A rapid and accurate method was developed for the determination of silicon by diffusion method. The recommended procedure is presented as follows.
To a weighed sample (Si :075μg) in platinum crucible, minimum of nitric (1+1) or hydrochloric acid (1+1) is added, followed by 1.0 ml of sulfuric (1+1) or perchloric acid. After the reaction and decomposition is accomplished, evaporate until white fumes appear. The crucible is cooled to room temperature, and 1.0 ml of water is added, followed by 0.6 ml of 25% hydrofluoric acid. The crucible is immediately covered with a platinum lid coated with 1.0 ml of 70% sodium hydroxide solution and kept at 110120°C for 90 minutes. During this period, silicon is volatilized as the tetrafluoride from a perchloric or sulfuric acid medium and absorbed in the sodium hydroxide on the inner surface of the lid. The sodium hydroxide solution which has absorbed the silicon is dissolved in 40 ml of 4% boric acid solution and neutralized with sulfuric acid (1+3) using pH test paper and further 1.0 ml of sulfuric acid (1+3) is added. The solution is diluted to about 50 ml with water, and 5.0 mlof 8% ammonium molybdate solution is added, and keep it about 25°C for 10 minutes. To the resulting solution, 3 ml of 20% tartaric acid solution is added, the solution is mixed, and 3 ml of ANS reducing solution is added. The solution is diluted to 100 ml with water. After the molybdenum blue color is developed, the absorbency is measured at 815 nm using 1-cm cells. After deducting the absorbency of a corresponding blank, the silicon content is calculated from the standard calibration curve.
The recovery of silicon was about 95% by this method and the relative errors stood within 4%. The present method was applied to the determination of 0.0080.17% silicon in aluminum and steels, and the analytical results obtained agreed with the standard values within experimental error. The elimination of most interfering ions in the analysis of silicon was accomplished by this procedure.

View full abstract

In Cu-Zn solutions, the intensity ratio (I_Cukα/ I_Znkα) against the concentration ratio (C_Cu/C_Zn) experimentally observed is linear in the range of the Cu-Zn composition of normal brasses. Furthermore, this straight line was almost unaffected either by iron, tin, lead, and manganese which had been added as the third constituents or by nitric and/or phosphoric acid used. Furthermore, the physical conditions such as the surface of the miler film upon the liquid sample cell also had no influence on this line. This linear correlation was utilized for the direct determination of copper and zinc in brasses. Samples, dissolved in nitric acid or an admixture of nitric and phosphoric acid were used. The concentration ratio of the sample solution α, which obtained experimentally is written by :
α=x/y …(1)
Similarly, the concentration ratio, β, of another sample which was prepared by an addition of a known quantity of zinc to the same solution, is written by:
β=W·x/100/ W·y/100+ ω…(2)
where W is the weight of sample, ω is the weight of zinc added, x and y are percentages of copper and zinc, respectively, in the sample. From equations (1) and (2), we can calculated x and y. The results obtained by this method for practical samples such as high strength brass, etc. were in agreement with the results of chemical analysis within an error of less than 0.5 per cent. This method may be applied to other copper alloys. However, in the case of the samples containing more than approximately 0.5 per cent of nickel, this method was not successful because the intensity ratio varied with the change of the amount of nickel.

View full abstract

In fire assaying of Au and Ag in ore, we certified the extents of errors caused by some coexistent elements. Samples were treated by standard crusible fusion and the lead buttons obtained were cupelled and parted ordinarily. Results are as follows.
(1) Suitable weight of lead button is 2535 g. Smaller button results in smaller recovery of Au and Ag.
(2) The influence of Cu content in lead button less than 2 g is negligible. But the existence of Cu more than 3 g makes the cupellation impossible. Metallic Cu in sample melts mostly into lead button.
(3) The content of Zn in sample increases the slagloss. Zn is practically not contained in lead button.
(4) The considerable portion of Ni content is absorbed in lead button. The content of Ni in lead button less than 0.025 g causes negligible error, but that more than 0.03 g makes the cupellation impossible.
(5) Te shows rather complex influences. The most part of Te remains in lead button. The existence of Te in lead button increases the cupellation-loss and that more than 1 g makes the formation of bead impossible. On the other hand, some of Te remains in bead when the content is more than 0.2 g.
(6) Se similarly decreases the recovery of Au and Ag and the content more than 5 g makes impossible the formation of bead. But its influence less than 0.5 g is negligible.
(7) The Bi content in sample less than 0.2 g gives negligible influence, but that more than 0.2 g increases cupellation-loss.
(8) The content of As and Sb in lead button less than 0.5 g does not cause any error.

View full abstract

X-Ray fluorescene spectrometry was applied to the determination of mercury in brine purification mud. An internal standard method was used, because this mud contains diverse elements. Bismuth trioxide was selected as an internal standard. The X-ray intensities of Bi(L_α1) and Hg(L_α1) were measured for Bi and Hg respectively, and the background intensity was measured at 1194Å.
In the case of measurement by W-target tube, the reproducibility of the HgL_α1 intensity was poor. It may be attributed to the overlapping of HgL_α1 (1239 Å) over WL_β2 (1245Å). When the spetra of mercury by Mo-target tube and Cr-target tube were compared, the spectra by Mo-target was about three fold as intense as the one by Cr-target tube. For all following measurements therefore the Mo-target X-ray tube was used.
The mud may contains mercury in several forms: mercuric oxide, mercurous chloride, mercuric chloride, and metal. The mercury contents in standards were made 0.22% by the addition of these forms of mercury in calcium carbonate which is a main component of the mud. When these prepared samples were grinded and mixed in a vibration mill for 15 min, the reproducibility of the Hg(L_α1) intensity was improved remarkably. This result may be due to that the distrubution of mercury was made more uniform by this grinding and mixing procedure. The grain size of about 80 wt. % of the mud sample became less than 74 μ ( = 200 mesh) under these grinding and mixing conditions, the reproducibility of the Hg(L_α1) intensity was satisfactory similar to the prepared samples.
Known amounts of mercuric oxide were added to calcium carbonate with bismuth trioxide (100 mg) and starch (1 g) as a binder, and the total weight was made 10 g. After the mixture had been grinded and mixed, it was prepared into briquettes under pressure of 25t/40 mmφ. The intensity ratio I_HgLα1/I_BiLα1 were measured for this known samples. The results were plotted to the contents of Hg. A linear relationship between I_HgLα1/I_BiLα1 and Hg content was obtained.
The matrix effect of the following elements on the mercury determination could be neglected by using this internal standard method; NaCl(10%), Sr(NO₃) ₂(10%), Fe₂O₃(5%), MgO(5%), CuSO₄(0.5%), CrO₃ (0.1%), and PbO₂(0.1%).
The coefficients of variation were 5.0, 2.7, and 1.0% for the mud samples containing 0.06, 0.22, 0.78% mercury respectively. The results obtained by this method were almost identical to those of the chemical analysis. The time for an analysis including that for the preparation of a sample was less than 30 min.

View full abstract

The author had already pointed out that the mixed packing gas chromatography has the advantage of the ease of the preparation of the stationary phase suitable to separate the solutes considering their retention values. However, the quantitatively sufficient separation of two or more solutes obviously depends on the column efficiency as well as on the difference between their retention values. In this paper, therefore, the column efficiency in the mixed packing gas chromatography was evaluated by means of HETP for the typical solutes on the mixed columns of squalane (SQ) coated-support/polyethylene glycol 600 (PEG) coated-support and also the reducing effect of tailing of the mixed packing was investigated. HMDS-treated Chromosorb W or acid-treated Chromosorb P was used as the solid support and various mixed packing columns were prepared by combining two types of packing with different coating ratio of liquid, with different kind of solid support, or with different grain size in the same way as was reported previously.
It was recognized from the measured HETP for the typical solutes in the mixed packing columns that the column efficiencies, except with the column packed with the mixture of the packings with different kind of solid supports, were generally intermediate between the single liquid phase gas chromatography with nonpolar liquid and with polar liquid, or the depreciation of the efficiency due to the mixing of two types of the packing did not occur. H-u curves obeyed well the the van Deemter's equation and the change of HETP with the liquid composition depended on mainly the variation of the partition ratio of the solute with the composition, but there was not a great change of HETP with the liquid composition in the medium region of the composition. On the other hand, in the case of the combination of the different kind of solid supports, the maximum in HETP-composition curve for nonpolar solute was shown. Perhaps this phenomenon may depend on the difference of the surface structure or the physical property of the support.
When a small portion of the polar liquid (PEG) coated-support was added to the non-polar liquid (SQ) coated-support, the tailing of the polar solute was remarkably reduced. From the fact that the reducing effect of tailing remained still on the nonpolar liquid coated-support sifted from the mixed packing and the solute was retained more strongly than on the original single liquid packing, it became clear that the tailing was reduced by covering the absorption sites existed on the non-polar liquid coated -support with the polar liquid which had transferred from the polar liquid coated-support to non-polar liquid coated-support as these supports came in touch with each other in the process of mixing.

View full abstract

Spectrophotometric determination of mercury by using Ruhemann-purple was studied. Mercury-Rehemann-purple red complex was extracted into benzene in the pH range 47, but Ruhemann-purple (RP) was not above pH 6. Different from dithiozonate extraction system, this reagent was soluble in water though not soluble in organic solvents, and so reagent blank was negligible. Extracting procedure was carried out by adding mercury solution, 5 ml of 10^-4 M RP solution, 2 ml of 0.5M acetate buffer solution (pH 6) and water up to 10 ml of aqueous phase, and then 10 ml of benzene was added. Absorbance of extract was measured after dehydration by anhydrous sodium sulfate in a test tube. Reduction of color intensity was observed when the extract was contacted with anhydrous sodium sulfate for a while, and thus the dehydration process in a test tube was carried out within a minimum time interval. The complex was extracted into benzene by shaking for 5 min and the extract was stable for 2 hrs. The complex had the absorption maximum at 530 nm in benzene, and the composition of the extract was confirmed to be Hg : RP=1 : 2 by the continuous variation method and the mole ratio method. The composition was not changed when either RP or mercury was excessive. About 98% of mercury in the aqueous phase was extracted into organic phase by a single extraction. Beer's law was held over the range of 030 μg Hg/ 10 ml in the aqueous phase. Apparent molar absorption coefficient of the complex was 6.4 × 10⁴, and this was almost comparable with that of dithizonate. Not like as complexes of organo-mercury compound (such as ethylmercury and phenylmercury), mercury -RP complex was not extracted into ethylether. Cadmium, zinc, nickel and copper ions interfered the determination of mercury by increasing the absorbance but anions such as iodide and bromide ions, which form stable mercury complex, depressed the absorbance. A large amount of sodium ion precipitated RP as sodium salt and interfered the determination. Stability of RP in aqueous solution was also studied and ca. pH 7.5 was found to be most suitable for the storage of RP solution.

View full abstract

Analytical results of titanium obtained by redox titration method using an aluminum reduction show an under estimation when a platinum dish or crucible is used for the sample preparation. Especially, the decrease of the analytical results is very large when samples contain alkali or alkaline-earth elements. As the result of an investigation, it is found that a reduction of titanium(IV) by aluminum is strongly inhibited by a small amount of platinum even when the concentration is 0.03 ppm and the platinum contamination is estimated from the decreasing rate of the analytical results to be about 10 to 40 ppm. This report describes the platinum interference and its elimination.
The decrease of reducing rate of titanium (IV) is due to the fact that platinum deposited by aluminum is in contact with an aluminum fragment and the dissolution of aluminum by hydrogen ion is accelerated by this platinum because the hydrogen voltage of platinum is higher than that of aluminum. An addition of lead ion before the aluminum reduction is effective to eliminate the platinum interference even when no prior separation of platinum has been made. The reason is that platinum particles deposited by aluminum reduction is coated with deposited lead which has a hydrogen over voltage larger than that of platinum. An addition of copper or tin ion is also effective. However, if a large amount of copper is added, the end point of the titration becomes indistinct due to the color of copper ion and a skill is required in determining the end point of the titration when tin is used because tin (II) ion reacts slowly with the iron (III) of the titrant. Thus an addition of lead ion is recommended and an addition of 0.5 mg of lead ion is enough in order to prevent the interference by ordinary platinum contamination.
Procedure : Take a sample solution containing about 200 mg of TiO₂ into a 500 ml Erlenmeyer flask and add lead solution containing 0.5 mg of lead. Adjust the volume of the sample solution to 130 ml and the acidity to 3.69 N HCI. Boil for five minutes on a hot plate. Remove the flask from the hot plate and add 1.7 g of pure aluminum foil to the flask. Put a rubber cork with a delivery tube and insert the delivery tube in saturated sodium bicarbonate solution in a 300 ml conical beaker. Allow the reaction to proceed until the reaction ceases. Cool the sample solution to room temperature. Remove the rubber cork, add 2 ml of saturated ammonium thiocyanate indicator solution immediately and titrate reduce titanium (III) with 0.1 N standard ferric ammonium sulfate solution to a light straw color end point.

View full abstract

A rapid and simple complexometric determination of fluorine in organo-fluorine compounds is described. The sample was ignited by Schöniger method using a quartz flask containing 5 ml of 0.01 M lanthanum chloride as an absorption liquid. The lanthanum chloride solution was prepared by an addition of 3 ml of conc. hydrochloric acid to about 1.63 g of lanthanum oxide which had been moistened by a small amount of water and a subsequent dilution to 1 l with water. One ml of this solution having its pH at 2.22.5 was equivalent to 570 μg of fluorine.
After combustion of the sample, the flask was stood for 10 min in order to absorb the combustion gas in the liquid and the flask was shaken and opened. The platinum basket and the ground glass neck were rinsed in water. The flask containing the absorption liquid was heated on a water bath at 70°C under stirring for 10 min. Since lanthanum ion, one of the rare earth metal ions, reacts not only with fluoride ion but also with carbon dioxide forming lanthanum carbonate which may interfere with the titration, the lanthanum carbonate in the liquid should be decomposed and at the same time carbon dioxide had to be drived out by, heating at 70°C under stirring. The temperature control was essentially important because at higher temperature, La(OH)₂⁺ or La(OH)₂⁺ ions were produced by hydrolysis of lanthanum chloride and these ions hardly reacted with EDTA.
The absorption liquid was transferred to a 100 ml beaker and 5 ml of hexamine-HCl buffer solution (pH 5.5) was added. The buffer solution was prepared by the following way : 28 g of hexamine was dissolved in 500 ml of water, the solution was adjusted at pH 5.5 with approximately 10 ml of 10% hydrochloric acid and it was finally made up to 1 l with water.
The fluoride in the solution was determined by backtitration of excess lanthanum with 0.01 M EDTA solution to the canary-yellow end point of methylthymol blue by making sure that the color tone was no more changed by further addition of a few drops of EDTA.
The titration was carried out by forming no precipitates and it was presumed that lanthanum reacted with fluoride forming LaF₃(H₂O)₃ complex which was stable in the hexamine-HCl buffer solution. This was partly evident by burning the fluorine compound in the combustion flask where the hydrogen ion concentration of the absorption liquid increased in proportion to the amount of decomposed fluorine.
Effect of some interfering elements are shown in Table I where a series of analysis of satndard fluorobenzoic acid with accompanying organic substances containing foreign elements has been listed. Presence of large amounts of bromine, iodine, nitrogen and iron caused no interference. Iron was oxidized to tri-iron tetroxide which did not dissolve in dilute hydrochloric acid. Some metal oxides such as mercury, barium and magnesium were partly dissolved by the acid in the absorption liquid and these ions reacted with EDTA. Boron also reacted with fluoride forming very stable complex in the absorption liquid. Phosphorus and sulfur interfered seriously by forming lanthanum phosphate and lanthanum sulfate, respectively.

View full abstract

A solvent extraction-atomic absorption spectrophotometry for the determination of trace amount of iron has been established. Various extraction methods of iron using such reagents as cupferron, acetylacetone, ammonium pyrollidine dithiocarbamate, sodium diethyldithiocarbamate, oxine and sodium benzoate, were compared and the oxine-methyl-iso-butylketone (MIBK) extraction system was adopted. Trace amount of iron is extracted as its oxinate from large amount of aqueous layer buffered by acetic acidsodium acetate into small amount of MIBK layer, and determined by the atomic absorption method, using an air-acetylene flame. Linear calibration curves were obtained within the concentration ranges of 100500, 20100 and 1050 ppb of iron, when the extractions were carried out with 100, 500 and 1000 ml of aqueous and 10, 20 and 30 ml of organic layers respectively. In the last case, the sensitivity of the method was 0.36 ppb of Fe in the aqueous layer for 1% absorption. The sensitivity of various iron lines were compared by spraying oxine-MIBK extracts obtained from 500 ml of aqueous layer containing 50 ppb of iron, and 2483 Å was selected as the most sensitive line.
Optimum concentration of oxine in the organi clayer was found to be 10^-3 M. Large amount of excess oxine in the organic layer decreased the absorbancy of iron. The same phenomenon of decreasing of absorbancy were observed when other kind of organic compounds was dissolved in the extract. Effect of various substances on the iron determination was examined. No serious interference was observed when the extraction was carried out in the presence of 100 ppm of Ca (II), 50 ppm of Mg(II), 1 ppm of Cu(II), Zn (II), Pb (II) Mn (II), Al (III), Si (IV) or 150 ppm of Cl^-. The interference of phosphate was relatively large but it was eliminated by using 10^-1 M oxine-MIBK solution as an extracting solvent.
The proposed method was applied to the determination of 50 ppb of iron in the model sample waters. Good recovery was obtained using 10^-3 M and 10^-1 M oxine-MIBK solutions. Then, the iron amount in several underground water and tap waters were determined by the proposed method. Good agreement was observed between the results obtained by the proposed method and the colorimetric method using 1, 10-phenanthroline as a reagent.

View full abstract

Glyoxal-bis(2-hydroxyanil) (GHA) forms a redcolored chelate with calcium ion in an alkaline solution. Because of superiority in selectivity of the reaction for calcium, many papers on spectrophotometric determination of calcium with GHA have been reported, but the disadvantage of the GHA method was that the reagent and its calcium chelate are considerably unstable in the alkaline aqueous solution. To overcome the instability, the color development in a mixture of water and alcohol or extraction of the chelate into a mixture of alcohol and chloroform was tried, all the attempts, however, were not always satisfactory.
In this paper, good color-stability has been achieved by extraction of the chelate into 1, 2-dichloroethane with an addition of zephiramine. The recommended procedure is as follows :
To a 20-ml sample solution containing calcium less than 1.25μ moles add 2 ml of borate buffer (a mixture of 0.1 M sodium hydroxide and 0.1 M sodium borate solution, pH 12.8) and 3 ml of 5 × 10^-3 M GHA ethanol solution in that order in a 100-mlseparatory funnel. After 1 min, add 1 ml of 5 × 10^-3 M zephiramine solution and 20 ml of 1, 2-dichloroethane, and shake the solution for 1 min. Separate the organic phase and desciccate it with anhydrous sodium sulfate. Measure the absorbance at 530 mμ against water as a reference.
The absorption spectrum of calcium-GHA chelate extracted into 1, 2-dichloroethane shows an absorption maximum at 530 mμ where the reagent blank has no absorption against water. Beer's law is held up to 1.25μmoles of calcium in the 20-ml extract and the molar absorption coefficient is 1.3 × 10⁴. The extracted chelate is very stable and no change in the absorbance is observed for at least 6 hours (Fig. 2).
The experiments showed that a constant absorbance was obtained in the ranges 12.713.0 of pH (Fig. 1), 15 min of the shaking time (Fig. 3) and 12 min in the time between the addition of GHA and the extraction (Fig. 4). Variations from 1.3 to 2.0 M in alcohol concentration (Fig. 5), (48)10^-4 M in GHA (Fig. 6), and (13)10^-4 M in zephiramine (Fig. 7) also showed the same absorbance.
It is emphasized that the color-instability of calcium-GHA chelate is completely eliminated by extraction of the chelate into 1, 2-dichloroethane in the presence of zephiramine. The effect of zephiramine on the extraction will be discussed elsewhere.

View full abstract

A quantitative method for the determination of pentachlorobenzaldoxime (Minokol) and its formulations has been developed by thin layer chromatographic (TLC) separation and polarography.
Minokol gave a polarographic reduction wave in the acidic media, the number of electrons involved in the reduction of Minokol was coulometrically estimated 4 and probably Minokol was reduced in the same manner of other oximes according to the following equation.
>C=N-OH+4H⁺+4_e^-→>CH-NH₂+H₂O
The height of the polarographic wave at -0.595 volt vs. S.C.E. was proportional to the concentration of Minokol in the range of 0.08 to 0.25 mg of Minokol/ ml. Pentachlorobenzaldehyde sometimes included in Minokol also gave a polarographic wave and was removed by TLC.
The recommended procedure is as follows. One ml of tetrahydrofuran solution containing about 8 mg of Minokol is spotted on a Kieselgel HF₂₅₄(Merck) plate with a thickness of 500μ and dimension of 20 × 20 cm. After development with n-hexane-acetone (4 : 1), the spot (R_f-value about 0.4) is located visually with a UV lamp and quantitatively scraped from the plate. After 25 ml of N, N-dimethvlfobmnamide is added, Minokol is eluted by shaking and the elute is centrifuged. To 15 ml aliquot of the supernatant, 5 ml of 0.5 N hydrochloric acid and 1 ml of 0.5% Triton X aqueous solution are added and the mixture is diluted to 25 ml with water. The oxygen dissolved in the solution is removed by passing nitrogen and the polarogram is obtained between 0 to -1.4 volt against mercury pool. Minokol in the formulations is determined after extracted with tetrahydrofuran.
The values of technical products fo Minokol containing more than 90% of Minokol were identical those analysed without TLC.

View full abstract

It was confirmed for various compounds that the energy difference between the K_{β1, 3} line and the K_β', line from Al, Si, P, S and Cl was dependent on the kind of atoms (C, N, O or F) in the nearest neighbor. The K_{β1, 3} lines from these elements were considered to be due to the transition of an electron from the 3p(X-ray emitting atom)-2p(ligand atom) valence band levels to the 1s vacancy (X-ray emitting atom), and the K_β' from the 2s (ligand atom) levels to the 1s vacancy (X-ray emitting atom), respectively. Consequently, the energy difference between K_{β1, 3} and K_β' was expected to be nearly equal to the energy difference between the 2p and the 2s levels of ligand atoms. The observed results were consistent with the consideration described above. In addition, the results of the molecular orbital calculation for SO₄^2-and ClO₄^- showed a similar tendency. Therefore, it was concluded that a chemical state of Al, Si, P, S or Cl, i.e., the kind of the nearest atoms, can be determined by measuring the separation between K_{β1, 3} and K_β' Chemical state analysis of Al in the AlN-H₃PO₄polymer was carried out using the phenomenon described above. A mixture of A1N and H₃PO₄ formed a kind of an inorganic polymer at 100200°C. Although the X-ray diffraction pattern of AlN disappeared, chemical analysis showed that nitrogen was kept in the polymer. The AlK_β spectrum from the polymer was found to be quite similar to that from crystalline AlN. Therefore, the conclusion was derived that no change existed in the nearest atoms of Al in the polymer. The change in AlK_α due to the coordination was also investigated and the coordination number of Al was found to remain unchanged. Instrumental conditions are given in the following. Samples were excited by X-rays from a Cr target tube. The spectrometer used was a commercially available unit (Toshiba AFV 104) equipped with EDDT (Al, Si) or Ge (111) (P, S, Cl) analyzer crystal. Spectra were recorded with a rate meter. K_{β1, 3}-K _β' separations were obtained graphically. The separations ranged from 3 to 18 eV and the reproducibility was about 0.3 eV.

View full abstract

A method for the spectrophotometric determination of iron was studied in which pivaloylacetylmethane (PAM) was used as an extractant and a photometric reagent. Iron was extracted quantitatively with exactly 10ml of 0.1M PAM in benzene from an aqueous solution at pH 2.54.5 by shaking for 10 minutes.
The complex had an absorption maximum at 435 nm. The calibration curve followed Beer's law in the range 0 to 120μg/10 ml of iron. The molar extinction coefficient was 4.3×10³. Iron can be selectively extracted in the presence of a large number of cations.

View full abstract

A rapid and simple spectrophotometric method for determining boron in glasses is devised.
A 0.100 g sample is decomposed with acid mixture of H₂SO₄, HF and H₂O₂ in a teflon beaker by heating on a water bath for 20 min. The solution is diluted to 20 ml when the content of boron is expected to be less than 10^-3%, and to 200 ml when the content is 10^-210^-3%. To 20 ml of the sample solution in a polyethylene separatory funnel, Methylene Blue solution and dichloroethane are added and the mixture is shaken. After washing the organic phase with silver sulfate solution, its absorbance is measured at 660 nm. When the content of boron is a few %, a 0.010 g sample is taken and decomposed. After diluting the solution suitably, H₂SO₄ and HF are added anew to 10 ml of the sample solution, and its solution is diluted to 20 ml with water. Then, the determination can be carried out in the same way.
Various substances usually present in glasses do not interfere with the determination of boron. It was possible to determine 2 × 10^-5a few % of boron in glasses.

View full abstract

A new analytical technique was developed previously for the photometric determination of a trace of iron in water and in high-purity aluminum with sodium sulfide in alkaline media by the present authors. In the previous paper, hydrazin sulfate was used as the reducing reagent for iron (III), but no experiment was carried out by using other reducing reagents. The present paper describes the experiments carried out by using various reducing reagents, i.e. ascorbic acid, sodium hydrogen sulfite, hydroxylamine hydrochloride and hydrazine sulfate. The results showed that the use of 2 ml of 5% hydroxylamine hydrochloride gave the best results for the determination of iron in water because of the highest stability of the color and the sensitivity, and that the use of 2 ml of 0.5%hydrazine sulfate described in the previous paper or 3 ml of 5% sodium hydrogen sulfite was better than other reducing reagents for the determination of iron in aluminum. Analytical results of well and city waters obtained by the recomended procedure using hydroxyl-amine hydrochloride and those by the ο-phenanthroline method are also described.

View full abstract