Analyst, May, 1983, Vol. 108, $9. 615-620 615 Simple Method for the Determination of Gypsum, with Some Observations on the Solubilities of Gypsum, Anhydrite, Calcite and Dolomite Henry A. Foner and Sarah Ehrlich Geochemistry Department, Geological SUYVEY of Israel, 30 Malchei Israel Street, Jerusalem. 95 501, Israel A rapid method is presented for the determination of gypsum and related minerals in natural and synthetic materials, based on the extraction of calcium from the matrix material using slightly ammoniacal water as a solvent with subsequent complexometric titration of the calcium using EDTA. The results compare reasonably well with those obtained by the conventional gravimetric procedure. The interferences due to other sparingly soluble calcium com- pounds, such as calcite and dolomite, commonly found in conjunction with gypsum were examined, as were the solubilities of anhydrite, calcite and dolomite in both water and sodium chloride solution.The method is particu- larly suitable for industrial control and deposit assessment purposes. Keywords : Gypsum detevnziwation ; anhydrite ; calcite ; dolomite ; solubility Gypsum is an important mineral both geologically and industrially. It is used as the basis for the manufacture of plaster and plaster boards, as the raw material for the preparation of plaster of Paris and as an additive to Portland cement. In nature, calcium siilphate is found in two principal forms: gypsum (CaS0,.2H20) and anhydrite (CaSO,). The principal form used in industry is the hemihydrate, plaster of Paris (CaS0,.0.5 H,O), which is made by heating gypsum.After addition of water, the plaster sets and reverts to gypsum. The usual method for determining the amount of gypsum (or other forms of calcium sulphate) in a rock, mineral or industrial product is to dissolve the material in acid and then to determine the sulphate dissolved.lS2 The determination is usually carried out gravimetrically using barium chloride solution to precipitate barium sulphate. This is the classical method for the determination of sulphate and can be traced back in the literature at least as far as 1927.l Although the gravimetric method for sulphate determination is a very common procedure, it is a time-consuming operation involving filtration and weighing and requires a skilled operator for good results. To obtain correct sulphate values, the recommended conditions have to be carefully observed as the method depends on the compensation of errors due to the solubility of barium sulphate in hydrochloric acid on the one hand and the adsorption of extraneous ions on the precipitate on the 0ther.~94 Various other methods of determining sulphate have been suggested but they have generally not been very successful and the gravi- metric method still holds its own. The determination of calcium sulphate in a given sample depends on the assumption that the only acid-soluble sulphate present is that of calcium.This is indeed generally so and the only common salts that would interfere in the determinations are sodium and magnesium sulphates. Fortunately, these are rarely present in natural gypsum deposits or in industrial products.Now gypsum (and the other forms of calcium sulphate) is moderately soluble in water (see Table I), and the question arises as to whether it is possible to determine gypsum by determin- ing the water-soluble calcium. To us this seemed to be a potentially attractive method as in naturally occurring deposits of gypsum and anhydrite, and a€so in most industrial products, there are usually no other water-soluble calcium salts present. The determination of calcium with EDTA is well known, convenient and rapid. To the best of our knowledge, it has not previously been used for this application. The method developed in this work is particularly suited to simple laboratories as it avoids the use of cyanide as a complexant in the calcium titration.The method is not suitable for the determination of gypsum in Portland cement because of the presence of readily soluble calcium silicates.616 FONER AND EHRLICH : DETERMINATION OF GYPSUM, AND Analyst, VoZ. 108 Experimental Many of the experiments described below were carried out on a series of natural gypsum- containing samples obtained from the quarries of Kibbutz Gesher in Israel. The deposits range from almost pure gypsum rock (98% gypsum) to an overburden containing only a few per cent. of gypsum with major amounts of calcite, dolomite, clays and quartz. A small amount of magnetite (Fe,O,) is also present. Other samples containing anhydrite, gypsum, calcite or dolomite were obtained from a variety of sources, including deep drillings. Analytical Method Reagents Distilled water and analytical-reagent grade reagents were used throughout.EDTA solutioiz, 0.05 M. Weigh 18.612 g of EDTA, disodium salt. dilute to 1 1. Indicator. Potassium hydroxide bufer solution. 2 N . Triethanolamine solution (1 + 1). Dissolve in water and Hydroxynaphthol blue (Mallinckrodt, No. 5630). Dissolve 112 g of potassium hydroxide in 1 1 of Add 500 ml of triethanolamine to 500 ml of water and water. stir until homogeneous. Procedure Accurately weigh about 1 g of sample into a 600-ml beaker, add 450 ml water and 1.0 ml of ammonia solution (sp. gr. 0.880) and boil for 15 min. Allow to cool, transfer the solution (including solid) into a 500-ml calibrated flask and dilute to volume. Allow to settle, take a 50- or 100-ml volume of the clear supernatant solution (depending on the sample) and transfer it into a 250-ml conical flask.If necessary, filter the supernatant liquid before pipetting. Add 5 ml of triethanolamine solution (1 + l), 10 ml of 2 N potassium hydroxide solution and approximately 100 mg of the indicator. Titrate with 0.05 M EDTA solution until the red solution turns blue. Use a similar procedure to the above but with 0.25 g of sample. Leave the solid in contact with the liquid for 24 h before diluting to volume (see next section). For samples containing gypsum and hemihydrate. For samples containing anhydrite. Stirring in the cold for 30 min is slightly preferable to boiling. 1.00 ml of 0.05 M EDTA solution = 2.004 mg of Ca. Development of the Method Solubility of anhydrite in water The solubilities of the various forms of calcium sulphate in water are shown in Table I.TABLE I SOLUBILITIES OF VARIOUS FORMS OF CALCIUM SULPHATE IN WATER Data from “Handbook of Chemistry and Physics.”& Solubility/mg 1-1 r- Mineral Foriiiula 2O0C l O 0 O d . Gypsum . . .. . . CaS04.2H,0 2 400 2 200 Hemihydrate . . . . CaS04.0.5H,0 3 000 - Anhydrite . . .. . . CaSO, 2 loo* 1600 * At 30 “C. Anhydrite is generally considered to be difficult to dissolve. To check that it is indeed soluble under the conditions suggested in this paper, a number of samples of naturally occurring anhydrites were analysed using the above procedure. In most instances the samples dissolved completely after stirring for 30 min in the cold (0.25 g of sample in 350 ml of water).One sample would not dissolve under these conditions and required standing for 21 h for completeM a y , 1983 SOLUBILITIES OF GYPSUM, ANHYDRITE, CALCITE AND DOLOMITE 617 dissolution. The solubility of synthetic anhydrite at 20 “C under these conditions was 2590 mg 1-l. The solubility of anhydrite, unlike that of gypsum, decreases with increasing temperature.61’ A series of experiments were carried out to study the effect of dissolving anhydrite in both hot and cold water. Recoveries were slightly worse when the sample was boiled in water at the start of the dissolution procedure. The conclusion to be drawn from the above experiments is that to be absolutely sure of completely dissolving anhydrite it is necessary to leave the samples in contact with cold water for 24 h.Solubility of calcite amd dolomite in water calcite and dolomite. Other than calcium sulphate, the only calcium salts present in the samples examined were The solubilities of these compounds, as quoted in the “Handbook of Chemistry and Physics,”5 are shown in Table 11. TABLE I1 SOLUBILITIES OF CALCITE AND Mineral Formula Calcite. . . . . . CaCO, Dolomite . . . . CaCO,.MgCO, DOLOMITE I N WATER Solubility/mg 1-1 Cold Hot 14 (25 “C) 18 (75 “C) 320 (18 “C) - An attempt was made to check the accuracy of these data. The solubility of calcite obtained experimentally (after standing for 24 h to attain equilibrium) was 13.5 mg l-l, which agrees well with the figure quoted above. However, it soon became apparent that the solubility of dolomite was completely at variance with that quoted above.Indeed, geological experience indicates that dolomite should be less soluble than calcite.8 I t is clear from the figures shown in Table I1 that the presence of a substance (dolomite) with a solubility of 130/, of that of gypsum would seriously limit the use of the suggested method, particularly at low gypsum concentrations. Solubility of dolomite. Any investigation of the solubility of dolomite is complicated by the fact that dolomite has never been precipitated from solutions in the laboratory under normal conditions of temperature and pre~sure.~ Garrels et aZ.1° studied the solubility of dolomite in water under a carbon dioxide pressure of 1 atm. They found that finely ground dolomite had a higher solubility than relatively coarsely ground rock and suggested that the very fine material dissolved incongruently, yielding a magnesium-rich solution.This phenomenon was attributed either to the disordering of the crystals on the grain surfaces and hence preferential solution of magnesium, or to super- saturation from excessive fineness of the dolomite particles. Yanat’eval1,l2 found that at 25 “C and a normal atmospheric partial pressure of carbon dioxide, dolomite dissolves in- congruently to yield calcite and a magnesium-rich solution. Hsu,13 on the other hand, casts doubt on the validity of Garrels et al.’s low figure for dolomite solubility and calculated it from a study of ground-water chemistry. At all events, all of the figures quoted in the studies mentioned above show that the solubility of dolomite is lower than that of calcite.The compositions of a number of dolomitic rock samples are shown in Table 111, which also gives an indication of the amount of calcite present. The latter information was obtained from X-ray diffractograms of the samples. Only samples B 1, YD 614 and BR 2 were free from calcite and of these only YD 614 had the theoretical molar [Mg] : [Ca] molar ratio of 1.00. Dolomites with up to 5 mol-% of structural calcite have been described by Goldsmith and Graf.l4 Israeli dolomites with up to 11 mol-% of excess of structural calcium carbonate (i.e., [Mg]: [Ca] molar ratio = 0.80) have been investigated by Katz.l5 One-gram samples of these rocks were boiled in 200 ml of distilled water for periods of 15 and 30 min.The solutions were allowed to cool to room temperature, filtered through Whatman No. 42 filter-papers and diluted to 250 ml. The calcium and magnesium contents of the solutions were determined by atomic-absorption spectrophotometry. Table IV shows the results for the samples boiled for 15 min. The samples boiled for 30 min gave identical results.618 FONER AND EHRLICH: DETERMINATION OF GYPSUM, AND Analyst, Vol. 108 In every instance the calcium in solution from the “dolomite” is less than that due to the solubility of calcite alone, even when the latter mineral is present, e.g., samples YD 411 and H 2. I t is also clear that dolomites are less soluble than calcites under these conditions. TABLE I11 COMPOSITION OF SOME “DOLOMITE” SAMPLES Sample B 1 .. .. H 2 .. YD 409 .. YD 411 .. YD 614 . . B R 1 .. B R 2 .. MgCO3, Yo . . 31.7 . . 34.6 . . 36.2 . . 37.6 . . 43.5 . . 38.3 . . 38.9 Total CaCO, + CaCO,, % MgCO,, % 44.2 75.9 60.8 95.4 51.0 87.2 54.4 92.0 51.9 95.4 55.0 93.3 56.7 95.6 [Mg] : [Ca] molar ratio 0.85 0.67 0.84 0.82 0.99 0.83 0.81 Calcite Trace Major Minor Present Trace Present Absent E$ect of pH on the solubility of calcite, dolomite and gypsum in water Theoretically, increasing the pH of the solvent should decrease the solubility of both calcite and dolomite. A series of experiments were carried out to confirm this supposition; the results are shown in Table IV. In each instance 1 g of finely ground rock sample was boiled in 200 ml of alkaline water for 15- and 30-min periods.A l-ml volume of ammonia solution (sp. gr. 0.880) was added to the mixture before boiling. The solutions were cooled to room temperature, filtered through a Whatman No. 42 filter-paper and the filtrate was diluted to 250 ml. The calcium and magnesium in the solutions was then determined. Only the results for the 15-min boilings are shown in Table IV, as the amount of alkaline earths in solution was greater for the 30- than for the 15-min boiling period. This is because ammonia is expelled during prolonged boiling, causing a decrease in pH. In all instances the total concentration of calcium + magnesium is less than that obtained in similar experiments without the addition of ammonia. The solubility of anhydrite under these experimental conditions was 2560 mg l-l, i.e., the same as in water (see Solubility of anhydrite in water).An attempt to work at a fixed pH of 10 using an ammonium chloride - ammonia buffer was abandoned when it was found that the solubility of calcite increased dramatically in this medium to a value of 68.5 mg 1-l of CaCO,. A standard method for the determination of gypsum in gypsum products2 is based on the high solubility of gypsum in ammoniacal am- monium acetate solution. Experiment showed that in this solution too, the solubility of calcium carbonate was much higher than in water (approximately 250 mg 1-1 of CaCO,). This, of course, limits the method to samples that contain no calcium carbonate or dolomite. TABLE IV CALCIUM AND MAGNESIUM CONCENTRATIONS DUE TO THE SOLUBILITY OF SOLUTIONS (AFTER BOILING FOR 15 MIN) CALCITE AND VARIOUS DOLOMITIC ROCKS IN WATER AND IN ALKALINE Solubility in water Sample B 1 .. .... H 2 .. .. .. YD 409 .. .. YD 411 .. .. YD 614 .. .. B R 1 ,. .. .. B R 2 .. .. .. Natural . . .. Reagent . . .. Type Dolomitic Dolomitic Dolomitic Dolomitic Dolomitic Dolomitic Dolomi tic Calcite Limestone r- Mg -7 mg I-’ mmol 1-1 1.1 0.046 3.1 0.129 1.4 0.058 0.55 0.023 1.1 0.046 0.6 0.025 0.9 0.037 - - - - 7 mg 1-l 2.6 2.3 2.2 2.2 1.4 2.8 2.0 3.6 4.3 1 Ca - mmol 1-I 0.065 0.058 0.055 0.055 0.035 0.070 0.050 0.09t 0 . l l t Solubility in alkaline solution 7 h - F Mg/mg 1-l Ca/mg 1-I 0.6 2.0 0.4 2.4 0.6 1.4 0.2 1.1 0.5 1 .0 0.7 2.4 1 . 1 0.8 - 1.8* -- 2.2 * Corresponds to 4.5 mg 1-1 of CaCO,. t Non-equilibrium value.May, 1983 SOLUBILITIES OF GYPSUM, ANHYDRITE, CALCITE AND DOLOMITE 619 Solubility of calcite, dolomite and calcium sulphate in sodium chloride solution Both anhydrite and gypsum approximately double their solubilities when the salt concentration of the dissolving brine is about 7% ; at higher salinities their solubility decreases again.6Yl6 Unfortunately, both calcite1' and dolomite also show increasing solubility with increasing salt content of the solvent. Table V shows the effect of increasing salt concentration in the solvent brine on the solubility of calcite and dolomite.The increased supersaturation of calcite and dolomite solutions with increasing salinity has been discussed by Sass.l8 The increased solubility of gypsum in salt solution is well known.6p16 TABLE V EFFECT OF SODIUM CHLORIDE CONCENTRATION IN THE SOLVENT ON THE SOLUBILITY OF CALCIUM IONS FROM CALCITE AND DOLOMITE Solubility/mg 1-1 of Ca I \ Sample Type In water In 1.4% NaCl In 3.5% NaCl Reagent .. . . Calcite 3.6 8 17 YD 614 . . . . Dolomite 1.4 5 9 RR 1 . . . . Dolomite 2.8 8 14 Interestingly, experiment shows that the [Mg] : [Ca] molar ratio in the dolomite solution (sample YD 614) decreases from 1.3 for water to 1.02 for 1.4% sodium chloride solution and to 0.80 for 3.5% sodium chloride solution. Similar behaviour was noticed in other dolomites. This indicates that calcium ion is preferentially dissolved with increasing salt concentration of the solvent. Because of the increased solubility of the carbonate minerals in brines, it was decided not to develop 1;his method further for the analysis of gypsum and related minerals. Results A comparison of the results of the analysis of various samples by the proposed method and The results are averages of by the conventional gravimetric method is shown in Table VI.from two to five individual determinations. TABLE VI COMPARISON OF (a) SUGGESTED AND (b) CONVENTIONAL METHODS FOR THE DETERMINATION OF CALCIUM SULPHATE Results as % SO,. Sample SM 114 .. . . A . . D 7 (29-30) . . D 7 (2-4) . . Reagent. . .. Synthetic mixture 704-3 . . D 7 (0-2j ' D 7 (23-24) . . . . . . . . . . . . . . .. .. . . . . Description Theoretical Gypsum ore Phosphate rock Dolomitic rock Gypsum ore Pure gypsum 46.51 Pure gypsum - calcite - dolomite 26.6 Gypsum rock Gypsum rock Gypsum rock Water dissolution 16.45 3.16 6.3 25.23 46.52 (4 26.8 26.5 36.30 14.70 -7 BaSO, gravimetric 16.53 2.93 6.5 25.23 46.52 (b) 26.6 27.17 36.23 15.35 (b) / ( a ) 1.005 0.93 1.03 1 .ooo 1.000 0.991 1.03 0.998 1.04 Accuracy I t is difficult to assess the accuracy of the suggested'method, which depends partly on the proportion of calcium sulphate in the sample and partly on the blank value due to impurities. Table VII lists the maximum blank values obtained when samples of pure analytical-reagent grade calcium carbonate and limestone were processed by the suggested methods, both with and without the use of ammonia.These values represent maxima due to calcite solubility.620 FONER AND EHRLICH TABLE VII MAXIMUM BLANK VALUES ON PURE CALCIUM CARBONATE AND LIMESTONE WITH AND WITHOUT AMMONIA CaCO, in solution/ Equivalent gypsum Titrationlml per Equivalent gypsum mg 1-1 in solution/mg 1-1 100-ml aliquot in original sample, yo In water .. .. 14 24.1 0.30 1.3 In ammonia . . .. 6 10.3 0.12 0.52 With samples containing only calcium sulphate and no calcium carbonates, the accuracy is determined by the size of the titration blank, which is approximately 0.05 ml of 0.05 M EDTA solution (equivalent to 0.2% of gypsum). The last column in Table VI illustrates the accuracies obtained on some actual samples; these are well within the requirements of an industrial analysis. Conclusion A simplified method for the determination of gypsum and related minerals in rocks and industrial products has been developed. The method is based on the fact that calcium sulphate is essentially the only soluble calcium salt in these materials.The solubility of other calcium salts in the extraction medium is reduced by increasing the pH of the latter. The method is particularly suited to industrial control and mineral deposit assessment purposes. The solubilities of gypsum, anhydrite, calcite and dolomite in water, ammoniacal water and brines have been studied. Because of increased calcite solubility in the brines, further investigation of this method was abandoned. It was shown that by addition of ammonia to the extraction medium it was possible to reduce the amount of calcium entering the solution from calcite and dolomite to negligible proportions. The solubility of dolomite in water was shown to be much lower than that quoted in the “Handbook of Chemistry and physic^."^ The authors thank Dr, Y.Druckman of the Geological Survey of Israel for kindly supplying samples of gypsum, anhydrite and dolomite. They also thank Prof. E. Sass of the Geology Department, The Hebrew University, and Drs. Y. Nathan and Y. Druckman of the Geological Survey of Israel for much helpful discussion. We are grateful to Kibbutz Gesher for kindly supplying samples. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Scott, W. W., and Furman, N. H., “Standard Methods of Chemical Analysis,” Fifth Edition, Van “Standard Methods for the Analysis of Gypsum and Gypsum Products,” ASTM Standard C 471-75, Hillebrand, W. F., Lundell, G. E. F., and Bright, H. A., “Applied Inorganic Analysis,” Second Vogel, A. I., “A Textbook of Quantitative Inorganic Analysis,” Third Edition, Longman, London, Weast, K. C., Editor, “Handbook of Chemistry and Physics,” Sixtieth Edition, CRC Press, Boca Deer, W. A., Howie, R. A., and Zussnian, J., “Rock Forming Minerals, Volume V, Non-silicates,” Posnjak, E., Am. J . Sci., 1938, 35A, 247. Palache, C., Berman, H., and Frondel, C., “Dana’s System of Mineralogy,” Seventh Edition, Volume Deer, W. A., Howie, R. A., and Zussman, J . , “Rock Forming Minerals. I‘olume V, Non-silicates,” Garrels, I<. M., Thompson, M. E., and Siever, R., Am. J . Sci., 1960, 258, 402. Yanat’eva, 0. K., Izv. Akad. Nauk SSSR, Otdel. I<hi?n. Nauk, 1954, 6, 1119. Yanat’eva, 0. K., Zh. Neorg. Khim., 1955, 1, 1473. Hsu, K. J., J . Hydrol., 1963, 1, 288. Goldsmith, J . R., and Graf, D. L., J . Geol., 1958, 66, 678. Katz, A., PhD Thesis, Hebrew University, Jerusalem, 196s (in Hebrew). Posnjak, E., Am. J . Sci., 1940, 238, 559. Miller, J . P., Am. J . Sci., 1952, 250, 161. Sass, E., J . Sediment. Petrol., 1965, 35, 339. Nostrand, New York, 1939, p. 214. American Society for Testing and Materials, Philadelphia, 1975. Edition, John Wiley, New York, 1953, p. 716. 1961, p. 462. Raton, 1979. Longmans, London, 1962, p. 209. 11, John Wiley, New York, 1951, p. 213. Longman, London, 1962, p. 281. Received July 20th, 1982 Accepted Decenibev 14th, 1982