Analyst, October, 1980, Vol. 105, @. 955-964 Improved Extraction Method for the Spectrophotometric Determination of Trace Amounts of Boron in River Water with 1,8-Dihydroxynaphthalene-4-suIphonic Acid and Removal of the Excess of Reagent 955 Takashi Korenaga,* Shoji Motomizu a n d Kyoji Toei Department of Chemistvy, Faculty of Science. Okayama Cnivevsity, Tsushima-naka, Okayama-shi 700, Japan A simple method for removing the excess of co-extracted reagent in the ion- association extraction of metal complex anions with quaternary ammonium salts has been applied successfully to the spectrophotometric determination of boron at the parts per log level in river water with 1,8-dihydroxynaphthalene 4-sulphonic acid (DHKS) and tetradecyldimethylbenzylammonium chloride (zephiramine) , The procedure using DHKS described here greatly improves the previous method using chromotropic acid.Acetate buffer (pH 3.8), EDTA and DHYS are added to the sample solution (less than 50 ml) and the pH of the resulting solution is adjusted to 10.2. Then sodium chloride is added and the mixture is shaken with 5 ml of a 2 x M solution of zephira- mine in 1,2-dichloroethane (DCE). The organic phase is washed once with 10 mi of a back-washing solution (1.0 h.~ in sodium chloride, pH 10.2) and the absorbance of the organic phase is measured in a quartz cell. The boron complex with DHNS is extracted quantitatively into DCE and its apparent molar absorptivity in DCE is 2.4 x lo4 1 mol-l cm-l a t 341 nm. The detection limit and precision achieved with the method are 1 p g 1-1 and 5%, respectively.EDTA allows most interferences caused by metals to be suppressed, and other sources of bias due to the effect of co-extractable anions are almost eliminated by adding relatively large amounts of sodium chloride to the extraction system. Parts per lo9 amounts of boron present as boric acid in river water are determined spectrophotometrically, and the results obtained are successfully compared with those obtained by the methylene blue method. Keywords; Boron trace determination; river water awalysis; spectrophotometry ; ion-association extraction In a previous paper, we have proposed a spectrophotometric method for the determination of boron in natural waters involving solvent extraction with 1,8-dihydroxynaphthalene-3,6- disulphonic acid (chromotropic acid).l In that method, a large amount of reagent (250-fold excess of boron) was used in order to complete the complex formation and the principle presented earlier by the authors2 was applied to the removal of a large excess of co-extracted reagent (chromotropic acid) from the organic phase.Recently, we developed 1,8-dihydroxy- naphthalene-4-sulphonic acid (DHNS) as a reagent for the solvent extraction of boric acid.3 DHNS was synthesised in order to improve the extractability of the boron complex and was found to be much superior to chromotropic acid as an analytical reagent for the solvent extraction - spectrophotometric determination of boron. In that work,3 the method using DHNS was applied to the determination of boron a t the parts per million level in sea water and hot-spring water, but not a t the parts per billion (lo9) level in river water.In this work, the previous method using chromotropic acid1 was greatly improved by using DHNS and it could be used successfully for the concentration and determination of boron a t the parts per billion level in river water, as the extractability of the boron complex with DHNS is much superior to that with chromotropic acid in the ion-association extraction using tetradecyldimethylbenzylammonium chloride (zephiramine). This paper therefore presents a greatly improved method for the spectrophotometric determination of boron present as boric acid in river water with DHNS and zephiramine and removal of the excess of DHXS reagent in the organic phase. * Present address: School of Engineering, Okayama University, Tsushima-naka, Okayama-shi 700, Japan.956 Reagents KORENAGA et al.: EXTRACTION FOR SPECTROPHOTOMETRY OF BORON Analyst, V o l . 105 Experimental All of the reagents used were of analytical reagent grade. 1,8-Dihydroxyna$hthalene-4-sul$honic acid (DHNS) solution, 0.01 M. Dissol\re 1.311 g of DHNS, sodium salt, in 500 ml of de-ionised water and transfer the solution obtained into an amber polyethylene bottle. The solution can be kept for at least 1 month in a refrigerator. Prepare a stock solution by dissolving 154.6 mg of boric acid in 250 ml of de-ionised water. Store the solution in a polyethylene bottle and dilute aliquots of it with de-ionised water to give working solutions of the required concen- trations.Dissolve 369.0 mg of dried zephiramine4 in 500 ml of DCE (analytical-reagent grade, used as received) to give a 2 x 10-3 M solution. Store the zephiramine solution in a glass bottle ; it is stable for at least 6 months. (1) An aqueous solution 0.01 M in EDTA and 0.25 M in acetic acid - sodium acetate (pH 3.8) and (2) an aqueous solution 4 M in sodium chloride and 0.8 M in sodium hydrogen carbonate - sodium carbonate (pH 10.2) were used. An aqueous solution 1.0 M in sodium chloride and 0.2 M in sodium hydrogen carbonate - sodium carbonate (pH 10.2) was used. Standard boron solution, 1.000 X M. Extraction solvent, 1,2-dichloroethane (DCE), 2 x M in zefilzira~nzine. Bufler solutions. Back-w'ashing solutiort. Apparatus A Hitachi, Model 139, spectrophotometer and a Hitachi, Model EPS-3T, recording spectro- photometer were used for measuring absorbances in a quartz cell of 1-cm path length.A Hitachi-Horiba, Model F-Bss, pH meter equipped with a combined electrode (6026-05T) was used for pH measurements. An Iwaki, Model KM, shaker (frequency 250 strokes min-l) was used for shaking separating funnels and stoppered test-tubes. Recommended Procedure Transfer by pipette the sample solution (less than 50 ml), which is not acidified and is first filtered with a membrane filter of 0.45-pm pore size, into a 100-nil polyethylene separating funnel. Dilute the solution to 50 ml with de-ionised water and add 5 ml of acetate buffer solution (1) and 5 ml of 0.01 M DHNS solution, in that order. Mix the solutions thoroughly and allow the mixture to stand for 30 min.Then add 5 ml of carbonate buffer solution (2) to adjust the pH of the resulting solution to the extraction pH, and also add 5 ml of 1.5 M sodium sulphate solution to accelerate the phase separation Add 5 ml of 2 x M zephiramine in DCE solution. Shake the separating funnel for 30 inin and allow two phases to separate. Transfer the organic phase into a stopped 25-ml glasi; test-tube and add 10 ml of the back-washing solution. Shake the test-tube for 10 min and then allow it to stand for about 30min to remove the excess of reagent from the orgaiiic phase. Measure the absorbance of the organic phase in a quartz cell of 1-cm path length a t 341 nm against a reagent blank as reference. Prepare a calibration graph by using standard boron solutions corresponding to 0 4 x 10-5 M of boron in the organic phase.Results and Discussion In the extraction of the ion associate formed between the boron - ;DHNS complex anion and the zephiramine cation, the extraction was previously carried out with 5 ml of aqueous solution and 5 ml of DCE s ~ l u t i o n . ~ In this work, however, as i he extraction was carried out with 70 ml of aqueous phase and 5 ml of organic solution, the choice of the extraction solvent (DCE or chloroform) was re-examined. The equilibrium constants for exchange of chloride (log F s C ) between DCE and water were determined for DHNS reagent and its boron compIex in previous work,3 but those between chloroform and water were not. Therefore, the constants for chloride between chloroform and water were ,also determined in this work, and shown in Table I for comparison with those between DCE and water.From Table I, the difference in the constants between HR2- and BR,% and the extractability of theOctober, 1980 IN RIVERS WITH 1,8-DIHYDROXYXAPHTHALENE-4-SULPHONIC ACID TABLE I EXCHANGE EQUILIBRIUM CONSTANTS FOR CHLORIDE OBTAINED AT 25 "c Extraction solvent 1 r--------_h------ Constant* 1,2-Dichloroethane Chloroform Log KZi- 2.51 j, 0.05 1.01 i: 0.02 Log KZ'i- 4.39 & 0.07 3.55 j, 0.06 Log K22i- 9.77 & 0.04 6.42 0.11 957 * The exchange equilibrium constant for chloride (Ki2C) refers to the following equation2.3: A"-(&) + nZCl(o) nCl-(a) + Z,A(o) KlEL- = ([Cl -l"(a)[Z,Al(o)) / ([A" -](a) [ZC11"(0)) where An- and Z+ are the n-valent anion and the zephiramine cation, respectively, and subscripts a and o refer to the aqueous and organic phase, respectively.boron - DHNS complex using DCE were found to be larger than those obtained when chloro- form was used. As shown in the previous the exchange constants for four univalent anions (chloride, bromide, nitrate and iodide) also indicated that chloride gave the highest concentration range for the removal of the excess of reagent and therefore caused very effective salting-out. In this work, DCE and chloride were therefore used as the extraction solvent and the univalent anion for the removal of the excess of reagent from the organic phase, respectively. For reference, the percentage extraction of reagent and boron complex with DHNS, which was calculated by using the equilibrium constants listed in Table I, is shown in Fig.1. It is obvious that the reagent in the form HR2- is removed more easily than that in the form H,R-. DCE was therefore suitable for the separation of complex and reagent. 100 80 8 . 60 0 ? 40 .- c 4- X w 20 0 -2 -1 0 1 2 Log [CI- 1 a Fig. 1. DCE against chloride concentration of aqueous solution for 2 x zephiramine. in the form BRS3- a t pH 9 ; and 3, DHNS in the form H,R- a t pH 3. Percentage extraction of reagent and boron complex into M 1, DHNS in the form HR2- a t pH 9 ; 2, boron complex De-ionised water, obtained from a common de-ionisation apparatus, must be used in this work, as the absorbance of the reagent blank with de-ionised water is more constant and smaller than that with distilled water obtained from a commercially available glass distilla- tion apparatus.The reason why the distilled water blank is higher and more variable than the de-ionised water blank is assumed to be based on the dissolution of boron present in the glass of the distillation apparatus. Hence, the former blank might decrease the accuracy, precision and detection limit of the proposed method in an actual determination.958 XORENAGA et d. : EXTRACTION FOR SPECTROPHOTOMETRY OF BORON AnUlySt, T’d. 105 Effect of pH on the Formation and Extraction of the Boron Complex The effective pH range for the formation of the boron - DHNS complex in aqueous solution without EDTA was found to be 3-10.3 ’The effect of pH on the formation of the complex with EDTA was also examined. The results obtained indicate that the optimum pH range is 3.5-9 in the presence of EDTA (Fig.2) and the absorbance of the complex decreases at pH above 9. The formation of the boron complex was therefore ctarried out a t about pH 3.8 by using an acetate buffer. 0.6 I 1 I I 0 2 4 6 8 1 0 1 2 PH Fig. 2. Effect of pH on complex formation. 1, Complex obtained with 2 x M boron, measured against reagent blank; and 2, reagent blank as 1 but no boron present, measured against DCE:. The effect of pH on the extraction of the boron - DHNS complex with EDTA is shown in Fig. 3. Constant absorbance was obtained at pH 4-12, the optimum pH range being 7-11. Because the absorbance of the reagent blank was constant at pH above 7, but not re- producible a t pH 12, the extraction of the boron complex was carried out a t about pH 10.2 by using carbonate buffer. Accordingly, the complex was formed a t pH 3.8 in the presence of EDTA and then extracted with zephiramine into DCE a t pH 10.2 with addition of chloride.0 2 4 6 8 1 0 1 2 PH Fig. 3. Effect of pH on the extraction of the boron complex into DCE. 1, Complex obtained with 2 x M boron, measured against reagent blank; and 2, reagent blank as 1 but no boron present, measured against DCE. The effect of pH on the back-washing of the organic phase, which was transferred into a When the concentration of chloride ion in the The back- stoppered 25-ml test-tube, was examined. back-washing solution was 1 M, the optimum pH region was found to be 5-12. washing was therefore carried out a t about pH 10.2.Effect of Chloride Concentration on the Extraction of the Boron Complex and on the Back-washing of the Organic Phase The effect of chloride concentration on the extraction of the boron complex was examined at pH 10.2 (Fig. 4 ) . The optimum concentration of chloride was found to be about 0.3 M. To complete and accelerate the phase separation in the ion-association extraction, 5 ml of 1.5 M sodium sulphate solution were also added as the salting-out agent in the extraction system.October, 1980 IN RIVERS WITH 1,8-DIHYDROXYNAPHTHALENE-4-SULPHONIC ACID 0.6 , I 959 0 0.1 0.2 0.3 0.4 0.5 0.6 iCI-1 ,/M Fig. 4. Effect of chloride concentration on the extraction of the boron complex. 1, Complex obtained with 2 x M boron, measured against reagent blank: and 2, reagent blank as 1 but no boron present, measured against DCE.Effect of pH on the Back-washing of the Organic Phase The effect of chloride concentration on the back-washing of the organic phase, which was transferred into a stoppered 25-ml test-tube, was examined at pH 10.2, and the results obtained are shown in Fig. 5. The optimum concentration of chloride was about 1.3 M. 0 0.2 0.4 0.6 0.8 1.0 1.2 “21-1 ,/M Fig. 5. Effect of chloride concentration on the back-washing of the organic phase. 1, Complex obtained with 2 x M boron, measured against reagent blank; and 2, reagent blank as 1 but no boron present, measured against DCE. The effect of the volume of the back-washing solution on the back-washing of the organic phase was examined. The volume of the back-washing solution taken being varied from 5 to 20 ml, and 10 ml was found to be the most effective.The efficiency of the back-washing with sodium chloride solution was examined a t pH 10.2. When 13 ml of back-washing solution were used, the absorbances of the boron complex and reagent blank were 1.289 and 0.764 (no back-washing), 0.605 and 0.119 (one back-washing) and 0.564 and 0.102 (two back-washings), respectively. Accordingly, the back-washing was carried out once with 10 ml of solution. Effect of Concentration of DHNS and Zephiramine The effect of the DHNS concentration on the formation of the boron complex was examined. Fig. 6 shows that the amount of 0.01 M DHNS solution necessary for the complete reaction was more than 4 ml when the concentration of boron was about 1.4 X M in 70 ml of aqueous solution and 5 ml of 2 x l o 3 M zephiramine in DCE solution was used as the extracting solvent.The zffect of the zephiramine concentration on the extraction of the boron complex was examined by using 0.01 M aqueous zephiramine solution4 (Fig. 7). A 1-ml volume of this960 solution was found to be necessary for the complete extraction of the boron complex. fore, 5 ml of 0.01 M DHNS solution and 5 ml of 2 x (equivalent to 1 ml of 0.01 M aqueous zephiramine solution) were used in this work. KORENAGA et al. : EXTRACTION FOR SPECTROPHOTOMETRY OF BORON Analyst, Vol. 105 There- M zepkiramine in DCE solution 0.6 1 0.6 7- , 0 2 4 6 8 1 0 Volume of 0.01 M DHNS/ml 0 0.4 0.8 1.2 1.6 2.0 Voluime of 0.01 M zephiramine/ml Fig.6. Effect of the concentration of Fig. 7. Effect of the concentration of DHNS. 1, Complex obtained with 2 x zephiramine. 1, Complex obtained with M boron, measured against reagent 2 x loe5 M boron, measured against reagent blank: and 2, reagent blank as 1 but no blank; and 2, reagent blank as 1 but no boron boron present, measured against DCE. present, measured against DCE. Effect of Time The time necessary for the complete formation of the boron complex was examined (Fig. 8). When the concentrations of boron (present as boric acid) and DHNS reagent were 1.4 x and 7.1 x lo-* M a t pH 3.8, respectively, the complete reaction was found to be achieved within 30 min. The time necessary for the complete extraction of the complex into DCE was examined. Fig. 9 shows that a suitable shaking time was 30 min when 5 ml of 2 x M zephiramine in DCE solution were used.I 0 10 20 30 40 50 60 0 10 2'0 30 40 50 60 Ti m e/m i n Time/rnin complex. Absorbance measured against reagent Fig. 9. Effect of 'shaking time. Absorbance blank. measured against reagent blank. Fig. 8. Effect of time on the formation of boron The shaking time necessary for the complete back-washing of the organic phase when 10 ml of the back-washing solution were used was examined. A shaking time of 5 min was found to be sufficient, so the shaking time for the back-washing was fixed a t 10 min. The separation of the two phases was immediate on the extraction of the complex from 70 ml of aqueous solution into 5 ml of DCE solution, and standing for a t most 30 min was found to be preferable to the complete separation of the two phases on the back-washing of the organic phase before measurement of the absorbance. Absorption Spectra and Calibration Graph The absorption spectra of the boron complex with DHNS and the reagent blank in DCE are shown in Fig.10 (solid lines 1-3). When the excess of co-extracted reagent in the organic phase is removed, the absorption maximum is obtained at 341 nm with a minimum reagentOctober, 1980 IN RIVERS WITH 1,8-DIHYDROXYNAPHTHALENE-4-SULPHONIC ACID 961 n v - 300 320 340 360 Wavelengthinm Fig. 10. Absorption spectra in DCE. 1, Complex obtained with 2 x M boron (0.216 pg of boron), measured against DCE in reference beam; 2, complex obtained with 1 x M boron (0.108 pg of boron), measured against DCE in reference beam; 3, reagent blank as 1 and 2 but no boron present, measured against DCE in reference beam; 4, solution obtained with 25 ml of Mino sample, measured against reagent blank in reference beam; 5, solution obtained with 50 ml of Mino sample, measured against reagent blank in reference beam; and 6, solution obtained with 25 ml of Mino sample plus 0.108 pg of boron, measured against reagent blank in reference beam.blank. The absorbance in DCE was measured in a quartz cell of l-cm path length at 341 nm against the reagent blank as reference. The calibration graph at 341 nm was a straight line that passed through the origin and obeyed Beer’s law over the range 0-3.6 x M boron present as boric acid in aqueous solution (corresponding to 0-5 x M boron in DCE).The apparent molar absorptivity in DCE calculated from the slope of the calibration graph was 2.4 x lo4 1 mol-l cm-l (con- verted into E = 3.4 x lo5 1 mol-l cm-l in aqueous solution) at 341 nm. The reagent blank at 341 nm was 0.119 & 0.006 (mean value of seven determinations), but reproducible when de-ionised water was used. Consequently, the detection limit and pre- cision achieved with the method are 1 pg 1-1 (difference in absorbance 0.03) and 5%, res- pectively. Determination of Boron in River Water with DHNS Interferences and masking agent DHNS reacts with some metal ions, such as aluminium, copper, iron, titanium and moly- bdenum. The tolerance limits of these metals and other ions generally present in river water were examined without EDTA according to the recommended procedure.The tolerance limit is defined as the concentration level at which the interferent causes an error of not more than 50/,. When EDTA was used as a masking agent for metals, the pH was adjusted to about 3.8 with acetate buffer. As shown in Table 11, the tolerance limits with EDTA were also deter- The results obtained are shown in Table 11.962 KORENAGA et al. : EXTRACTION FOR SPECTROPHOTOMETRY OF BORON Analyst, VoZ. 105 TABLE I1 TOLERANCE LIMITS FOR DIVERSE IONS WITH AND WITHOUT EDTA Ion Tolerance limit*/= .. I t Na+, K+, Sot- . . .. .. .. .. .. .. HCO- H,PO, .. .. .. .. .. .. .. .. 0 . l t Caz+, Sr2+, Baa+, Br-, NO;, Si'0:- .. .. .. .. 2 x 10-3 UOg+, A13+, Cr3+, F -, SCN -, I - . . .. .. .. .. 2 x 10-4 Fe3+, Ti4+, &foe+, ClOa, dodecylbenzenesulphonate .. .. .. 2 x 10-5 NO; .. .. . * .. .. .. .. 2 x 10-8: Ag+, Mna+;'Coa+,'Nia+, Cu2+, Zn2+, Cda+ . . .. .. . * 1 x io-3tf Hg2+, PbZ+, UOZ,+, AP+, Cra+, Fea+, Ti4+, Mo*+ .. .. 1 x 10-4s M g Z a ' .. .. .. .. .. .. .. 0.0lt Ag+, Mn*+, Fez+, &a+, Ni2+, Cuz+, ZAa+, Cda+, Hg2+, Pba+, * The tolerance limit is defined as the concentration level a t which the interferent causes an error of not more than 6% (precision of the method). t Maximum tested. If samples are acidified to pH below 2 by adding more than 0.5 ml of concentrated sulphuric acid per litre, nitritk ion may be almostcompletely removed as nitrogen monoxide gas. Add 5 ml of 0.01 N EDTA solution to 70 ml of the aqueous test solution. mined when 5 ml of 0.01 M EDTA solution were added to 70 ml of aqueous solution a t pH 3.8.Hence, the interferences of these metals and other metals commonly present in river water are eliminated by adding EDTA to the extraction system a t a concentration of about 10-3 M. Contamination from glassware In order to test for possible contamination from the glassware, DHNS, EDTA and acetate buffer were allowed to stand a t pH 3.8 in a 100-ml glass separating funnel for 2 h. The solution was examined by the proposed procedure and little boron was found (absorbance 0.130). However, when the solution was allowed to stand a t pH 10.2 (carbonate buffer) for 2 h, the absorbance was found to be higher. Polyethylene separating funnels were therefore used in order to prevent any possible contamination caused by the solution standing for a long time in glass.Polyethylene bottles were also used for storage of sample and reagents solutions. Pre-treatment of samflle solution When the sample was immediately treated with 0.5 ml of concentrated sulphuric acid per litre after collection, the boron content was found to be identical with that obtained without acidification. The stability of boron in sample solutions was examined with and without sulphuric acid. The results obtained for boron content did not vary for a t least a week in both instances, The sample therefore need not be acidified. When the membrane filter (0.45-pm pore size and 47-mm diameter circle) was used, no loss or gain of boron was found and reproducible results were obtained. The effect of sample acidification was examined.The effect of sample filtration was examined. Results of determination The results obtained by the recommended procedure for the determination of boron in water samples from the River Asahi, Okayama Prefecture, Japan, ai-e given in Table 111. An example of absorption spectra in the determination of boron in river water is shown in Fig. 10 (broken line 4). Boron present in sample solutions as boric acid and as tetraborate can be determined by this method, but not boron present as fluoroborate. In generad, river water contains boron in the form of boric acid and seldom in the form of tetraborate and fluoroborate. Accordingly, the total boron in river water can be determined by the use of the recommended procedure. In order to check the results of the determination of boron in river water, two series of experiments were carried out.In the first experiment, the amount of sample solution taken was varied between 10 and 50 ml and de-ionised water was added to each so as to give a constant volume. For all samples, linear graphs were obtained and the lines could beOctober, 1980 I N RIVERS WITH 1,8-DIHYDROXYN.4PHTHALENE-4-SlJLPHONIC ACID TABLE I11 DETERMINATION OF BORON I N RIVER ASAHI WATER Boron contentt/,pg 1-1 Sample source* Shimotokuyama . . . . Hatsuwa . . . . . . Katsuyama , . . . . . Ochiai . . . . . . Nishikawakami . . . . Eyomi . . . . . . Asahigawa-damu . . . . Shinada . . . . . . Kanagawa . . . . . . Ohara . . . . . . Mino . . . . . . Xanokaichit . . . . Hot-spring a t Misasa . . Hot-spring a t Yubara .. Seto Inland Sea a t Shibukawa Seto Inland Sea a t Teshima Pacific Ocean a t Tanohama Japan Sea a t Aoya . . . . Distance from estuary/ km . . 137 . . 122 . . 89 . . 76 . . 68 . . 62 . . 54 . . 49 . . 32 . . 16 . . 12 . . 4.6 963 Proposed method 8 . 8 & 0.3 8.4 & 0.2 8.7 i: 0.3 9.4 & 0.4 11.5 rt 0.5 8.6 i 0.4 11.4 i: 0.3 10.4 i: 0.3 11.4 0.1 9.3 & 0 . 4 11.7 rt 0.4 360 & 20 Methylene blue method; 8.4 7.6 8.7 9.7 10.9 8.1 11.5 9.8 10.8 10.0 11.6 i: 0 . 9 400 34 & 2 38 2300 i: 100 1900 4400 & 100 4 200 4500 i: 200 4300 4400 i: 100 4 600 3900 200 4 200 * Samples from the River hsahi were sampled on May 3rd and 4th, 1978. t Average values of four determinations i: standard deviations. $ This sample contained 0.043% of chloride, probably caused by sea water.The boron contents of hot- The values without standard deviations spring and sea waters are also given; the samples used were the same as those in previous are averages of two determinations. extrapolated to the same point, which coincided with the point obtained for 50 ml of de-ionised water. It was concluded that the determination of boron in river water was quantitative and de-ionised water could be used as the reagent blank. An example of absorption spectra in the experiment is shown in Fig. 10 (broken lines 4 and 5 ) . In the second experiment, the recovery of boron was examined by adding various amounts of boron to the sample solutions. All of the results obtained were linear and the slopes of the graphs were identical with those of the calibration graph obtained with de-ionised water. An example of absorption spectra in this experiment is shown in Fig.10 (broken lines 4 and 6). Comparison zeith the conventional wtethylene blue method The analytical values obtained by the proposed method were compared with those obtained by the methylene blue m e t h ~ d . ~ The results obtained by the latter method, conventionally available in Japan, are shown in Table 111. From Table 111, two methods were found to be comparable and the sample from the River Asahi at Mino showed that the relative standard deviation obtained when using the proposed method was smaller than that with the methylene blue method. The correlation coefficient of two methods was 0.95 in the determination of boron in 11 samples from the River Asahi, except for the sample from Nanokaichi ( a = - 0.05 and b = 0.98 in the equation y = a + bx, where x is the value obtained by the proposed method and y value obtained by the methylene blue method).Consequently, the results of the proposed method using DHNS are almost identical with those of the conventional methylene blue method5 so that the accuracy of the method is good in the practical analysis of river water samples. Conclusion The method proposed here could be applied to the spectrophotometric determination of parts per billion amounts of boron in river water with satisfactory results. This method of determining trace amounts of boron with DHNS is a great improvement on the previous method employing chromotropic acid,l and possesses the following advantages: (a) the complex formed with DHNS has a large molar absorptivity, which is about 1.7 times greater than that with chromotropic acid; (b) the extractability of the complex with DHNS is higher964 KORENAGA, MOTOMIZU AND TdEI than that with chromotropic acid, so that the concentration of boron by extraction from aqueous solution into DCE is complete and precise; (c) there is a llarge difference in the ex- change equilibrium constants between the complex and reagent, i.e., the removal of the excess of reagent can easily be achieved without loss of the boron complex; (d) the procedure is simple and back-washing is preferably carried out only once; and (e) the synthesis and purification of DHNS reagent are simple.As described here, the simple method for the removal of the elccesj of reagent in the organic phase must lead to a greatly improved sensitivity in a given ion-association extraction system, as the measurement can be carried out at the most sensitive wavelength of the complex, which cannot be measured without removal of the excess of reagent. As a sufficient amount of the reagent and a cationic surfactant can be added, the complex can easily be formed and extracted over a wide pH range. Moreover, the addition of relatively large amounts of salts such as sodium chloride causes very effective salting-out, so that the phase separation becomes more rapid. Also, the absorbances of the complex and reagent blank become more constant and reproducible on adding salts t o the ion-association extraction system, because the inter- ferences caused by co-existing anions are effectively eliminated. References 1. 2. 3. 4. 5 . Korenaga, T., Motomizu, S., and TBei, K., Analyst, 1978, 103, 745. Motomizu, S., and TBei, K., Anal. Chim. Acta, 1977, 89, 167. Korenaga, T., Motomizu, S., and TBei, K., Anal. Chinz. Acta, in th.e press. TBei, K., and Kawada, K., Bunseki Kagaku, 1972, 21, 1610. Japan Industrial Standard (JIS), K 0102, 1974. Received December loth, 1979 Accepted June 2nd, 1980