DECEMBER, 1966 T H E ANALYST Vol. 91, No. 1089 The Determination of Iron(I1) Oxide in Silicate and Refractory Materials Part I. A Review* BY H. N. S. SCHAFER (Division of Coal Research, CSIRO, P.O. Box 175, Chatswood, New South Wales, Australia) SUMMARY OF CONTENTS Introduction Acid decomposition in a sealed tube Acid decomposition in an inert atmosphere Acid decomposition in the presence of an oxidant Evolution method Fusion method Combustion method Conclusion THE determination of iron(I1) oxide presents a number of difficulties. Two of special im- portance are the method of bringing the sample into solution and the prevention of oxidation of iron(I1) in the process. With materials that are readily soluble in sulphuric or hydro- chloric acids, the determination should present no great difficulty because it is possible to carry out decomposition in glass apparatus, and maintain an atmosphere of a suitable gas (carbon dioxide or nitrogen) to prevent oxidation.This is not so with materials resistant to a simple acid attack, and some procedures designed to overcome the difficulties associated with the determination of iron(I1) oxide in such materials are dealt with here. A third difficulty inherent in all methods is the possibility of the oxidation of iron(I1) oxide during sample preparation, especially if grinding in air is used. According to Hillebrand et aZ.,l if on simple crushing of the material a product is obtained that leaves little or no residue when gently boiled with hydrofluoric acid, then no further preparation of the sample is needed.If, on the other hand, there is considerable residue, then the sample should be ground under alcohol only long enough to yield a powder that leaves little or no residue. The main characteristics of each of the methods considered here are used as the basis for the arbitrary headings under which the methods are discussed. Usually the differences in the methods relate to the initial decomposition stage. ACID DECOMPOSITION IN A SEALED TUBE The method described by Mitscherlich2 y 3 involves the decomposition of the material by heating with sulphuric acid in a sealed glass tube at temperatures up to 200" C. In this procedure, exclusion of oxygen (air) should present no difficulty, but complete dissolution of some of the materials is not always achieved.In Hillebrand's modification1 of the original method, a weaker sulphuric acid solution is used and the air in the tube is replaced with carbon dioxide. The determination of iron(I1) oxide is subject to errors arising from the presence of sulphides and carbonaceous matter. With sulphides the error can be quite large because of the reduction of the iron(II1) that may be present. Pyrite is normally resistant to boiling dilute acids, but under the conditions of increased temperature and pressure that prevail in the sealed-tube method, and in the presence of iron(TII), it will go into solution with reduction of the iron(II1) to iron(I1). These difficulties with the method were fully discussed by Hillebrand, who concluded that, in the absence of sulphides and when the material is fully decomposed, the sealed-tube method is almost ideal; otherwise its application is limited.* Reprints of this paper will be available shortly. For details see Summaries in advertisement pages. 755756 SCHAFER: DETERMINATION OF IRON(II) OXIDE IN [Analyst, Vol. 91 For the decomposition of resistant silicate minerals, Riley4 proposed a method that can be regarded as an extension of the sealed-tube method of decomposition. The resistant mineral is decomposed with hydrofluoric and perchloric acids in a Teflon bomb, which is placed in a steel vessel fitted with safety valves and containing water. The apparatus is maintained at 150" C in an oil-bath for 3 to 4 hours. Riley and Williams5 suggested that this method of decomposition is suitable for the micro determination of iron(I1) oxide in resist ant minerals.ACID DECOMPOSITION IN AN INERT ATMOSPHERE This method generally (but not always) implies the use of hydrofluoric acid to decompose the sample. As originally devised by Cooke6 it involved decomposition of the sample with hydrofluoric and sulphuric acids in a platinum crucible heated in a specially designed water- bath. The crucible was surrounded by an atmosphere of carbon dioxide to prevent aerial oxidation of the iron(I1) during decomposition of the sample. The iron(I1) was titrated with permanganate. One of the difficulties associated with Cooke's method is the tendency for iron(I1) to be readily oxidised in air in hydrofluoric acid solution, and this can occur during the titration. Another is the oxidation of bivalent manganese by permanganate in the presence of hydro- fluoric acid.For reliable results the latter difficulty must be overcome by removal of the hydrofluoric acid or the elimination of its effect. Barnebey' used boric acid to eliminate the effect of the hydrofluoric acid by the formation of undissociated fluoroboric acid. Pratts dispensed with Cooke's apparatus, and, to prevent oxidation of iron(II), carried out the de- composition in a large platinum crucible fitted with a perforated cover under which carbon dioxide was introduced. Hillebrand modified the method by introducing carbon dioxide only at the beginning, relying upon the evolution of steam from the boiling acid mixture to prevent entry of air during the dissolution of the sample.Various other modifications designed to prevent oxidation have been proposed. Tread- wellg carried out the decomposition in a platinum crucible supported in a lead box in which an atmosphere of carbon dioxide was maintained. The box was heated in a paraffin-bath, and the temperature at the end of the decomposition period could be raised to 120" C to remove hydrofluoric acid and so prevent it from interfering in the permanganate titration. BarnebeylO dispensed with carbon dioxide as used in the Cooke apparatus and relied upon the generation of steam to displace air. Instead of permanganate, Sarverll used dichromate to titrate iron( 11), with diphenylamine as internal indicator and a platinum crucible fitted with a tight transparent bakelite lid for the sample decomposition.The lid was provided with a bakelite tube and funnel to enable the carbon dioxide and acid to be introduced into the crucible, The decomposition procedure used was as in Pratt's method. After cooling in a current of carbon dioxide, a measured excess of dichromate was added; the solution was then transferred to a beaker containing boric acid, and the excess of dichromate back-titrated with iron(I1) sulphate with diphenyl- amine as indicator. Soule13 dispensed with the use of platinum ware and carried out the decomposition with hydrofluoric and sulphuric acids in a Pyrex flask in a current of carbon dioxide. He applied a correction for the glass dissolved by conducting a blank determination on ammonium iron(I1) sulphate. In a later paper14 he discussed possible errors in the determination arising from the presence of arsenic in the glass, and claimed that these could be avoided by titrating with cerium(1V) sulphate instead of permanganate.In considering the effects of sulphides in met hods in which hydrofluoric acid decomposition is used, Hillebrand suggested that hydrogen sulphide evolved from soluble sulphides would probably be expelled without reducing iron( 111), but with sulphides containing iron the iron(I1) oxide content would be in error by the amount of iron(I1) they contained. This would be especially significant for pyrrhotite. The insoluble sulphide, pyrite, is resistant to attack by hydrofluoric acid in the absence of air, but Stokes15 showed that in the presence of iron(I1) it is attacked, with the oxidation of the sulphide and reduction of iron(II1).The extent of this reaction depends on the amount of iron(II1) present, and on the degree of fine- ness of the pyrite. Hillebrand considered the influence of pyrite on the iron(I1) oxide determination in most rocks to be negligible, a view that was shared by Dittler.16 The latter maintained that carbon dioxide should, liowever, be introduced into the bottom of the crucible to remove any trace of hydrogen sulphide that might be formed. Schollenberger12 applied the technique developed by Sarver.December, 19661 SILICATE AND REFRACTORY MATERIALS. PART I 757 The effect of organic matter can be minimised if the titration of iron(I1) is carried out with dichromate instead of permanganate, because, according to Sarver,ll dichromate reacts much less readily with organic matter. Densem,17 in modifying Pratt’s method, developed a rather complex apparatus designed to eliminate air during sample decomposition and during the subsequent titration of iron(I1).This apparatus was subsequently modified by Harris.18 In an attempt to overcome oxidation during sample decomposition, Smirnov and Aidinyanlg proposed dissolution of the sample with hydrofluoric and sulphuric acids while it was under a layer of toluene. They claimed that this could be achieved without “bumping” and without forming reaction products that would affect a permanganate titration of iron(I1). Modification of procedures in which hydrofluoric and sulphuric acids are used for sample decomposition has been made to enable micro or semi-micro determinations to be made of iron(I1) oxide. Titrimetric methods were developed by Shioiri and Mitsui,20 Das Gupta21 and Meyrowitz.22 Riley and Williams5 used a spectrophotometric method in which the sample is decomposed in a stoppered Teflon tube heated in a boiling water-bath, the iron(I1) being determined with dipyridyl.In a similar procedure, Shapi1-0~~ decomposed the sample with hydrofluoric and sulphuric acids in the presence of o-phenanthroline in a small plastic bottle. This reagent was designed to react with the iron(I1) as it was released, with the formation of the iron(I1) - o-phenanthroline complex, and so minimise aerial oxidation. The fact that heating affects the stability of the colour seems to militate against obtaining reliable results.N i c h o l l ~ ~ ~ determined iron( 11) oxide in carbonaceous shales by decomposing the sample in sulphuric acid - hydrofluoric acid solution, adding boric acid, and pouring the resulting solution into hydrochloric acid containing iodine monochloride. The iodine liberated was extracted into carbon tetrachloride and titrated with potassium iodate. I t is claimed that the method is applicable in the presence of organic matter equivalent to up to 4 per cent. of carbon. The decomposition of materials that cannot be brought into solution with hydrofluoric acid has been achieved with phosphoric acid. Konopicky and Caesar25 determined iron(I1) oxide in chrome ore by decomposing the sample by heating it with phosphoric acid in an atmosphere of carbon dioxide, the resulting solution being titrated with permanganate.For the determination of iron(I1) oxide in ferrites Kleinert and Funke26 applied a similar method in which nitrogen was used to prevent oxidation. Clemency and Hagner2’ departed from the general practice of determining iron( 111) as the difference between iron(I1) and total iron; instead they determined it by using coulo- metrically generated titanium(II1) ion as titrant, with an automatic spectrophotometric apparatus devised by Malmstadt and Roberts.28 For the determination of iron(II1) in the rocks G-1 and W-1, dissolution of the sample was achieved by heating it to just below the boiling-point with hydrofluoric and concentrated sulphuric acids. Initially, this decomposition was carried out in a specially constructed box in which an inert atmosphere could be main- tained.The results were erratic and always low, but this was evidently not because of the incomplete dissolution of the sample, as a total-iron analysis showed a virtually complete recovery of iron. The low results were presumed to he caused by reduction of iron(III), perhaps by some constituent of the rocks. Subsequently the decomposition was carried out in a covered crucible. The heating time proved critical, 3 to 4 minutes being the optimum. With heating times in the range of 4 to 5 minutes results were low and erratic, andreduction of iron(II1) appeared to occur. Heating times in excess of 5 minutes gave gradually increasing iron( 111) values, presumably because o f the predominance of air oxidation over the reduction observed.Evidently, therefore, in this instance the problems associated with the determination of iron(I1) are not overcome by carrying out a determination of iron(II1) instead. ACID DECOMPOSITION IN THE PRESEKCE OF AN OXIDANT To overcome difficulties with iron( 11) oxide determinations arising from aerial oxidation, many methods have heen proposed for decomposing the sample with acids in the presence of an oxidising agent. A known amount of oxidant is used, which oxidises the iron(I1) as it is brought into solution, and the excess of oxidant, or the product of its reduction, is then determined by a suitable (usually titrimetric) method. This was suggested (but never tried) by H a ~ k l , ~ ~ who proposed decomposing the sample with hydrofluoric acid in the presence758 SCHAFER: DETERMINATION OF IRON(II) OXIDE IN [Analyst, Vol.91 of potassium dichromate and subsequently titrating the excess of dichromate. The first practical application of such a procedure was carried out by Shein,30 who used it for the analysis of iron(I1) oxide in chromite. He decomposed the sample by heating (360” to 380” C) with phosphoric and sulphuric acids in the presence of a weighed amount of vanadium pentoxide, and then titrating the excess of vanadate with iron(I1) sulphate. Modifications of Shein’s method have been used by many other workers for determining iron(I1) oxide, generally in chromite and chrome - magnesite materials. These materials are not decomposed by hydrofluoric and sulphuric acids but can be brought into solution with phosphoric acid.31 to 38 Most of the applications of acid decomposition in the presence of an oxidant reported in the literature are based on the use of phosphoric acid and vanadium pentoxide, as in Shein’s method.However, other oxidants have been used. G o ~ w a m i ~ ~ claimed that a considerable amount (16 to 20 per cent.) of vanadium pentoxide underwent decomposition during sample dissolution, and proposed the use of cerium(1V) sulphate as oxidant. Although cerium(1V) sulphate undergoes some decomposition, Goswami obtained consistent and reproducible results for the determination of iron(I1) oxide in chromite. The volume of iron(I1) sulphate required for the cerium(1V) sulphate added was multiplied by a derived factor, which pre- sumably allows for the slight decomposition of cerium(1V) sulphate under the conditions used.Ingamells40 based his method on the stability of both bivalent and tervalent manganese in phosphoric acid - pyrophosphate mixtures. A measured excess of potassium permanganate is incor- porated in the phosphoric acid reagent containing bivalent manganese, which is oxidised by the permanganate to tervalent manganese. After addition and dissolution of the sample, the excess of tervalent manganese is titrated with iron(I1) sulphate. Nagato41 developed a spectrophotometric method based on the decomposition of the sample with phosphoric acid in the presence of manganese(II1) oxide; the optical density of the resulting solution containing tervalent manganese was read at 525 mp.used an oxidant prepared from ammonium cerium( IV) nitrate treated with phosphoric acid. For the determination of iron(I1) oxide in ferrites he decomposed the sample with phosphoric acid containing a measured amount of the phosphatocerate reagent and determined the excess of reagent by titration with iron( 11) sulphate. Cheng claimed that phosphatocerate was stable at the temperature and for the time recommended for sample decomposition. For samples decomposed by cold hydrochloric acid or hydrochloric - hydrofluoric acids mixture, Hey43 used iodine monochloride as the oxidant. The iodine formed by reaction with iron( 11) was determined by titration with potassium iodate. Ishibashi and Kusaka4* decomposed silicates by heating with hydrofluoric and sulphuric acids in the presence of ammonium metavanadate, and titrated the excess of vanadate with iron(I1) sulphate.Wilson45 used the same approach but carried out the decomposition at room temperature, extended periods being required for the dissolution of some samples. Under these conditions vanadate was considered to be stable in hydrofluoric acid. Later,46 he adapted the method for micro-scale work by titration, as well as by introducing a colorimetric method based on the regeneration at pH 5 of iron(I1) when the reaction V4+ + Fe3+ + V5+ -1- Fe2+ proceeds to the right. The removal of iron(I1) by the formation of a complex with dipyridyl assisted the reaction to proceed in the desired direction. The colour developed was then measured spectrophotometrically. Wilson’s volumetric method has been adapted by Jackson47 on a semi-micro scale for determining iron(I1) in the ash and slag from pulverised- fuel boilers. Potassium dichromate as the in situ oxidant was used by Reichen and FaheV8 for deter- mining iron(I1) oxide in rocks and minerals.Decomposition was effected with hydrofluoric and sulphuric acids, at a temperature between 65” and 70” C. However, under these conditions dichromate was found to decompose slightly by reaction with hydrofluoric acid. This effect could be minimised by adding an iron(TT1) salt. The destruction of dichromate appeared to be proportional to the amount in excess, and so a correction based on the simple subtraction of a blank could not be applied. The alternative procedure was therefore adopted of multi- plying the volume of iron(I1) solution used to titrate the excess dichromate by the ratio of the volume of dichromate to iron(I1) solution used in the blank determination.The determination of iron(I1) oxide in materials that can be decomposed by acids in the presence of an oxidant, although attractive in principle, may in practice give rise to a It must be assumed that the extent of decomposition is reproducible. More recently He termed the resulting product “phosphatocerate.”December, 19661 SILICATE AND REFRACTORY MATERIALS. PART I 759 number of difficulties, the most important of which is the stability of the oxidant under the conditions used. If decomposition of oxidant occurs it must be reproducible, and therefore rigid control of the conditions is required to ensure the same degree of decomposition in both sample and blank determinations.In addition, the possible effect of other constituents in the sample on the decomposition of an oxidant should not be overlooked. Some workers have tried to eliminate the possibility of oxidant decomposition by carrying out the acid treatment at room temperature. The disadvantage here is that while some materials are readily dissolved, others require long periods of digestion. A possible advantage is that oxidation of organic matter, such as carbonaceous material, may be less likely at room temperature. SUMMARY OF METHODS IN WHICH ACID DECOMPOSITION IS USED IN THE PRESENCE OF ADDED OXIDANT Material Acid decomposition Silicates, Heating with including H F - H,SO, rocks and minerals room temperature HCl or fICJ - HE' a t Heating with H F a t room tempera- H F a t room tempera- H F - H2S04 ture ture Heating with H3P04 - NaH,P04 Heating with HF - H,S04, 65" to 70" C Chromite Heating with (and chrome - H,PO, - H,SO,, magnesite 360" to 380" C refractories) H,PO, - H,SO, heated a t 360" C Heating with Heating with H3P04 - H2S04 H,PO, - H2S04 heated a t 290" to 300" C HSPO, - H2SO4 HSPO, - H2SO4 Heating with H3P04 - H2S04 Heating with H3P04 Ash and slags HF at room tcmpera- ture Tourmaline Heating with H3PO4 - H2S04 Ferrites H,PO4 a t 250' C H,P04 a t 250" to 300" C Transition H,SO, metal oxides Oxidant Method of determination Reference K2Crz07 Titration of exccss di- H a ~ k l ~ ~ chromate potassium iodate with iron(I1) sulphate NH4V0, Exccss vanadate titrated Wilson45 with iron(1I) sulphate KH,VO, Micro titration of excess Wilson46 vanadate, and also spectrophotometric Tervalent Titration of excess man- Ingamell~4~ manganese ganese with iron(I1) IC1 Iodine formed titrated with Hey43 V,O, Excess vanadate titrated Ishibashi and Kusaka44 sulphate chromate with iron(I1) sulphate K,Cr,O, Titration of excess di- Reichen and Fahey* V205 Excess vanadate titrated Shein30 with iron(I1) sulphate V,O, Titration of excess vana- Samanta and Sen36 V,O, Titration of excess vana- Balyuk and Mirakyan31 V,05 V4+ formed titrated with Nagaoka and Yamazaki3s date date permanganate Cerium (IV) Excess cerium( 1V) titrated Goswami39 sulphate with iron(I1) sulphate V,O, Titration of excess vana- Sasuga and Iidaa7 V,O, Amperometric titration of Kondrakhina el aZ.34 date excess vanadate with iron( IT) sulphate V,O, Titration of cxcess vana- DippeP2 date NaVO, Excess vanadate titrated Jacksond7 with iron(1I) sulphate V,O, Titration of excess vana- Gekht and Putoka3 date with iron (I J) sulphate Mn2O3 Spectrophotometric deter- Nagato,' mination of excess tervalent manganese Phosphato- Excess cerate titrated Cheng'* cerate with iron(I1) sulphatc VzO, V4+ titrated with per- Wickham and WhippIe3B manganate760 SCHAFER: DETERMINATION OF IRON(II) OXIDE IN [AaaZyst, Vol.91 The use of phosphoric acid enables many refractory materials to be brought into solution, but at the temperatures used (up to 380" C) there may be extensive decomposition of the oxidant as well as oxidation of organic matter.Another difficulty that can affect iron(I1) oxide determinations is the presence of sulphides, which may result in consumption of oxidant and thus in high values for the iron(I1) oxide content. A summary of methods involving acid decomposition in the presence of an oxidant is given in Table I. EVOLUTION METHOD Sei149 proposed a completely different approach to the determination of iron( 11) oxide ; he suggested decomposing the finely ground material in phosphoric and sulphuric acids in a stream of carbon dioxide, and passing the resulting vapours through absorption tubes containing a measured amount of standard dichromate solution. According to Seil, sulphur dioxide was evolved by reaction of iron(I1) oxide and the acids. This reduced the dichromate, and the excess of dichromate could then be determined by titration.More recently Tikhomi- rova et aL50 applied this method to the determination of iron(I1) in chromium ores and slags. They thought that the reduction of dichromate was due to absorption of sulphur dioxide and phosphine formed by reaction 01 iron(T1) oxide with the acid mixture. I t is apparent that other substances such as sulphides and organic matter, which could result in the reduction of sulphuric or phosphoric acid with the evolution of the gases mentioned above, must be absent. FUSION METHOD For the decomposition of refractory silicates and the subsequent determination of iron( 11) oxide Rowledge51 investigated the application of a number of fluxes, the most satisfactory being a fluoroborate (NaF),.B,O,.He heated the sample which was mixed with the flux in a sealed glass tube at 900" C, dissolved the melt in dilute sulphuric acid and titrated the iron(I1) with permanganate. He applied the method to refractory silicates such as staurolite, tourmaline, axinite and garnet, and compared the values obtained by decomposition in hydrofluoric and sulphuric acids with those obtained by using the fluoroborate fusion procedure. Because these refractory materials were only partly decomposed by the acid treatment, the iron( 11) oxide contents were determined by weighing the undecomposed material and calculating from the weight of mineral decomposed (obtained by difference). This procedure may be acceptable for homogeneous minerals, but it is unsatisfactory because of doubt about the composition of the undecomposed material.The comparisons made by Rowledge are rather confusing. With tourmaline the results obtained by the two methods agreed ; for garnet the fusion method gave a value of about 1 per cent. higher than that by acid decom- position, and for staurolite and axinite the method gave a value of about 1 per cent. lower. In an attempt to apply the Rowledge fusion method on a micro scale, Hey43 encountered difficulties arising from the oxidation of iron(I1) by air enclosed in the fusion tube, as well as those arising during the slow process of dissolving the melt in sulphuric acid. He overcame the former by conducting the fusion in an evacuated tube, and the latter by dissolving the melt in an excess of iodine monochloride in hydrochloric acid in a stoppered flask and titrating the liberated iodine with potassium iodate.By carrying out the fusion in a platinum boat placed in a silica tube at 950" C through which a current of carbon dioxide was passed, Groves52 excluded the possibility of oxidation. The iron(I1) was then determined by permanganate titration or, better still, by the iodine monochloride method of Hey. Groves regarded the combination of fusion in carbon dioxide, followed by dissolution of the melt in hydrochloric acid containing iodine monochloride, as probably the best method available for determining iron( 11) oxide in refractory materials. More recently Mikhailova et aZ.53 used the method of fluoroborate fusion in an atmosphere of carbon dioxide as developed by Groves for determining both bivalent and tervalent iron in rocks that are difficult to decompose.The solution obtained by dissolving the melt in a mixture of 10 per cent. sulphuric acid and saturated sodium oxalate was subjected to polaro- graphic reduction. The heights of the waves corresponding to Fe2+ and Fe3+ were measured and the content of each form calculated. S~bsequently~~ the fusion was carried out in a platinum crucible into which carbon dioxide was introduced.December, 19661 SILICATE AND REFRACTORY MATERIALS. PART I 761 COMBUSTION METHOD This is based on the oxidation of iron(I1) oxide to iron(II1) oxide when the former is heated in a current of oxygen. In the method originally developed by S h e i r ~ , ~ ~ the sample was heated in a current of dry oxygen or air in an electric furnace maintained at 1000" C.The water and carbon dioxide evolved were collected in an absorption tube and weighed. The oxygen used in the oxidation of iron(I1) oxide was then calculated from the increasein weight of the ignited residue. In a later modification the sample was heated first in pure nitrogen to expel volatile matter; it was then weighed, and finally re-ignited in oxygen. The iron(I1) oxide content could then be calculated from the increase in weight. An investigation of Shein's method as applied to chrome ores by Sasuga and Iida37 with a thermobalance showed that, for samples low in aluminium and magnesium, the increase in weight reached a maximum at 800" to 900" C and that the weight decreased at higher temperatures.They concluded that the temperature at which the weight increase is recorded must be changed according to the sample analysed. De Wet and van Niekerk56 adapted the method on a semi-micro scale for the analysis of iron(I1) oxide in South African chromites. In a study of the rapid analysis of chromite and chrome ores, D i r ~ n i n ~ ~ compared iron(I1) oxide determinations made by the Sliein com- bustion method, the Shein method with vanadium pentoxide as oxidant , and the evolution method of Seil. In a different approach, Habashy5* determined the iron(I1) oxide content of material that is difficult to dissolve, by ignition at 1000" C in a limited, known volume of oxygen. He then calculated the iron(I1) oxide content from the volume of oxygen used in the oxidation of iron(I1) oxide to iron(II1) oxide.He claimed that interferences could be overcome by applying suitable corrections. CONCLUSION The diversity of publications on the determination of iron(I1) oxide in silicate and refractory materials is indicative of the extent of the problem, and it would be unwise to assume that it has been solved. There does not appear to be any one method applicable to all types of materials. All methods suffer from some disadvantages, but perhaps the one most closely approaching the ideal is that of Rowledge, as modified by Groves. However, its application to some materials would amount to the use of unnecessarily severe conditions for sample decomposition that could be readily achieved by some simpler method. The extent of interference arising from the presence of organic matter in a sample would depend on the method of sample decomposition used.I t could be quite serious with decom- position in the presence of an added oxidant, especially if heating were required. With other titrimetric methods the degree of interference would be dependent on the type of organic matter and the oxidising power of the titrant used for the iron(I1). With materials containing sulphides, particularly iron sulphides, errors in the deter- mination of iron(I1) oxide seem unavoidable. Their magnitude is dependent on several factors, such as type and amount of sulphide, the iron(II1) content and the method of sample decomposition. Apart from the likely sources of errors mentioned above, interference with iron( 11) oxide determination could occur because of the presence of oxidising substances such as manganese dioxide, which would lead to low results arising from the oxidation of iron( 11) during sample dissolution.On the other hand, Hillebrand has cited the interference of tervalent vanadium, which would consume the titrant used for iron(I1) and lead to high results. These two sources of error are unlikely to be of common occurrence. However, given an awareness of the problems involved, and a knowledge of the range of methods available for the determination of iron(I1) oxide, it should be possible to select the method best suited to the type and number of samples to be analysed. The author thanks Dr. D. J. Swaine for helpful discussions. REFERENCES 1. 2. 3.___ , Ibid., 1861, 83, 445. 4. 6. 6. Hillebrand, W. F., Lundell, G. E. F., Bright, H. A., and Hoffman, J. I., "Applied Inorganic Mitscherlich, A., J . 9rakt. Chem., 1860, 81, 116. Riley, J. P., Analytica Chim. Ada, 1958, 19, 418. - , and Williams, H. P., Mikrochim. Acta, 1959, 516. Cooke, J. P., Amer. J . Sci., 1867, 44, 347. Analysis," Second Edition, John Wiley and Sons Inc., New York, 1953, p. 907.762 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. SCHAFER Barnebey, 0. L., J . Amer. Chem. SOC., 1915, 37, 1481. Pratt, J. H., Arne:, J . Sci., 1894, 48, 149. Treadwell, F. P., Barnebey, 0. L., J . Amer.Chem. SOC., 1916, 38, 374. Sarver, L., Ibid., 1927, 49, 1472. Schollenberger, C. J,, Ibid., 1931, 53, 88. Soule, B. A., Ibid., 1928, 50, 1691. -, Ibid., 1929, 51, 2117. Stokes, H. W., Bull. U.S. Geol. Surv., 1901, 186. Dittler, E., Z . anorg. allg. Chem., 1926, 158, 264. Densem, N. E., J . SOC. Glass Technol., 1936, 20, 303. Harris, F. R., Analyst, 1950, 75, 496. Smirnov, V., and Aidinyan, N., Dokl. Akad. Nauk SSSR, 1937, 14, 353. Shioiri, M., and Mitsui, S., Mikrochim. Acta, 1938, 3, 291. Das Gupta, J., Sci. Cult., 1941, 6, 677. Meyrowitz, R., Amer. Miner., 1963, 48, 340. Shapiro, L., Prof. Pap. U . S . Geol. Surv., 1960, [400-B], 496. Nicholls, G. D., J . Sedim. Petrol., 1960, 30, 603. Konopicky, K., and Caesar, F., Ber. dt. keram. Ges., 1939, 20, 362. Kleinert, P., and Funkc, h., Z , Chemze., 1962, 2 , 155.Clemency, C. V., and Hagner, A. F., Analyt. Chem., 1961, 33, 888. Malmstadt, H. V., and Roberts, C. B., Ibid., 1956, 28, 1884. Hackl, O., 2. analyt. Chem., 1925, 67, 197. Shein, A. V., Zav. Lab., 1937, 6, 1199. Ralyuk, S. T., and Mirak’yan, V. M., Ibid., 1949, 15, 1004. Dippel, P., Silikattechnik, 1962, 13, 51. Gekht, I. I., and Putok, S. I., Vest Akad. i\Tauk. Kazakh. S.S.R., 1960, 9, 68. Kondrakhina, E. G., Egorova, L. G., and Songina, 0. A., Izv. Akad. Nauk. Kazakh. SSR, 1957,1, 45. Nagaoka, T., and Yamazaki, S., j a p a n -4nalyst, 1954, 3, 408. Samanta, H. B., and Sen, N. B., Traws. Indian Ceram. Soc., 1946, 5, 97. Sasuga, H., and Tida, Y., Japan A4nalyst, 1958, 7, 248. Wickham, D. G., and Whipple, E. R., Talanta, 1963, 10, 314. Goswami, N., Sci. Cult., 1957, 22, 398. Ingamells, C. O., Talanta, 1960, 4, 268. Nagato, H., Japan AnaZyst, 1961, 10, 985. Cheng, I<. L., Analyt. Chem., 1964, 36, 1666. Hey, M. H., Mineralog. Mag., 1941, 26, 116. Ishibashi, I. M., and Kusaka, Y., J . Chem. Sor. Japan, 1950, 71, 160. Wilson, A. D., Bull. Geol. Survey, Gt Brit., 1955, 9, 56. -, Analyst, 1950, 85, 823. Jackson, P. J., J . A p p l . Chem., Lond., 1957, 7, 605. Reichen, Laura E., and Fahey, J . J., Bull. U.S. Geol. Surv., 1962, 1144-B. Seil, G. E., Ind. Engng Chem. Analyt. Edn, 1943, 15, 189. Tikhomirova, 0. F., Strebulaeva, E. N., and Sazonova, 2. V., Sb. Trud. Tsent. Naztchne-Issled. Rowledge, H. P., J . Proc. Roy. Sor. West. Aust., 1934, 20, 165. Groves, A. W., “Silicate Analysis,” Second Edition, George Allen and Cnwin Ltd., London, 1951, Mikhailova, 2. M., Mirskii, R. V., and Yarushkina, A. A., Z h . Analit. Khim., 1963, 18, 856. Mikhailova, 2. M., Yarushkina, A. A,, Mirskii, R. V., and Shil’dkrot, E. A., Trudy Kuibyshev. Shein, A. V., Zav. Lab., 1937, 6, 505. Wet, J . F, de, and h’iekerk, J. N. van, J . Chpm. Metall. M i n . SOC. S . Afr., 1952, 53, 10. Dinnin, J., Bull. U.S. Geol. Surv., 1969, 1084-R. Habashy, M. G., Analyt. Chem., 1962, 34, 1015. Kurzes Lehrbuch analytischen Chemie,” 1913, Volume 2, p. 425. Inst. Chern. Mettall., 1963, 31, 180. pp. 88 and 181. Gos. n’aulhnc.-Issled. Inst. Neft. Prom., 1963, 20, 124. Received October 28th, 1965