Salt error

Salz-farbe,/. metallic dye, dye salt, -fehler, m. salt error, -fleisch, n. salt meat, -fiuss, m.. saline flux salt rheum, eczema, salz-fdrmig, a. saliniform. -frei, a. free from salt, salt-free. 

Particularly in autoanalyser methods this wide variation in chloride content of the sample can lead to serious salt errors and, indeed, in the extreme case, can lead to negative peaks in samples that are known to contain ammonia. Salt errors originate because of the changes of pH, ionic strength and optical properties with salinity. This phenomenon is not limited to ammonia determination by autoanalyser methods it has, as will be discussed later, also been observed in the automated determination of phosphate in estuarine samples by molybdenum blue methods. 

Figure 6.2. (a) The effects of salinity on the sensitivity of standard additions of ammonia in laboratory mixed waters ( ) and in waters from the Tamar estuary (A) expressed as percentage of response in river water. For comparison, the salt error curves reported by Loder and Gilbert [3] are also shown (... and —, respectively), (b) Contribution of reactive index and organic absorbance to the optical blacks in the Chemlab Colorimeter. = River water-seawater mixture, o = De-ionized water-seawater mixture. Source [2]... 

The spectophotometric methylene blue method for anionic surfactants has been applied to seawater. In one version, the surfactants are collected in ethyl acetate. The solvent is then evaporated, the surfactants put back in solution in water, and the standard spectrophotometric methylene blue method is applied to this solution. In this manner, the salt error introduced by seawater is eliminated [195]. A similar method, with the methylene blue-surfactant complex extracted into chloroform, and measured directly was proposed by Hagihara [192]. 

These constants determine the titration exponents pH and the best indicators for the successive hydrions. The acid can be titrated as dibasic, using methyl yellow, methyl orange or bromophenol blue, and as tetrabasic using phenolphthalein, thymolphthalein or thymol blue in the presence of a moderate excess of soluble barium salt. The values of pH in the partly neutralised acid were corrected for the salt error, and the constants Kz and jfiT4 which prevail in solutions of low concentration were thus deduced —6... 

A standard of pH 52 may be prepared by mixing 4.2 1 of 0.2 AT acetic acid with 158 cc. of o 2 AT sodium aceta while the 5.4 standard contains 29 cc. of the acid to 1 cc of sodium acetate solution Bromo-cresol purple may used as indicator The range 5.2 to 5.4 was determined wit out correcting for the protein or salt error. For details pH determinations the student is referred to Clark s book. 

Determination of pH With Buffer Solutions.—If a series of buffer solutions of known pH, which must lie in the region of the pH to be determined, is available the estimation of the unknown pH is a relatively simple matter. It is first necessary to choose, by preliminary experiments, an indicator that exhibits a definite intermediate color in the solution under examination. The color produced is then compared with that given by the same amount of the indicator in the various solutions of known pH. In the absence of a salt error, to which reference will be made later, the pH of the unknown solution will be the same as that of the buffer solution in which the indicator exhibits the same color. Provided a sufficient number of solutions of known pH are available, this method can give results which are correct to about 0.05 pH unit. 

For a given color tint, i.e., corresponding to a definite value of a/(l — a), the actual pH will clearly depend on the value of the ionic strength of the solution at low ionic strengths, e.g., less than about 0.01, the neutral salt error is negligible for most purposes. 

The actual neutral salt error is less than estimated by equation (24) because the experimental values of pXin are generally based on determinations made in buffer solutions of appreciable ionic strength. For equation (24) to be strictly applicable the indicator exponent pifin should be the true thermodynamic value obtained by extrapolation to infinite dilution. 

Figure 15.7. Stoichiometric correlations among nitrate, phosphate, oxygen, sulfide, and carbon. The correlations can be explained by the stoichiometry of reactions such as equation 3 concentrations are in micromolar, (a) Correlation between nitrate nitrogen and phosphate phosphoms corrected for salt error in waters of the western Atlantic, (b) Correlation between nitrate nitrogen and apparent oxygen utilization in same samples. The points falling off the line are for data from samples above 1000 m (Redfield, 1934, p. 177). (c) Correlation between nitrate nitrogen and carbonate carbon in waters of the western Atlantic, (d) Relation of sulfide sulfur and total carbonate carbon in waters of the Black Sea. Numbers indicate depth of samples. Slope of line corresponds to AS /AC = 0.36. (From data of Skopintsev et al., 1958, as quoted in Redfield et al., 1966.) (e) Correlation of the concentration of nitrogen to phosphate in the Atlantic Ocean (GEOSECS data). The slope through all the data yields an N/P ratio close to 16.

Silicate is determined as a reduced silicomolybdate dye 14). The method is sensitive to the reductant used, either stannous chloride or ascorbic acid. The reaction kinetics with stannous chloride are much faster than with ascorbic acid. The analysis with stannous chloride is more sensitive, therefore, and has a detection limit of 0.5 pM compared to 1.0 pM when ascorbic acid is used. Eighty determinations can be made per hour at these detection limits. A detection limit of 0.1 pM can be obtained if the sampling rate is decreased to 50 per hour 14). The analysis with stannous chloride has a smaller salt error than with ascorbic acid, but the interference due to... 

The properties of most of these indicators are known only incompletely (purity, salt error, protein error), and they must be employed with great care. It is advisable for practical purposes to select a list of indicators which have proven to be most useful. Such a list is found in the accompanying table. 

The introduction of polar groups, such as the sulfonic acid or carboxyl group, increases the solubility sufficiently to yield rather useful indicators. We shall see in Chapter Ten that these water-soluble azo dyes are especially valuable because their salt error is negligible. 

The commercial preparation may be recrystallized from water. Stock solutions should contain 0.1% of the indicator in water, and two drops of the indicator solution are required per 10 c.c. The transformation interval is between pH 1.3 and 3.0, the color changing from red to yellow-orange. This indicator is very useful, possessing a small salt error. 

Commercial preparations may be purified by adding hydrochloric acid to an aqueous solution in order to precipitate the indicator, which is then washed and dried. A small quantity of alkali, insufficient to dissolve all of the indicator, is added and the solution evaporated to yield the sodium salt. A 0.1% aqueous stock solution is prepared, and 1-3 drops used per 10 c.c. of solution under investigation. The interval occurs at pH 1.3-4, the color going from blue-violet to red. This substance has a high salt error and protein error, and is not recommended for use as an indicator. 

J. F. McClendon used it for determining the pH of gastric juice. The properties of the substance were studied later on in greater detail by I. M. Kolthoff. Commercial preparations are available in the form of a dark red-black powder. The indicator is insoluble in water but dissolves in alcohol producing a dark red solution. A stock solution, containing 0.1 % of indicator in 95% alcohol, should be stored in dark containers. Aside from having a rather large salt error, the indicator is satisfactory. 

Stock solution 0.1% in 70% alcohol store in dark colored containers, and prepare anew every three months. The transformation range is 6.8-8.0, from red to yellow-orange. About 1-3 drops of indicator solution may be used per 10 c.c. The salt error is small, and the compound is a satisfactory indicator. 

This brief and necessarily incomplete review of fluorescence indicators must sufl ce. A more quantitative discussion of the subject would be premature at present because too many factors, such as salt error, effect of indicator concentration, etc., remain yet to be studied exhaustively. Chlorides, for example, diminish considerably the fluorescence of the quinine cations and of the naphtholsulfonic acid anions. Fluorescence indicators are as yet... 

The salt content of the buffer solutions of McIlvaine is rather high, and therefore the salt error of the indicator in these solutions will be much different from the error in other mixtures. 

E. Beesslau has improved the procedure to permit good results with small quantities of liquid and in weakly buffered solutions. To eliminate the salt error (Chapter Ten) as much as possible, Beesslau employed indicator solutions which were more dilute than those prescribed by Michaelis. Because the colors of the various alkaline indicators are the same, it is possible to use 15 to 18 permanent comparison solutions to include the pH range 2.6-S.9. The indicators mentioned in the following tables have the same colors at the pH s recorded. 

The ratio/i /2 in (16) increases with larger salt concentrations much more rapidly than the corresponding/o /i ratio in (14). Hence the salt error of dibasic indicator acids is much greater than of monobasic acids (such as the nitrophenols). Indicator bases may be treated in analogous fashion. 

The above discussion leads us to expect that the benzeins will have a smaller salt error than the corresponding sulfone-phthaleins. 

At smaller ionic strengths (m < 0.05) the influence of the ionic diameter may be neglected for all practical purposes. Thus we may expect all dibasic indicator acids to have the same salt error in dilute electrolyte solutions provided that the buffers used in the pH measurement all have the same ionic strength and contain the same kind of ions. The buffer solutions in colorimetric pH determinations are, however, always considered as standards, and the pH of the buffer solution is taken as a standard value. Hence it follows immediately that the sign of the salt correction will be... 

For these reasons the calculated salt error is erroneous especially at high ionic strengths, and we must content ourselves with the results of empirical measurements. The agreement between the calculated and experimentally determined values, however, is satisfactory at low ionic strengths. One should always remember that the salt correction depends not only upon the ionic strength but upon the nature of the buffer solution used for comparison as well. Workers in this field should make it a point to mention in their reports the kind and composition of the buffers they employ. 

In these tables are to be found a description of the particular electrolyte media employed as well as the ionic strengths at which salt errors were determined. 

Aside from methyl orange and methyl red, thymol blue (in the acid region pH 1.3-2.8) and tropeolin 00 also have negligible salt errors. Clearly the indicators thymol blue (pH 1.3-2.8), tropeolin 00, methyl orange, and methyl red are unusually well suited for the colorimetric determination of paH because they yield reliable results, at not too high ionic strengths, which need not be corrected. 

Alizarine yellow and salicyl yellow must be used with discretion because they show a very large salt error. It would be exceedingly desirable to have available better indicators for measuring the pH in the alkaline region (10-13). 

The salt errors of the most important indicators are summarized in the succeeding table. It is assumed here that the comparisons were all made with buffer solutions having an ionic strength of 0.1, as is usually the condition obtaining in practice. If the ionic strength deviates considerably from this value, the... 

Congo red, azolitmin, and litmus must be rejected as indicators for colorimetric pH determinations because of their very large salt error. The author has found, for example, that the salt error of congo red in 0.5 N sodium chloride is — 0.5, and — 1.0 in 1 N NaCl solutions. Azolitmin in 0.5 N NaCl has an error of - 0.55. 

---