fuming T h e complex was observed to be relatively strong, so that no perchloric acid solution of the sample add about 1 gram of sodium chloride. If red fumes of chromyl chloride are observed, great excess of reagent should be required t o develop full color.
ANALYTICAL CHEMISTRY
1628
Xitson and Mellon (IO) commented that the reagents used by previous workers were usually too concentrated, and merely served t o contribute t o background absorbance. Some means was needed t o determine t h e optimum concentration or range of concentrations out of the various recommendations which have been mentioned in the literature. A factorial experiment was designed to study the effect of the following variables on the color development: acid concentration a t levels of 0.2, 0.4, 0.6, and 0 . 8 X ; vanadium concentration a t levels of 0.4, 1.0, and 2.0mM; and molybdenum concentration a t levels of 0.01,0.02, 0.04, and 0.06M.
Table IV.
Two-way Tables fur Absorbance of Blanks Acid Molarity
310, .M
Av.
0 2 0 . 0797 0.1068 0.1228 0.1370 0 1116
HC104, .lf
0.4
1.0
2.0
0.2 0.4 0.6 0.8
0.0794 0.0505 0.0365 0.0258 0 0481
0.1088
0.1466 0.1048 0.0825 0,0634 0 0993
0.01
Av. of 6 analyses a t 3 vandium levels
0.02
0.04 0.06
0 4
0 6
0 8
0 . 0285 O,O4S5
0 . 0 2 ~ 8 n. 0227 0:0380
0.1065 0.1142 0.0744
0.0688 0.0892 0.0557
0.0325 0.0572 0,0657 0.0445
0.04M molybdenum, and 0.4 to lOmM vanadium was prepared.
The absorbance of these solutions was determined against a n appropriate blank a t 400 mp (0.055-nun. slit). A region of constant absorbance was observed over the range of 1.0 to 4.OmM vanadium. As a result of these studies it was decided to w e final concentrations of 0.444' perchloric acid, 0.04X molyhdenuni, and 2 OmJl vanadium in the recommended procedure. Order of Addition of Reagents. There seems to be no doubt that if reagents are added separately, the order of addition should be arid, vanadate, and molybdate as recommended by Kitson and Mellon ( I O ) . Any variation in this order will cause the formation of other complexes or precipitates which are only sloivly converted to the molybdovanadophosphoric acid. Color Stability. Using the recommended procedure, the maximum color intensity was developed a t once, decreased about 1.3% in 15 minutes, and then remained constant for at least 2 hours. Table V.
Two-way Tables for Absorbance of Samples
Vanadium Concn., m M
Av. of 8 analyses a t 4 molybdenum levels
Av. V, m.11 Av. of 8 analyses a t 4 acid levels
0.4 1.0 2.0
4v.
0.0680
0.0481 0,0444 0 0673
Acid ,Molarity Av. of 6 analyses a t 3 vanadium levels
JIolybdenum Concn., .M 0 01 0 02 0 04 0.06 0.0261 0.0410 0.0511 0.0394
0.0342 0.0510 0.0841 0 0564
0.0595 0.0813 0.1258 0.0889
0,0723 0,0960 0.1363 0.1015
T o run a determination of each combination of these variables required 48 samples and 48 blanks (4 levels of acid X 3 levels of vanadium X 4 levels of molybdenum). Duplicates were run a t each condition thus raising the number of solutions required to 192. T h e samples all contained 2 mg. of phosphorus pentoxide per 100 ml. of final volume The absorbance of both the samples and the blanks mas measured a t 400 mp (0.055-mm. slit) using deionized water as the reference standard so that the effect of change in concentration of reagents could be observed on both samples and blanks. The results of the experimental work are presented in Tables I, 11, and 111. The variables are summed over one a t a time-Le., the results for all conditions of one variable are summed a t each combination of the remaining tivo-and the data are grouped in the form of two-way tables in Tables IV, V, and VI. Without any further statistical analysis, the following conclusions can be reached from the results: ABSORBIXCEOF BLIXK. Increasing acidity causes the absorbance of the blank to diminish, while increasing molybdenum and vanadium concentration cause the absorbance t o increase. ABSORBANCE OF SAMPLE. Again, increasing acid concentration causes a decrease in absorbance, xhile increasing vanadium and molybdenum concentration causes an increase in absorbance. . ~ B S O R B A N C EOF SAMPLE ?\IISUS BLANK. When the absorbance is measured versus a reagent blank, maximum absorbance is obtained in 0.4.11 acid, and increasing the vanadium concentration a t least up to 2,OmM causes an increase in absorbance. However, over the range of 0.02 t o 0.06M molybdenum there is no difference in the average absorbance of the samples even though the acid and vanadium concentrations were varied over the entire range chosen for the factorial experiment. Obviously the optimum molvbdenum concentration liea anywhere in the range of 0.02 t o 0.06X. The optimum acid concentration is 0.4M, but the optimum vanadium concentration TI as not explicitly revealed in the experiment. I n order to determine the optimum concentration of vanadium a series of solutions containing 20 p.p.m. of of phosphorus pentoxide, a final concentration of 0.4M acid,
310, ,If
0.2
0.01 0.02 0.04 0.06
0.8048 0.8732 0 8783 0.8873
0 4 0.7867 0.8130 0.8638 0.8810
0 6 0.7140 0.7928 0,8195 0.8405
0.8 0.5178 0.7705 0.8135 0 8190
AV.
0.8609
0.8361
0,7917
0.7302
Vanadium Concn.. 1n.V __ HClOI, .M Av. of 8 analyses a t 4 molybdenum levels
0 2 0 4 0 6
0 8 Av
.
V, m M Av. of 8 analyses a t 4 acid levels
Table VI.
0 4 8263 8036 7451 6750 0.7625 0 0 0 0
1.0 8518 8317 7872 7423 0,8032
0 0 0 0
0 0 0 0
2.0 SO4G 8730 8427 7732
0.8483
3Iolybdenum Concn., M 0.01 0.02 0.04 0.06
0.4 1,O 2.0
0.6413 0.7198 0,7563
0.7790 0.8093 0.8488
Av.
0.7058
0.8124
0.8075 0.8353 0.8886 0.8438
0.8221 0.8480 0.8998 0.8566
Two-way Tables for Absorbance of Samples Minus Blanks
Av. of 6 analyses a t 3 vanadium levels
M
0.01 0.02 0.04 0.06
0.2 0.7235 0.7662 0.7555 0.7513
Av.
0.7491
MO,
Acid Molarity 0 4 0.6 0,7581 0.7643 0.7573 0,7678
0.6738 0.7541 0,7506 0.7516
0.7619
0.7325
0.8 0,4916 0.7355 0,7563 0,7498 0.6833
Vanadium Concn., m M HC10d.M Av. of 8 analyses a t 4 molybdenum levels
0.2 0.4 0.6 0.8
0.4 0.7475 0.7537 0.6996 0.6476
1.0 0.7431 0.7637 0.7391 0.6978
2.0 0.7567 0.7682 0.7590 0.7045
Av.
0.7121
0.7359
0.7471
V, m M Av. of 8 analyses a t 4 acid levels
0.4 1.0 2.0
Av.
Molybdenum Concn., .%.I 0.01 0.02 0.04 0.06 0.6078 0.7437 0 7480 0.7490
0.6988
0.6787
0.7582 0.7631
0.7540 0.7828
0.7828 0.7636
0.6618
0.7550
0.7649
0,7551
Temperature Dependence of Color. Several investigators who app:ied the molybdovanadophosphoric acid method t o steel analjsis ( 5 , 8) found the color intensity to be temperature dependent. However, Hill (9) demonstrated that the change in color with change in temperature is actually due to the background absorbance of iron in solution, and the complex itself is temperature stable a t least a t normal room temperature. Optimum Wave Length. According t o Barton ( 2 ) the spectra of the complex acid and reagent blank both show increasing absorbance as the wave length is decreased toward the ultraviolet. The largest spread between the two spectra is a t 400 mp. However, in steel analysis a longer wave length was usually
V O L U M E 27, NO. 10, O C T O B E R 1 9 5 5 chosen t’o diminish intcrfewnce from iron. The authors’ observations on the spectra confirmed those of Barton, so that all absorbance measurements were made a t 400 inh in this study. Calibration Data. Using the recommended procedure, duplicate solutions of known concent,ration in the range 2.5 t o 50 p.p.m. of phosphorus pentoxide were prepared. Beer’s Ianwas obeyed over thP entire range. The average absorbance index (per parte per million) was 0.03i2 and the coefficient of variation was iO.D‘$. The range of most accurate spect,rophotometric measurements was 3 to 20 p.p.m. of phosphorus pentoxide. Study of Interferences. The possible interference of 60 different ioiis \$‘asevaluated tiy Kit,son and >fellon (ZO), and several other authors have examined the interferences. In general, it may be said that’ reducing agents should be avoided, hecause they may reduce the ciomplex t o molybdenum blue. Ions which complex molykidenurn, such as oxalate, tartrate, and citrate, tend to bleach the color.. These interferences are all removed in the recommended sample preparation. Kitson and RIellon (10) claimed that rhloride and fluoride also bleach the color, but Gee and Dietz ( 6 )and RlcCutchen, Robinson, and House ( 1 2 )reported that chloride and fluoride do not interfere. Tests in this laboratory indirate that chloride does not interfere up to conpentrations of 500 p.p.m. but that fluoride does interfere, causing erratic readings at concentrations of 10 p.p.m. or more. Therefore, any fluoride ii~etliri the decoinpovition of the sample should be removed I)y evaporation :i:d fuming with perchloric acid.
1629 available. Therefore, the method was tested by two other approved analytical techniques. The first technique was to analyze leach liquors from typical ores treated in this laboratory, and then add known concentrations of phosphorus pentoxide to the leach liquors t o determine whether an accurate analysis of the amount added could be obtained. This procedure afforded a test of precision and accuracy, as without reproducibility the amount of phosphorus pentoxide added could not be determined accuratrly. The results of theqe analyses are summarized in Table T’III. The added amount of phozphorus pentoxide was determined tvith an accuracy within 1 to 2%, which is acceptable for a spectrophotometric procedure. The second test of the method was a comparison of the proposed spectrophotometric method with the “alkalimetric volumetric” method for phosphorus. Although a comparison with a gravimetric procedure might have been more significant, the volumetric method was being used in the laboratory a t the time. A series of leach liquors \vas analyzed by both procedures and the results are compared in Table I X . The sver:tge deviation between the two methods \vas 2.8%.
Table VIII. ~
Saniple 1 2
3 4
Table \ 11. I Z r ~ r r i o ~ aof I Dichromate Interference PsOa
Taken, 1 s CrzOv
P205 Found, big.
Test of Precision and .4ccuracy
Concn., ~ a hfg. Duplicates 0.187, 0 . 1 9 2 14 5 , 1 4 . 5 15.1, 1 5 . 4 25 8. 2 5 . 6 z
per XI. A T z d , Av. Mg. 0190 1.00 14.5 1.00 I 5 2 25 7
100 1 00
_ _ _ _ ~ ~ ~ ~ ~ _ _ _ _ ~ ~
PlOk
Recovered, Duplicates 101,1.01 1.02,1.00 1.01,101
hfg.-
%
Av.
Error f1.0 +1.0
1 04.0.99
1.01 1.01 1.01 1.02
+%.O
~. ~
‘Table IX. Comparison of F‘olumetric and Spectrophotonietric Methods
Treatments Necessary
Grams ~-_ _ P?Os per Liter ~~
Tliv interference of dichromate is well known ( I O ) aiid results from the fact that, the color of this ion resembles very closely that of the molybdovanadophosphoric acid. This interference was found to be directly proportional t o dichromate concentration with cnch part per million of dichromate equivalent to 0.14 p.p.m. of phosphorus pentoxide. Since the interference of dirhromate is ioiq a study was made to devise a means of separating dichromate and phosphate. The simplest procedure seemed to lie :I volatilization of the undesired chromium as chronij-1 chloride from hot perchloric acid. I n order t o determine rvhether this proposed method is applicd~le,mixtures of 1.00 mg. of phosphorus pent oxide and various amounts of dichromate were prepared and treated according t o the recommended procedure. Holvever, ivhen the samples came to dense perchloric acid fumes, about 1 gram of sodium chloride crystals was added t o the fuming acid. Red fumes of chroinyl chloride were observed. The sides of the beakws containing the mixtures were rinsed down, the solutions were evaporated to fumes of perchloric acid again, and the treatment v i t h sodium chloride was repeated. This procedure TT’SS contiriued until no dichromate color v a s visihle in the solutions :tnd no cahroniyl chloride color was visible in the fumes. Then the color-developing agents were added according to the recommended procedure, and the solutions were spectrophotometrically nnal:-zed for phosphorus pentoxide. The results of this experiment, shown i n Tahle VII, indicate satisfactory removal of dichromate. Accuracy and Precision. Some indication of the precision of the method was given in t,he low coefficient of variation for the average absorbance index obtained in the calibration data. I n order t,o test the accuracy of any method it is usually necessary to aiialyze samples of known composition. S o standard samples appi,o\.ini:itirig the mnterials analyzed in this laboratory. were
+1.0
Sample S E R 10F S E R 103 J B B 651 J B B 015 J B L 801 JBI. 807 J B B 704 J B B 636 XBS 120b 0
b C
Volumetric
2.07 1.78 0.002 0 479 0.425
Spectrupiiototnetric 2 no 1.71 0 . d‘J:i 0 458 0 416 0 704 10.6 0 . 7 31 34 7 %
0.806 10.9 0,722 3 5 . 3 % P2OSC PlOS Toliitiietric analysis taken as standard. National Bureau of Standards phosphate rock. Cprtified value.
%
Deviation0 -3.4 -3.9 -1.4 -4.4 -2.1 -0.C) -3.(i f4.0 -1,
LITERATURE CITED (1)
(2) (3) (4)
(5) (6) (7)
(8)
(9) (10)
(11) (12) (13) (14)
Anderson, B., and Wright, W. H., Jr., U. S. Atomic Energy Commission, Y-900 (August 1952). Barton, C . J., ANAL.CHEM.,20, 10% (1948). Bogataki, G., Arch. Eissenhuttewu., 12, 195 (1938). Rrabson, J. A., Karchmer, J. €I., and Kata, AI, S., IND.EKQ. CHEAT., .IK.&L. ED., 16, 553 (1944). Center, E. J., Willard, H.11.. Ibid., 14, 287 (1942). Gee, h..and Diete, V. R..A I s ~CHEX., < ~ . . 25, 1320 (1953). Getaov, B. B., Zacod8kaga Lab.. 4 , 349 (1935). Hinwes, C. C.. Am. Inst. LIining AIet. Engrs., Tech. Pub. 1794 (1945). Hill, U. T.. .h.&:,. C H E l r . . 19, 318 (1947). Kitson. R. E.. and llellon. 11.G., IXD. EKG.CHEM.,AXAL.ED., 16,379 (1944). Koenig, 11. A , . and Johnson, C. R., Ibid., 14, 155 (1942). 1lcCutrheii. 11. L.. Robinson, J. W., and House, H. P., U. 8. Atoinie Energy Commission, Y-E131-138 (August 1952). Misson, G.. Chem.-Ztg., 32, 633 (1908). Murray. K.31.,and Ashley, S.E. Q . , ISD. ESG. CHEM.,ANAL.
ED.,10,l(1938). (15) Racicot, E. L., As.\i.. Cxirv.. 23, IS73 (1951). (16) Schroder, R.. SIahZ u . Eise,,. 38, 316 (1918). (17) Wadelin. C.. and IIellon. 11. G . , .\s.