ATOMIC ABSORPTION ANALYSIS OF PHOSPHATES IN WATER Calvin O. Huber University of Wisconsin Milwaukee, Wisconsin October 1973 DISTRIBUTED BY: National Technical Information Service U. S. DEPARTMENT OF COMMERCE 5285 Port Royal Road, Springfield Va. 22151 This document has been approved for public release and sale. ------- BPA-670/2-73-079 October, 1973 ATOMIC ABSORPTION ANALYSIS OF PHOSPHATES IN WATER by Calvin 0. Huber Department of Chemistry University of Wisconsin-Milwaukee Milwaukee, Wisconsin 53201 Project 16020 DHD Program Element » 310401 (1972), IB1027 (1973) Project Officer: Mr. Robert L. Booth Analytical Quality Control Laboratory National Environmental Research Center Cincinnati, Ohio 45268 Prepared For: OFFICE OF RESEARCH AND MONITORING U.S. ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 -670/2-73-079, October, 1973. StLECTED WATER RESOURCES ABSTRACTS INPUT TRANSACTION FORM ue Atomic Absorption Analysis of Phosphates in Water 7. Aatbot(l) Huber, C.O. Department of Chemistry. -— 10. Pro/tctlfo. /403LO II. Coatrtct/GraatMo. is. AtMrict The project investigatwf»utilization of phosphate, silicate, and sulfate chemical inhibition effects in magnesium atonic absorption spectrometry for the purpose of determination of these anions. The variables found to be of greatest significance in the inhibition pro- cesses were flame temperature and solution stoichiometry. The procedure termed 'atomic absorption inhibition titration1 (MIT) was developed. It provides useful, new determination methods for single anions and also for simultaneous determination of phosphate, silicate, and sulfate in a single sample. The methods have been evaluated and applied to water and waste water samples. 17*.Descriptors *AluBili. Absorption, 'Inhibition Titfation., *Phosphates, *Water i X -* */) fln ~l- i <• Analyses, Sulfates, Silicates. nb.idtatiSa, *Pheii|jlinl8t, 'Inhibition Titration, Wa«o> S Wactac ^ Uc. CO WRR Field A Group £) 5 /\ vailability ? Trnm flic Naf)n 19. (Rtpcrt) -••• paper cop/, |1.45 in mlgmfl.k. t JTO.O/ , Send To; 33. Price WATER RESOURCES SCIENTIFIC INFORMATION CENTER US DEPARTMENT OF THE INTERIOR ; , WASHINGTON. D. C 2O24O Robert L. Booth AQCL/NERC, USEPA wRsrc 102 (nev ------- ABSTRACT The project investigated utilization of phosphate, silicate and sulfate chemical inhibition effects in atomic absorption spectrometry for the purpose of determination of these anions. The magnesium atomic absorption signal was used. The variables found to be of greatest significance in the inhibition processes were flame temperature and solution stoichiometry. As a result of the latter, the procedure termed atomic absorption inhibition titration, AAIT, was developed. It involves constant rate addition of magnesium to the stirred sample solution while monitoring the latter solution for magnesium atomic absorption. Rather remarkable titration curve shapes are obtained, especially for solutions containing all three anions. Based on such titration curves, useful new determination methods were developed for single anions and also for simultaneous determination of the three anions present in a single sample. The methods developed as applied to water and waste water have been investigated and found to be suitable in most respects and superior to present methods in some. This report was submitted in fulfillment of Research Grant * 16020 DHD by the University of Wisconsin-Milwaukee ii under the partial sponsorship of the Office of Research and Monitoring, U.S. Environmental Protection Agency. Work was completed as of July, 1973. iii ------- TABLE OF CONTENTS Page ABSTRACT ii LIST OF FIGURES ....... V LIST OF TABLES vi ACKNOWLEDGMENTS vii CONCLUSIONS viii RECOMMENDATIONS ............. . x INTRODUCTION .......... 1 EXPERIMENTAL 4 RESULTS ..... ....... 6 DISCUSSION ......... 32 REFERENCES 34 LIST OF PUBLICATIONS 35 LIST OF FIGURES Pages Figure 1 Inhibition and Gas Flow Hates 9-10 2 Titration of Blank and 8 ppm Orthophosphate . .12-13 3 Effect of Air:Hydrogen Flow Ratio, R, on Titration Curve 25-26 ------- LIST OF TABLES Table I II III IV V VI VII Page Titration PO Mole Ratios for J. .L l^, VI..I.WA1 ^u./lly rawXC rUil-4.UEl 1.UX Solutions Containing Phosphate and Sulfate ... 16 River Hater Analysis 18 Determination of Phosphate in Detergent Products by AMT 19 Determination of Silica in Artificial Drinking Hater 21 Determination of Sulfate by AAIT 23 Data Comparing AAIT and Gravimetric Methods . .28 Simultaneous Determination of SiO,, PO., and S04 in Haste Hater 31 ACKNOWLEDGMENTS Mr. Walter Crawford, Dr. Robert Looyenga, and Dr. Chuang-I Lin were the co-investigators in this work. Their contributions were absolutely indispensible to its success. The help and liaison provided by project officers, Mr. Robert L. Booth of the Analytical Quality Control Laboratory, and Dr. George W. Bailey of the Southeast Water Research Laboratory, is gratefully acknowledged. The project was supported by the Office of Research and Monitoring, U.S. Environmental Protection Agency. The cooperation and assistance of The Center for Great Lakes Studies of this campus was helpful. ------- CONCLUSIONS 1. New and improved analytical methods for analysis of common anions in water are achieved to meet the challenges of speed, accuracy, convenience, sample deterioration, precision, and cost. 2. Flame spectroscopic chemical inhibition effects can be a basis for such improved methods. 3. Flame temperature and solution ion ratios are the most significant parameters for utilizing the chemical inhibition methods. Solution ion ratios are most readily regulated by use of a titration technique. 4. Phosphate, silicate, and sulfate can be determined in real water and waste water samples, separated or in combination. A procedure requiring less than ten minutes allows the determination of these three anions in a single fifty milliliter sample. 5. Chemical inhibition processes in flames are shown to be rate controlled in the depleting droplet and resulting particle rather than controlled by solution equilibria. viii 6. The success in achieving these improved analytical methods indicates that further research should be pursued to exploit the above advances for water analysis problems of various types. ix ------- RECOMMENDATIONS An advantageous new technique has been made available for water analysis. As a result, the following specific courses of action regarding it are recommended. 1. The method should be applied by a number of analytical chemistry laboratories for increased efficiency and for in-the-field evaluation of its capability. 2. The method should be completely automated by means of automatic sample handling and data acquisition and treatment computer equipment. 3. Simple minimal adequate design flame photometric apparatus should be developed or applied to exploit the cost and dependability advantages of the inherent simplicity of this new method. 4. The method should be extended to other anions, flame emission and fluorescence signals, and analysis of cations. 5. Finally, the technique provides a new tool for investigation of flame chemical inhibition reactions and very high temperature rate processes and stoichiometry. These capabilities should be applied to fundamental studies in these areas of chemistry. x INTRODUCTION The importance for phosphate and other anion analysis is well-stated in an American Chemical Society report on pollution abatment: Progress in analytical chemistry and instrumentation is vital to both water pollution research and survellance and research on eutrophication is a notable example of the need. The identification and measurement of limiting nutrients would be greatly eased by improved analytical methods for low levels of the forms of phosphorous compounds, the various forms of nitrogen compounds, trace metals, and trace organic growth factors in water. Accordingly, the project objective was stated as follows: The initial aim is to determine phosphate in natural and waste waters by atomic absorption with high sensitivity, speed and specificity. The linear absorbance decrease for metals due to flame compound formation will be the observed quantity. The investigation will necessarily include studies of flame reaction stoichiometries and rates and should provide improved understanding of compound formation in flames. This will be used to establish a specific determination for the various phosphates as well as for other nonmetallic anions. The extension of the advantages of atomic absorption to analysis of many anions at concentrations present in natural waters is the final objective. Present analytical methods for measurement of anions as outlined in Public Health Services' "Standard Methods of Water and Waste Water Analysis" are usually laborious and sometimes empirical. Interferences by silicate, arsenate, and other anions are troublesome. Because the methods are relatively slow errors sometimes enter during storage due to 1 ------- micro organism action and adsorption effects. The above reference points out the desirability of fast field methods whenever possible. A vast literature exists aimed at removing or minimizing chemical interference effects in flame spectroscopy. These effects of inhibiting anions due to formation of refractory species in the flame remain as probably the most important limitation on accuracy for most practical analysis of metals. In contrast, little has been reported on the use of these inhibitions as a means of analyzing the corresponding anions. The most common inhibiting anions are coincidentally also among the anions of most interest in natural and waste water chemistry and pollution studies, i.e., phosphate, sulfate, and silicate. Studies in our laboratory have been aimed at obtaining specific inhibitions for each of these anions. Among the parameters investigated were: pH, beam height, fuel:oxidant ratio, burner type, total gas flov, rate, and metal ; cation:inhibiting anion ratio. Among these parameters, fuel:oxidant ratio (i..e. , flame temperature) and metal cation:anion ratio U.e., solution stoichiometry) were found to be most useful. To provide for convenient manipulation of solution stoichiometry as well as for an accurate analysis method 2 a titration procedure was adopted. We have termed the technique AAIT—atomic absorption inhibition titration. In this procedure the titrant, typically 100 yg/ml MgCl2 is added by an infusion pump to 100 ml of the sample solution which contains 0.1 to 25 yg/ml of anion while the sample solution is being sampled by aspiration into the flame and the magnesium atomic absorption is monitored. Magnesium is selected as the monitor metal because of its sensitivity and its high extent of inhibition by refractory forming anions. Atomic absorption inhibition titration (AAIT) has been successfully employed to determine phosphate, silicate, and sulfate as well as simultaneous determination of any combination of these anions in drinking and waste water. ------- EXPERIMENTAL Methods based on inhibition effects require the use of relatively cool flames in which the compounds formed in the evaporating droplets are stable. The hydrogen-air flame provides conditions at which inhibition of these anions (Si02, PO4, and SOj) is realized without significant interference from less stable compounds. A number of experimental adjustments influence inhibition by anions. Inhibition ordinarily increases with excess of either fuel or oxidant gas flow, but decreases as beam height or total gas flow rate increases. These variables establish flame temperature in the light path. The optimized conditions for silicate determination has the combination of hydrogen flow rate of 10 ft3/hr (30 psi), air, 10 ft3/hr (40 psi) and for sulfate, 30 ft3/hr (30 psi) for hydrogen and 10 ft3/hr (30 psi) for air. The previous statement indicates a possibility to determine both silicate and sulfate anions at certain combinations of instrumental parameters with minimum interference from each other. To obtain well characterized titration curves it is necessary to exchange cations with a hydrogen form ion exchange resin for standard as well as sample solutions, i.e., sodium and potassium ions impair sharpness of 4 titration signals. Such separations are accomplished in jost a few minutes via a batch process because the ion concentrations involved are so low. ------- RESULTS Phosphate Determination by AAIT: The anions selected in this study are ortho-, pyro-, tri-, tetra-, and hexa- phosphates. Observations were made with both total consumption and pre-mix slot burners. A hydrogen-air flame was used in order to have the relatively low flame temperatures and the flame temperature control which is necessary. The total consumption burner was a Jarrell-Ash "HETCO" model. The pre-mix burner had a 0.5 x 100 mm slot. For manual titration an ordinary glass buret was used with 0.027 in. I.D. plastic tubing leading into the titration solution in a 250-ml beaker stirred with a magnetic stirrer. All solutions were prepared from reagent grade material using deionized distilled water (unless specified otherwise). Sodium hexametaphosphate and tetraphosphoric acid were commercial materials (Benlo Chemical Co.). To remove cations, the solution was treated with cation exchange resin (Dowex 50 x 8, 20-25 mesh) using a separatory funnel by a batch method. The cation exchange 'resin was exposed to dilute acid solution between uses. Several detergent product brands were selected randomly from among those sold locally. Before beginning a titration, gas flow rates were adjusted for maximum inhibition using a magnesium plus phosphate solution. The atomic absorption signal was observed at intervals during the titration by dipping the aspirator tube into the titration vessel. Steps were taken to insure that the titration solution level with reference to the burner did not change and that solution uptake rates were not sufficient to introduce significant errors. A complete manual titration curve was obtained in less than one-half hour. Semi-automatic titration requires less than five minutes. Results for the total consumption burner inhibitions by the three anions, orthophosphate, silicate, and sulfate, were similar. The extent of inhibition was proportional to the magnesium absorption signal and generally followed the temperature profile of the flame. Rather wide variation of hydrogen to air flow ratios (0.5 to 4.0) and total gas flow rates (15 to 50 ft /hr) showed little effect on the relative amounts of inhibition among the three anions. The degree of inhibition is similar for phosphate and silicate and is somewhat less for sulfate. The inhibition effect is relatively linear with anion concentration for mole ratios of inhibitor (anion) to ------- absorber (Mg) less than unity and anion concentrations down to 0.5 ug/ml. This linear response suggests the possibility of relatively precise and sensitive indirect measurement of these anions by their inhibiting effect. The specific analysis of one anion in the presence of the others is unlikely, however, because the relative inhibitions of sulfate, silicate, and phosphate are independent of vertical position, fuel to oxidant ratio, total gas flow rate, and anion concentration. Data taken using a pre-mix burner are significantly different from those with the total consumption burner due to the smaller range of droplet sizes. Although with the latter, phosphate and silicate inhibited similarly. Figure 1 shows that with the pre-mix flame a progression in inhibition is observed. Inhibition for various fuel to oxidant flow rates is also shown in Figure 1. Several features of these plots are significant. Absorption by magnesium is at a maximum value and inhibitions are at or very near minimum values at hydrogen to air flow ratios near 2:5. This ratio corresponds to a stoichiometric mixture for combustion of hydrogen with air and thus a maximum in flame temperature. . At lower or higher ratios the cooler flame results in more inhibition for all three anions, but also less absorption by magnesium. Observations for many more combinations of gas 8 Figure 1. Inhibition and Gas Flow Rates Pre-mix burner Burner height, 10 mm; X = 285.2 nm; pH = 7 ------- OT % ABSN OB O 8 o o 3 O «• 2 ro O • ot o IN5 O •o •O 3 O o m (J> o r—r—i ro O w o o c flow rates all yielded a similar pattern. When total gas flow rate is increased while maintaining constant hydrogen to air flow ratio the inhibition decreases due to the increased thermal input to the flame. Inhibition for all three anions increases moderately with pH and with decreasing beam height in the flame. These trends were not distinctive for any one of the anions and were not considered suitable for application to selective determination of anions. Data obtained here, in agreement with other workers, show that when the mole ratio of inhibiting anion to metal exceeds approximately unity the extent of inhibition no longer increases on increasing the concentration of anion. At anion to magnesium mole ratios less than unity the occurrence of discontinuities in solution stoichiometry vs_. absorption was observed for phosphate and silicate solutions. Titration permits convenient variation of the solution stoichiometry. Titration also allows easy maintenance of other experimental conditions during measurements. An example of a curve obtained when titrating phosphate with magnesium is shown in Figure 2. The remarkable shape of this curve forms the basis of the new measurement technique for low concentrations of phosphate. It represents a unique type of titration chemistry in that high temperature 11 ------- Figure 2. Titration of Blank and 8 ppm Orthophosphate Titrant, MgCl2 solution Gas flow rates: Hydrogen, 25 ft /hr (60 psi) Air, 5 ft3/hr (80 psi) 12 FIGURE 2 100 90 80 70 to CD 0 LO 2,0 3.0 4.0 5.0 60 7.0 ftO 9.0 10.0 ml 200 ppm Mg 13 ------- processes on small bits of solution is the observed quantity. Reverse titrations of metals with inhibiting anions might also be advantageous in certain cases. On the titration curve shown in Figure 2, points suitable for application to phosphate determinations are designated as A, B, and C. The shape of the curve combined with the titration data indicates that any of these points may be used for convenient endpoint designation. The three titration endpoints suggest refractory compound formations with approximately two, three, and four magnesium atoms per phosphorus atom. Such stoichiometries do not correspond to magnesium phosphate compounds or complexes of ordinary solution chemistry, but rather reflect high temperature processes occurring in the evaporating droplets. The sample droplet is subject to drastic reactions including formation, dehydration, and volatilization of dry aerosol particles. All these processes occur in about one millisecond. Stoichiometries quite different from those of ordinary solution chemistry are to be expected. The decreasing signal before point C can be assigned to a rate process involving formation of a more stable refractory species from a less stable initial species. The effect is even more prominent for solutions containing sulfate. This titration method allows investigation into the inhibition mechanisms. 14 Table I summarizes results for titrations of solutions containing sulfate as well as phosphate. The data show general agreement with titrations where no sulfate is present. The effects of sulfate on the titration curve were to decrease the absorption throughout the titration and to enhance the minimum at point C. Polyphosphate Titrations: Polyphosphate titration curves had the same general shapes as those for orthophosphate. Titration stoichiometry was examined under the same instrumental settings as for phosphate titration. The endpoints used corresponded to point A of Figure 2. The magnesium to phosphorus atom ratio appears to approach a constant atom ratio of unity with increasing size of the polyphosphate, whereas for orthophosphate the ratio is near two. These results suggest a kinetically determined inhibiting process in the droplets. Silicate can be expected to interfere when present in samples. It must either be first separated or it must be 4 determined by AAIT and subtracted from total silicate and phosphate results. Analysis of Milwaukee River water was performed in order to examine the utility of the analysis system. The slightly acidified sample was filtered through paper and 15 ------- TABLE I Titration PO4/Mg Mole Ratios for Solutions Containing Phosphate and Sulfate ppm P04 5 5 10 10 10 10 15 ppm S04 0 10 0 5 10 15 10 A 0.551 0.613 0.556 0.527 0.551 0.538 0.602 B 0.319 0.376 0.297 0.321 0.366 0.334 C 0.261 0.285 0.244 0.263 0.289 0.248 0.562 0.336 0.262 16 then a 0.4 micron membrane filter. It was then passed through a cation exchange column in the acid form to replace naturally occurring magnesium, calcium, and other cations. The sample was diluted ten-fold for convenience. Titrationa were performed on river water, river water plus standard addition of phosphate, and standardized phosphate solutions. The results are shown in Table II. Accuracy and precision are good considering the speed and sensitivity of the method. The absence of measurable silicates in the samples titrated was confirmed by observing a negligible endpoint for a titration at conditions under which only silicate would be detected. Thus, either the river water contained negligible dissolved silicate or silicate was quantitatively held in the cation exchanger. For best results, a standard addition must be titrated with the same flame adjustment and instrument settings used for the samples. Phosphate determination was also performed on four brands of commercial detergent products (Table III). The detergent solutions were either boiled for one hour or aged for several days before titration. Determination of sample No. 1 of the series of water analysis standard solutions offered by the Analytical Qua- lity Control Laboratory yielded 0.050 (s = 0.001) ppm phosphorus compared with the stated value by AQCL of 0.05 ppm. 17 ------- TABLE II River Water Analysis 2 ppm PO4 (ml) 0.73 0.59 0.50 0.51 0.57 :un-i .ruim River (ml) 1.50 1.23 1.08 1.05 1.29 River + 2 ppm P04 (ml) 2.24 1.80 1.60 1.61 1.84 uoncen P04 -in River Water (ppm) 4.14 4.10 4.40 4.32 4.46 trataon Added PO4 Determined (ppm) 1.97 1.93 2.08 2.16 2.12 x = 4.28 x = 2.05 s = 0.16 s = 0.10 TABLE III Determination of Phosphate in Detergent Products by AAIT Brand Number 1 2 3 4 POa Found— (ABIT), 36.2 ± 0.5% 36.1 ± 0.2% 36.1 ± 0.5% 28.9 ± 1.0% PO. Found (Std Method) 35.8% 36.5% 36.6% 24.6% ^Standard deviations based on five replicate titrations. Percent PO. in commercial material as received. 18 19 ------- In conclusion, the results of this investigation indicate that atonic absorption inhibition titration (AAIT) is applicable to phosphate determination in surface and waste waters. A pre-nix burner is necessary for the AAIT method. Hhen the phosphate solution is titrated with magnesium, three break points were found. At these points the magnesium to phosphate ratio remain essentially constant, and therefore can serve as the endpoints for titration, even in the presence of sulfate. These details for AAIT determinations of phosphates are cited in reference S. Silicate Determination by AAIT: AAIT titration curves for silicate are of the same shape as those for phosphate. With use of a flame of sufficiently high temperature, silicate inhibits magnesium atomic absorption, but phosphate and sulfate do not. Accordingly, silicate can be determined in the presence of these ions in water samples. Typical data for such determina- tion is given in Table IV. A report on silicate determination by this new method developed here is already in print*. TABLE IV Determination of Silica in Artificial Drinking Water SiO, added Tppm) SiO, det'nd tpp»> Number of runs 0.50 1.00 2.00 3.00 4.00 0.54 1.00 2.06 3.09 3.77 2 2 1 1 1 X (mean yg SiO2 found/yg SiO2 present) = 1.02 s = 5.1% 20 21 ------- Sulfate Determination by AAIT; Of the three anions considered here, sulfate has least tendency to undergo inhibiting reactions with magnesium in the droplets. Therefore, the AAIT determination of this anion requires use of a. relatively cool flame. This was accomplished by using a fuel rich flame with hydrogen flow rate, 30 ft /hr, and air flow rate, 10 ft /hr. Inlet pressure to the flow meters was 30 psi. Under these conditions a simple titration curve is obtained in which a linearly increasing absorbance is observed beyond the titration endpoint. A linear extrapolation of this segment to the base line conveniently defines the endpoint. Typical data is indicated in Table V. Determinations over a pH range of 1-9 resulted in no significant change in results obtained. Precision measurements on a 5 ppm SO, indicate a standard deviation of 0.2% for six determinations. Such titrations using 100 ppa Mg as titrant required less than 1.5 min to perform. Chloride, acetate, nitrate, and fluoride do not interfere; however, phosphate and silicate do interfere and must be removed. The latter two anions are much more basic than sulfate and so previous separation of them (e.g_. , ion exchange) is readily accomplished. A report describing sulfate determinations by AAIT as developed in this project is already in print and can be 22 TABLE V Determination of Sulfate by AAIT ppm S04 added ppm SO^ found 1.00 2.00 4.00 8.00 16.0 20.0 0.86 1.94 4.00 8.00 15.9 20.1 23 ------- consulted.. Simultaneous Determination of Anions by AAIT: Studies of AAIT curve shapes with more than one of the anions present indicated that with proper choice of flame conditions remarkably useful signals were obtained. Figure 3 shows titration curves obtained for solutions containing one yg/ml silicate and 25 vg/ml sulfate. Points A and B serve as endpoints for silicate and C minus B serves as the endpoint for sulfate. Figure 3 also shows data obtained for optimizing hydrogen to air flow ratios. Up to R = 0.5, this small amount of silicate does not show a titration signal, whereas at ratios greater than 1.5 the silicate signal predominates. In order for simultaneous determination of silicate and sulfate together, the background intensity after the silica peak must be as low as possible without sacrificing the clarity of the silicate endpoint. The best condition for our work is hydrogen, 10 ft /hr (30 psi) and air, 7.5 ft /hr (30 psi). In the course of optimization we have used both the "Tri-Flame" pre-mix laminar flow burner head and the newer, improved teflon-lined burner head (Jarrell-Ash 82-374). It is found that the newer burner is more efficient in nebulizing solution. 24 Figure 3. Effect of Air:Hydrogen Flow Ratio, R, on Titration Curve Hydrogen flow rate, 10 ft3/hr (30 psi) 25 ------- FIGtTKE 3 9Or 80- R=0.8 = 0.75 Simultaneous determination of silicate-sulfate system requires calibration standards for both components. These standards are least square fitted and show a linear correlation coefficient of 0.998 for 0.50 to 4.00 pg/ml SiO2 in the presence of 20 pg/ml SO, and 0.984 for 5.00 to 30.0 ug/ml SO. in the presence of 1.0 iig/ml SiO2. Thus, linear calibration for simultaneous determination of silicate and sulfate is possible. This method has been applied to determine silicate and sulfate in raw water from Lake Michigan and drinking water from the Milwaukee Water Purification Plant. The results are summarized in Table VI. Data obtained by their laboratory using standard gravimetric methods are included for comparison. It is important to point out that the recovery of added one ug/ml silica in raw and effluent waters is almost 100% in the AAIT method even though the calibration curve was obtained in deionized distilled water. The small discrepancy between the standard gravimetric and AAIT methods was further investigated by synthesizing drinking water according to the data published by the American Chemical Society. The high accuracy and precision shown in Table IV lead to the conclusion that the AAIT method is valid for Si02 at these levels. The gravimetric data apparently contains 27 TIME, min. ------- TABLE VI Data Comparing AAIT and Gravimetric Methods Sample Raw H2O Raw H2O -Concentration of SO. Concentration of SiO.— (ppm) (ppm) AAIT Gravimetric AAIT Method- Methods Method- 21.5±0.1 21.7±0.5 20.7 1.0 ppm SiO2 25.3±0.2 1.39*0.04 2.41+0.05 1.0 ppm SiO2 Drinking HjO 24.8±0.1 26.0 0.95±0.01 Drinking H. 2.03±0.03 Gravimetric Method^. 1.2 1.0 ^Deviations shown indicate range for two or three determinations. -As reported by Milwaukee Water Purification Plant. 28 negative error. Table IV indicates that the common anions present in drinking water do not interfere in the silica determination. A series of simultaneous titrations of samples' containing various relative amounts of silicates and sulfate shows that the method is valid whenever the ratio of sulfate to silicate is larger than 2. A survey of data for municipal water supplies indicates that many fall into this category. Addition of known amounts of sulfate provides easy accomodation of those which do not. Further investigations were aimed at even more general application to waste waters and at including phosphate in the analysis. Data obtained for solutions containing silicate, sulfate, and phosphate showed that relatively linear relationships between points A, B, and C on the titration curve of Figure 3 were obtained as follows: A = k1[Si02) + b (1) B - k2(Si02] + k3[P04l + b (2) C = k4[Si02l + k5[P04] + kg[S04] + b (3) where bracketed terms indicate concentration. A computer-assisted multiple linear regression analysis treatment of an array of experimental data resulted in evaluation of the coefficients in the set of equations: 29 ------- [S04] KgC - K?B + KgA + Kg (4) (5) (6) Thus a single titration for analysis of all three anions of interest is possible. Such titrations ordinarily require about five minutes to perform. Actual titration time is usually less than three minutes. Data obtained for artificial waste water show that when total yg/ml of SiO2 and PO, exceed one-half of ug/ml SO. the.error for SiCX, and PO. determination becomes 4 24 excessive. This is attributed to merging of points B and C of the titration curve (Figure 3). This limitation, however, does allow determination of most surface waters and many waste waters (See Table VII). Again, a full report of the new method developed for simultaneous determination of the three anions in a single sample has been published , 30 TABLE VII Simultaneous Determination of SiO2, PO4, and SO4 in Waste Water Aaaea Si02 1. 1. 4. 10. 1. 50 50 00 00 00 I PO 3. 2. 2. 5. 4. Mg/m 4 00 00 00 00 00 -U so4 20. 20. 20. 20. 20. 0 0 0 0 0 Si02 1.44 1.60 3.48 7.30 1.06+0.15- Pi 3 1 IMg/aij.;- .04 .56 2.38 9 .56 4.04+0.10- SO, 20. 20. 20. 20. 21.3±0 5 3 3 0 „ S3- ^Mean and estimated standard deviation for five replicate titrations. 31 ------- DISCUSSION The analytical technique developed in this project is characterized by speed and convenience, especially when compared to alternative methods for determining the three anions (POj, SiOj, and SOj) in a single sample. The lower determination limits are comparable to present standard methods. The apparatus required, even for the semi-automatic titration, is not extremely costly, i.e., an atomic absorption spectrometer and an infusion pump. The first of these items is already part of the equipment of many water and waste water laboratories. Several of the advantages of atomic absorption are thus made available for these common important anions. An additional advantage is that since procedurally the method is a titration, the demands on the spectrometer are not high, i^.e., only changes in signal are observed. Hopefully instrument suppliers will exploit these features to provide a relatively low cost unit which will allow for determination of magnesium and other common metals by atomic absorption or emission and these anions by AAIT. The apparatus used in the method is suitable for automation. Although a beginning in the procurement of equipment for this task was possible under the budgetary limitations of the project, this effort is not yet complete. 32 At least two benefits of the program to chemistry have occured. One is that AAIT provides a new tool for observing inhibition processes in flame spectrometry. This provides elucidation of inhibition mechanisms. In general these studies have shown that the inhibition processes are rate determined processes as opposed to solution equilibrium regulated processes. Finally, it should be noted that, although the method is based on inhibition effects, it is not an indirect method but rather employs the procedural advantages of direct titrations and atomic absorption measurements. 33 ------- REFERENCES 1. Anon. Cleaning Our Environment. The Chemical Basis for Action. American Chemical Society, Washington, D.C., 1969, pp 109* 152. 2. Alkemade, C. Th. J., Flame Emission and Atomic Absorption Spectroaetry, J.A. Dean and T.C. Rains, Eds., New York, Marcel Dekker, 1969. Vol. I, Chpt. 4. 3. Smith, R., Stafford, C.M., and Winefordner, J.D. Anal. Chem. 41:946, 1969. 4. Looyenga, R.W. and Huber, C.O. Determination of Silicate in Wastewater by Atomic Absorption Inhibition Titration. Anal. Chem. 43:498-501, 1971. 5. Crawford, W.E., Lin, C.I. , and Huber, C.O. Atomic Absor- ption Inhibition Titration of Orthophosphate and Polyphos- phates. Anal. Chim. Acta. 64:387-395, 1973. 6. Looyenga, R.W. and Huber, C.O. The Determination of Sulfate by Atomic Absorption Inhibition Titration. Anal. Chim. Acta. 55:179-183, 1971. 7. Lin, C.I. and Huber, C.O. Determination of Phosphate, Sili- cate, and Sulfate in Natural and Wastewater by Atomic Absorption Inhibition Titration. Anal. Chem. 44:2200-2204, 1972. LIST OF PUBLICATIONS "Determination of Silicate in Waste Water by Atomic Absorption Inhibition Titration," R. W. Looyenga and C. O. Huber, Analytical Chemistry, 43, 498-501 (1971). "The Determination of Sulfate by Atomic Absorption Inhibition Titration," R. W. Looyenga and C. O. Huber, Analytica Chimica Acta, 55, 179-183 (1971). "Determination of Phosphate, Silicate, and Sulfate in Natural and Waste Water by Atomic Absorption Inhibition Titration," C. I. Lin and C. O. Huber, Analytical Chemistry, 44, 2200-2204 (1972). "Atomic Absorption Inhibition Titration of Orthophosphate and Polyphosphates," W. E. Crawford, C. I. Lin, and C. 0. Huber, Analytica Chimica Acta, 64, 387-395 (1973). 35 ------- |