ISSN:0003-2654    

DOI:10.1039/AN98005FX029

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BX031

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005FP089

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98005BP097

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Vol. 105 No. 1253 Preservation of Inorganic Arsenic Species at Microgram Levels in Water Samples Venghuot Cheam and Haig Agemian Canada Centre for Inland Waters, Special Services Section, Water Quality Branch, P.O. Box 5050, Burlington, Ontario, Canada A preservation study was carried out to determine the stability of inorganic species of arsenic as arsenic(II1) and arsenic(V) in distilled and natural waters for a duration of 125 d. Various conditions of storage, such as pH, level of arsenic, type of container material and size of container, were studied. I t was found that arsenic(II1) and arsenic(V) are satisfactorily preserved for this time in both distilled and natural water samples a t 1 and 10 pgl-1 levels a t room temperature if they are stored in polyethylene or Pyrex bottles with 0.2% V / V sulphuric acid.Keywords: AYsenic(III) and arsenic( V ) ; presevvation and stability of solutions; distilled and natural water samples; P H ; container sizes Preservation conditions that maintain integrity and stability of environmental water samples are a prerequisite to the successful execution of analytical determinations and inter- laboratory quality control studies. Our laboratory has initiated a series of studies on key parameters that have recently been of environmental importance. These include mercury,l seven phenoxy herbicidal acids2 and ~elenium.~ These studies have served as a basis for the confident preparation of stable aqueous solutions in National Interlaboratory Quality Control This study was undertaken with the same purpose as the above and deals with the preservation and stability of arsenic in water. I t was believed for a long time that arsenic exhibited great toxicity but it has recently been suggested to have, along with selenium, nutritional value owing to its electron-transfer ~apability.~ Arsenic is one of the many constituents routinely analysed by most analytical- services laboratories.It is, therefore, an important parameter to be investigated so that the optimum conditions are found that will demonstrate its stability in environmental samples. These optimum conditions are not only necessary for the planning and designing of inter-laboratory comparison studies and for accurate and precise determinations, but are also vital information for the preparation of bulk reference samples. There is no comprehensive documentation on the stability of arsenic at natural levels and it is not clear whether a preservative is needed.The Analytical Methods Manual of the Water Quality Branch8 makes no special recommendation for the preservation and storage of arsenic. This procedure satisfactorily maintains arsenic in solution and at the same time prevents the formation of algae that occurs when unacidified natural samples with bacterial activity are stored at room temperature. As will be shown later, algae are a serious interferent. Low-temperature storage is, however, impractical as it requires special facilities and could be expensive for large numbers of samples. Furthermore, transportation of samples becomes a real problem. Samples must be kept preserved in transit from field sampling stations to the laboratory.Also, inter-laboratory quality control samples must retain their original integrity from the time of shipment to participating laboratories until time of analysis. The United States Environmental Protection Agencyg has recommended preservation with nitric acid to a pH of 2. Nitric acid, however, has been reported to be a serious inter- ferentl0P1l in the reduction reactions commonly used in arsenic determinations. In inter- laboratory quality control studies, where many different methodologies are used by partici- pating laboratories, the use of nitric acid as a preservative could result in a negative bias 737 In the past, samples have been stored unacidified but at 4 "C.738 CHEAM AND AGEMIAN PRESERVATION OF INORGANIC Analyst, vd.105 for those laboratories using methods sensitive to the above interference. Therefore, this acid could not be considered. Hydrochloric acid has no interfering effects but it is very volatile and somewhat impractical for use in the field. Sulphuric acid was chosen because it is a common acid used in analytical chemistry, and is stable and non-volatile and therefore is easy to use in the field. Furthermore, it has no serious interfering effects on the reduct ion reactions used in the analysis of arsenic. In our Water Quality Laboratories, arsenic and selenium are normally analysed from the same solution, as their analytical methodologies are similar. Cheam and Agemian3 reported that 0.2% V/V sulyhuric acid satisfactorily stabilises aqueous solutions (distilled and natural waters) of selenium(1V) and selenium(V1) for the duration of 4 months in both glass and plastic containers.This study was designed in a similar way so that common preservation conditions could be found for the two metals. I t can be seen that 0.2y0 V/V sulphuric acid is a common preservative for the two metals at the parts per bil-lion (lo9) levels (micrograms per litre) and for seven phenoxy herbicidal acids too.2 Optimum conditions are discussed in terms of pH, water types and container material and size. Two types of water are used: distilled water representing very clean sample types and Hamilton Harbour water, which is heavily loaded with nutrients, bacterial activity and major ions, which is at the other extreme.Most water samples analysed in Water Quality laboratories lie within these two extremes. The preservation conditions recommended in this study take into account practical considerations for both routine monitoring of a large number of samples as well as large scale inter-laboratory quality control studies. Experimental Reagents Analytical-reagent grade chemicals were used throughout. Stock solutions of arsenic, 1000 mg 1-l. Prepared using the sodium salts NaAsO, and The Na2HAs0,.7H,0 (Baker Analyzed Reagents and Ventron Corporation/Alfa Division). solutions were preserved with 1% V/V sulphuric acid. Sodium tetrahydroborate(ll1). Fisher Scientific Co. Sodium hydroxide pellets. Concentrated sulphuric acid. Concentrated hydrochloric acid. Fisher Scientific Co.Baker Analyzed Reagent. Baker Analyzed Reagent. Containers Two sizes of containers were considered : 25-gal poly- ethylene barrels with spigots, purchased from CANBAR Products Ltd., and 500-ml Nalgene and Pyrex bottles. Pyrex and 9olyethylene containers. Pyrex calibrated flasks. All containers were cleaned with chromic acid, and rinsed five times with hot tap water For the stock, intermediate and standard solutions. and finally three times with de-ionised distilled water. Hamilton Harbour Water A comparison of some chemical composition data for Hamilton Harbour water with other major water systems is given in Table I. The data substantiate the fact that Hamilton Harbour water represents the highly contaminated extreme of water sample types used in Water Quality Laboratories.Bulk and specijic sample preparation The bulk preservation samples were prepared using eight 25-gal barrels, four of which were filled with de-ionised distilled water and the other four with unfiltered Hamilton Harbour water. They were then spiked with arsenic(II1) and arsenic(V) species at concentrations near 1 and 10 p.p.b. Each solution was then homogenised by a closed circuit mixing for 3 h using a magnetic drive pump (Fasco Industry Inc.). A control barrel containing de- ionised distilled water was monitored for possible arsenic release; after 3 months no arsenic was detected, and therefore further monitoring or discussion was deemed unnecessary.August, 1980 ARSENIC SPECIES AT MICROGRAM LEVELS IN WATER SAMPLES 739 Specific preservation samples were prepared by sub-sampling each spiked bulk sample into 3 Pyrex and 3 polyethylene 500-ml bottles.The pH of these samples were then adjusted to 1.5 (corresponding to about 0.2% V/V sulphuric acid), 5.4 (the approximate pH of distilled water) and 7.2 (the approximate pH of Hamilton Harbour water). ,411 preservation samples were kept a t room temperature throughout the study. TABLE I SOME WATER QUALITY PARAMETERS REPORTED FOR HAMILTON HARBOUR LOWER AND UPPER GREAT LAKES AND WORLD LAKE AND RIVER WATER Constituent Mean of ~~..... ~. World lake Hamilton Lake Lake and river Harbour12* Ontario" Lake Erie13 Lake Huron13 Superiorls waters" Calciumlmg 1-1 . . . . . . . . 54 40.1 38 26.0 13.1 15 Magnesiumlmg 1-l . . . . . . 11.6 8.2 8 7.2 2.8 4.1 Sodium/mg 1 - l .. . . . . . . 30 13.1 11.4 3.1 1.2 6.3 Potassiumlmg 1-I . . . . . . . . 5.37 1.4 1.2 0.8 0.6 2.3 Chloridelmg 1-1 . . . . . . . . 61.2 28.3 24.5 5.6 1.2 7.8 Specific conductance/pohm-' cm-l . . 511 344 292 20 7 97 149 Sulphate/mg 1-l .. . . 63 28.6 25.0 16.0 2.7 11.2 Hardness, total CaCOJmgi-l . . . . 183 134 128 95 44 54 (Calculated, ref. 12) 0.23 Ironlmg 1-1 . . . . . . . . 0.28 0.018 0.003 0.003 0.002 2 0.67 - - - - Complexing capacitylwg 1-' Cu . . . . 20016 5515 Total phosphorus/Wgl-' . . . . . . 73 24 28 5.5 5.0 - Nitratelmg 1-l N 1.89 0.14 0.337 0.282 0.308 Ammonia/mg 1-1 N ' 1: :: :: 1.13 0.01 - 0.003 0.002 Manganeselmg 1-1 . . . . . . . . 0.09 0.0006 0.000 5 0.000 4 0.000 3 - Copperlmg 1-1 . . . . . . . . 0.02 0.001 2 0.002 5 0.001 O.Ol0t Zinclmg 1-1 .. . . . . . . 0.06 0.002 2 0.010 0.001 0.003 T 0.01ot - 0.000 5 * The values reported for Hamilton Harbour are mean results observed in 1975. t Mean copper content of ordinary fresh waters is about 0.010 mg 1-l (ref. 14, p. G-47). Mean zinc content of ordinary lake and river waters is about 0.010 mg I-' (ref. 14, p. G-48). Arsenic Monitoring The monitoring of inorganic arsenic was carried out for a period of 4 months using the automated technique reported by Agemian and Cheam.16 The relative standard deviations at 1 and 10 pg 1-1 of arsenic were estimated to be around 10 and 574, respectively. During the study many runs were made, and each consisted of about 60 actual analyses in addition to those for calibration graphs, bracketing about every 12 analyses. The intermediate and standard solutions were prepared freshly on the day each run was carried out.Results and Discussion The levels of arsenic in the North American Great Lakes17 are in the range <0.1- 1.20 pg I-!. Schramel et aZ.18 have reported a range of 2.2-8.1 pg 1-1 of arsenic in some locations in Germany. Arsenic levels as high as 70 pg 1-1 have been reported, such as in the Allegheny River, Pittsburgh.lg In this study the stability of 1 and 10 pg 1-1 levels of arsenic were studied, as this range represents the lower end of the range of natural levels. As the higher the level of an analyte the greater is its solution stability, it would follow that if low-level solutions preserve well then higher level solutions will also preserve well. Figs.1 (a) and (b) show the stability of 1 pg 1-1 of arsenic(II1) solutions under various conditions of storage. I t is apparent that pH 1.5 is the most favourable condition for preservation for both types of container and water. It is interesting to note that in distilled water and at the higher pH values of 5.4 and 7.2, preservation was satisfactory for 500-ml polyethylene containers while substantial losses occurred in 500-ml Pyrex bottles. Because distilled water contains no interfering ions and no species with complexing ability or bacterial activity and as volatilisation is not a common problem with arsenic under these conditions, the low recoveries observed above are predominantly due to a surface interaction. Com- parison of Figs. 1 (a) and (b) shows that at pH 5.4 and 7.2 the stability of arsenic(II1) in distilled water and in polyethylene bottles was satisfactory for the duration of the experi- ment, but the stability in Hamilton Harbour water decreased considerably after about 2 months.There was, however, an interesting observation in the middle of the experiment. Traces of algal growth were observed after 6 weeks and by 2 months had become substantial. The appearance of algae corresponds well with the drop in arsenic(II1) recovery at around 2 months. For the Pyrex containers the behaviour of the graphs is different for the two740 Analyst, VoZ. 105 types of water. In the distilled water a steady loss is observed and as indicated above is most probably due to a surface interaction. However, for the Hamilton Harbour water, glass containers are satisfactory for up to 2 months but stability declines as algae appear. CHEAM AND AGEMIAN : PRESERVATION OF INORGANIC I , 6 I I I I 50 t I 10 20 30 40 50 60 70 80 90 100 110 120 Contact time/d Fig.1 . (a) distilled water; ( b ) Hamilton Harbour water. A pH 5.4, ethylene barrel: x pH 5.4 for distilled water or 7.2 for Hamilton Harbour water. Time dependence of arsenic(II1) species, 1 pg l-l, remaining in solution: Polyethylene bottles: pH 1.5, Pyrex bottles: 0 pH 1.5, A pH 5.4, pH 7.2. Poly- pH 7.2. Evidently the higher ionic load of this water compared with distilled water removes the effect of the container material on the stability but the bacterial activity causes algal inter- ference after 2 months. For those containers preserved at pH 1.5, no algae were observed in the Hamilton Harbour water samples.The data indicate that acidification to pH 1.5 with sulphuric acid effectively stops bacterial activity and removes the above-mentioned interference as well as the effect of container material. 100 8 90 c' 80 + 70 2 60 .- - C m C m .- .- .- 100 .- ; 90 % 80 70 c I I I 1 1 I I I 1- 60 - 50 - I 1 I I 10 20 30 40 50 60 70 80 90 100 110 120 Contact time/d Fig. 2. Time dependence of arsenic(II1) species, 1 p g 1-l, remaining in solution: (a) distilled water; ( b ) Hamilton Harbour water. Polyethylene bottles: pH 1.5, A pH 5.4, pH 7.2. Pyrex bottles: 0 PIE 1.5, A pH 5.4, 0 pH 7.2. Poly- ethylene barrel: x pH 5.4 for distilled water or pH 7.2 for Hamilton Harbour water.August, 1980 ARSENIC SPECIES AT MICROGRAM LEVELS IN WATER SAMPLES 741 Figs.2 (a) and (b) show similar graphs to the above for arsenic(II1) at a 10-fold greater concentration. Polyethylene containers are again preferred to Pyrex bottles at high pH in distilled water. The negative effect of algae in Hamilton Harbour water is similarly seen after 2 months of storage at pH 5.4 or 7.2. Acidification with 0.2% sulphuric acid to a pH of 1.5 again inhibits bacterial activity and the consequent growth of algae, thus satisfactorily preserving arsenic(III), irrespective of container material or water type. Of these the former is probably the dominant state in the natural environment owing to the reducing effects of the abundant organic matter. It is, however, necessary to study the behaviour of both species in water samples in order to obtain a better understanding of their role in solution stability.Figs. 3 (a) and ( b ) show the behaviour of arsenic(V) at 1 pgl-1 for the same conditions as studied above. The effect of container material is similar to arsenic(III), where polyethylene con- tainers were shown to be better than those made of Pyrex. There is, however, a significant difference with the distilled water. While with arsenic(II1) and plastic bottles satisfactory stability is obtained at pH 1.5, 5.4 and 7.2, for arsenic(V) this is only true at pH 1.5 at the end of the experiment. This signifies that the stability of arsenic(V) is much more dependent on pH than that of arsenic(II1). In the Hamilton Harbour water, the stability again deteriorates after 2 months when algae appear, but the loss is more severe than for arsenic(II1).Acidification to pH 1.5 removes all interfering effects and gives lOOyo recovery after 4 months. Arsenic has two stable oxidation states, three and five. 100 90 80 $ c- 70 60 .- - 54 C m C .- .- .g 100 g 90 L v 80 70 4 60 50 .- i 40 30 I I I I I 1 1 I I 10 20 30 40 50 60 70 86 90 100 110 120 Contact time/d Fig. 3. Time dependence of arsenic(V) species, 1 pg 1-I, remaining in solution: (a) distilled water; (b) Hamilton Harbour water. Polyethylene bottles: pH 1.5, A pH 5.4, Poly- ethylene barrel: x pH 5.4 for distilled water or pH 7.2 for Hamilton Harbour water. pH 7.2. Pyrex bottles: 0 pH 1.5, A pH 5.4; 0 pH 7.2. Figs. 4 (a) and (b) show the behaviour of arsenic(V) a t a concentration of 10 pg 1-1 under the above conditions.An increase in concentration has a slight stabilising effect for all conditions, unlike arsenic(III), for which there was no detectable change. The behaviour of arsenic(V) at 10 pg 1-1 is otherwise similar to that of 1 pg 1-1 solutions. The experiments discussed so far were carried out in 500-ml bottles, which is a typical size for routine monitoring or for inter-laboratory quality control studies. Some experi- ments were also carried out in 25-gal polyethylene barrels to determine the effect the volume to surface area ratio has on the stability of unacidified arsenic solutions. It is apparent from Figs. 1 4 (polyethylene barrel) that no deterioration in stability occurred after 4 months of storage for any of the various conditions studied.Even at pH 5.4 for distilled water, or 7.2 for Hamilton Harbour water, no significant loss occurred. A calculation of the742 100 8 90 80 3 70 .- + - 2 C 0 7 c C cu .- .- .- E 5) 100 .- :: 90 2 80 a 70 60 50 CHEAM AND AGEMIAN : PRESERVATION OF INORGANIC Analyst. VoZ. 105 Contact time/d Fig. 4. Time dependence of arsenic(V) species, 1 p g l-l, remaining in solution: (a) distilled water; (b) Hamilton Harbour water. Polyethylene bottles: pH 1.5, A pH 5.4, H pH 7.2. pH 5.4, 0 pH 7.2. Poly- ethylene barrel : x pH 5.4 for distilled water or pH 7.2 for Hamilton Harbour water. Pyrex bottles: 0 pH 1.5, surface area to volume ratio showed that this ratio is 7 times higher for the 500-ml bottles than for the 25-gal barrels. Therefore, container size effectively changes the available surface area and substantially affects solution stability.Even for Hamilton Harbour water where algae form, no loss occurred. Evidently the algae surface area is not large enough to cause any appreciable loss. It is therefore obvious that while a pH of 1.5 achieved by adding sulphuric acid is indeed necessary for proper preservation and for the inhibition of algal growth in 500-ml bottles, no preservative is required for 25-gal barrels. This type of observation was also made earlier with selenium specie^,^ which were stable in the barrels but unstable in the 500-ml bottles for both distilled and natural waters. The above observations on the effect of container size could lead to some practical implica- tions. For large scale inter-laboratory quality control studies large bulk samples must be obtained, homogenised, spiked and stored until sub-divided into smaller bottles for shipment to participating laboratories.The above data indicate that 25-gal barrels are suitable as storage vessels without any chemical treatment and thus are directly applicable in such round-robin studies. The 0.2% sulphuric acid preservative, by inhibiting algal growth, makes the samples suitable for analysis by an automated system. If samples are not acidified and algae are produced in the sample, the only way that a meaningful determination could be carried out would be to digest and oxidise the whole content of the bottle and release all the arsenic back into solution. The inclusion of such a step is not desirable as this leads to a manual step being introduced, which is more expensive and time consuming to carry out, introduces possible errors and a blank value due to the reagents used, and destroys the whole sample.The preferred and suggested preservation procedure is acidification with 0.2% V/V sulphuric acid (pH 1.5) and storage at room temperature in polyethylene bottles. Although glass is equally suitable under these conditions, it is excluded for practical reasons. This preservation method was tested in a National Interlaboratory Quality Control Study.6 Ten samples, concentrations ranging from 0 to 1000 pg lF1, were distributed to 41 laboratories. Statistical analyses of all returned analytical data support the effectiveness of the preserva- tion system described, particularly for those data determined by the hydride generation technique. We thank 0.Elkei for permission to use his unpublished da.ta and Carm Pacenza for her secretarial help.August, 1980 ARSENIC SPECIES AT MICROGRAM LEVELS I N WATER SAMPLES References 743 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Carron, J., and Agemian, H., Anal. Chim. A d a , 1977, 92, 61. Chau, A. S. Y., and Thomson, K., J . ASSOG. OH. Anal. Chem., 1978, 61, 1481. Cheam, V., and Agemian, H., Anal. Chim. Acta, 1980, 113, 237. Aspila, K. I., and Carron, J. M., “Interlaboratory Quality Control Study No. 16-Total Mercury in Natural Waters,” Report Series No. 53, Inland Waters Directorate, Water Quality Branch, Ontario Region, Burlington, Ontario, Canada, 1978.Agemian, H., and Chau, A. S. Y., “Interlaboratory Quality Control Study No. 24-Eight Phenoxy Acid Herbicides in Fresh Natural Water,” Report Series No. 67, Inland Waters Directorate, Water Quality Branch, Ontario Region, Burlington, Ontario, Canada. Cheam, V., and Aspila, K. I., “Interlaboratory Study No. 26-Arsenic and Selenium in Water,” Inland Waters Directorate, Water Quality Branch, Ontario Region, Burlington, Ontario, Canada. Approved for publication. Canada, 1974. tion of Water and Wastewater,” EPA-60014-76-049, September 1976. Frost, D. V., Chem. Eng. News, 1978, 4, 2. “Analytical Methods Manual,” Water Quality Branch, Inland Waters Directorate, Ottawa, Ontario, United States Environmental Protection Agency, “Handbook for Sampling and Sample Preserva- Pierce, F. D., and Brown, H. R., Anal. Chem., 1976, 48, 693. Pierce, F. D., and Brown, H. R., Anal. Chem., 1977, 49, 1417. “Hamilton Harbour Study 1975,” Ontario Ministry of the Environment, Water Resources Branch, Elkei, O., Great Lakes Water Quality Data, personal communication, 1976. Livingstone, D. A., “Data of Geochemistry,” Sixth Edition, “Chemical Composition of Rivers and Lakes,” Geological Survey Professional Paper 440-G, Geological Survey, US Department of the Interior, 1963. Toronto, Ontario, Canada, August 1977. Cheam, V., Mudroch, A., Sly, P. G., and Lum-Shue-Chan, J., Gveat Lakes Res., 1976, 2, 272. Agemian, H., and Cheam, V., Anal. Chim. A d a , 1978, 101, 193. Traversy, W. J., Goulden, P. D., Sheikh, Y. M., and Leacock, J . R., “Levels of Arsenic and Selenium in the Great Lakes Region,” Scientific Series No. 58, Inland Waters Directorate, Ontario Region, Water Quality Branch, Burlington, Ontario, 1975. Schramel, P., Samsahl, K., and Pavlu, J., Int. J . Environ. Stud., 1973, 5 , 37. Morette, A., and Divin, J., Ann. Pharm. Fr., 1965, 23, 169. Received September 28th. 1979 Accepted Mavch 6th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500737

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

744 Analyst, August, 1980, Vol. 105, $9. 744-750 Automated Procedure for the Determination of Soluble Arsenic Using Hydride Generation Atomic-absorption Spectroscopy M. H. Arbab-Zavar and A. G. Howard* Department of Chemistry, The University, Southampton, Hampshire, SO9 5NH An automated procedure for the determination of soluble arsenic, using hydride generation atomic-absorption spectroscopy, is described and opti- mised operating conditions are derived. Interferences are observed in the presence of silver( I), gold(II1) , iron( III), platinum( IV), antimony( III), strontium(II), fluoride and sulphide but can be overcome by suitable pre- treatment procedures. The detection limit (based on twice the standard deviation for 15 blank measurements) is 0.90 ng ml-l of arsenic for arsenic(III), arsenic(V) and methylarsenic species.A 10% negative bias of results is observed in the determination of dimethylarsenic species. Keywords : Arsenic determination ; JLydride generation ; atonzic-absorption spectroscopy ; interferences Arsenic is a widespread element that has found extensive use in medicinal, agricultural and industrial fields. Although arsenic does not appear to be accumulated by man1 it has been implicated in the development of hyperkeratosis and skin cancer,:! lung cancer3 and arterio- ~clerosis.~ In natural waters it is considered to represent a significant health hazard5 and as such it is important that suitable sensitive methods of analysis are available for its routine determination. Although many techniques have been applied to the analysis of arsenic,6 only spectro- photometry and atomic-absorption spectroscopy have been widely used.The standard method of the American Public Health Association7 is based on the colorimetric reaction between arsine and silver diethyldithiocarbamate (SDDC) . This method is slow, compara- tively insensitive, of poor precision in inexperienced hands and is limited to the determina- tion of inorganic arsenic(II1) and arsenic(V) species. Although other colorimetric procedures are available,s-ll few are superior to the SDDC procedure. Owing to the inherent insensitivity of the flame atomic-absorption spectroscopy of arsenic,12 the conversion of arsenic compounds into arsines, followed by atomisation in a heated atomiser tube, has been widely adopted.13-15 Using the hydride generation techniques very high sensitivity can.be achieved, but only at the expense of a comparatively long analysis time and constant attention from the operator.Procedures for the automation of arsenic analysis by hydride generation atomic-absorption spectroscopy have been described by several workers.16-18 Major interferences have been observed from those elements which consume reductant and thereby reduce the efficiency of arsine production,lg but little information is available on overcoming such problems. In this paper the optimised automation of the hydride generation atomic-absorption spectroscopic determination of arsenic is described ; interferences have been investigated and treated. The procedure retains much of the inherent sensitivity of the manual hydride generation technique whilst allowing the efficient processing of large numbers of samples.Experimental Reagents and Glassware Unless otherwise stated all chemicals were of analytical-reagent grade. Glassware was soaked in nitric acid (1 + 9), rinsed with distilled water and dried before use. Stock 1000 pg ml-l solutions of arsenic were prepared from arsenic trioxide, disodium hydrogen arsenate, methylarsonic acid (disodium salt) and dimethylarsenic acid (sodium salt), and were checked for total arsenic content by atomic-absorption spectroscopy using * To whom correspondence should be addressed.ARBAB-ZAVAR AND HOWARD 745 an air - acetylene flame. Solutions of lower concentration were prepared immediately before use. Control of pH was achieved using potassium dihydrogen orthophosphate (O.€ M) - orthophosphoric acid for pH 1-3, sodium acetate (0.2 M) - acetic acid for pH 4-7 and borax (0.25 M) adjusted to give pH values between 8 and 11.NaBH4 F’ &-- Waste Proportioning pump (2.5 rnl min-’ ) Nitrogen 28-turn coil separator qua rtz-tu be atom iser 4 Waste Fig. 1. Schematic diagram of the arsine generation manifold. Apparatus The apparatus consisted of a Hook and Tucker A40 Mk I1 automatic sampler, a Technicon AutoAnalyzer proportioning pump fitted with 2.18 mm i d . silicone-rubber tubing giving a flow-rate of approximately 2.5 ml min-l on all channels, and standard AutoAnalyzer com- ponents. The manifold arrangement, the major role of which is the reduction of the arsenic compounds to arsines and separation of the gas phase from the reactants, is shown schema- tically in Fig.1. Samples from the autosampler are mixed with hydrochloric acid and then with sodium tetrahydroborate(II1) solution. Rapid evolution of hydrogen from the decomposition of the sodium tetrahydroborate(II1) leads to the formation of gas pockets that segment the liquid stream whilst it travels through the 28-turn mixing coil (Technicon No. K105-0083). A flow of nitrogen carrier gas takes the liquid - gas mixture rapidly to a preliminary gas - liquid separator (Technicon No. K116-0122) and from there to the separa- tion chamber, where residual solvent droplets sediment out. The gas stream is then swept into a quartz tube (14 cm x 8 mm i.d.) atomiser aligned in the optical path of a Varian- Techtron AA175 AB background-corrected atomic-absorption spectrometer.In order to achieve atomisation of the arsine, the quartz tube atomiser is heated with an air - acetylene flame.20 The following spectrometer conditions were used throughout the work : lamp current, 7.0 mA; wavelength, 197.3 nm; spectral band pass, 1 nm; and damping, B. Sample Digestion Acid procedure A sample aliquot (less than 1 g) is boiled with 10 ml of concentrated sulphuric acid until the sample is dry. After cooling, concentrated sulphuric acid (10 ml) and concentrated nitric acid (10 ml) are added and the mixture is heated almost to boiling. The digestion is completed by the careful dropwise addition of 10 ml of hydrogen peroxide solution (30% m/V) to the sample and evaporation of the sample to a final volume of approximately 5 ml.Alkaline fusion sodium hydroxide (1 g ) . 10 ml of distilled water. An accurately weighed sample aliquot (less than 1 g) is fused in a zirconium crucible with The melt is heated for 10 min and, after cooling, is dissolved in746 ARBAB-ZAVAR AND HOWARD : AUTOMATED PROCEDURE AnaZyst, VOZ. 105 Results Optimisation of Experimental Conditions Reduction PH The redox potential of sodium tetrahydroborate(II1) is highly dependent on pH and the reduction of the four arsenic species was therefore studied over the pH range 1-9 and at high acid concentrations. For these experiments the “acid” reagent reservoir was filled with buffer solution of the appropriate pH or with dilute hydrochloric acid, and the sodium tetrahydroborate(II1) concentration was maintained at 2y0.The results of these experiments are shown in Fig. 2 as a graph of response due to 200 ng ml-l arsenic solutions versus acidity. The pH of samples of low buffer capacity had no effect on the response of the instrument. \ ,./- . . \ \ \ \ \ I 1 I I I 6 4 2 2 4 6 8 Hydrochloric acid strength/mol dmP3 PH Fig. 2. Effect of reduction acidity on arsine yield, 200 ng ml-1 arsenic solutions with sodium tetra- A, Arsenic(V) ; B, methylarsenic; C, dimethylarsenic; hydroborate(II1) concentration held at 2% m/ I/. and D, arsenic(II1). Sodium tetrahydro borate(III) conce.ittration The concentration of the sodium tetrahydroborate(II1) solution was varied from 1 to 5% m/V in distilled water, whilst maintaining the acid concentration at 3 M.The peak heights due to 200 ng ml-1 arsenic solutions were monitored as a function of sodium tetrahydro- borate(II1) concentration and are presented in Fig. 3. 2 I c g x m 0.)- C 8 - 6 - 4 - R *- n 1 2 3 4 5 a Sodium tetrahydroborate ( I I I ) concentration, % mIV Fig. 3. Effect of sodium tetra- hydroborate(II1) concentration on arsine yield using 200 ng ml-I arsenic solutions and 3 M hydrochloric acid. A, Arsenic(II1). arsenic(V) and methylarsenic; B, dimethylarsenic. Carrier gas JEow-rate The effect of altering the carrier gas flow-rate through the atomiser tube is influenced by both the instrument response time and the steady-state arsenic concentration in the atomisa-August, 1980 FOR SOLUBLE ARSENIC BY HYDRIDE GENERATION AAS 747 tion cell.In order to investigate the effect of the nitrogen carrier gas flow-rate on instru- ment sensitivity the nitrogen flow-rate was varied from 0 to 900 ml min-l and the response due to 200 ng ml-l arsenic solutions was monitored (Fig. 4). 1 1 200 400 600 200 400 600 Nitrogen flow rate/rnl min-' Arsenic concentratiodpg ml-' Fig. 4. Influence of carrier gas flow-rate on peak height obtained for 200 ng ml-l arsenic solutions. Fig. 5. Typical calibration graphs. A, Arsenic (I I I), arsenic (V) and A, Arsenic(III), arsenic(V) and methyl- methylarsenic : B, dimethylarsenic. arsenic : B, dimethylarsenic. Optimased conditions The following conditions were considered to give optimum instrument response : sampling time, 30 s; wash time, 40 s ; sodium tetrahydroborate(II1) concentration, 2% m/V; hydro- chloric acid concentration, 3 M ; and nitrogen flow-rate, 100 ml min-l.A typical calibra- tion graph is shown in Fig. 5. When samples containing relatively high arsenic concentrations are interspersed between the low-concentration samples, a small amount of carry-over can be observed using the 40-s wash interval. Under such conditions, wash intervals of 150 and 250 s have been found to be sufficient to overcome carry-over from 10 and 100pgml-l arsenic samples, respec- tively. Based on 20 replicate analyses, the relative standard deviation of the technique was 1.7% for 100 ng ml-1 arsenic samples, increasing to 8.1% for 10 ng ml-l arsenic samples. The detection limit (based on twice the standard deviation of 15 blank measurements) is 0.90 ng ml-l of arsenic as arsenic(III), arsenic(V) or methylarsenic. Interferences Identijcation Possible interference effects were assessed by the analysis of standard 200 ng ml-l arsenic solutions in the presence of an excess of foreign ions.The results of these analyses are presented in Table I. Interference effects were observed in the presence of silver(I), iron(III), gold(III), platinum(IV), antimony(III), strontium(II), fluoride and sulphide. No effects (Le., less than 2% enhancement or depression of response) were observed for up to 10 pg ml-l of aluminium(III), boron( 111) , bismuth (111) , calcium( 11) , cadmium (11) , cobalt (11) , chromium (VI) , copper (I I), germanium(1V) , mercury( 11) , potassium (I), lanthanum (111) , magnesium(II), manganese(II), molybdenum(VI), nickel(II), lead(II), selenium(VI), tin(II), tellurium(IV), tellurium(VI), vanadium(1V) or zinc(I1) or 100 pg ml-l of nitrite, bromide, iodide or cysteine hydrochloride.In addition, 1000 pg ml-1 of calcium, potassium, hydrogen carbonate, chloride, nitrate, perchlorate and sulphate did not affect the analytical result. Treatment solution of the sample. extract the sample (pH 2) with a 0.005 M solution of dithizone in dichloromethane. Antimony(III). Add 5 ml of 0.5 M pyrogallol solution to 25 ml of a 0.5 M sodium tartrate GoZd(1II). Alternatively, Leave the sample for 2 h prior to analysis. Make the sample solution (pH 2) 2.5 mM in thiosemicarbazide.748 Au(II1) . . Fe(II1) . . Pt(1V) .. Sb(II1) . . Sr(I1) .. F- . . . . sa- .. . . ARBAB-ZAVAR AND HOWARD : AUTOMATED PROCEDURE TABLE I AnaZyst, VoZ. 105 ASSESSMENT OF INTERFERENCES Concentration/ pg ml-l .. 0.1 1.0 .. 0.1 1.0 .. 0.1 1.0 .. 0.1 1 .o .. 0.1 1 .o . . 0.1 1.0 10 10 10 10 10 10 .. 10 . . 1 Depression of signal due to 200 ng ml-l of arsenic, yo Arsenic( 111) Arsenic(V) Methylarsenic Dimethylarsenic r A \ 5 10 32 43 16 41 17 19 50 19 10 10 8 9 10 18 25 60 60 57 68 34 30 50 45 77 84 72 83 17 48 52 50 55 18 20 17 22 PZatinum(IV) and strontium(I1). Add 1% m/V 1,lO-phenanthroline to the 3 M hydro- For all species except dimethylarsenic, 0.1% m/V 1,lO-phen- Iron(I1I). Add 1 ml of 0.1 M cupferron to 10 ml of the sample solution (pH 1). Fluoride. Add 0.01 M lanthanum nitrate and 0.01 M alizarin fluorine blue to the 3 M SuZphide. Fit a lead acetate-impregnated glass-wool scrubber to the exit tube of the chloric acid or sample.anthroline solution is sufficient. hydrochloric acid or sample. separation chamber. Arsenic(V) Methylarsenic Dimethylarsenic 1 Fig-. 6. Treatment of interferences. 1, Arsenic alone; 2, arsenic + interferents; and 3, arsenic + interferents after treatment.August, 1980 FOR SOLUBLE ARSENIC BY HYDRIDE GENERATION AAS 749 The efficiency of these treatments was assessed by the analysis of solutions containing 200 ng ml-l of arsenic together with 10 pg ml-l of silver(I), gold(III), platinum(IV), antimony(II1) and fluoride with and without treatment. Those analyses carried out with treatment involved extraction of the sample (pH 2) with dithizone followed by the addition of a reagent “cocktail” containing 1 ,lo-phenanthroline, pyrogallol and alizarin fluorine blue reagents to the sample.As washing powders are a potentially large source of arsenic in effluent and river waters, the analysis procedure has been applied to the determination of arsenic in several washing powder samples. The analyses were carried out using both acid and alkali fusion digestions and the results are presented in Table 11. Typical traces are shown in Fig. 6. TABLE I1 ANALYSIS OF WASHING POWDERS Values are the means of 5 determinations. Arsenic contentlpg g-1 I I Sample Acid digest Alkaline fusion A . . . . 1.23 f 0.03 1.25 0.04 B . . . . 1.70 f 0.05 1.67 f 0.05 c . . . . 0.24 f 0.03 0.27 f 0.02 Discussion and Conclusions Optimised conditions have been developed for the automated determination of arsenic by atomic-absorption spectroscopy following conversion of arsenic compounds into their corre- sponding arsines.Below pH 1, reduction and atomisation of arsenic(III), arsenic(V) and methylarsenic are essentially complete and the instrument response is related to concentra- tion irrespective of the arsenic species tested. Dimethylarsenic compounds, however, give a lower response than would be expected based on arsenic content and may represent incomplete reduction, condensation of dimethylarsine in the transport tubes or incomplete atomisation. The optimum sodium tetrahydroborate(II1) concentration is 2% m/ V , efficient reduction of dimethylarsenic being obtained whilst the intrinisic alkalinity of the sodium tetrahydroborate(II1) is sufficiently low to maintain effective reduction of the other arsenic species.Although at high sodium tetrahydroborate( 111) concentrations the responses due to the four arsenic species converge, destruction of dimethylarsenic by digestion prior to analysis is recommended. The nitrogen carrier gas plays an important role in determining the sensitivity and speed of analysis. As can be seen from Fig. 4, an increase in the flow-rate decreases the sensitivity by reducing the residence time of atomic arsenic in the absorption cell. However, at the same time high carrier gas flow-rates give rise to improved equilibration of the system towards sample changes and help to cool the gas transit lines. A compromise between these two factors led to the selection of a carrier gas flow-rate between 100 and 200 ml min-l.Interferences that have so far been identified are silver(I), gold(III), iron(III), strontium(II), platinum(IV), antimony(III), fluoride and sulphide. By the addition of a “cocktail” of pyrogallol, thiosemicarbazide, 1,lO-phenanthroline, cupferron, lanthanum nitrate and aliazarin fluorine blue to the sample and by fitting a lead acetate scrubber to the separation chamber, all of the identified interference effects can be overcome without a significant increase in the reagent blank. The technique described provides a rapid and sensitive method for the assessment of arsenic levels in a wide range of sample types. With samples believed to contain dimethyl- arsenic, and where a maximum under-estimate of approximately 10% is unacceptable, preliminary destruction of dimethylarsenic species must be carried out prior to the deter- mination step. Although most wet-ashing procedures fail to destroy organoarsenic com- pounds efficiently, two digestion procedures that have proved suitable are fusion with sodium hydroxide and wet ashing with nitric acid - sulphuric acid - hydrogen peroxide.750 1.2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. ARBAB-ZAVAR AND HOWARD References Schroeder, H. A,, and Balassa, J., J . Chronic Dis., 1966, 19, 85. Fierz, U., Dermatologica, 1965, 131, 41. Osburn, H. S . , Cent. A f r . J . Med., 1957, 3, 215. Heydorn, K., Clin. Chim. Acta, 1970, 28, 349. Louria, 11. B., Joselow, M. M., and Browder, A. A., Ann. Intern. Med., 1972, 76, 307. Talmi, Y., and Bostick, D. T., J . Chromatogr. Sci., 1975, 13, 231. American Public Health Association, American Water Works Association and Water Pollution Control Federation, “Standard Methods for the Examination of Water and Waste Water,” Thirteenth Edition, American Public Health Association, New York, 1971, p. 62. Kellen, G. J., and Jaselskis, B., Anal. Chem., 1976, 48, 1538. Cordebard, H., and Louis, L., Anal. Chim. Acta, 1954, 27, 204. McChesney, E. W., Anal. Chem., 1949, 21, 880. Association of Official Agricultural Chemists, “Official and Tentative Methods of Analysis,” Eighth Parsons, M. L., Smith, B. W., and Bentley, G. E., “Handbook of Flame Spectroscopy,” Plenum, Chu, R. C., Barron, G. P., and Baumgarner, P. A. W., Anal. Chem., 1972, 44, 1476. Smith, R . C., and Van Loon, J. C., Anal. Chim. Acta, 1977, 93, 61. Shaikh, A. V., and Tallman, D., Anal. Chem., 1977, 49, 1093. Vijan, P. N., and Wood, G. R., A t . Absorpt. Newsl., 1974, 13, No. 2, 33. Pierce, F. D., Lamoreaux, T. C., Brown, H. R., and Fraser, R. S., .4ppl. Spectrosc., 1976, 30, 38. Fishman, M., and Spencer, R., Anal. Chem., 1977, 49, 1599. Pierce, F. D., and Brown, H. R., Anal. Chem., 1976, 48, 693. Thompson, K. C., and Thomerson, D. R., Analyst, 1974, 99, 595. Edition, Association of Official Agricultural Chemists, Washington, D.C., 1955. New York, 1975. Received October 24th, 1979 Accepted March 13th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500744

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Analyst, August, 1980, Vol. 105, pp. 751-755 751 Arsenic Speciation: Limitations with Direct Hydride Analysis T. A. Hinners” US Environmental Protection Agency, Analytical Chemistry Branch (MD-78), Research Triangle Park, N.C. 27711, USA Speciation of inorganic arsenic(II1) by hydride evolution directly into an atomic-absorption system was found to be subject t o error when organic forms of arsenic (such as dimethylarsinic acid) were present. Organic forms of arsenic can produce an underestimation of total arsenic when the hydride response from concentrated acid is quantitated against inorganic arsenic. Hydride responses from solutions of various acidities are reported for di- methylarsinic acid, monomethylarsonic acid, inorganic arsenic (111) and inorganic arsenic(V). Even simultaneous equations did not provide a means of resolving mixtures of the four arsenic species investigated by using direct analysis of the evolved hydrides.Keywords : Arsenic speciation ; arsine generation ; atomic-absorption spectro- Photometry ; environmental analysis Aggett and Aspelll reported the application of an arsenic speciation method for the deter- mination of inorganic arsenic(II1) and total arsenic in environmental water samples. In our experience the presence of methylated arsenic compounds can produce errors in speciation methods, like that described by Aggett and Aspel1,l where hydrides of arsenic are generated directly into a detector. Some environmental water samples have been reported2-4 to contain methylated arsenic compounds in amounts as high as 1644% of the total arsenic.Even where met hylated arsenic compounds have not been used agriculturally, inorganic arsenic can be converted into methylated forms in the envir~nment.~-~ Most of the inorganic arsenic ingested by humans is r e p ~ r t e d l y ~ v ~ ~ excreted in the urine as methylated forms, which could add organic arsenic to environmental water systems. Dimethylarsinic acid is expected by some investigators* to be “the most abundant methylated arsenic compound in both freshwater and seawater.” Experimental Apparatust A commercial hydride generation cellll (Instrumentation Laboratory Inc.) was used with a Perkin-Elmer, Model 403, atomic-absorption spectrophotometer (set to 193.7 nm) equipped with a deuterium arc background corrector, a nitrogen-entrained air - hydrogen flame and a 10-mV chart recorder with a rapid pen response (requiring less than 0.5 s for full-scale response).A three-slot burner (length 11.2 cm) was used with flow-rates of 4 1 min-l of hydrogen and 23 1 min-l of nitrogen. The slit width on the spectrophotometer was set to 0.7 nm. Reagentsf Arsenious oxide for inorganic arsenic(ll1). Arsenic pentoxide for inorganic arsenic( V ) . Disodium salt of monomethylarsonic acid (MMA). Sodium salt of dimethylarsinic acid (DMA). Sodium tetrahydroborate(II1). Fisher Scientific, S-678. de-ionised water were prepared just before use. USA. US National Bureau of Standards, 83c. Fisher Scientific, A-54. Alpha Products, 12134. Fisher Scientific, S-257. Solutions of 1 and 2% m/V in * Present address: US Environmental Protection Agency, OAD, P.O.Box 15027, Las Vegas, Nev. 89114, t Use of brand names does not constitute endorsement by the US Environmental Protection Agency.752 HINNERS : ARSENIC SPECIATION : LIMITATIONS Analyst, VoL 105 Acetate bufer solution, 2 M (PH 4.8). A 28.6-m1 volume of glacial acetic acid (Fisher Scientific, A-38) plus 68.04 g of sodium acetate trihydrate (Fisher Scientific, S-209) were made up to 500 ml with de-ionised water. Concentrated hydrochloric acid. Redwing solution mixture. ACS reagent grade. An 8.30-g amount of potassium iodide (Matheson , Coleman and Bell, CB 626) plus 5.0 g of L-ascorbic acid (Fisher Scientific, A-61) were made up to 50 ml with de-ionised water (after Siemer et a1.12). Procedure For hydride generation at pH 4.8, 4 ml of acetate buffer were combined in the hydride cell with the sample solution and water to provide a total volume of 8ml.With a septum covering the top port of the hydride generation cell, nitrogen was directed through the solution in the hydride generation cell for 30 s to remove air. The flow of nitrogen was then directed so as to bypass the generator solution. With a PTFE-coated magnet stirring the solution in the hydride generation cell and with the chart recorder pen tracing, 2 ml of the 1 yo sodium tetrahydroborate(II1) solution were injected (using a syringe with a bevelled 23-gauge needle) through the septum in 2-3 s. At 30 s after the start of this injection, the flow of nitrogen was quickly directed through the hydride generating solution for 5 s before the nitrogen was re-directed to bypass the hydride generating solution.With the recorder pen still tracing, the injection of 1 yo sodium tetrahydroborate( 111) solution was repeated, and the response recorded after 30s by again flushing the hydride generating solution with nitrogen. The generator solution was drained and the hydride generation cell was rinsed with de-ionised water. The two absorbance peak-height measurements for each solution were corrected for reagent blank contributions before being summed. Electronic integration values for these transient signals were not as reproducible as the peak heights. Use of a single injection of more sodium tetrahydroborate(II1) solution instead of the two injections produced a premature release of arsenic hydrides into the nitrogen bypass stream to the detector.Purging the hydride generating solution with nitrogen while the sodium tetrahydroborate(II1) solution was injected reduced the analytical sensitivity. The optimum signal to noise ratio for arsenic hydrides was obtained by positioning the burner just below the height at which light from the arsenic lamp began to be obstructed. Treatment with the reducing agent mixture involved mixing 0.5 ml with a 2-ml portion of sample and 1.5 ml of 6 M hydrochloric acid, and allowing at least 30 min for reaction at room temperature before analysis. For hydride generation from hydrochloric acid solution, varous acidities were obtained by using different volumes of concentrated hydrochloric acid combined with the sample solution and water, and the generation procedure was the same as for the studies using acetate buffer except that 2% sodium tetrahydroborate(II1) solution was used instead of the 1 Yo solution (and no acetate buffer was used).Results and Discussion With the system described above, a detection limit of 3 ng was obtained for inorganic arsenic(V) in hydrochloric acid, which agrees with the results of other workers11~13~14 obtained using flame atomic-absorption analysis. For a reagent blank standard deviation corre- sponding to 0.9 ng (n = lo), a detection limit of 3 ng excludes 99.90/, of the individual blank values. For the pulse-purged system described (providing an acceptable detection limit), reaction competition and the generator design prevented complete reaction of arsenic with a single treatment with sodium tetrahydroborate(II1) solution, and the double treatment improved the precision of the measurement.With the pulse purge of the hydride generating solution, a 5-s flush was sufficient because the arsenic hydrides were driven into the detector beam rapidly, as indicated by the return of the recorder pen to the base-line level in 2 s. Aggett and Aspelll attributed the response from environmental water samples buffered near pH 5 solely to inorganic arsenic( 111) when the contribution from inorganic arsenic(V) was insignificant. They evaluated total arsenic in environmental water samples with the response from 5 M hydrochloric acid using an inorganic arsenic(II1) calibration. Braman and Foreback15 reported the use of pH adjustment in the selective hydride generation for inorganic arsenic(II1).August, 1980 WITH DIRECT HYDRIDE ANALYSIS 753 Table I shows the relative hydride-forming behaviour observed for the arsenic compounds investigated.With the system described the hydride response from acetate buffer (pH 4.8) for the dimethylarsinic form (DMA) was found (Table I) to be 85 & 2% of the response for an equal amount of arsenic in the form of inorganic arsenic(II1). A mixture prepared to contain 12 ng ml-l of arsenic in each of the forms inorganic arsenic(III), inorganic arsenic(V) and MMA, plus 60 ng ml-l of arsenic as DMA, produced a hydride response from acetate buffer, indicating the presence of 66 ng ml-l of arsenic in the form of inorganic arsenic(II1) when the response was attributed solely to this arsenic species. TABLE I ATOMIC-ABSORPTION RESPONSES FOR HYDRIDES FROM FOUR ARSENIC SPECIES Linear regression slopes f standard errors for several days and levels.Generating solution pH 4.8* . . 2MHCI . . 4MHC1 . . 6~ HC1 . . 8~ HC1 . . 1 0 ~ HC1 . . Response relative to inorganic arsenic( 111) = 100% , 1 As( 111) W V ) MMA DMA . . 100 f 2 3.3 j, 0.07 12 j, 1 85 j, 2 . . 100 99 99 44 . . 100 101 88 15 . . 100 f 6 99 f 3 69 j, 2 7 j, 0.3 . . 100 96 50 6 . . 100 & 2 103 & 2 29 f 0.3 6 0.2 * Acetate buffer at 1 M in total acetate after dilution by sample. Both our measurements and the measurements by Aggett and Aspelll near pH 5 were conducted with strong buffering. Not more than a 0.25 pH unit change occurred during hydride generation according to Aggett and Aspe1l.l The buffering condition described by Braman and co-workers2J5 is not comparable in strength to those used by us or by Aggett and Aspe1l.l When weak buffering was used, the hydride response from DMA relative to inorganic arsenic(II1) was reduced from 84% to 12y0, but both the precision and magnitude of the inorganic arsenic( 111) response suffered with weak buffering.The buffering capacity is relevant because the sodium tetrahydroborate(II1) reagent forms a basic solution at pH 11 or more (by hydrolysis when an alkali metal hydroxide is not added). Hydride formation from DMA (with pK, 6.19)2 at a stable pH near 5 is consistent with the contention by Braman and Foreback15 that the undissociated molecules are reactive.DMA may not be the only organoarsenic compound that can contribute large amounts of hydride, relative to inorganic arsenic(III), from a strong buffer at pH 4.8. When MMA was treated with the reducing agent mixture before hydride generation from acetate buffer, the response at 103 & 2.4% (n = 26) relative to inorganic arsenic(II1) differed markedly from the 12 & 1% for untreated MMA shown in Table I. Neither inorganic arsenic(II1) nor DMA showed a significant change in hydride response from acetate buffer after treatment with the reducing agent mixture. It is conceivable (but not proved) that the change in behaviour observed with MMA is the result of the conversion of the pentavalent arsenic in MMA to the trivalent (and less dissociated) arsonous [CH,As(OH),] or arsenosomethane (CH,AsO) forms.The hydride response from hydrochloric acid (4 M and above) for IjMA was found (Table I) to be 15% or less of the response for an equal amount of arsenic in the form of inorganic arsenic(II1). The mixture specified three paragraphs above (with a total arsenic concentra- tion of 96 ng ml-l) produced a hydride response from 10 M hydrochloric acid, indicating the presence of only 37 ng ml-1 of total arsenic when the response was attributed solely to inorganic arsenic, i.e., evaluated with an inorganic arsenic calibration. When sufficient sodium tetrahydroborate(II1) reagent is used, inorganic arsenic( 111) and inorganic arsenic(V) respond similarly from hydrochloric acid in our system. The use of 10 M hydrochloric acid provided a means of minimising the hydride contribution from the methylated arsenic com- pounds as part of a speciation investigation.Judging from Table I, the results for the determination of total arsenic in samples containing methylated arsenic using the hydride response from 5 M hydrochloric acid would also be an underestimate when evaluated with an inorganic arsenic calibration. Table I shows that the relative hydride response for bothAnalyst, Vol. 105 methylated arsenic species decreased with increasing acidity of the hydride generating solu- tion. While the decreasing hydride response with increasing acidity of the hydride generating solution may simply result from proton competition for the sodium tetrahydro- borate(II1) reagent, the formation of arsenic cations by protonation may also play a role.Even samples subjected to acid digestion may still contain DMA, as it is reportedly16 ‘ra remarkably stable compound, remaining undecomposed by the action of fuming nitric acid, aqua regia or potassium permanganate even upon heating.” If DMA is not decomposed to inorganic arsenic by a digestion procedure, underestimation of the total arsenic could occur when the evaluation is based on an inorganic arsenic calibration. Other investigators17 have commented on the digestion procedure suggested by Aggett and Aspelll for plant materials in terms of the inorganic arsenic species. When synthetic mixtures were limited to the two inorganic arsenic species, the practicality of compensating (by calculation) for the hydride contribution from inorganic arsenic( V) in determining inorganic arsenic(II1) from acetate buffer responses was confirmed by a statistical evaluation.Inorganic arsenic(V) was evaluated for these two-component mixtures by the difference between the total arsenic (from the hydrochloric acid response) and the inorganic arsenic(II1). Although Table I shows that the hydride response from acetate buffer for inorganic arsenic(V) is only 3.3% of the inorganic arsenic(II1) response, the bulk of the hydride response from acetate buffer can stem from inorganic arsenic(V). For example, 60% of the hydride response from acetate buffer observed for a mixture prepared to contain 18 ng ml-l of inorganic arsenic(II1) and 600 ng ml-l of inorganic arsenic(V) was attributable to the latter species.The hydride response from acetate buffer for inorganic arsenic(V) was not merely contamination of the inorganic arsenic(V) by inorganic arsenic( 111) because the response did not decrease with additional treatments with sodium tetrahydroborate(II1) solution. Some other inve~tigatorsl~~ (but not all) have observed a contribution from inorganic arsenic(V) under the conditions used to determine inorganic arsenic(III), and the extent of the contribution from inorganic arsenic(V) may depend on the buffering capacity of the hydride generating solution. With weak buffering, the pH of the hydride generating solution can increase by several units on addition of the sodium tetrahydroborate(II1) reagent, which can alter the hydride response by altering the dissociation (and thereby the reduction potential) of the arsenic species.While overlapping hydride responses for mixtures limited to inorganic arsenic( 111) and inorganic arsenic(V) did not prevent speciation, resolving mixtures of all four arsenic com- pounds (specified in Table I) by simultaneous equations (using four generating conditions) proved unreliable, even though the statistical analysis showed less than 1 yo interaction among the arsenic species. 754 HINNERS : ARSENIC SPECIATION : LIMITATIONS Conclusion The warning offered in this paper concerns the application of a method, not the method per se ( i e . , in the absence of methylated arsenic compounds the published methodl could provide acceptable data). Arsenic speciation method^^,^,^^ that involve separation of the hydrides before the measurement step are not subject to the potential errors described here, except that any inorganic arsenic(V) contributions under the conditions used to determine inorganic arsenic(II1) are not physically separable from the inorganic arsenic( 111) because both inorganic arsenic species yield the same hydride.Acknowledgement is made to Dr. Paul Mushak for the MMA, to D. B. Ray for the septum- bearing stopper and for technical assistance, and to J. C. Suggs for the statistical analysis. References 1 . 2. 3. 4. 5 . 6. Aggett, J . , and Aspell, A. C., Analyst, 1976, 101, 341. Braman, R. S., Johnson, D. L., Foreback, C. C., Rmmons, J . M., and Bricker, J . L., Anal. Chem., Andreae, M. O., Anal. Chem., 1977, 49, 820, Shaikh, A. U., and Tallman, ID. E., Anal. Chem., 1977, 49, 1093. Wood, J . M., Science, 1974, 183, 1049. Braman, R . S., in Woolson, E. A., Editor, “Arsenic Pesticides,” ACS Symposium Series No. 7, 1077, 49, 621. American Chemical Society, Washington, D.C., 1975, p. 108.August, 1980 WITH DIRECT HYDRIDE ANALYSIS 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 755 Woolson, E. A., Environ. Health Perspect., 1977, 19, 73. Ridley, W. P., Dizikes, L. J., and Wood, J. M., Science, 1977, 197, 329. Smith, T. J., Crecelius, E. A., and Reading, J. C., Environ. HeaEth Perspect., 1977, 19, 89. Crecelius, E. A., Environ. Health Perspect., 1977, 19, 147. Hwang, J . Y., Ullucci, P. A., Mokeler, C. J., and Smith, S. B., Jr., Am. Lab., 1973, 5, 43 Siemer, D. D., Koteel, P., and Jariwala, V., Anal. Chem., 1976, 48, 836. Fernandez, F. J., A t . Absorpt. Newsl., 1973, 12, 93. Knudson, E. J., and Christian, G. D., Anal. Lett., 1973, 6, 1039. Braman, R. S., and Foreback, C. C., Science, 1973, 182, 1247. Raiziss, G. W., and Gavron, J. L., “Organic Arsenical Compounds,” American Chemical Society, Diamondstone, B. I., and Burke, R. W., Analyst, 1977, 102, 613. Talmi, Y., and Bostick, D. T., Anal. Chem., 1975, 47, 2145. Chemical Catalog Co., New York, 1923, p. 79. Received October 24th, 1979 Accepted February 25th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500751

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

756 Analyst, August, 1980, Vol. 105, $$. 756-761 Simultaneous Determination of Trace Concentrations of Arsenic, Antimony and Bismuth in Soils and Sediments by Volatile Hydride Generation and Inductively Coupled Plasma Emission Spectrometry Behrooz Pahlavanpour, Michael Thompson and Laurence Thorne Applied Geochemistry Research Group, Department of Geology, Imperial College of Science and Technology, London, SW7 2BP Trace amounts of arsenic, antimony and bismuth in soils and sediments are determined simultaneously by an inductively coupled plasma - volatile hydride method after rapid attack with concentrated hydrochloric acid in sealed tubes. The samples are treated with the acid a t 150 "C for 2 h in capped test-tubes. After addition of potassium iodide solution, the hydrides are formed by mix- ing the solution with sodium tetrahydroborate(II1) solution in a continuous- flow system, and are swept into the plasma by a stream of argon for deter- mination by atomic-emission spectrometry.Acceptable precision and ac- curacy are obtained, and the detection limits for all three analytes are about 0.1 pg g-l. Approximately 200 samples can be analysed by one person in a 2-d cycle. Keywords : Arsenic, antimony and bismuth determination ; hydride genera- tion ; inductively coupled plasma emission spectrometry ; geochemical samples The analytical characteristics of a procedure for the simultaneous determination of arsenic, antimony, bismuth, selenium and tellurium in aqueous solution has been described previ- 0usly.1~2 The elements were reduced to their hydrides by mixing the test solution with a solution of sodium tetrahydroborate(II1) in a continuous-flow system.The hydrides were carried into an inductively coupled plasma (ICP) and determined by atomic-emission spectro- metry. The detection limits obtained were comparable to the best reported for atomic absorption - volatile hydride methods, but the linear calibration ranges were greater. A few transition metal ions (mainly copper and nickel) were found to interfere with the reduction of bismuth, selenium and tellurium, causing reduced efficiency in the production of the hydrides, whereas arsenic and antimony were virtually free from interference problems. The initial purpose of the work partially reported here was to produce a rapid ICP - volatile hydride method by which the five analytes could be determined simultaneously in rock, soil and sediments.However, no rapid method of sample attack has been found that was suitable for bringing the elements into solution simultaneously. A variety of mineral acids and mixtures have been investigated. Strongly oxidising acids are required to effect the dissolution of selenium and tellurium, which would otherwise be reduced to the elements [by trace amounts of reducing agents such as organic matter or iron(II)] and subsequently lost. However, the use of oxidising acids was found to give rise to very low recoveries of antimony, possibly owing to the production of a hydrolysed species of antimony(V) that was not amenable to reduction and hydride formation. In view of this difficulty, the two periodic groups of elements have been treated separately.The determination of selenium in soils and sediments has already been reported3; this paper reports a simple and rapid method of sample attack for the simultaneous determination of arsenic, antimony and bismuth. Experimental Equipment Hydride generator - plasma system generator and an Applied Research Laboratories 29000B quantometer. conditions and procedure were as previously described.l The system consisted of a continuous-flow hydride generator, a Radyne R50 plasma The operatingPAHLAVANPOUR, THOMPSON AND THORNE 757 Screw-cap#ed test-tubes Sample attacks were carried out in Sovirel screw-capped borosilicate-glass test-tubes (160 x 16 mm), with specially made cap liners consisting of a layer of silicone-rubber (Esco type SR70, 3.2 mm thickness) covered with a chemically resistant film (Du Pont Teflon FEP Type A, 0.36 mm thickness).These were more effective than the original liners in preventing leakages, and were designed by Dr. J. Lovell, Barringer Research Inc., Toronto, Canada. Heating block of about 65 mm. blast of air from a domestic vacuum cleaner. Batches of tubes were heated in a thermostatically controlled aluminium block to a depth The Bakelite caps and the tops of the tubes were kept cool by means of a Materials Analytical-reagent grade chemicals and purified water were used throughout. Sample Attack The powdered samples (0.250 g) were weighed into test-tubes and 5 ml of concentrated hydrochloric acid were added to each tube.(For samples containing carbonate minerals, the acid should be added cautiously, in small portions, and the tubes set aside until evolution of gas has ceased.) The tubes were capped and placed in the heating block a t 150 "C for 2 h, then they were removed cautiously, cooled rapidly in a cold water-bath and unsealed. A 5-ml volume of 0.2% m/V potassium iodide solution was added to each tube and the contents were mixed by thorough shaking. The solid residue was allowed to settle (for about 4 h) and the solutions were used directly for the determinations. Determination The arsenic, antimony and bismuth were determined simultaneously in the analyte solution. A calibrating solution was made containing the analytes [as arsenic(III), antimony(II1) and bismuth(III)] at a concentration of 100 ng ml-l in 1 + 1 hydrochloric acid, and was run, together with a blank solution, after every tenth sample solution.Sample solutions containing an analyte at a concentration above the limit of linear calibration (arsenic 800 ng ml-l, antimony 1500 ng ml-l, bismuth 500 ng ml-l) were diluted 10-fold and re-run. Safety Precautions Although no instance of a tube bursting has occurred in the operation of this procedure, the possibility must be guarded against. The operator must be protected from flying glass and fumes while the tubes are hot and care must be exercised to prevent the temperature of the hot block exceeding the required level. Results and Discussion Selection of Sample Digestion Conditions In primary magmatic rocks, arsenic, antimony and bismuth are highly concentrated in sulphide phases, although bismuth has some tendency to replace calcium in apatite and certain silicates.However, in sedimentary rocks, soil and sediment, the three elements are largely bound to precipitated iron(II1) oxide minerals. Hot, concentrated hydrochloric acid is a good solvent for this type of material, but has not been used for the determination of arsenic and antimony because of the volatility of their lower chlorides. A wider range of minerals including many silicates are decomposed by hydrochloric acid at elevated tempera- tures and pressures, but the normal methods for carrying out the attack (in PTFE bombs or heavy-walled sealed glass tubes) are not suitable for the rapid, high-throughput work required in applied geochemistry.The use of hydrochloric acid as a reagent for decomposition of minerals has been reviewed by Dolezal et al.* and Bock.5758 PAHLAVANPOUR et aZ. : DETERMINATION OF As, Sb AND Bi IN SOILS Analjst, VoZ. 105 Sovirel capped test-tubes, when used as described above, can withstand the pressure generated by concentrated hydrochloric acid up to a block temperature of at least 180 "C. Above this temperature the caps have an increasing tendency to fail. The liquid phase adopts a temperature slightly lower than block temperature: with the block at 150 "C the liquid equilibrates at 140 "C within a few minutes. With the cap liners described and a moderate finger torque for fastening, loss of the analytes from the tubes is insignificant. When the tubes were tested under the conditions of the proposed method with 5-ml portions of concentrated hydrochloric acid containing 25, 250 or 1250 ng ml-l of the analytes, no significant loss (ie., greater than the uncertainty of the determination) could be detected.Where the sealing of the tube is imperfect, an obvious loss of acid volume occurs, which enables the analyst to test for adequacy of sealing. Thorough shaking after the attack is necessary as arsenic tends to accumulate in the cool upper part of the tube. The effect of different block temperatures for a 2-h attack with concentrated hydrochloric acid was tested for a wide variety of samples within the range 25-180 "C. All samples show increasing proportions of the analytes up to 150 "C, with no further increase at 180 "C.How- ever, at the lower temperatures the proportions of the elements extracted varied markedly from sample to sample, presumably depending on the mode of occurrence of the analytes and the resistance to dissolution of the host minerals. Effect of the Attack on Various Minerals The efficacy of the attack on a variety of minerals was tested by subjecting samples to the procedure described, where necessary subjecting the separated residues to identification by X-ray diffraction. I t was found that the iron minerals magnetite, haematite, siderite and geothite, in which the analytes would normally be expected to be concentrated, were virtually completely dissolved, as was the sample of apatite. Some silicates (montmorillonite, biotite) were also completely destroyed, leaving only a residue of amorphous silica.In other silicate minerals (pyroxene, kaolinite, hornblende, muscovite, albite, orthoclase) the lattice was apparently unaffected by the treatment, although Foster6 has sh.own that hydrochloric acid (under milder conditions) is an effective reagent for the extraction of trace amounts of a number of base metals from some of these minerals. The attack was also attempted on small (about 20-mg) samples of various sulphide minerals, although these are not normally encountered in stream sediments or soils. It was found that galena, sphalerite, chalcopyrite and pyrrhotite were alrnost completely dissolved, whereas pyrite and marcasite were hardly affected. Effect of Organic Matter on Recovery of Bismuth Initial work suggested that low recoveries of bismuth were occurring from samples that contained appreciable amounts of organic matter.However, no reliably analysed standard samples with suitable levels of bismuth and organic carbon were available to test this possi- bility. Accordingly, three samples of soils containing a range of organic carbon contents were treated by the standard method, both alone and after spiking with 0.25-, 2.5- or 12.5-pg amounts of arsenic, antimony and bismuth. The solutions were then analysed in the normal manner. A low recovery of any analyte would be indicated by the differences between the original samples and the respective spiked sample, compared with the concentration added. It was found that recoveries of arsenic and antimony were consistently good, no significant losses being observed.Bismuth, however, gave low recoveries, the loss being roughly in proportion to the organic matter content, and almost independent of the concentration of bismuth. This loss amounted to about 50% for a soil sample containing 19.9% of organic carbon. Addition of 0.10 ml of bromine to the reaction mixture before the heating stage resulted in much better recoveries of bismuth, suggesting that bismuth may be partially reduced to the element in the presence of the organic matter. However, the use of bromine brought about very low recoveries of antimony and therefore was not adopted for general use. Reaction conditions milder than those in the proposed method might be suitable for optimising the recovery of bismuth in organic-rich samples.August, 1980 AND SEDIMENTS BY HYDRIDE GENERATION AND ICP SPECTROMETRY 759 Effect of Potassium Iodide on Formation of the Hydrides Potassium iodide has the effect of slightly enhancing the analytical response of arsenic(II1) and antimony(II1) and slightly suppressing the response of bismuth(II1) in the ICP - hydride system.2 It also brings about the reduction of arsenic(V) and antimony(V) to the +3 state, which is necessary in this system because of the diminished response given by the higher oxidation states, especially antimony(V) .Finally, the potassium iodide is required to reduce the interference with bismuth due to copper. Arsenic and antimony can be determined by the ICP - hydride system with no interference from the other elements at concentrations likely to be encountered in the analysis of soils and stream sediments.2 Bismuth is affected by copper in concentrations above 2 mg 1-1 in the test solution, equivalent to 80 mg kg-l of copper in the solid sample when it is attacked by the method given.In the presence of potassium iodide, however, the critical level of copper (at which the onset of interference is apparent) is increased to about 80 mgl-1 in solution (Fig. l), equivalent to 3200 mg k g l in the sample. This level of copper is rarely exceeded in soil and sediment samples. L (D P Q a 0.1 1 .o 10 100 1 000 Cu concentration/mg I-’ Fig. 1. Effect of copper concentration on the recovery of bismuth from A, solutions containing 0.1 yo potassium iodide (closed circles) and R, solutions with no added potassium iodide (open circles).Accuracy and Precision of the Method The method described has been applied to a variety of standard analysed materials, mostly soils and rocks. Table I shows a comparison of values obtained by the proposed method and the values obtained for a series of soil standards from the Canadian Certified Materials Project (CCMP) . The proposed method produced comparable although generally slightly lower values than those currently recommended, but the number of laboratories returning data for arsenic, antimony and bismuth was small, and most of the results obtained for bismuth are close to the detection limit of the proposed method and apparently the CCMP method. TABLE I RESULTS PRODUCED ON THE CANADIAN CERTIFIED MATERIALS PROJECT STANDARD SOILS COMPARED WITH THE RECOMMENDED VALUES SO-1, SO-2 and SO-3 are single determinations; SO-4 is the mean and standard deviation of eight determinations on separate portions of the material.Sample Result so-1 . . . . Found s o - 2 . . . . Found SO-3 . . . . Found SO-4 . . . . Found Recommended Recommended Recommended Recommended As/mg kg-l 1.94 1.9 f 0.3 0.77 1.2 f 0.2 2.32 2.6 f 0.1 6.45 & 0.07 7.1 f 0.7 Sb/mg kg-1 Bi/mg kg-1 0.12 0.24 0.2 0.5 0.1 0.03 0.1 0.1 0.22 0.03 0.3 0.1 0.7 0.1 0.25 & 0.03 0.19 & 0.02760 Aatalyst, vol. 105 Results obtained on some of the US Geological Survey (USGS) rock standards are com- pared with recommended values’ in Table 11. Although the proposed method might be expected to produce low values on igneous rocks (because of the resistance of many silicates to the attack), agreement is reasonably close for arsenic and antimony, considering the low levels involved.Values obtained for bismuth are around the detection limit and not strictly comparable to the recommended values, which were obtained by a neutron-activation method with a much lower detection limit. PAHLAVANPOUR et al. : DETERMINATION OF As, Sb AND Bi IN SOILS TABLE I1 RESULTS OBTAINED BY THE PROPOSED METHOD ON SOME USGS STANDARD ROCKS COMPARED WITH ACCEPTED VALUES’ Sample Result As/mg kg-l Sb/mg kg-1 Bi/mg kg-1 w-1 . . . . Found 0.89 1.18 <0.04 G-1 .. . . Found 0.50 0.55 0.12 G-2 . . . . Found 0.22 0.12 0.12 Accepted 1.9 1.0 0.046 Accepted 0.5 0.31 0.065 Accepted 0.25 0.1 0.043 GSP-1 .. Found Accepted 0.12 3.20 0.18 0.09 3.1 0.037 The method has also been applied to the Geochemical Exploration Series (GXR) standards, which were -produced by the USGS in conjunction with the Association of Exploration Geochemists, and which contain a higher range of analyte values. The GXR standards consist of a variety of rocks and soils. In Table I11 the results obtained are compared with those of other workers. However, the comparisons are riot completely satisfactory because of the dearth of reliable analyses. The “extracted values’’ cited are based on a tabulation of results produced by many laboratories on a world-wide basis.8 They were produced by one of the authors (M.T.) by a weighted average of the median result obtained for each given method. In view of the large variations between laboratories and between methods in the tabulation, especially for antimony and bismuth, the “extracted values” must be regarded as suspect.The other results that are cited are independent of the tabula- tion. TABLE I11 COMPARISON OF THE RESULTS OBTAINED BY THE PROPOSED METHOD AND BY OTHER WORKERS ON THE GEOCHEMICAL EXPLORATION SERIES STANDARD ROCKS AND SOILS Results by the proposed method show the mean and standard deviation (in parentheses) of eight replicated samples taken through the entire procedure. A11 results in mg kg-l. Standard sample Element Arsenic . . . . Antimony .. Bismuth . . .. Source of data This work “Extracted value” This work “Extracted value” Hannaker and Hugh9 This work “Extracted value” Hannaker and Hugh9 VietslO GXRl GXR2 395 (6.4) 23.4 (0.30) 390 22 159 (5.5) 46.5 (0.19) 115 41 154 49 3419 (70) 0.59 (0.09) 1600 5 939 49 I725 0.4 GXR3 GXR4 4638 (124) 107 (2.9) 4 200 95 66.0 (2.7) 4.6 (0.05) 30 4.!j 8.1 - 0.29 (0.06) 17.8 (0.07) 20 24 16 <0.2 21.2 - GXR5 GXR6 10.7 (0.26) 398 (10.8) 11 306 1.15 (0.04) 3.36 (0.08) 1.5 3 3.4 2.3 0.32 (0.04) 0.27 (0.06) 7 11 3 1 0.4 0.2 Agreement with the “extracted value’’ is close for arsenic, less so for antimony and negligible for bismuth.For antimony the results of Hannaker and Hughsg (produced by solvent extraction and flame atomic-absorption spectrometry) agree reasonably well, but less well for bismuth. Apart from the value for GXRl the bismuth results obtained by the proposed method agree with those of Viets,l* also produced by solvent extraction and flameAztgust, 1980 AND SEDIMENTS BY HYDRIDE GENERATION AND ICP SPECTROMETRY 76 1 atomic-absorption spectrometry. The high result produced by the proposed method on GXRl is noteworthy.GXRl is a sample of jasperoid, high in iron and silica, and possibly more amenable to the attack with hydrochloric acid than to the attacks used by other workers. The precisions obtained by replicate analysis of the GXR samples (Table 111) and one of the soil samples (Table I) are satisfactory for most purposes, and suggest that the practical detection limits for all three elements is of the order of 0.1 pg g-l. Approximately 200 samples can be analysed by one person in a 2-d cycle. This work was part of a project funded by a grant from the Natural Environment Research Council. References 1 . 2 . 3. 4 . 5 . 6. 7 . 8 . 9. 10. Thompson, M., Pahlavanpour, B., Walton, S. J., and Kirkbright, G. F., Analyst, 1978, 103, 568. Thompson, M., Pahlavanpour, B., Walton, S. J., and Kirkbright, G. F., Analyst, 1978, 103, 705. Pahlavanpour, B., Pullen, J. H., and Thompson, M., Analyst, 1980, 105, 274. Dolezal, J ., Povondra, P., and Sulcek, Z., “Decomposition Techniques in Inorganic Analysis,” Bock, R., “Decomposition Methods in Analytical Chemistry,” translated by Marr, I. L., Blackie, Foster, J . R., Can. Min. Metall. Bull., 1973, 66, 85. Flanagan, F. J., “1972 Compilation of Data on USGS Standards,” U S Geol. Surv. P ~ o j . Pup., 1976, Allcott, G. H., and Lakin, H. W., “Tabulation of Geochemical Data Furnished by 109 Laboratories for 6 Geochemical Exploration Reference Samples,” U S Geol. Surv. Open File Rep., 1978, No. translated by Hughs, D. O., Floyd, P. A., and Barratt, M. S., Iliffe Books, London, 1968. Glasgow, 1979. NO. 840, pp. 131-183. 78-163. Hannaker, P., and Hughs, T. C . , J . Geochem. Explor., 1978, 10, 169. Viets, J . G., Anal. Chem., 1978, 50, 1097. Received March 4th, 1980 Accepted March 26th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500756

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

762 Analyst, August, 1980, Vol. 105, pp. 762-767 Specific and Sensitive Spectrophotometric Determination of Cobalt with 3-(2'-Thiazolylazo)- 2,6=diaminotoluene F. Garcia Montelongo, C. R. Tallo Gonzalez and V. Gonzalez Diaz Department of Analytical Chemistry, University of La Lagunn, Tenerife, Caqiary Islands, Spain Cobalt and 3- (2'-thiazolylazo)-2,6-diaminotoluene react in an acidic sodium acetate medium in the presence of vanadate to give several complexes. The soluble blue complex in a strong perchloric acid medium ( H , = -0.5) obeys Beer's law between 0.05 and 0.60 p.p.m., with a molar absorptivity of 9.74 x lo4 1 mol-l cm-l a t 590 nm, a Sandell sensitivity index of 0.60 ng and a relative error of 0.13%. The method is highly specific and has been applied to the spectrophotometric determination of cobalt in low-alloy steels, hydro- fining catalysts and high-purity nickel salts.Keywords : 3-(2'-Thiazolylazo)-2,6-diaminotoluene reagent ; cobalt determina- tion ; spectrophotoinetry Cobalt (11) and cobalt (111) react readily with organic compounds containing donor atoms such as oxygen, nitrogen or sulph~r.l-~ In recent years very sensitive reagents derived from o-aminoazopyridine have been intr~duced,~ the cobalt complexes of which have molar absorptivities ranging from 1.07 x lo5 1 mol-l cm-l for 4-(2'-pyridylazo)-l,3-diaminobenzene to 1.42 x 105 1 mol-l cm-l for 5-(3',5'-dichloro-2'-pyridylazo)-2,4-diaminotoluene. The spectrophotometric determination of cobalt using these reagents suffers from interference mainly by iron and palladium, re~pectively.~,~ In a previous paper7 we have described the synthesis of 3-(2'-thiazolylazo)-2,6-diamino- toluene (2,6-TADAT) and studied its reaction with palladium.Cobalt reacts with 2,6- TADAT only after oxidation to cobalt(II1). It has been shown8 that when solutions of cobalt(I1) and 2,g-TADAT are left in solution at pH > 11 for at least 30 mins, and then acidified with perchloric acid, several 1 : 3 complexes are formed, the most interesting origi- nating at Ho = -1.5 (€590 = 8.50 x lo4 1 mol-1 cm-l). However, the spectrophotometric determination of cobalt by this method suffers interference from those ions which form precipitates in very alkaline media and are difficult to redissolve, and from ions that complex strongly with the cobalt. In order to overcome these problems, this work establishes the experimental conditions required for the determination of cobalt in an alkaline complexing medium, where the cobalt is oxidised by atmospheric oxygen, or in an acidic medium, where sodium vanadate is used as a mild oxidant.Very sensitive and selective methods have been developed for the spectro- photometric determination of cobalt. One of these methods has been used to assay several cobalt-containing materials. Experiment a1 Apparatus The apparatus used included a Beckman 25 and a Hitachi Perkin-Elmer 200 recording spectrophotometer with 1-cm path length glass or silica cells. ,4 Radiometer PHM25 pH meter with glass and saturated calomel electrodes, a Pye Unica.m 1900 atomic-absorption spectrophotometer and a Sartorius MPR35 balance were also used.Reagents further purification. Chemicals of analytical-reagent grade and de-ionised water were used throughout, without 3-(2~-Thiazolylazo)-2,6-diaminotoluene sol.ution, Cobalt(l1) nitrate solution, 10-1 M. M in 1 M perchloric acid. Standardised by EDTA t i t r a t i ~ n . ~MONTELONGO, GONZALEZ AND D ~ A Z 763 Sodium vanadate solution, 10-1 M. Sodium acetate solution, 1 M. These solutions were diluted with water as required. Recommended Procedures for the Determination of Cobalt Method 1 To a solution containing 2.5-17.5 pg of cobalt in a 25-ml calibrated flask, add 2 ml of a M solution of 2,6-TADAT in 1 M perchloric acid and 5 ml of 1 M ammonia solution. Measure the After 1 h add 2 ml of 70% perchloric acid and make up to volume with water.absorbance of the solution at 590 nm against a reagent blank, in l-cm path length cells. Method 2 To a solution containing 1.25-15.0 pg of cobalt in a 25-ml calibrated flask, add 2 ml of a M solution of 2,6-TADAT in 1 M perchloric acid, 2 ml of 1 M sodium acetate and 1 ml Wait for 1 h and then add 2 ml of 70% perchloric acid Measure the absorbance of the solution at 590nm of and make up to volume with water. against a reagent blank, in l-cm path length cells. M sodium vanadate solution. 0.700 0.600 0.500 0) 6 0.400 e 2 n a 0.300 0.200 0.100 500 600 500 600 Wavelengt h/n m 1 I 500 600 Fig. 1. Influence of initial pH, vanadate concentration, Cv, and time, t , on the formation of the (a) Influence A, pH = 1.5; B, pH = 2.9; C, pH = 3.5; D, pH = 4.2; (b) Influence of the vanadate concentration: CV = A, 0.0; B, 8.0 x 1 0 - 6 ~ ; C, 4.0 x 1 0 - 5 ~ ; D, 2.0 x 1 0 - 4 ~ ; E, 4.0 x M.A, 0 min; B, Co(II1) - 2,6-TADAT complex. of initial pH: CV = 4.0 x 1 0 - 4 ~ , 1 = 1 h. E, pH = 5.2; F, pH = 8.4; R = reagent alone at pH = 4.0. initial pH = 4.85; t = 1 h. 16min; C, 30min; D, 60min; E, 120min. CL = 4.0 x lod5 M, CC,, = 8.0 x M, final Ho = -0.5. M ; F, 8.0 x M. (6) Influence of time: initial pH = 4.85; Cv = 4.0 x Determination of cobalt in steels and hydro$ning catalysts Weigh accurately 0.1-1 .O g of the steel or catalyst sample and dissolve in hydrochloric acid (1 + 1) while heating on a water-bath. Dissolve the residue in 1 M perchloric acid in a 100-ml calibrated flask and make up to volume with 1 M perchloric acid.Evaporate the solution to dryness. Assay suitable aliquots of the final solution as described under Method 2.764 Determination of cobalt in high-purity nickel salts calibrated flask. MONTELONGO et al. : DETERMINATION OF COBALT WITH Analyst, Vol. 105 Weigh accurately 1-10 g of the .nickel salt, dissolve and dilute to 100 ml with water in a Assay 5-ml aliquots of the final solution as described under Method 2. Results and Discussion Reaction with Cobalt In ammoniacal medium The absorption spectra, graphs of absorbance v m u s hydrogen concentration and the stoicheiometry of the solutions obtained when cobalt and 2,6-TADAT are mixed in an ammoniacal medium of pH about 10, left for 1 h, and then acidified with perchloric acid until H , = -0.5, are all similar to those obtained by the same procedure when mixing is carried out in sodium hydroxide solution at pH > 11.8 If method 1 is followed, then in a final medium of H , = -0.5 (1.35 M perchloric acid) and at 590 nm, Beer's law is obeyed for 0.10-0.70 p.p.m.of cobalt. The minimum error range of 0.18-0.47 p.p.m. was evaluated from a Ringbom plot. The molar absorptivity is 8.83 x 104 1 mol-1 cm-1, with a Sandell sensitivity index of 0.66 ng cm-2. The relative error of the method is 0.13% (95% confidence interval). The effects of several cations and anions were examined by using method 1 to assay solutions containing 0.237 p.p.m. of cobalt and various concentrations of other ions (Table I). Solutions are stable for more than 5 d. In acidic medium In order to avoid the interferences found when the determination of cobalt with 2,6- TADAT is commenced in alkaline media, it would be necessary to begin the determination in acidic media.However, when cobalt(I1) and 2,6-TADAT are mixed in acidic media a reddish brown colour develops, which becomes bluish violet on acidifying to 1-3 M in perchloric acid. Lower and non-reproducible absorbance values are obtained in this way, probably owing to incomplete oxidation of cobalt. Therefore, the use of an oxidising agent for cobalt(I1) in acidic media would appear to be necessary. 0.5 0.4 Q, C 0.3 a a a 0.2 0.1 " -4 -3 -2 -1 0 1 2 3 H , V P H Fig. 2. Influence of the final acidity on the formation of the Co(II1) - 2,6-TADAT complex. CL = 4.0 X M, cc0 = 8.0 X ~ O - B M , initial pH = 4.85, Cv = 4.0 x 10-4 M, t = 1 h.Action of Oxidising Agents As 2,6-TADAT is destroyed by some oxidising agents, several were tried in order to find one mild enough to oxidise cobalt(I1) to cobalt(II1) in acidic media, without destroying theAugust, 1980 3-(2’-THIAZOLYLAZO)-2,6-DIAMINOTOLUENE 765 reagent. A 2 x M solution of 2,6-TADAT is quickly decolorised by peroxidisulphate, bromate, periodate, etc., even in the cold; however, vanadate appeared satisfactory and was therefore chosen for subsequent use. Initial pH, Vanadate Concentration and Time Experiments were performed on solutions with a 2,6-TADAT concentration of C, = 4.0 x M. As the complex develops in 0.5-2.0 M perchloric acid medium, in which cobalt and 2,6-TADAT do not react directly, in this work the initial pH and vanadate concentration to be studied were adjusted, and the mixture left for a measured time, and then the required volume of 70% perchloric acid was added so as to give a final perchloric acid concentration of 1 M.The spectra were then recorded. From Fig. 1, it is clear that to obtain the highest molar absorptivities it is necessary to start with an initial pH near to 4.8, with a vanadate to cobalt ratio of about ten, and to wait for 1 h before adding the perchloric acid. M and a cobalt concentration of C,, = 8.0 x Final Acidity In Fig. 2, the variation of absorbance with the final hydrogen ion concentration is shown. It is clear that the measurements of absorbance must be made between pH = 0.5 and H , = -1.0. At pH > 3 solutions develop a slight turbidity owing to the formation of less soluble complex species similar to those previously described.8 Absorption spectra and Job’s plots show that CoH,L, (z+3)+ is the bluish violet soluble complex8 (x being the number of protonated amino groups in the reagent).TABLE I DETERMINATION OF COBALT IN THE PRESENCE OF OTHER IONS The concentration of cobalt present in each determination was 0.237 p.p.m. Ion c1- . . NO,- . . so*2- . . CH,COO- Cu(I1) . . Ni(I1) . . Zn(I1) . . Pb(I1) . . Hg(I1) . . Pd(I1) . . Ca(I1) . . Ba(I1) . . Mg(I1) . . Mn(I1) . . Fe(I1) . . Fe(II1) . . Al(II1) . . Cr(II1) . . Mo(V1) .. Pt(I1) . . V(V) .. W(V1) .. .. .. . . .. . . . . .. . . . . . . . . . . .. . . . . .. .. . . . . . . .. . . Molar ratio, ion : Co(I1) . . 150 .. 150 . . 150 .. 150 .. 20 100 . . 150 . . 150 .. 100 . . 1 10 .. 1 . . 1 . . 100 .. 100 .. 100 . . 10 .. 100 . . 100 . . 100 . . 1 100 . . 50 .. 100 .. 1 25 Co(I1) found, p.p.m.* Alkaline medium Acidic medium 0.238 0.237 0.236 0.236 0.237 0.237 0.237 0.237 0.239 0.238 0.240 0.319 0.232 0.234 0.240 0.235 0.238 0.237 0.239 0.235 0.205 0.238 0.103 0.113 0.402 0.438 0.233 0.236 0.234 0.237 0.234 0.237 0.230 0.239 0.239 - 0.236 0.235 0.232 0.238 0.180 0.238 - 0.237 0.230 - 0.234 0.239 0.098 0.238 - 0.240 I A 7 * Mean of three determinations. Determination of Cobalt In solutions with final H , = -0.5 (1.35 M obeys Beer’s law in the ran& 0.05-0.60 p.p.m. - A Ringbom plot shows that 0.144.44 p.p.m. of cobalt is the optimum concentration range. The molar absorptivity is 9.74 x perchloric acid) and at 590 nm, the complex766 MONTELONGO et al.: DETERMINATION OF COBALT WITH TABLE I1 Analyst, VoZ. 105 DETERMINATION OF COBALT IN LOW-ALLOY STEELS Certified cobalt Cobalt found, * Steel content, yo % BCS 251 (Mn 0.160/,, Ni 5.15%) . . . . 0.070 0.070 BCS 252 (Mn 0.016%, Ni 4.10%) . . . . 0.040 0.041 BCS 254 (Mn 0.525%, Ni 2.08%) . . . . 0.029 0.019 * Mean of three determinations. 104 1 mol-f cm-1, with a Sandell sensitivity index of 0.60 ng cm-2, and the relative error of the method is 0.14% (95% confidence interval). The influence of several ions was examined by applying Method 2 to the assay of solutions containing 0.237 p.p.m. of cobalt and various concentrations of other ions (Table I). Solutions are stable for at least 5 d.TABLE I11 DETERMINATION OF COBALT I N HYDROFINING CATALYSTS Cobalt found,* % Estimated cobalt t o r-'------, Sample nickel ratio 2,6-TADAT AASf 1 0.02 0.053 0.056 2 65 2.080 2.085 3 75 2.355 2.355 * Mean of three determinations. t Atomic-absorption spectrophotometry. Determination of Cobalt in Different Materials A comparison of the two methods developed for the spectrophotometric determination of cobalt with 2,6-TADAT, shows that method 1 is simpler, but that method 2 offers higher sensitivity and specificity (except for copper). Consequently, the determinations of cobalt in steels, hydrofining catalysts and high-purity nickel salts described below, were carried out by method 2. The results obtained for three determinations of cobalt in each of the BCS 251, BCS 252 and BCS 254 low-alloy steels are shown in Table 11.The interference due to manganese in the BCS 254 steel is apparent because the manganese to cobalt ratio is greater than the limit previously established. TABLE IV DETERMINATION OF COBALT IN HIGH-PURITY NICKEL SALTS Salt Cobalt found,* 74 2,6-TADAT AASt -7 NiSO,.GH,O . . . . . . 0.38 x 10-3 0.41 x 10-3 NiC1,.6H20 . . . . . . 0.59 x 10-3 0.50 x 10-3 Ni(NO,),.GH,O . . . . . . 0.65 x 10-3 0.70 x 10-3 Ni(CH3C00),.4H,O . . . . 0.88 x 10-1 0.90 x 10-1 * Mean of three determinations. t Atomic-absorption spectrophotometry. Spectrophotometric determinations of cobalt in several hydrofining catalysts, containing aluminium (4045y0), molybdenum (&lo%), nickel (0.05-3y0), and cobalt (0.05-2.5%) were carried out.Results are shown in Table I11 where they are compared with those obtained by atomic-absorption spectrophotometry.1° The direct determination of cobalt in high-August, 1980 3- (2’-THIAZOLYLAZO) -2,6-DIAMI NOTOL UE N E 767 purity nickel salts may also be carried out using 2,6-TADAT. Results obtained for several nickel salts are shown in Table IV where they are compared with results obtained by atomic- absorption spectrophotometry.1° 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Sandell, E. B., and Onishi, H., “Photometric Determination of Traces of Metals, General Aspects,” Snell, F. D., “Photometric and Fluorimetric Methods of Analysis,” John Wiley, New York, Part 1, Tbei, K., and Motomizu, S., Analyst, 1976, 101, 497. Shibata, S., Furukawa, M., and Kamata, E., Anal. Chim. Acta, 1974, 73, 107. Shibata, S., Furukawa, M., and Goto, K., Talanta, 1973, 20, 426. Kiss, E., Anal. Chim. Acba, 1973, 66, 385. Garcia Montelongo, F., GonzAlez Diaz, V., and Tallo GonzAlez, C. R., Analyst, 1979, 104, 1091. Tallo Gonzhlez, C. R., Garcia Montelongo, F., and Gonzhlez Diaz, V., An. Quim., in the press. Schwarzenbach, G., and Flaschka, H., “Complexometric Titrations,” Methuen, London, Second Pinta, M., “SpectromCtrie d’ Absorption Atomique,” Tome 11, Masson, Paris, 1971. John Wiley, New York, Fourth Edition, 1978. 1978, p. 931. Edition, 1965, p. 244. Received January 21st, 1980 Rccepted February 26th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500762

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

768 Analyst, August, 1980, Vol. 105, pp. 768-773 Spectrophotometric Determination of Anionic Surfactants in River Waters Using 144-N itrobenzyl)-4=(4=diethylaminophenylazo)- pyridinium Bromide Keiro Higuchi, Yasuaki Shimoishi, Haruo Miyata and Kyoji Toei" Japan Department of Chemistry, Faculty of Science, Okayama University, Tsushima-naka 3-1-1, Okayama-shi 700, and Tadashi Hayami College of Liberal Arts, Okayama University, Tsushima-naka 2-1-1, Okayama-shi 700, Japan Nitro, broino and methyl derivatives of l-(benzyl)-4-(4-diethylamino- pheny1azo)pyridinium bromide were synthesised and evaluated as new cationic reagents for the determination of anionic surfactants. These reagents were very stable and reacted with anionic surfactants, such as alkylbenzene- sulphonate and alkylsulphate, to form a 1 : 1 stable ion associate, which was extracted into chlorobenzene in a single extraction.The apparent molar absorptivity of the ion associate of the 4-nitro derivative with sodium di- (2-ethylhexyl)sulphosuccinate was 6.10 x lo4 1 mol-l cm-l (at 573 nm) in chlorobenzene. 1- (4-Nitrobenzyl)-4-( 4-diethylaminopheny1azo)pyridinium bromide was used in the determination of pgl-1 levels of anionic surfac- tants in river water. The results were compared with the methylene blue method (Japanese Industrial Standard method) for river waters. This method is designed to determine anionic surfactant concentrations in solution. Keywords : Anionic surfactant determination ; water analysis ; spectrophoto- metry ; l-(benzyl)-4- (4-diethylaminopheny1axo)pyridinium bromide deriva- tives A minute amount of anionic surfactants in river waters has been determined by an extrac- tion - spectrophotometric method with Bindschedler's Green derivatives1 The dibutyl derivative can react with an anionic surfactant to form an ion associate that is extracted into toluene at pH 7 and the absorbance a t 730 nm can then be measured.The apparent molar absorptivity is 7.1 x lo4 1 mol-l cm-l. However, the reagent is not very stable and the colour fades gradually, especially in the presence of an oxidant such as iron(II1). To improve the stability of the reagent, some pyridinium azo compounds were synthesised and used for the determination of anionic surfactants. Of these compounds, 1-(4-nitrobenzy1)-4- (4-diethylaminophenylazo)pyridinium bromide (NDP) forms a stable 1 : 1 ion associate with anionic surfactants, such as sodium dodecylsulphate, sodium dodecylbenzenesulphonate (DBS) and sodium di(2-ethylhexyl)sulphosuccinate (DESS), which are extracted into chloro- benzene by a single extraction.This paper describes a simple and rapid method for the determination of pg 1-1 levels of anionic surfactants in river waters with NDP, with 97-100y0 recovery levels. Experimental Apparatus EPS-3T, spectrophotometers with 10-mm glass cells. shaking with an Iwaki, Model KM, shaker. Hitachi-Horiba, Model I;-~SS, pH meter. Model CPN-005, centrifuge. Spectrophotometric measurements were carried out using Hitachi, Model 139 and Model Extractions were carried out by The pH measurements were carried out with a The centrifugation was 'performed with a Shimazu, * To whom correspondence should be addressed.HIGUCEEI, SHIMOISHI, MIYATA, T ~ E I AND HAYAMI Reagents Syntheses of 1-(benzy1)-, 1-(4-rnethylbenzyl)-, 1-(4-bromobenzyl)- and 1-(4-nitrobenzyl)-4- (4-diethylaminop~enylazo)pyr~d~n~um bromides.4-(4-Diethylaminophenylazo)pyridine (DP)2 was obtained by coupling diazotised 4-amino- pyridine with NN-diethylaniline in phosphoric acid. A solution of 1.8 g (0.019 mol) of 4- aminopyridine in 10 ml of 85% phosphoric acid and 5 ml of concentrated nitric acid were mixed at 0 "C; 1.4 g (0.02 mol) of sodium nitrite and 25 g of ice were then added successively. The solution was poured into 20 ml of 30% phosphoric acid containing 3 g (0.02 mol) of NN-diethylaniline. After the reaction, the solution was neutralised with sodium carbonate, and the precipitate was filtered off and recrystallised from an ethanol - water solution (1 + 4) (yield 51%).The crystals were reddish yellow plates, n1.p. 182-184 "C. Elemental analysis: found, C 69.87%, H 7.09% and N 21.30%; calculated (for CISH,,N,), C 70.84%, H 7.13% and N 28.03%. The DP was quaternised by refluxing with benzyl, 4-methylbenzyl, 4-bromobenzyl or 4-nitrobenzyl bromide in benzene over an oil-bath. The reagents obtained, respectively, 1 -(benzyl)-4-(4-diethylaminophenylazo)pyridinium bromide (BDP) , 1-(4-methylbenzy1)-4- (4-diethylaminopheny1azo)pyridinium bromide (MDP) , 1-(4-bromobenzy1)-4-(4-diethylamino- pheny1azo)pyridinium bromide (BrDP) and l-(4-nitrobenzyl)-4-(4-diethylaminophenylazo)- pyridinium bromide (NDP), were washed with benzene until the washings were no longer coloured and then dried at reduced pressure and at 50 "C to a constant mass.A 2.6 x M solution of each reagent was prepared by dissolving each one in distilled water. The anionic surfactant used was sodium di(2-ethylhexyl) sulpho- succinate (DESS), certified as 96.3% by the Japan Oil Chemists' Society. I t was dried a t 50 "C under reduced pressure (about 3 mmHg) before weighing. Prepare the buffer solution by adding 2 M sodium dihydrogen orthophosphate solution, with mixing to 2 M potassium monohydrogen orthophosphate solution until a pH of 6 is obtained as measured by a pH meter. 769 Anionic surfactants. Bivfer solution (pH 6). Procedure M), in a 100-ml separating funnel, add 5 ml of phosphate buffer solution (pH 6) and 1 ml of 2.6 x lo-* M NDP solution.Add 5 m l of chlorobenzene and shake the funnel mechanically for 5min, 2-3 times per second, in order to extract the ion associate that has formed; stand for 10 min. Transfer the chlorobenzene into a test-tube and centrifuge for 1 min a t 2000 rev min-l. Measure the absorbance of the organic phase at 573 nni against chlorobenzene using 10-mm glass cells. Place 100 ml of a sample solution (surfactant content below 7.0 x Results and Discussion Molar Absorptivity of the Ion Associate extracted completely into chlorobenzene. containing 6.3 x absorption occurs at 573nm. ion associate are shown in Table I. anionic surfactants. at ionic strength, I = 0.1, adjusted using potassium nitrate.The ion associate formed between the reagent cation and the anionic surfactant was Fig. 1 shows the absorption spectrum of a solution The wavelength of maximum The molar absorptivities of the reagent solution and the NDP is the preferred reagent for the determination of The pK, values of the reagents were determined spectrophotometrically M DESS using the described procedure. Extraction Solvents Generally, as the dielectric constant of the solvent increases, the amount of reagent extracted increases, e.g., the order of increasing extractability of the solvents is chloroform, chlorobenzene, o-dichlorobenzene, 1,2-dichloro- ethane and nitrobenzene, and the last (dielectric constant 35) can extract all the reagents. Although the dielectric constant of chlorobenzene (5.5) is between chloroform and o-dichloro- benzene, the reagent was little extracted into this solvent.On the other hand, solvents useful for the extraction of the ion associates are carbon tetrachloride, benzene, toluene, Eleven organic solvents were examined.770 HIGUCHI et al. : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 105 TABLE I pK, AND MOLAR ABSORPTIVITY OF THE REAGENTS AND THEIR ION ASSOCIATES Ion associate with DESS in Reagent in water chlorobenzene 7 7 7-- -7 Molar absorptivity/ Wavelength/ Molar absorptivity/ Wavelength1 A Reagent pK, 1 mol-' cm-l nm 1 mol-1 cm-1 nm BDP . . . . 3.8 6.60 x 1 0 4 576 5.27 x 1 0 4 566 MDP . . . . 3.8 6.47 x 1 0 4 574 5.50 x 104 $60 BrDP . . 3.8 6.90 x 1 0 4 576 5.70 x 1 0 4 565 NDP . . . . 3.8 7.36 x 1 0 4 578 6.10 ;< 104 573 xylene, diisopropyl ether, isoaniyl acetate and chlorobenzene.From the results of the solvents tested, chlorobenzene was preferred because it gave a small blank value and the largest absorbance difference at 573 nm between the ion associate and the blank. Stoicheiometry of the Ion Associate amount of NDP is more than twice that of the DESS. associate in chlorobenzene. NDP to DESS was found to be 1 : 1. a 1 : l . DESS reacts with NDP rapidly and quantitatively to form the ion associate when the Job's curves were plotted for the ion The results are shown in Fig. 2 and the stoicheiometric ratio of The ratio of NDP to DBS or DS was also found to be 0 0.2 0.4 0.6 0.8 1.0 Wavelength/nm Fig. 1. Absorption spectrum of DESS - NDP ion associate.Concentration of DESS 6.3 x ~ O - ' M . Ion associate extracted into chlorobenzene at pH 6. [DESS] [NDPI + [DESS] Fig. 2. Composition of DESS - NDP ion associate by the continuous variation method. [DESS] 4- [NDP] = 1.30 x M. Effect of pH on Extraction range (Fig. 3). the ion associate into chlorobenzene was therefore carried out at pH 6. The absorbance was a maximum between pH 4 and 8 and was constant throughout this The extraction of The absorbance decreased above pH 9 and below pM 4. Effect of Reagent Concentration, Shaking and Standing Time The effect of an excess of the reagent was examined; the addition of a 2-30 times molar excess of NDP against DESS gave a constant maximum absorbance. The extraction of the ion associate into chlorobenzene was examined. The separating funnel was shaken mechanically and then left to stand.When the shaking time was varied between 3 and 30min, the absorbances obtained remained constant. The standing time was varied between 5 and 90 min, and the absorbances were found to be constant. A standing time of 10 min was therefore considered to be sufficient. The organic phase was still slightly turbid after standing when no phosphate buffer was present, so the buffer solution was needed to hasten the phase separation and the organic phase should additionally be centrifuged for 1 min at 2000 rev min-1.Azqyst, 1980 Effect of Volume of Water The absorbance of the ion associate in 5 ml of chlorobenzene was constant when the aqueous phase was varied between 50 and 500 ml (Fig. 4). From the results, the enrichment of anionic surfactants by 10-100 times was possible from sample waters into the organic phase.OF ANIONIC SURFACTANTS IN RIVER WATERS USING NDP 771 The effect of the volume of water on extraction was examined. 0.5 0.4 $ 0.3 0, Ll a 0.2 0.1 o.6 A - - - - B L + I S + I I * - 0.1 0 2 4 6 8 1 0 1 2 PH Fig. 3. Effect of pH on extraction. A, DESS - NDP ion associate; B, reagent blank; [DESS] = 4.83 x 1 0 - 7 ~ ; [NDP] = 2.55 X 10-6M. I Calibration Graph relationship over the range 0 to 1.3 x tivity of DESS with NDP was 6.10 x lo4 1 mol-l cm-l at 573 nm. The calibration graphs obtained by the procedure were found to show a good linear The molar absorp- M of an anionic surfactant. TABLE I1 INTERFERING IONS The test solution contained 4.83 x of water. M DESS in 100 ml Ion None .. .. c1- . . .. HC03- .. NO3- . . .. .. Si032- .. NHf .. Mg2+ .. Na+ . . .. K+ . . - . Ca2+ . . .. Fe3+ . . .. Fe3-+ + EDTA Sodium stearate .. .. .. .. .. .. .. .. * . .. .. .. .. .. Humic acid* : 1 g 1-' 0.1 g 1-1 0.01 81-1 Ion concentration/ M - .. .. 10-3 .. 10-3 . . 5 x 10-4 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-3 .. 10-4 - .. 10-6 .. .. .. Absorbance of ion associate 0.582 0.580 0.585 0.582 0.599 0.583 0.584 0.580 0.585 0.582 0.578 0.579 0.538 0.582 0.965 0.582 0.678 0.602 0.582 * Humic acid concentrations are expressed as grams per litre of soluble matter.772 HIGUCHI et al. : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vol. 105 Effect of Diverse Ions The interferences due to various ions were examined in 100 ml of sample solution con- taining 4.83 x l o - 7 ~ of DESS.The concentration of the diverse ions tested was 10-100 times higher than the normal content of the ions in river waters in Japan. The results obtained are shown in Table 11. The common cations and anions present in river water did not interfere. Nitrate ion above M and stearate ion above lod5 M caused positive errors. Iron(II1) above M caused a negative error but the interference was effectively masked with EDTA. Humic acid solution at a concentration of more than 0.01 g 1-1 of soluble matter caused a positive error. TABLE I11 PRECISION OF MEASUREMENTS WITH THE PROPOSED METHOD Results obtained using a 100-ml river water sample. (a) Repeatability tests- Absorbance of ion associate 0.235 0.246 0.231 0.228 0.235 0.244 0.230 0.238 0.239 0.246 Meanvalue .. .. . . Standard deviation . . .. Relative standard deviation Ion associate content as DESS/pg 1-’ 85.2 89.2 83.8 82.7 85.2 88.5 83.4 86.3 86.7 89.2 .. 86.0 .. 2.3 . . 2.7% (b) Recovery tests- DESS concentration/pg 1-] A I 7 Added Found Recovered Recovery, yo None 86 - - 89 177 91 102 89 176 90 101 89 175 89 100 89 172 86 97 89 176 90 101 Accuracy and Precision The accuracy of the procedure was evaluated by recovering experiments, in which known amounts of DESS were added to river water and the samples were treated as described in the procedure. In all instances, theoretical recoveries were obtained, within experimental error (Table 111). The precision of the procedure was evaluated by analysing ten samples of river water. The mean result obtained was 86.0 p.p.b.with a standard deviation of 2.3 p.p.b. and a relative standard deviation of 2.7%. 0.6 I I Fig. 5. Loss of anionic surfactant on storage. 0, Glass container; 0, polyethylene container. Results obtained using a 100-ml sample of river water.August, 1980 OF ANIONIC SURFACTANTS IN RIVER WATERS USING NDP 773 Loss of Anionic Surfactants on Storage A sample solution of river water was filtered and stored and possible losses of anionic surfactants owing to adsorption on to the walls of glass or polyethylene containers were examined; the results are shown in Fig. 5 . There was no loss within 12 h in either container, but a loss of about 15% was found after 24 h. Determination of Anionic Surfactants in River Water The concentration of anionic surfactants in river water can be measured by the above procedure. The river water samples should be filtered through a membrane filter (0.45 pm), and 1 ml of 0.1 M EDTA solution added to remove the interference due to iron(II1). Table IV shows results for anionic surfactants in river waters compared with those by the methylene blue method (Japanese Industrial Standard m e t h ~ d ) .~ Each value was the average of three determinations. The recovery test [Table 111, part ( b ) ] was carried out in each instance, and the recovery was found to be from 97 to lOOY-,. Determination by the methylene blue method was impossible with samples 1 4 , because their concentrations were below the lower limit of the methylene blue method.With samples 5-7, the results obtained by the proposed method agreed approximately with those obtained by the methylene blue method. TABLE IV COMPARISON OF RESULTS OBTAINED BY THE PROPOSED METHOD WITH THE METHYLENE BLUE METHOD FOR RIVER WATERS IN OKAYAMA PREFECTURE Values in parentheses are recoveries (per cent.), determined as in Table 111. Proposed method Methylene blue method River water sample No. 1 2 3 4 5 6 7 r Absorbance 0.079 0.064 0.172 0.149 0.226 0.422 0.552 f A > Absorbance Anionic surfactants/* PLg I-' Pg I-' 29 (98) 23 (97) 62 (98) 54 (98) 82 (98) 0.027 70 153 (98) 0.053 130 200 (100) 0.076 190 * The amounts of anionic surfactants were calculated from the calibration graph of DESS. Conclusion A new cationic azo dye, l-(4-nitrobenzyl)-4-(4-diethylaminophenylazo)pyridinium bromide, can react with anionic surfactants, such as alkylbenzenesulphonate and alkylsulphate, to form ion associates that can be extracted into chlorobenzene by a single extraction. The ion associates formed are very stable in the organic phase and the stoicheiometric ratios are 1 : 1. The calibration graphs are straight lines over the range 0 to 1.3 x RI of DESS, and the apparent molar absorptivity is 6.10 x lo4 1 mol-l cm-l at 573 nm. The method is simpler and faster than the methylene blue method for the determination of anionic surfactants in river waters. References 1. 2. 3. Tbei, K., Miyata, H., Motomizu, S., and Tesumoto, J . , Bunseki Kagaku, 1978, 27, 138. Faessinger, R. W., and Brown, E. V., Trans. Ky. Acad. Sci., 1963, 24, 106. Japanese Industrial Standard, K 0102, 1974. Received February 18th, 1980 Accepted March 26th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9800500768

出版商:RSC    

年代:1980

数据来源: RSC