944 Analyst August 1983 Vol. 108 pp. 944-951 Spectrophotometric Determination of Phosphorus and Arsenic in Steel by Solvent Extraction of their Heteropolyacids with Ethyl Violet Shoji Motomizu Toshiaki Wakimoto and Kyoji Toei Department of Chemistry Faculty of Science Okayama University Tsushima-naka Okayama-shi Japan Under the same conditions orthophosphate and orthoarsenate react with molybdate to form molybdophosphate and arsenomolybdate which are extracted into a cyclohexane - 4-methylpentan-2-one mixture (1 + 3) with ethyl violet. The absorption spectrum of each ion pair extracted into the organic phase is almost the same in the visible region and the molar absorp-tivity of each ion pair in the organic phase is 2.8 x lo5 1 mol-l cm-l a t 602 nm. In determining phosphate arsenate can be masked with thiosulphate and hydroxylamine.The arsenate concentration was obtained by subtracting the phosphate concentration from the total concentration of phosphate and arsenate. Steel samples (less than 0.25 g in mass) were dissolved in aqua regia and the solution was diluted to 11 with distilled water. In the deter-mination of phosphorus (about 0.003y0) and arsenic (about 0.007~0) in steel, the relative standard deviations were 2.5 and 2.1 yo respectively. Fifteen standard steel samples were analysed and the results obtained for phosphorus and arsenic were in good agreement with their certified values. The results for the recovery test were also good. The limit of detection for both phosphorus and arsenic is about O . O O l ~ o in steel.Keywords Phosphorus and arsenic determination ; solvent extraction; hetero-polyacid; steel analysis ; ethyl violet Almost all of the methods for the spectrophotometric determination of phosphorus depend on the formation of the heteropolyacid which is formed in an acidic medium by the reaction between orthophosphate and molybdate. A reduced species of molybdophosphate (molyb-denum blue) and of molybdophosphovanadate have also been used. The molar absorptivities of such heteropolyacids are less than 3 x lO41mol-1 cm-1 when these are determined spectrophotometrically in aqueous solution or in an organic phase by solvent extraction. Molybdophosphate has been extracted into an organic phase with cationic dyes such as safranine T,112 crystal violet or iodine green,3 methylene blue4 and rhodamine B.5 Recently, we studied the solvent extraction of molybdophosphate with ethyl violet.6 The method has the following advantages (1) high sensitivity (the molar absorptivity in cyclchexane -4-methylpentan-2-one mixture at 602 nm is 2.8 x lo5 1 mol-1 cm-1); (2) a simple and less time-consuming procedure (the procedure involves a single extraction and does not require washing the organic phase after extraction) ; (3) good reproducibility; (4) relatively low absorbance of the reagent blank (about 0.1) ; and (5) few interferences from co-existir g ions except for arsenate.In the previous work,6 several parts per 1 0 9 (p.p.b.) of phosphorus in waters were determined. In the course of further study for the extraction of the heteropoly-acid we found that arsenate could be extracted into an organic phase with ethyl violet under the same conditions as phosphate and the absorption maximum and molar absorptivity at 602 nm of the ion pair of molybdophosphate - ethyl violet were the same as that of arseno-molybdate - ethyl violet.It was also found that arsenite did not interfere in the deter-mination of phosphate and arsenate could be easily reduced to arsenite with sodium thio-sulphate. In the simultaneous determination of phosphate and arsenate by the molybdic acid - spectrophotometric method Johnson' and Johnson and Pilsons masked arsenate with sodium thiosulphate and sodium disulphite which reduce arsenate to arsenite. In this work sodium thiosulphate and hydroxylammonium sulphate were used and phosphorus and arsenic in steel were determined simultaneously by solvent extraction -spectrophotometry MOTOMIZU WAKIMOTO AND T6EI Experimental 945 Apparatus The absorptiometric measurements were made on a Hitachi 139 spectrophotometer with glass cells of 10-mm path length.An Iwaki Model V-S Type KM shaker was used for the horizontal shaking of the 25-ml stoppered test-tubes. Reagents A stock solution of phosphate was prepared by dissolving disodium hydrogen orthophosphate (anhydrous) which had been previously dried at 50 "C under reduced pressure (about 5 mmHg) until a constant mass had been reached in distilled water to give a The working solution was prepared daily by diluting the stock solution accurately . Dissolve 0.19 g of arsenic(II1) oxide in 20 ml of 5% sodium hydroxide solution.Acidify with 3 M sulphuric acid. Boil and add 1% potassium permanganate solution until the solution is slightly coloured. Neutralise with 5% sodium hydroxide solution and dilute to 100 ml with distilled water. The working solution was prepared daily by the appropriate dilution of the stock solution. MoZybdate solution. Ammonium molybdate [(NH,),(Mo,O,,) .4H,O] (31 g) was dissolved in distilled water and the solution was made up to 1 1. Ethyl violet solution. Commercially available ethyl violet was recrystallised twice from water. The crystals (0.13 g) were dissolved in distilled water and the solution was made up to 200 ml. Redacing (masking) a g e d . Hydroxylammonium sulphate [(NH,OH),SO,] (0.41 g) was dissolved in distilled water and the solution was made up to 50 ml (0.05 M).Sodium thio-sulphate pentahydrate (0.025 g) was dissolved in distilled water and the solution was made up to 50 ml (2 x These solutions were prepared daily. Ethylenediaminetetraacetic acid disodium salt (dihydrate) (3.7 g) was dissolved in distilled water and the solution was made up to 500 ml. A mixture of 100 ml of cyclohexane and 300 ml of 4-methylpentan-2-one was prepared and was used as the extraction solvent. Standard Phosphate solution. M solution. Standard arsenate solution. A stock solution of arsenate was prepared as follows. M ) . EDTA solution. Extraction solvent. Preparation of the samfle solution Add about 3 ml of aqua regia and dissolve the steel by heating on a hot-plate (at about 100 "C).After dissolu-tion add 10 ml of 2.5 M sulphuric acid and heat for about 30 min on a hot-plate at about 150 "C in order to remove the excess of nitric acid. The solution is then accurately diluted to the required volume with distilled water. Weigh the required amount of the steel sample into a 50-ml beaker. Procedure f o r the determination of phosphorus and arsenic Transfer an appropriate aliquot (<lo ml) of the sample solution containing up to 1 pg of phosphorus into a stoppered test-tube and dilute with distilled water to 10 ml. Pipette 0.1 in1 of 0.05 M hydroxylammonium sulphate solution into it and mix then pipette 0.1 ml of 2 x 10-3 M sodium thiosulphate solution into it and mix again. Mix thoroughly and leave for about 10 min. Pipette 1 ml of EDTA solution 1 ml of ethyl violet solution and 5 ml of the extraction solvent into it.After shaking for 25 min measure the absorbance of the organic phase at 602 nm in a 10-mm glass cell against a reagent blank. The phosphate concentration was calculated from the calibration graph for phosphate which had been obtained previously according to the same procedure. Transfer an appropriate aliquot of the sample solution (<lo ml), containing up to 0.03 pmol of total orthophosphate and orthoarsenate into a stoppered test-tube and dilute with distilled water to 10ml. Add the same amounts of sulphuric acid, molybdate EDTA ethyl violet solution and extraction solvent as those for procedure A and carry out according to this procedure. The total concentration of phosphate and arsenate was calculated from the calibration graph for either phosphate or arsenate.The arsenate Procedure ( A ) f o r phosphorus. Pipette 1 ml of 5 M sulphuric acid and 1 ml of molybdate solution into it. PTocedure (B) f o r arsenic 946 MOTOMIZU et aZ. SPECTROPHOTOMETRY OF P AND As IN STEEL Analyst VoZ. 108 concentration was obtained by subtracting the phosphate concentration from the total concentration. Results and Discussion Absorption Spectra and Calibration Graphs Each absorption spectrum of the ion pair formed between molybdophosphate and ethyl violet and between arsenomolybdate and ethyl violet in the organic phase was obtained by procedure B. Both spectra were identical and the maximum absorbance occurred at 602nm. The calibration graphs were linear in the range 0-3 x 1 0 - 6 ~ of phosphate and arsenate and the molar absorptivity calculated from the slope of the graph was 2.8 x lo5 1 mol-l cm-l.Effect of the Reducing Agent on the Reduction of Arsenic(V) to Arsenic(II1) Of these sodium dithionite and sodium thiosulphate can effectively reduce arsenic(V) to arsenic( 111). However when sodium dithionite was used the absorbance of the reagent blank became larger than the absorbance obtained with sodium thiosulphate. In view of this further experiments were carried out using sodium thiosulphate. In the range 0-2pmol of sodium thiosulphate added the absorbances of the reagent blank were almost equal and were about 0.105. In the range below 0.1 pmol of sodium thiosulphate, the greater the concentration of the sodium thiosulphate the smaller the absorbance of the ion pair extracted into the organic phase becomes and the differences in absorbance between the solutions with and without arsenate are almost equal in the range 0.1-2.0 pmol of sodium thiosulphate; this small difference in absorbance may be due to the phosphate in arsenic solution resulting from impurities in such reagents as sulphuric acid sodium hydroxide and potassium permanganate.As in this work iron and steel samples are dissolved in aqua regia iron may exist as iron(II1) in the sample solution. Iron(II1) may consume sodium thiosulphate. As shown in Fig. 1 the addition of large amounts of sodium thiosulphate causes an increase in the absorbance of the reagent blank. As can be seen from Table I, hydroxylammonium sulphate hardly reduces arsenic(V) to arsenic( 111) ; this reagent can effectively reduce iron(II1) to iron(II).g The effect of hydroxylammonium sulphate is shown in Table 11.It can be seen that more than 90% of the arsenate can be masked with only 0.2 pmol of sodium thiosulphate even when 36 pmol of iron(II1) are present. Such a result is interesting but is not explained on the basis of standard oxidation - reduction potentials of thiosulphate arsenate and iron(II1) ions. At present we cannot give a reasonable explanation for “the selective masking of arsenate with thiosulphate.” Also it is found Several reducing agents were examined (Table I). The effect of sodium thiosulphate can be seen in Fig. 1. TABLE I REDUCING AGENTS FOR REDUCTION OF ARSENATE TO ARSENITE Reducing agent None NaBH (0:0013’mol) .. Zn (0.0015 mol) + NaH (0.001 3 mol) . . Zn (0.001 5 mol) . I . . Zn (0.0015 moll + KI’ i0.01 Zn (0.0015 mol) + KI (0.01 pmol) + NaB€€ (0.0013 ,hi) KI (0.01 pmol) + Na,SO (0.0001 mol) . . . . Sb (0.001 mol) + KI (0.01 pmol) . . . . . . (NH,OH),SO (60 pmol) + KI (0.01 pmol) . . L-Ascorbic acid (60 pmol) + KI (0.01 pmol) SnC1 (0.05 pmol) + KI (0.01 pmol) Na,S,O (2 pmol) + KI (0.01 pmol) . . Na,S,O (2 pmol) . . . . HCOONa (2 pmol) . . . . (COONa) (2 pmol) . . ,. Na,S,O (2 pmol) . . . . . . . . * Reference solvent. Reference reagent blank. As(V) 0.012 pmol. Absorbance of reagent blank* . . 0.106 . . 0.119 . . 0.209 . . 0.178 . . 0.092 . .0.162 . . 0.185 . . 0.205 . . 0.165 . . 0.149 . * 0.110 . . 0.168 . . 0.109 . . 0.184 . . 0.101 . . 0.150 Absorbance? 0.672 0.458 0.225 0.299 0.344 0.281 0.125 0.542 0.629 0.519 0.457 0.005 0.013 0.355 0.331 0.03 August 1983 BY EXTRACTION OF HETEROPOLYACIDS WITH ETHYL VIOLET 947 Sodium thiosulphate/ymol 2 4 I 1 I I t 0 0.05 0.1 0.2 0.4 0.6 0.8 1.0 Sodium thiosulphate/pmol Fig. 1. Effect of sodium thiosulphate on reduction of arsenic(V) to arsenic(II1). (1) and (2) reagent blank reference solvent; and (3) and (4) 10 ml of 1.2 x M As(V) reference reagent blank. that the absorbance becomes gradually smaller as the hydroxylammonium sulphate content is increased and is almost constant in the region above 5 pmol of hydroxylammonium sulphate.The absorbances obtained in the region of 5-30 pmol of hydroxylammonium sulphate are in good agreement with those expected for the phosphate content. In deter-mining phosphorus in the presence of iron(II1) ion arsenate was masked with sodium thio-sulphate and hydroxylammonium sulphate. Here the absorbance of the reagent blank was about 0.04 larger than that in the absence of the reducing agents. This increase in the reagent blank may be caused by the phosphate contained in the agent. TABLE I1 EFFECT OF HYDROXYLAMMONIUM SULPHATE ON THE REDUCTION OF IRON(III) TO IRON(II) Sample 36 pmol of iron alum [Fe,(S04),(NH4),S04.24H,0] + 0.008 pmol of phosphate + 0.008 pmol of arsenate. This iron alum contains 0.008 pmol of phosphate per 1 g of salt that is 36 pmol of iron alum contains 0.0027 pmol of phosphate.Sodium thio-sulphate content 0.2 pmol. Amount of (NH,OH),SO,/pmol 0 1 2 3 4 5 10 20 30 Absorbance of reagent blank* Absorbancet 0.124 0.621 0.130 0.620 0.126 0.614 0.139 0.608 0.140 0.602 0.136 0.590 0.138 0.583 0.150 0.580 0.155 0.578 * Reference solvent. t Reference reagent blank. Effect of Co-existing Ions In determining phosphorus and arsenic the following ions produced absorbances contri-buting less than 0.01 unit Mg2+ (0.4 mg) ; Ca2+ (0.6 mg) ; Co2+ (0.9 mg) ; Zn2+ (1 mg) ; Cd2+ (0.7 mg) ; Mn2+ (0.8 mg) ; Cu2+ (1 mg) ; Ni2+ (0.9 mg) ; A13+ (0.4 mg) ; Ba2+ (1.4 mg) ; .Cr3+ (0.1 mg); Fe3+ (0.1 mg); Pb2+ (20 pg); W(V1) (1 pg); Sn2+ and Sn4+ (1 pg); and Si(1V) (14 pg).The effects of these ions were examined by using the metal chlorides which are commercially available. As these chlorides are more or less contaminated with phosphat 948 MOTOMIZU et al. SPECTROPHOTOMETRY OF P AND As IN STEEL Analyst vol. 108 0.6 I RI 0.4 2 0.2 - ~~ 0 2 4 6 8 10 Volume of sample solution/ml Fig. 2. Effect of volume of sample taken. (1) and (2) NBS 55e 0.496 g l-l (1) for phosphate (2) for phosphate and arsenate; (3) JSS 232-3 0.1982 g l-l for phosphate; (4) FeC13.6H,0 0.0536 g per 100 ml for phosphate; and (5) Fe,(SO,),.xH,O 0.0801 g per 100 ml for phosphate. and arsenate the results obtained by using such chlorides do not necessarily indicate the upper limit of a tolerable concentration of the co-existing ion. In order to examine the interference of iron the relationship between the absorbance and the amount of iron was obtained by using several iron(I1) and iron(II1) salts and standard steel samples.The examples are shown in Fig. 2. From graphs 1 and 2 in Fig. 2 it seems that in the region below about 4 mg of iron in procedure A and 3 mg of iron in procedure B the graphs are linear and each intercept of the y-axis is in good agreement with each reagent blank. Other graphs are also linear and each intercept is in good agreement with the reagent blank. From these results it can be seen that amounts below 3 mg do not interfere in the determination of phosphorus and arsenic. TABLE I11 DETERMINATION OF PHOSPHORUS AS PHOSPHATE IN IRON (11) AND IRON (111) SALTS Solution/ Sample Grade* Procedure g per 100 in1 FeSO,.H,O .. a A 0.055 8 FeS04(NH4),S04.6H,0 . . . . a A 0.0782 Fe,(S04),.xH,0 . . a B 0.080 1 Fe,(S04),(NH4),S04.24H,0 . . a A 0.096 9 a B 0.0964 FeC13.6H,0 e A 0.053 6 Solution takenlml 6 8 10 6 8 10 6 8 10 6 8 10 6 8 10 6 8 10 Absorbance? 0.011 0.014 0.026 0.021 0.035 0.042 0.047 0.067 0.082 0.021 0.026 0.035 0.012 0.018 0.023 0.061 0.076 0.103 Phosphorus content yo 2.3 x 10-4 2.1 x 10-4 3.0 x 10-4 3.1 x 10-4 3.7 x 10-4 3.5 x 10-4 6.9 x 10-4 3.1 x 10-4 2.6 x 10-4 2.6 x 10-4 1.4 x 10-4 1.5 x 10-4 1.2 x 10-3 1.2 x 10-3 6.8 x 6.6 x 1.6 x lo-' 1.3 x * a and e denote analytical-reagent grade and extra-pure reagent respectively.t Reference reagent blank August 1983 BY EXTRACTION OF HETEROPOLYACIDS WITH ETHYL VIOLET 949 Determination of Phosphate in Commercially Available Iron Salts The iron salts were dissolved in 0.01 M sulphuric acid and used as the sample solution. Phosphate was determined by procedure A. As the arsenate content in iron(II1) salts is less than one tenth of the phosphorus content the arsenate content cannot be determined accurately by procedure B. The results obtained are shown in Table 111. Iron salts contain 10-4-10-3 yo of phosphorus as phosphate. Preparation of Steel Sample Solution Dissolution 1 Steel samples (0.1-0.2 g) were dissolved by adding 10 ml of sulphuric acid (2.5 M) and 1 or 2 ml of about 10% hydrogen peroxide at room temperature. After dissolution potassium permanganate solution was added until the solution was coloured.An excess of perman-ganate was decomposed by adding hydrogen peroxide and boiling for 30 min. The solution was diluted with distilled water to 11. Phosphate and arsenate were determined by pro-cedures A and B. From Table IV it is found that not all of the phosphorus and arsenic in steel is converted into orthophosphate and ortho-arsenate. The values obtained are shown in Table IV. TABLE IV RESULTS OBTAINED BY DISSOLUTION WITH DILUTE SULPHURIC ACID AND HYDROGEN PEROXIDE (DISSOLUTION METHOD 1) AND NITRIC ACID (DISSOLUTION METHOD 2) Sample solution taken 10 ml. Certified value % (-*- Concentration of Steel sample* P As steel/g 1-1 JSS 230-4 . . 0.0124 0.096 7 JSS 231-3 .. 0.0223 0.099 6 JSS 030-2 . . 0.0179 0.0046 0.2030 0.2054 JSS 050-3 . . . . 0.0186 0.0113 0.1966 0.196 8 0.198 8 JSS 061-2 . . 0.0104 0.0056 0.201 0 Found % Dissolution (-*-, method P As 1 0.0100 1 0.0156 1 0.009 4 0.002 5 2 0.0136 0.0037 1 0.0095 0.0054 2 0.0186 0.0089 1 0.0068 0.002 7 2 0.0074 0.002 9 * Japanese Standards of iron and steel. Dissolution 2 Steel samples (0.1-0.2 g) were dissolved by adding 5 ml of dilute nitric acid (about 4 M). After dissolution 10 ml of 2.5 M sulphuric acid and potassium permanganate solution were added until the solution was coloured. The solution was boiled for 5 min 1 ml of 15% hydrogen peroxide was added and the solution was boiled for about 30 min and diluted with distilled water to 11. Phosphate and arsenate were determined by procedures A and B.The contents of phosphorus and arsenic obtained were about 40% lower than the certified values (see Table IV). Dissolution 3 Steel samples (masses less than 0.25 g) were dissolved in 3 ml of aqua regia on a hot-plate (at about 50 "C). After dissolution 10 ml of 2.5 M sulphuric acid were added and the solution was boiled on a hot-plate for about 20 min to remove the excess of nitric acid. The solution was then diluted with distilled water to 1 1. In this way the preparation of sample solution can be made within 2 h. Determination of Phosphorus and Arsenic in Steel Samples Sample solutions were prepared using dissolution 3 and phosphorus and arsenic were determined by procedures A and B. The results obtained are shown in Tables V VI and VII 950 MOTOMIZU et al.SPECTROPHOTOMETRY OF P AND As IN STEEL Analyst VoZ. 108 TABLE V REPEATABILITY OF THE EXPERIMENTS Sample NBS 55e; certified values phosphorus 0.003% and arsenic 0.007%. Absorbance Sample solution/ ,-A-, g per 500 ml P As 0.097 1 0.101 0.084 0.0974 0.106 0.086 0.099 8 0.102 0.090 0.100 2 0.104 0.092 0.097 1 0.106 0.086 0.096 5 0.103 0.088 0.098 0 0.101 0.087 0.099 0 0.106 0.086 0.100 5 0.107 0.086 Mean : Relative standard deviation : Found, v P 0.003 1 0.003 2 0.003 0 0.003 0 0.003 2 0.003 1 0.003 1 0.003 2 0.003 2 0.003 1 2.5 % I As 0.006 3 0.006 4 0.006 5 0.006 6 0.0064 0.006 6 0.006 4 0.0063 0.006 2 0.0064 2.1 In Table V the result for repeatability of the method proposed in this work is shown.The relative standard deviations for the determination of o.oo3y0 phosphorus and o.007y0 arsenic in the NBS 55e sample are 2.5 and 2.1yo respectively. In Table VI the results obtained for phosphorus and arsenic in steel samples are shown. The results obtained for phosphorus and arsenic are in good agreement with the certified values except for JSS 232-3. Although the result obtained for JSS 232-3 is about 25% smaller than the certified value for phosphorus the values obtained vary as narrowly as the values obtained for other steel samples. Further as shown in Fig. 2 the graph of absorbance against volume of sample solution taken shows good linearity from 0 to 10 ml, and the result for the recovery test of phosphorus is 96-101y0 (see Table VII).Hence the value for the phosphorus content of JSS 232-3 of 0.028y0 obtained in this work would appear to be the correct value. In Table VII the results for the recovery test are shown. To 5 ml of sample solution known amounts of phosphorus as phosphate and arsenic as arsenate were added and the recovery of phosphate or arsenate was examined. The results for this test are recoveries of 96-101y0 for phosphorus and 95-101y0 for arsenic respectively. TABLE VI DETERMINATION OF PHOSPHORUS AND ARSENIC IN STANDARD STEEL SAMPLES Steel sample JSS 230-4 . . JSS 231-3 JSS232-3 JSS 030-2 JSS 050-3 * . JSS 061-2 JSS 159-3 . . JSS 160-3 . . JSS 161-3 . . NBS 19g . . NBS lOle . . * . NBS 362 . . NBS 364 .. NBS 126b . . High purity steel . . . . ,. Certified value % & P As 0.0124 0.022 3 0.0389 0.0179 0.0046 0.0186 0.0113 0.0104 0.0056 0.10 0.051 0.01 0.046 0.025 0.041 0.092 0.01 0.052 Found % P As r A \ 0.0135 (0.0002)* 0.022 4 (0.000 5) * 0.0284 (0.0004)* 0.0167 (0.0002)* 0.0048 (0.0003)* 0.0177 (O.OOOO)* 0.0091 (0.0001)* 0.0096 (0.0001)* 0.0040 (0.0001)" 0.010 0.096 0.004 6 0.047 0.001 4 0.0104 0.002 0.0442 (0.0003)* 0.0226 (0.000 1)* 0.006 1 0.039 8 0.0874 0.0099 0.049 8 0.0049 0.007 8 0.001 6 (0.000 1)* Not detected * These are the means of three determinations. The values in parentheses are the mean n A=l deviations which are given by Z I X - 8 I In where 8 is the mean August 1983 BY EXTRACTION OF HETEROPOLYACIDS WITH ETHYL VIOLET TABLE VII RESULTS OF RECOVERY TESTS ON PHOSPHORUS AND ARSENIC 951 A 5-ml volume of sample solution was taken and phosphate or arsenate solutions were added to it and diluted to 10 ml with distilled water.Steel sample Sample solution/g 1-1 NBS 55e . . 0.196 0 0.1980 0.201 0 JSS 232-3 0.2048 0.196 8 0.1962 NBS 126b . . 0.199 0 NBS 364 . . 0.099 6 NBS 362 . . . . 0.0394 JSS 159-3 . . 0.0373 JSS 161-3 . . 0.1005 JSS 160-3 0.099 4 Added/pg & P As 0.241 0.522 0.241 0.522 0.241 0.522 0.241 0.241 0.241 0.120 0.261 0.120 0.261 0.120 0.261 0.120 0.261 0.120 0.261 0.120 0.261 Recovery % Y-7 P As 98 95 96 95 101 96 98 101 96 99 97 100 98 96 97 98 98 96 101 96 95 Conclusion The proposed solvent extraction - spectrophotometric method for the determination of micro-amounts of phosphorus and arsenic was applied to the determination of phosphorus and arsenic in commercially available iron(II1) and iron(I1) salts and steel samples.The steel samples were advantageously dissolved in aqua regia. In determining phosphorus, co-existing arsenic(V) is masked with sodium thiosulphate and hydroxylammonium sulphate. In the absence of thiosulphate and hydroxylamine phosphate and arsenate each react with molybdate to form heteropolyacids which are extracted into a mixture of cyclohexane -4-methylpentan-%one with ethyl violet. The advantages of this method are as follows (I) high sensitivity; (2) simple and less time-consuming procedure ; (3) the procedure is scarcely subject to interferences by co-existing ions ; and (4) good reproducibility of determination of phosphorus and arsenic at the level in steel samples. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Ducret L. and Drouillas M. Anal. Chim. Acta 1959 21 86. Sudakov F. P. Klitina V. I. and Dan’shova T. Ya. Zh. Anal. Khim. 1966 21 1333. Babko A. K. Shkaravskii Yu. F. and Kulik V. I. Zh. Anal. Khim. 1966 21 196. Matsuo T. Shida J. and Kurihara W. Anal. Chim. Acta 1977 91 385. Kirkbright G. F. Narayanaswamy R. and West T. S. Anal. Chem. 1971 43 1434. Motomizu S. Wakimoto T. and TBei K. Anal. Chim. Acta 1982 138 329. Johnson D. L. Environ. Sci. Technol. 1971 5 411. Johnson D. L. and Pilson M. E. Q. Anal. Chim. Acta 1972 58 289. TBei K. Motomizu S. and Korenaga T. Analyst 1975 100 629. Received November 29th 1982 Accepted January 31st 198