506 Analyst, July, 1973, Vol. 98, pj4. 506-511 The Determination of Microgram Amounts of Sulphate by Emission Spectroscopy of Barium with a Nitrous Oxide - Acetylene Flame BY E. A. FORBES (Ruakura Soil lPesearch Station, Pvivate Bag, Hamilton, New Zealand) The sulphate content of aqueous solutions has been determined in- directly, in the ranges 0.5 to 5-0 and 1.0 to 10-0 p.p.m., from barium emission measurements. By using a slight excess of barium, the sulphate is precipi- tated in a 50 per cent. solution of propan-2-01 and its concentration is calcu- lated from the decrease in the barium content of the solution. The amount of barium in solution is determined from its emission at 553.55 nm in a nitrous oxide - acetylene flame. The only major flame interference detected is that from the band emission of calcium; 410 p.p.m.of calcium gave an emission intensity equal to that of 1 p.p.m. of barium. Sulphate has been determined in both pure solutions and in synthetic sample solutions containing other electrolytes. Major interferences were noted for potassium and ammonium oxalate, sodium orthovanadate, nickel chloride and to a lesser extent for sodium fluoride and perchloric acid. The sulphur content of biological material, digested by oxygen-flask combustion, has been determined satisfac- torily by using this method. THE determination of barium as a method of indirectly measuring microgram amounts of sulphate has, in the past, had limited application owing to interferences and the low sensitivity of the techniquesl-7 available for determining barium, which have made the method relatively unattractive.Some p i - ~ c e d ~ r e ~ , ~ ~ g while possessing sensitivity, have not been amenable to rapid routine analysis because of the need to use radioisotopes. However, recent worklo with the nitrous oxide - acetylene flame has errabled a sensitive, non-radiochemical method for measuring barium to be produced, which is comparatively free fron interferences. A simple procedure is dcscrihed here, which depends on the precipitation of barium sulphate in a 50 pcr cent. solrit.,on or" proym-2-o!. This nicdium was chosen because it reduces the solubility of barium sulphrzee to an aciep-tablte levcl, and because it can be sprayed into a nitrous oxide - acetykn.: urcd by sclecting an amount of barium sligiiliy in excess 0:' that which is r,)tyrilvnlcnt to ti1 : mctximum amount of sulphate.Following precipkLtion xiid cent\ Zfugation, Ltx unr d II.LI iurn content of the solution can be determined from its erni5s;on intensity at 553-35 n:n in the nitrous oxide - zcetylene flame. It has b x n found that this method can be US::^ l o - the measurement o€ sulphate concentrations in the prcsenice of an excess of many electrolytes. ithout c~usin:; institb.1 i ty. conn=eali-a, ;0.1s can be rn A particul:ar r~ngi: of s EXPERIMENTAL I K STR u M E N TA I,- A Techtmn AA4 spectrophotometer, with a reciproc:il clisymion o€ 3.3 nrn mmn-l, was operated with a flame-emission accessory and a slit width of 50 pm. The nitrous oxide - acetylene flame was formed on a Techtroa AB50 b w ncr.'Xhe ni trous oxide pressure was maintained at 26 p.s.i. whil:: the acety:?.lit: flow was adju;'i?d to give maximum barium emission. Measztvemcnt of cmissZo;?,-Aikr a &minute w:ixn-im rci i d the tcst solutions were introduced into the dame through a 'Techtron nebuliser and spray chamber. The flame was viewed lengthwise m d the emission intensity of the atomic i-e'anance line at 553-55 nm was measured without noise s u p p ~ ~ i o r . The ccnission signal iroin barium in a 50 per cent. solution of prop;)n-Z-ol was as stable as that originating from aqueous solution. However, the approach to a steady rcading following sample changes was considerably slower than that for an aqueous solution. @ SAC and the author.FORBES 507 REAGENTS- Analytical-reagent grade chemicals were used throughout.Standavd barium solution-Dissolve 0.9782 g of barium chloride diliydrate in distilled water and dilute to 1 litre to give a stock solution containing 550.0 p.p.m. of barium. Dilute 5 and 10-ml volumes of the stock solution to 100 ml so as to give working solutions containing 27.5 and 55.0 p.p.m. of barium, respectively. .Stock sulphate solution, 1000 p.p.m.-Dissolve 1.814 g of potassium sulphate in distilled water and dilute the solution to 1 litre. Prepare working solutions in the concentration range 0.5 to 10 p.p.m. of sulphate by appropriately diluting the stock solution. Chloroacetic acid - potassium hydroxide solution-Dissolve 47-25 g of chloroacetic acid in distilled water. Add 3.40 g of potassium hydroxide and dilute the solution to 250 nil.Filter the solution through Whatman No. 542 filter-paper before use. ProPaia-2-ol. RECOMMENDED PROCEDURE- Quickfit conical flasks with ground-glass stoppers were used as reaction vessels. Reaction volumes were prepared by mixing 10 ml of aqueous sulphate solution, 15 ml of propan-2-01, 2 ml of chloroacetic acid - potassium hydroxide solution and 3 ml of standard barium chloride solution. (The barium concentration of the standard solution is chosen so as to accommodate the required sulphate range.) For the ranges 0-5 to 5.0 and 1.0 to 10.0 p.p.m. of sulphate it is recommended that solutions containing 27.5 and 55-0 p.p.m. of barium, respectively, be used. The flasks were agitated for 15 hours, and the barium sulphate suspension was centrifuged at 3500g for 20 minutes.Aliquots of the supernatant liquid were then taken for emission analysis. Although the two ranges of sulphate concentration given above overlap considerably, the use of the weaker barium solution is advantageous for smaller sulphate concentrations, particularly at and below the 1 p.p.m. level, because of the nature of the difference method used. DEVELOPMENT OF THE METHOD ENHANCEMENT OF BARIUM EMISSION BY POTASSIUM- Results similar to those reportedlo for 1 p.p.m. of barium in aqueous solution were obtained for 1 p.p.m. of barium in a 50 per cent. solution of propan-2-01. The barium emission intensity obtained in the presence of 600 p.p.m. of potassium (the concentration of potassium obtained from the chloroacetic acid - potassium hydroxide solution) was 97 per cent.of that obtained in the presence of 1000 p.p.m. of potassium. EMISSION INTENSITY AS RELATED TO BARIUM CONCENTRATION- over a wide concentration range (Table I). for a zero addition of barium, relative to 2.42 and 22.3 p.p.m. of barium. The relative emission intensity was found to be linearly related to barium concentration The background emission at 553.55 nm originating from the flame and reagents is shown TABLE I VARIATION OF EMISSION INTENSITY WITH BARIUM CONCENTRATION Barium concentration, p.p.m. 0 0.17 0.34 0.81 1-20 1.61 2.42 Relative emission intensity* 6 13.5 20.2 37.5 54.5 70 100 Barium concentration, p.p.m. 0 1.72 3.44 5.15 8.58 12.0 17.2 22.3 Relative emission intensityt 0.8 9-8 17-5 26 43 58 80 100 * Relative t o 2.42 p.p.m.of barium = 100. t Relative to 22-3 p.p.in. of barium = 100.508 FORBES: DETERMINATION OF MICROGRAMS OF SULPHATE BY EMISSION [A?Zdyst, VOl. 98 ENHANCEMENT OF BARIUM EMISSION BY PHOSPHATE- The effect of phosphate (as potassium dihydrogen orthophosphate) on the emission intensity produced by 2.7 and 5.3 p.p.m. of barium was measured for amounts containing 0 to 10mg of phosphorus. The presence of 4mg of phosphorus in solution produced an enhancement of the order of 2 per cent. for both barium concentrations. A maximum enhancement of 8 per cent. was obtained from 10 mg of phosphorus in a solution containing 2.7 p.p.m. of barium. The total volume of each solution was 30 ml. The results given later in Table VII for added potassium chloride (22 mg) show that this enhancement is not due to the suppression of barium ionisation by potassium.CALCIUM EMISSION AT 553.55 nm IN THE NITROUS OXIDE - ACETYLENE FLAME- The emission intensity from calcium in a 50 per cent. solution of propan-2-01 in the presence of potassium was linearly related to concentration, but the intensity was reduced by a factor of 2.6 when the Techtron burner head AB40 was replaced with the Model AB50. While using the latter the net emission from 410 p.p.m. of calcium was identical with that produced by 1 p.p.m. of barium. EFFECT OF REACTION TIME ON PRECIPITATION- A factor that greatly affects the measurement of sulphate concentration is the rate of growth of barium sulphate crystals during precipitation. By using sulphate labelled with sulphur-35 it was possible to measure the efficiency of removal of sulphate from solution as barium sulphate.The efficiencies obtained by centrifuging at 3500 g are given in Table I1 as functions of sulphate concentration, reaction time and degree of agitation of the solution. It is evident from Table I1 that, by agitating the solutions during precipitation, the efficiency of sulphate separation at low concentrations was increased. A high efficiency of separation a t all concentrations in the range 0.5 to 5.0 p.p.m. of sulphate was obtained when the reactants were agitated for 15 hours. Maximum efficiency was achieved within 15 hours for all concentrationi in PERCENTAGE Reaction time/hours the range 1.0 to 10.0 p.p.m. of sulphate. TABLE I1 OF SULPHUR-% ACTIVITY REMOVED BY CENTRIFUGATION FOLLOWING PRECIPITATION Unstirred solution Agitated solution A r 7 15 * .... 14 90 0.5 2.5 Sulphate concentration, p.p.m. 0.5 75 92 44 67 94 1.01 88 95 58 84 96 2.01 91 97 74 93 98 4-03 95 97 -* -* 97 5-04 97 97 77 94 96 * Not determined. RESULTS SULPHATE DETERMINATION IN STANDARD SOLUTIONS- The sulphate contents of dilute potassium sulphate solutions were measured by using standard barium chloride solutions containing 24.9 and 49.7 p.p.m. of barium. The more dilute barium solution was used for the range 0-5 to 5.0 p.p.m. of sulphate. Table I11 gives the emission from the unreacted barium for various sulphate concen- trations. The theoretical emission intensities, calculated for stoicheiometric reactions in which all of the sulphate is precipitated, are given in Table I11 for comparison.As barium sulphate has an acceptably small but nevertheless finite solubility in a 50 per cent. solution of propan-2-01, the measured emission values (adjusted for flame background) should, in principle, be slightly greater than the theoretical values given in Table 111. EFFECT OF PHOSPHATE ON PRECIPITATION- The presence of phosphate (as potassium dihydrogen orthophosphate) during the pre- cipitation reaction caused, in some instances, a reduction in the residual emission intensity.July, 19731 SPECTROSCOPY OF BARIUM WITH A NITROUS OXIDE - ACETYLENE FLAME TABLE I11 RESIDUAL EMISSION INTENSITY AS A FUNCTION OF SULPHATE CONCENTRATION 509 Emission Sulphate* 7- Sulphatet concentration, Measured concentration, p.p.m. (adjusted) Theoretical p.p.m.0 100 100 0 0-81 84.5 84.6 1.22 1.22 76.5 76.6 2-03 2.03 60.0 61.1 4-05 4.05 19.5 22.4 4.87 4.87 6.2 6.7 6-07 9.33 * Precipitated with a solution containing 24.9 p.p.m. of barium. t Precipitated with a solution containing 49.7 p.p.rn. of barium. Emission 1 Measured (adjusted) Theoretical 89-0 88.3 82.0 80.5 62.5 61.1 54.5 53.3 42-5 41.7 11.0 10-4 100 100 By precipitating in the presence of 2.5 mg of phosphorus, the effect of the phosphate on the determination of sulphate in the ranges 0.5 to 5.0 and 1.0 to 10.0 p.p.m. of sulphate, by using solutions containing 26.6 and 53.2 p.p.m. of barium, respectively, was ascertained. The results are given in Table IV. TABLE IV EFFECT OF 2.5 mg OF PHOSPHORUS AS PHOSPHATE ON RESIDUAL EMISSION INTENSITY Residual emission intensity Sulphate* - Sulphatet concentration, Measured concentration, p.p,m.(adjusted) Theoretical p.p.m. 0 100 100 0 2.5 76.0 77-6 0.5 5.0 51.0 55-2 1.0 6.0 41.3 46.2 2.0 7.5 27.3 32.7 3.0 9.0 16.9 19.3 5.0 * Precipitated with a solution containing 53-2 p.p.m. of barium. t Precipitated with a solution containing 26.6 p.p.ni. of barium. 10.0 9.9 10.3 Residual emission intensity r------7 Measured (adjusted) Theoretical 90.3 91.0 81.2 82.1 62.8 64.1 45.6 46.2 12.5 10.3 100 100 The presence of 2.5 mg of phosphorus reduced the residual barium emission intensity obtained for the concentration range 1 to 10 p.p.m. of sulphate, the maximum reduction in intensity occurring at 6.0 p.p.m. of sulphate. In contrast, a negligible effect was observed for the range 0-5 to 5 p.p.m.of sulphate. The reduction in residual barium emission following a precipitation is in contrast with the enhanced barium emission in the presence of phosphate referred to earlier. It is believed that phosphate becomes involved in the precipitation process and that a small amount of barium, in addition to that removed stoicheiometrically with sulphate, is lost from solution. The influence of different amounts of phosphate on the residual emission intensity obtained for a single sulphate concentration in the range 1 to 10 p.p.m. of sulphate is given in Table V. In order to obtain the maximum variation in results an aliquot containing 6-15 p.p.m. of sulphate was allowed to react with 3 ml of a solution containing 55.21 p.p.m. of barium (see Table IV).Table V shows that 0.125 mg of phosphorus produces an effect of about 2 per cent. An increase from 0.125 to 1.00 mg of phosphorus resulted in a further reduction in emission intensity, while the change from 1.00 to 2-63 mg had no additional effect. By reducing the residual barium emission intensity, phosphate causes an overestimation of the sulphate concentration. At the sulphate concentration that gives maximum phosphate interference the presence of as much as 0.125 mg of phosphorus as phosphate can be tolerated. Even when 2.63 mg of phosphorus as phosphate is present the sulphate concentration is overestimated by only 9 per cent. The effect of phosphate on the determination of sulphate can be partially eliminated by preparing a standard graph from sulphate solutions containing phosphate.However, this standard graph will not be linear.510 FORBES: DETERMINATION OF MICROGRAMS OF SULPHATE BY EMISSION [Analyst, Vol. 98 TAUI-E V VARIATION OF RESIDUAL EMISSION INTEKSITY wIm PHOSPHATE CONTENT Phosphate contentlmg of phosphorus . . 0 0.125 0.25 0-50 1.00 2-63 Residual emission intensity . . . . 50.3 48.7 47.9 46-9 45.7 45.7 INFLUENCE OF OTHER ELECTROLYTES ON PRECIPITATION- Microgram amounts of sulphate were precipitated in the presence of different electrolytes in order to determine the possible effect of each on the measurement of sulphate concentrations by this procedure. The effect of each salt, with the exception of calcium nitrate, was examined for only one sulphate concentration (see Table VI). In Table VII the effect of several other salts, at a single concentration, is given so as to indicate interferences arising from reaction with barium alone or from an effect on the precipitation reaction.TABLE VI EFFECT OF ELECTROLYTES ON RESIDUAL BARIUM EMISSION INTENSITIES OBTAINED FROM THE PRECIPITATION OF 4.62 p.p.m. OF SULPHATE Sodium Residual barium fluoridefmg emission intensity 0 62.8 1.5 61.7 3 62-3 6 56.6 9 51.5 15 40.8 Perchloric Residual barium acid/mmol emission intensity 0 61.4 0.24 60.4 0.47 61.4 0.93 81.9 1.39 67-5 2.32 75.6 Potassium oxalatelmg 0 1.0 2.1 4.1 6.15 10.25 Residual barium emission intensity 61.4 59.4 57.9 55.8 54.3 51.S Calcium ztitvate-The effect of 2 mg of calcium [as Ca(N0,),.4H20] on the precipitation was determined for several sulphate concentrations in the range 1 to 10 p.p.m.of sulphate. It was found that, after subtracting the calcium component of the emission, the residual barium emission intensity obtained for each sulphate concentration was identical, within experimental error, with that given in Table 111. TABLE VII INDICATIONS OF THE EFFECT OF OTHER SALTS ON SULPHATE DETERMINATIONS Relative effect on Relative effect on barium alone* sulphate precipitation? Salt Amount/mg (emission units) (emission units) None .. . . .. . . - 50 62.3 Magnesium chloride hexahydrate 16 49 61.7 Cobalt chloride hexahydrate . . 20 49.6 61.2 Nickel chloride hexahpdrate . . 20 44 50.0 Sodium orthovanadate . . . . 6 (approx.) 46 51.0 Sodium acetate trihydrate . . 24 48.5 61-53 Sodium citrate dihydrate . . . . 8 48 58.7 Potassium chloride .... 22 50 62-8 Potassium iodide . . .. .. 6 49.7 59.7 Potassium carbonate . . .. 10 51 61-7 Potassium hydrogen tartrate . . 6 50.5 60.2 Ammonium chloride . . .. 15 49 59.7 Ammonium molybdate . . .. 6 49 60.7 Aluminium chloride hexahydrate 45 44-5 54.1 Citric acid . . .. .. .. 22 48-5 57.6 Ammonium oxalate monohydrate 20 48 47.5 * The salts were present in 50 per cent. solutions of propan-2-01 containing chloroacetic acid - potassium hydroxide and 5.52 p.p.m. of barium. The solutions were agitated for 14 hours and centrifuged a t 3500g. Aliquots of the supernatant solution were then analysed. ? The emission intensity values were obtained from the analysis of a 10-ml aliquot of a 4.62 p.p.m. sulphate solution by using 3 ml of 55.21 p.p.m.barium solution. The same procedure as that described in the above footnote was used. Sodium cfEuoride-Amounts of sodium fluoride up to 3 mg had no effect on the deter- mination of 4.62 p.p.m. of sulphate in the range 1 to 10 p.p.m. of sulphate. With amountsJuly, 19731 SPECTROSCOPY OF BARIUM WITH A NITROUS OXIDE - ACETYLENE FLAME 51 1 of sodium fluoride above 3 mg the residual barium emission intensity decreased linearly with concentration (see Table VI). Perchloric acid-The residual barium emission intensity was unaffected by the presence of 0-93 mmol of perchloric acid during the precipitation reaction. Greater amounts of perchloric acid apparently increase the solubility of barium sulphate in the medium (see Table VI). Potassiztm oxalate-A progressive decrease in the residual barium emission intensity was obtained for amounts of potassium oxalate increasing from 1 to 10mg (see Table VI).Sodi.um chloride and sodium nitrate-Amounts of these salts up to 20 mg of each had no effect on the residual barium emission intensity obtained for 4-62 p.p.m. of sulphate. ANALYSIS OF BIOLOGICAL MATERIAL- The sulphur content of sixteen samples of biological material including mixed pasture, lucerne, grass, onion, fruit tree leaves, dried blood and faeces has been determined. The samples were digested by oxygen-flask combustion followed by absorption in dilute sodium hydroxide solution. Aliquots of this solution were analysed by the above procedure. Eight samples were analysed in duplicate, six in triplicate and two in quadruplicate. The replicates were prepared from fresh sub-samples of material (not from the same solution) and at least one replicate was produced by a different analyst. The sulphur content of the dry matter ranged from 0.135 to 0-804 per cent. The over-all coefficient of variance was 3-3 per cent.and the standard deviation was 0.0126. CONCLUSION Barium emission in the nitrous oxide - acetylene flame provides a satisfactory means of determining microgram amounts of sulphate indirectly. Smaller solution volumes can be analysed by this procedure simply by reducing by a constant factor the sample volume and the reagent volumes given in this paper. The minimum volume is limited only by the amount required for an accurate emission reading. The sulphur content of biological material, digested by oxygen-flask combustion, has been satisfactorily determined by using this method.An alternative digestion procedure, which involves the use of nitric and perchloric acids, can be used provided the perchloric acid concentration in the digest is reduced sufficiently by dilution prior to the precipitation of barium sulphate. In principle, atomic-absorption spectroscopy could be applied in a similar way to determine the residual barium in solution after precipitation of the sulphate, but this possibility has not been explored in the present work. The author gratdully acknowledges the assistance given by the late Mr. J. E. Allan, who also suggested the idea on which this paper is bascd. The author also thanks Dr. J . H. Watkinson for providing the results for the determination of sulpliur in l~io’logical material. REFEREKES 1. 2. 3. -- - , Ibid., 1960, 85, 688. 4. 5. 6. 7. 8. 9. 10. Burriel-Marti, F., Rarnirez-Mufioz, J., and Rexach-M. cle Lizarduy, M. L., Analytica C h k . Acta, Cullutn, D. C., and Thomas, D. R., Analyst, 1959, 84, 113. Alt, b., Laadzo. Forsch., 1964, 16, 278. Strickland, R. D., and Maloney, C. M., Amer. J . Clin. Path., 1954, 24, 1100. Roe, D. h., Miller, P. S., and Lutwak, L., AylaZyt. Rimhem., 1966, 15, 313. Dunk, R., Mostyn, R. A., and Hoare, H. C . , Atom. Absovptmn Newsl., 1969, 8, 79. Picou, D., and Waterlow, J. C., Nature, Lond., 1963, 197, 1103. Goode, E. V., Analyst, 1968, 93, 663. Koirtyohann, S . R., and Pickett, E. I<., Spectvochim. Acta, 1968, 23B, 673. 1957, 17, 559. Received January 22nd, 1973 Accepted March Znd, 1973