298 A d y s t , May, 1968, Vol. 93, $$. 298-305 The Colorimetric Determination of Boron in Carbon and Stainless Steels BY D. BELL AND K. McARTHUR (Colville’s Ltd. Laboratory, Glengarnock Steel Works, Glengarnock, Ayrshive) A method proposed for the colorimetric determination of boron in carbon and alloy steels, and probably also in high-nickel alloys, is described. The procedure has been applied to several B.C.S. boron-containing steels with acceptable reproducibility and accuracy. A METHOD proposed for the colorimetric determination of boron in carbon and alloy steels is described. The colorimetric reagent used is a highly substituted hydroxyanthraquinone, l-hydroxy-4+toluidinoanthraquinone, which is more sensitive to boron than quinalizarin, and has the additional advantage of reacting in sulphuric acid solutions of lower concen- tration.Photo-electric measurement of optical densities is used to ensure maximum precision. The method has been applied to several boron-containing steels, with good results. Repro- ducibility in the boron range up to 0.003 per cent. is satisfactory at ~0.0002 per cent. for carbon steels. The procedure has been extended to include stainless steels. In the Rudolph and Flickingerl method for the colorimetric determination of boron in steel in the 0 to 0@06 per cent. range, a solution of quinalizarin in concentrated sulphuric acid is used as the boron colour reagent. The method has the attraction of great simplicity, because no preliminary separation of boron is required, and iron, in the iron(I1) state, does not interfere but, as it precipitates as sulphate in a very fine form, it obscures the reagent colour change for a lengthy period, and allowing the solution to stand overnight is usually prescribed to ensure satisfactory clarity.This investigation was primarily concerned with shortening the time required for the determination of boron in steel and obtaining better reproducibility by replacing visual assessment of the reagent colour change with photo-electric measurement in a suitable instrument. In the experimentd work that followed, it was found preferable to remove precipitated sulphates before adding the boron reagent, and the procedure adopted enables this to be done in a relatively short time. As the reaction between boron and the reagent used takes only a few minutes to complete, the total time required for the determination was reduced considerably.In the Rudolph and Flickinger procedure the optical density range is only 0.05 for 0.003 per cent. of boron, as measured in 1-cm cells with red filters. The low reagent concentration of 0*0005 per cent. was probably adopted to facilitate visual comparison against standards, as the reagent itself is strongly coloured. The optical density range can be improved by using a higher concentration of quinalizarin, a &fold increase about doubling it, but the resulting disproportionately large increase in optical density caused by excess of reagent prevents this procedure from being used. Trinder2 showed that 1-hydroxy-4-fi-toluidinoanthraquinone (HPTA) was superior to quinalizarin in sensitivity to boron and had the further advantage of reacting with boron in sulphuric acid solutions down to about 75 per cent.As the HPTA reagent reacts with boron in cold acidic solution, it appeared to have considerable advantages over quinalizarin, and was used in all subsequent experiments. The use of this anthraquinone derivative as a reagent for boron was first noted by RadleyS and investigated by Trinder.2 It is soluble in sulphuric acid solutions above about 65 per cent. v/v, giving a green solution that changes to blue in the presence of boron. The green colour deepens considerably with increasing sulphuric acid concentration, but boron sensitivity reaches a maximum at about 83 to 84 per cent. This is shown in Table I for a reagent concentration of 0-0025 per cent., which was the final concentration adopted for use 0 SAC and the authors.BELL AND MCARTHUR 299 in the proposed procedure.A Hilger photo-electric colorirneter, with tungsten lamp, 1 and 2-cm cells, heat-absorbing and Chance OY2 filters, was used throughout the investigation to measure optical densities. TABLE I EFFECT OF INCREASING SULPHURIC ACID CONCENTRATION ON OPTICAL DENSITY MEASUREMENTS Sulphuric acid concentration, Optical density of reagent Optical density of boron addi- per cent. v/v . . .. 65 70 76 80 82.6 86 87.6 90 95 (l-cm cell) . . .. . . 0.09 0.10 0-10 0.135 0.207 0.317 0.440 0.626 0.686 tion (0403per cent.) . . Nil 0.002 0.066 0.122 0.187 0.182 0.161 0-100 0-061 EXPERIMENTAL Because of the rapid increase in the optical density of the HPTA reagent with increasing acidity, small differences in the sulphuric acid content of ordinary supplies of concentrated sulphuric acid can result in appreciable variations between different batches of reagents prepared at different times, but when each series of tests and standards is processed with the same batch of reagents the accuracy of the boron determination is not affected.As Table I indicates, the change in optical density of the reagent between 82 and 85 per cent. acidity is about 0.09 for each 1 per cent. difference (with 2-cm cells and OY2 filters), while boron sensitivity is practically unaltered. As an example of the variations that can arise, several bottles of laboratory-grade concentrated sulphuric acid, taken at random, showed the following optical density blanks when added to aliquots of boron-free iron solution as pre- scribed in the adopted procedure: 0-603, 0.622, 0.402, 0-269, 0.344 and 0.454.As excessively high blanks are undesirable, selected batches of concentrated sulphuric acid can be advantageously reserved for use in the boron determination. Boron colour development, in the rather low concentration of less than 0.01 mg in a sample aliquot, is rapid and reaches a maximum in a few minutes, remaining stable thereafter for several hours, as shown in Table I1 (with 2-cm cells and OY2 filters). TABLE I1 RATE OF COLOUR DEVELOPMENT Optical densities for standing times of- r---- - -3 2 6 10 16 30 1 2 3 minutes minutes minutes minutes minutes hour hours hours Boron-free blank .. 0.424 0.423 0.420 0.418 0.418 0.421 0.419 0.422 With 0.003 per cent. 'Af boron added . . 0-751 0.766 0.764 0.763 0.765 0.766 0.763 0.768 Optical density bi boroh'addition 0.327 0.343 0.344 0.346 0.347 0.346 0.344 0.346 In the rather highly concentrated sulphuric acid medium required to keep certain anthraquinone derivatives in solution, many metal sulphates, especially those of iron(I1) and iron(III), are relatively insoluble and largely precipitate on raising the acidity to the prescribed level, but complete sedimentation may be rather prolonged. In the procedure adopted for use with the HPTA reagent, however, the precipitated sulphates separate fairly quickly and are removed before applying the boron colour test. As considerable heat is liberated when adding concentrated sulphuric acid to the sample aliquot, a protective glove should be worn during handling.Contraction on cooling should be allowed for. By the time the solution has cooled down to room temperature a slight cloudiness usually remains, but this can be removed completely by transferring some of the cloddy solution into a dry test-tube, warming it until the sulphates in suspension re-dissolve, then allowing it to cool again. Re-precipitation does not occur in the time required to complete the remainder of the procedure. To determine if any significant amount of boron was absorbed in the iron(I1) sulphate precipitate, the precipitates recovered from a boron calibration test were washed several times by stirring them in concentrated sulphuric acid and decanting off after settling; they300 BELL AND MCARTHUR : COLORIMETRIC DETERMINATION [~4,rtabSt, VOl.93 were then dissolved in 20 per cent. sulphuric acid and tested for boron by the proposed procedure. With the boron-free test as the standard of reference, the results given below show that none of the precipitates appears to have absorbed any measurable amount of boron. Optical densities were measured with 2-cm cells and OY2 filters. Original boron addition, per cent. . . Nil 0.001 0.002 0.003 0.004 0.006 0.006 Optical densities of sulphate precipitates 0.501 0.502 0.502 0.503 0602 0.603 0.603 The boron colour is developed by mixing 10 ml of the clear sample solution, after removal of the precipitated sulphates, with 10 ml of HPTA reagent in a stoppered test-tube or other suitable vessel, then measuring the optical density due to boron, as described in the procedure.A representative boron calibration, in which the Hilger photo-electric colorimeter was used, with 2-cm cells and OY2 filters, is given in Table 111. TABLE I11 CALIBRATION Optical densities r ~p Boron added, per cent. . . Nil O*OOP 0.002 0.003 0.004p’ As found .. .. . . 0-410 0.542 0.672 0-802 0.913 Less nil-boron blank . . - 0.132 0.262 0.392 0.503 The results given in Table IV show the reproducibility and accuracy of the proposed method, as applied to several B.C.S. standard steels. TABLE IV REPRODUCIBILITY OF RESULTS OBTAINED ON B.C.S. STANDARD STEELS AND COMPARISON WITH CERTIFICATE VALUES B.C.S. No. 273 277 326 327 328 329 330 Number of tests 14 9 23 26 23 6 6 Boron found, per cent.r 7 Range Average A 0.0024 to 0.0026 0.0025 0-0002 to 0.0004 0*0003 0*00106 to 0.0013 0.001 18 0*0029b to 0.0033 0.00306 0*00396 to 0.0042 0*00406 0.0081 to 0.0083 0-0082 0.0071 to 0.0073 0.0072 Boron, per cent. certificate value 0.0025 < 0.001 0.001 0.003 0.004 0.008 0.007 The above British Chemical Standard steels, which were supplied by the Bureau of Analysed Samples, Middlesbrough, are of the low alloy type, and have been accurately standardised for several elements, including boron. INTERFERENCES- None of the common residual elements, in the low concentrations normally occurring in carbon and low alloy steels, adversely affects the boron - HPTA colour reaction. Titanium interferes when present in appreciable amounts by altering the colour of the reagent, but this interference is negligible in samples containing less than 1 per cent.Elements that have coloured ions and are soluble in concentrated sulphuric acid solution, such as chromium, molybdenum and vanadium, increase the blank proportionately when present and would require appropriate compensation. Oxidising agents and nitrates, which either destroy the reagent or change its properties, should be absent. Reducing agents such as sulphites or iron(I1) sulphate do not interfere. Iron(I1) sulphate mostly precipitates in a dense form that settles out satisfactorily. Iron(II1) sulphate, on the other hand, separates in a voluminous form under the same conditions, so that only the minimum amount of iron(II1) should be present.The proposed method has been found to be applicable, even when considerable amounts of the more common alloying elements, in addition to iron, are present, provided compensation can be made for any increase in optical density by these elements. This is shown in Table V, all of the elements indicated being added in metallic form.May, 19681 OF BORON IN CARBON AND STAINLESS STEELS 301 TABLE V NET EFFECT OF LIMITED AMOUNTS OF CERTAIN ALLOYING ELEMENTS ON OPTICAL DENSITIES Optical densities Iron only + + + + + + + + + .. .. i.i,ofCi .. . . 20% of Cr . . .. 8yoofNi .. .. 0-5y0 of Ti . . .. 1% of Ti . . .. 1%of Mo . . .. 1%of v .. . . 1%of Nb .. .. l%of cu . . .. r Total 0-195 0.216 0.594 0.234 0.206 0.217 0.214 0.198 0-216 0.195 I Boron absent 1 Increase caused by added element - 0.021 0.399 0-039 0.01 1 0.022 0.019 0-003 0.02 1 Nil With 0.003% of boron addeci r Total 0.353 0.367 0-752 0-392 0-364 0-377 0.369 0.355 0-377 0.352 Net increase caused by added boron 0.168 0.151 0.158 0.158 0.158 0.160 0.155 0.167 0.161 0.157 Before the proposed method could be applied to alloy steels of unknown composition, it was necessary to find some way of obtaining a compensating blank, so that the net optical density due to the boron present could be obtained. Attempts to remove boron from the sample solution were either unsuccessful or inconvenient, but it was found that the presence of fluoride inhibited the boron colour reaction without affecting the optical density of the alloy solution.The addition of a soluble fluoride, before adding the HPTA reagent, therefore, allows a compensating optical density for the alloy components present to be obtained.Any pick-up of boron from the action of fluoride on the borosilicate glassware used is also inhibited. The optical density of the reagent is slightly altered by the fluoride addition and this varies to some extent between different batches of reagent but, as it can also be compensated for, the accuracy of the boron determination is not affected. The procedure adopted for obtaining a compensating blank for alloy samples is as follows. A 15-ml aliquot of the dissolved sample, after the sulphate precipitate has been removed, is transferred into a 50-ml test-tube containing 0-05g of sodium fluoride and heated, with occasional shaking, until. the sodium fluoride is dissolved.When cold, 10 ml of this fluoride- treated solution is mixed with 10ml of HPTA reagent solution, and the optical density measured under the same conditions as adopted for the boron colour test. The action of fluoride on the reagent is found by treating some of the nil-boron standard with sodium fluoride in a similar manner, and the difference in optical density before and after fluoride treatment applied as a correction to the compensating optical density. The application of this procedure to a stainless-steel sample is shown in Table VI, optical densities being measured in l-cm cells with OY2 filters. TABLE VI COMPARISON OF BORON CALIBRATIONS MADE IN PURE IRON AND IN STAINLESS-STEEL SOLUTIONS AFTER SODIUM FLUORIDE COMPENSATION Optical densities I h Boron added, yo .. . . No fluoride present . . . . Optical density due to boron addition . . .. .. After fluoride treatment . . Change in optical density of reagent after fluoride treat- ment .. .. . . Optical density due to boron after applying compensat- ing blank and reagent cor- rection . . .. .. Pure iron Stainless steei f A > r A -l Nil 0.001 0.002 0.003 Nil 0.001 0.002 0.003 0.230 0.300 0.368 0.436 0.645 0.714 0.784 0-854 - 0.070 0.138 0.206 - 0.069 0.139 0.209 0.276 0.276 0.277 0.277 0.690 0-692 0.691 0-692 + 0.046 + 0.045 - 0.070 0.137 0.206 - 0.067 0.138 0.207302 BELL AND MCARTHUR: COLORIMETRIC DETERMINATION [AutabSt, Vol. 93 The compensating optical density obtained for the stainless-steel sample in Table VI is 0.046 high because of the action of fluoride on the reagent but, as the same increase also occurs in the nil-boron standard simultaneously treated with fluoride, this correction is easily made.High-chromium steels are not readily soluble in dilute sulphuric acid and, in addition, the iron(I1) sulphate precipitate settles out very slowly so that the procedure, as applied to to these alloys, takes much longer to complete. High-nickel heat-resisting alloys may also contain boron and were included in the investigation, but as no actual samples were available synthetic metallic mixtures of equivalent composition were used instead. In these materials it was found that nickel sulphate, when in high concentration , does not separate satisfactorily, post-precipitation occurring on warming during fluoride treatment but, if when making the concentrated sulphuric acid addition to the sample aliquot the temperature is raised to between 120" and 130" C after half of the acid has been added, this difficulty is overcome.Alloys of this type are best processed on half-sample weight. In both high-chromium and high-nickel materials sulphate separation is rather slow, but filtration through a dry glass-fibre filter, after most of the precipitate has settled out, has been found effective in giving a clear solution in a reasonably short time. Table VII gives comparative calibrations for boron additions to pure iron, two B.C.S. stainless steels, two synthetic stainless steels and two synthetic high-nickel alloys. All of the tests were processed together with the same batch of reagents; optical densities were measured in l-cm cells with OY2 filters.TABLE VII COMPARATIVE BORON CALIBRATIONS IN A WIDE RANGE OF ALLOY COMPOSITIONS AFTER SODIUM FLUORIDE COMPENSATION Optical densities t \,, r L I Nil-boron blanks Boron added, yo A After fluoride Composition Direct* treatment? Difference# 0.001 0.002 0-003 B.C.S. No. 149/1 . . .. . . 0.192 0.230 +0.038 0.061 0.123 0.181 Cr, 18% + Ni, 8% + Fe, 74% . . 0.673 0.612 +0*039 0.068 0.116 0.176 Cr, 18% + Ni, 8% + Mo, 2.6% + Cu, 2% + Ti, 1% + Fe, 68.6%$ . . 0.661 0.693 +0*042 0.062 0.123 0.181 B.C.S. No. 236/2 (Cr, 18% + Ni, 8% + Ti, 3%) . . .. .. . . 0,635 0.673 +0*038 0.062 0.122 0.184 B.C.S. No. 261 (Cr, 17% + Ni, 13% + Nb, o-7y0) . . .. .. . . 0.718 0.759 +0*039 0-063 0-123 0.179 Ni, 76% + Cr, 20% + Ti, 2.6% (3 Ni- monic SOA)$II .. .. .. . . 0.382 0.425 +0*043 0.062 0.120 0.182 Ni, 66% + Cr, 20% + CQ, 20% + Ti, 2.6% (Nimonic 9O)JII. . .. . . 0.423 0.463 +0-040 0.058 0.118 0.178 Average boron calibration . . .. - - - 0.061 0.121 0.180 Average effect of fluoride on reagent . . - - +0.040 - - - * As obtained in the boron-free tests. ?As obtained in the sodium fluoride treated tests. $ Increase in optical density of reagent in the presence of fluoride. f All components added in metallic form. 11 Processed on half of the standard weight because of the high nickel and titanium contents, but received the same boron addition as the other tests. The boron calibrations obtained after applying the net compensating blanks are prac- tically the same in all of the above tests and confirm that the alloying constituents in the proportions indicated do not materially affect the determination of boron in complex nickel- chrome alloys by the proposed method.It is also apparent that when fluoride is used to obtain a compensating blank a boron reference calibration prepared from boron additions to pure iron is applicable to both carbon and alloy steels.May, 19681 OF BORON IN CARBON AND STAINLESS STEELS 303 As the four stainless-steel compositions quoted in Table VII show the same boron calibration, whether or not the component elements have been fused together, it is quite probable that the simulated Nimonic compositions will be the equivalent of the commercial alloys as far as the application of the proposed boron procedure is concerned.METHOD FOR THE COLORIMETRIC DETERMINATION OF BORON IN CARBON AND STAINLESS STEELS PRINCIPLE- After dissolution of the metal in dilute sulphuric acid, the acid concentration is increased to the prescribed level, and the precipitated sulphates are allowed to settle out. An aliquot of the cold clear solution is then mixed with an equal volume of the boron reagent solution and the optical density measured in a suitable photo-electric colorimeter. For alloy steels a compensating blank is obtained by treating a second aliquot with sodium fluoride to inhibit the boron colour reaction. RANGE- The range is 0 to 0*004 per cent. on the full sample weight. REPRODUCIBILITY- of +04002 per cent. APPLICATION- The method is applicable to all carbon steels and low alloy steels containing less than 1 per cent.of titanium. An extension of the method is applicable to highly alloyed steels of the stainless type and possibly also to heat-resisting high-nickel alloys. APPARATUS- any material absorption of boron occurring. 26-mm diameter test-tubes. A linear calibration graph is obtained up to 0-003 per cent., with a reproducibility Use preferably boron-free glassware, although borosilicate glassware can be used without Test-tubes-150 x 25-mm diameter nominal, marked at 10 and 15 ml, are suitable. Air condenser-About 60-cm lengths of glass tubing, with rubber stoppers to fit the Calibrated JEasks4O-ml capacity, with polythene stoppers. Specimen tubes-30-ml capacity, with leak-proof polythene caps, or any other suitable type of stoppered vessel.Photo-electric colorimeter-The Hilger “Spekker” instrument, with tungsten lamp, heat- resisting filters, 1 and 2-cm cells and Chance OY2 filters, or instruments of similar type, can be used. REAGENTS- Sulphuric acid solutions-2, 20, 50, 75 and 85 per cent. v/v and concentrated. Boron colour reagent-Prepare a 0@05 per cent. w/v solution of l-hydroxy+-toluidino- anthraquinone (HPTA) in 85 per cent. v/v sulphuric acid solution for the total-boron pro- cedure and in 75 per cent. v/v sulphuric acid for the acid-soluble boron procedure. Dissolve it in the cold, with shaking, and prepare freshly each day. Use the same batch of reagents with each group of tests and standards. It is made by several manufacturers; that made by I.C.I. Ltd., is sold under the trade name Waxoline purple A.l-Hydroxy-4-~-toluidinoanthraquinone is Solvent Violet 13 (Colour Index 60725). Ammonium persulphate solution-Prepare a 25 per cent. solution in distilled water. Iron(II) sulphate, analytical-reagent grade. Boron-free iron, B.C.S. 149/1 pure iron granules. Sodium carbonate, analytical-reagent grade, anhydrous. Sodium fluoride, analytical-reagent grade. Boric acid, analytical-reagent grade. Standard boron solutim A-Dissolve 0.0572g of boric acid in 1OOml of 20 per cent. sulphuric acid.304 BELL AND MCARTHUR : COLORIMETRIC DETERMINATION [Analyst, Vol. 93 Standard boron solution B-Dilute 10ml of standard boron solution A to 100ml with 20 per cent. sulphuric acid. 1 ml of solution contains 0~0o0010 g of boron. 1-5 ml of solution contains 0*001 per cent.on a 1.5-g sample weight. PROCEDURE FOR THE DETERMINATION OF TOTAL BORON CARBON AND LOW ALLOY STEELS- Preparation of sample solution-Transfer 1.5 g of drillings into a marked test-tube, add 13.5 ml of 20 per cent. sulphuric acid from a burette, fix the air condenser in position, then heat the tube gently until the drillings are dissolved. When dissolution is complete, add 0.25 ml of 25 per cent. ammonium persulphate solution to oxidise most of the carbonaceous matter, boil for a few minutes, then allow the solution to cool. Wash down the condenser with about 0 4 m l of distilled water, remove it from the test-tube, then make the volume of the solution up to the 15-ml mark and mix thoroughly. Filter through a dry No. 40 Whatman or similar filter-paper and collect it in a clry test-tube.Transfer a 10-ml aliquot to a 50-ml calibrated flask before iron(I1) salts begin to crystallise out. Now transfer all of the acid-insoluble matter on to the filter, wash it with 2 per cent. sulphuric acid, until free from iron salts, then with water until acid-free. Place 0.5g of sodium carbonate in the filter, transfer it into a platinum crucible, dry, ignite and fuse for 5 minutes. Dissolve the cold melt in 15 ml of 50 per cent. sulphuric acid containing 0.1 g of iron(I1) sulphate in solution to reduce chromates, etc., if present. Transfer 10 ml of this solution into the 50-ml calibrated flask containing the 10-ml aliquot of the main filtrate. Precipitation of iron(II) and other fairly insoluble sulphates-To the combined aliquots of the soluble and insoluble parts of the sample, representing 1-g sample weight, add concen- trated sulphuric acid, with occasional swirling, to the 50-ml mark, then add a further 2.5 to 3-0ml from a burette, depending on the temperature attained, to allow for contraction on cooling.Set aside to cool, by which time the precipitate should have settled out and the volume returned to the 50-ml mark. Decant off about 15 ml of the almost clear solution into a dry test-tube and warm it to re-dissolve any sulphate haze that may be present, then allow it to cool. Development of the boron colow-Transfer 10 ml of the clear sample solution into a 30-ml specimen tube or bottle, add 10 ml of the 0.005 per cent. HPTA reagent, replace the stopper or cap, mix the solution and allow it to stand for a few minutes for full colour development.Measure the optical density under standard conditions. Convert the optical densities thus obtained into percentages of boron by using a calibration graph prepared simultaneously with the tests. Preparation of a refeyence graph-Place 1.5 g of pure iron into five marked test-tubes, add in sequence, 0, 1.5, 3.0, 4.5 and 6-0ml of standard boron solution B, then make each solution up to exactly 13.5 ml with 20 per cent. sulphuric acid. These standards represent nil, 0-001, 0.002, 0.003 and 0.004 per cent. of boron, respectively. Carry this series through the prescribed procedure, together with the unknown sample. Measure the optical densities against water, with 2-cm cells and OY2 filters.Plot the optical densities obtained, after subtracting the boron-free blank reading, against the respective boron contents. The calibration graph is linear up to 0.003 per cent. For higher boron contents use a smaller sample weight and make up to 1*5g with pure iron. Stopper the flask and mix its contents thoroughly. HIGHLY ALLOYED STEELS- Use the same procedure as prescribed for carbon steels when the boron content is not more than 0.004 per cent. and the titanium content not more than 1 per cent. For higher boron and titanium contents use a reduced sample weight and make it up to 1.5 g with pure iron. Pass the sample solution, after most of the sulphate precipitate has settled out, through a dry glass-fibre filter to obtain a clear solution.To develop the boron colour add 10ml of HPTA reagent to a 10-ml aliquot of the prepared sample solution in a stoppered, glass test-tube or bottle, mix the solution and allow it to stand for a few minutes, then measure the optical density with 1-cm cells and OY2 filters.May, 19681 OF BORON IN CARBON AND STAINLESS STEELS 305 To obtain a compensating blank add 0.05 g of sodium fluoride to a 15-ml aliquot of the prepared sample solution, warm and shake it vigorously to assist dissolution, then allow to cool. Add 10ml of HPTA reagent to 10ml of this sodium fluoride treated solution, mix and measure the optical density, as for the boron colour. To obtain a correction for the action of sodium fluoride on the reagent, treat a 15-ml aliquot of the nil-boron calibration standard with sodium fluoride in exactly the same way, and simultaneously with the com- pensating blank, and apply the difference in optical density of the nil-boron standard, before and after fluoride 'treatment, as a correction to the compensating blank. Prepare a boron reference calibration graph in the same manner as for carbon steels. Refer the net optical density obtained, after deducting the corrected compensating blank, to the boron reference graph to obtain the amount of boron present in the alloy steel. HIGH-NICKEL HEAT-RESISTING ALLOYS (PROVISIONAL)- Treat as for highly alloyed steels, but use not more than half-sample weight because of the high nickel content. Make up to 1.5 g total weight with pure iron. When making the concentrated sulphuric acid addition, heat to between 120" and 130" C after half of the acid has been added to obtain a satisfactory separation of nickel sulphate. Continue thereafter as for highly alloyed steels. PROCEDURE FOR THE DETERMINATION OF ACID-SOLUBLE BORON- Use the same general procedure as described for carbon or highly alloyed steels, but discard the insoluble residue and omit the 10-ml addition of 50 per cent. sulphuric acid prescribed for dissolving the fused residue. Use a solution of the HPTA reagent in 75 per cent. sulphuric acid to balance the consequently higher acidity of the sample solution. REFERENCES 1. Rudolph, G. A., and Flickinger, L. C., A~zalyst, 1943, 68, 384. 2. Trinder, N., Ibid., 1948, 73, 494. 3. Radley, J. A., Ibid., 1944, 69, 47. First received June 24th, 1965 Amended May 9th, 1967