Determination of Heavy Metals and Silica in Chromate-Treated

TN THE analysis of chromate-treated cooling waters, the color and chemical reactivity of the chromate increase the difficulty of analyzing for certain...
0 downloads 0 Views 392KB Size
ANALYTICAL CHEMISTRY

1644 drying under vacuum. The 1% nitrogen range given in the last column of Table I is that stated on the labels of the samples. Blank values for this procedure increase gradually as the time of reaction is increased; after 30 minutes they are approximately double the 5-minute values. Consequently, the precision of the determination decreases somewhat for compounds which require more than 10 minutes for the liberation of all of their amino nitrogen under the conditions present in the reaction flask. 4CKNOWLEDGMEYI

The authors gratefully arknodedge the assistance of D.

Warren Stanger, whose comments and suggestions were especially helpful. LITERATURE CITED

Hershberg, E . B., and Kellwood, G. W., ISD. E S G . CHEX.,ASAL. ED.,9,303 (1937). (2) Xiederl, J. B., and Niederl, V., “Micromethods of Quantitative Organic Analysis,” 2nd ed., pp, i9-99. 301-10, S e w York, John TViley &- Sons, 1942. (3) Pomatti, Renato, 1x11. ENG.CHEM.,-2x.k~.ED.,18, 63 (1946). (4) Van Slyke, D. D., J . B i d . Chem., 9, 185 (1911): 12,278 (1912).

(1)

RECEIVED for review X a y 28, 1952.

Accepted June 17, 1952.

Determination of Heavy Metals and Silica in Chromate-Treated Cooling Waters SIDYEY SUSSMAN AND IRVING L. PORTNOY r a t e r Service Laboratories, Znc., ‘Yew York 27, h’.Y .

N T H E analysis of chromate-treated cooling waters, the color

I and chemical reactivity of the chromate increase the difficulty of analyzing for certain ions. In some cases, it is possible to remove this interference by reduction of the chromate ion to chromic ion, precipitation of the hydroxide, and separation of the precipitate by filtration. This process is laborious and timeconsuming. In recent years, ion exchange resins have proved useful tools for separating various components of solutions prior to analysis ( 2 ) . I t appeared likely that ion exchange could be utilized for the separation of chromates prior to analyzing cooling waters for silica and for heavy metals. This proved correct, and the authors have been able to develop procedures which are simple, rapid, and sufficiently accurate for the ordinary requirements of water treatment control. SEPARATION OF HEAVY METALS

It is often necessary to determine corrosion products, principally iron and copper, in treated cooling waters. Removal of these heavy metals from chromate-containing samples can be effected by passing the sample through a column of acidregenerated cation exchanger. The heavy metals, retained on the euchanger, can be recovered by elution with an acid.

+ C U + ++CUR + 2 H + CUR + 2 H + +HlR + CU”

H2R

minute followed by a distilled water rinse a t the same flow rate until the effluent became light pink-yellow to methyl orange indicator. The tube was now fully regenerated. This operation is necessary only when utilizing new ion exchange resin. In normal use, the acid elution of the heavy metals from the ion exchanger also regenerates the resin. To carry out a separation, a 50-ml. water sample was passed through the bed a t 10 ml. per minute. Distilled water was then introduced until the effluent was no longer yellow, and the effluent to this point was discarded. Twenty-five milliliters of 15% hydrochloric acid were then passed through the exchanger a t 10 ml. per minute followed by a 50 m1.-portion of distilled water to flush the acid through the ion exchanger. The effluents of both acid and rinse portions were combined. One milliliter of concentrated nitric acid was added to the combined effluent prior to evaporating to dryness. The residue was treated with 2 ml. of 3 to 2 hydrochloric acid and warmed to bring it into solution. Distilled water was added. The solution was boiled, cooled, and made up to 50 ml. This solution was used for determining the heavy metals by the usual procedures (1). With the small portions of the water handled, it was not necessary to backwash the exchanger bed for each sample. However, it is desirable to backwash with distilled water from time to time in order to prevent the accumulation of a dirt layer a t the top of the exchanger bed.

Table I.

(1)

Effect of Separation Procedure on Iron and Copper Content of Cooling Waters

(2)

Small amounts of organic matter, presumably leached from the exchanger, tend to interfere with the normal analytical determinations for the heavy metals in the acid eluate. The interference can be removed by evaporating the effluent solution to dryness, with addition of a small amount of nitric acid or other oxidizing agent. The residue is then dissolved with hydrcchloric acid and diluted as required for the usual analyses for the heavy metals. Apparatus and Reagents. A borosilicate glass tube 17 mm. in inside diameter X 300 mm. long was used as the exchange tube. This was equipped with a thin pad of glass wool to support the ion exchange resin, a small diameter outlet tube, and an S-trap t o prevent draining of the liquid in the tube below the upper surface of the ion exchanger. A screw clamp, inserted in a rubber section between the bottom of the ion exchange tube and the outlet of the S-trap, permitted regulation of the flow rate. An enlarged diameter section a t the top of the exchanger tube or a funnel set in a stopper inserted in top of the tube provided storage capacity for feed solutions. The cation exchange resin used in this work was Dowex-50 (Nalcite HCR, National Aluminate Corp., Chicago). This was used in the usual small spheres furnished for water conditioning. Presumably any other sulfonated polystyrene-type cation exchange resin could be used equally well for this separation. Procedure. Twenty-five milliliters of 15% hydrochloric acid were passed through the ion exchange tube a t 10 to 12 ml. per

Iron

Original Sample, P . P . X . Fe 0 8 9.5 10.5

Copper

11.5 12 13 13.5 14.5 17 23 31 P.P.M. c

After Separation of 500 P.P.M. NalCrOl, P.P.M. Fe 0 8 9: 10.5 11.5 12 11 13.6 10.5

13.5 19 29.6

u

P . P . M . Cu

n

n

6.7

0.9 2.4 4.2

2.2

4.2

Experimental Results. Typical experimental data are presented in Table I. In these examples, aater samples containing varying concentrations of iron and copper \\*ere analyzed by the usual procedures ( I ) , and sodium chromate was added to additional portions of the same samples to a 500 p.p.m. concentration. These chromate-containing samples were then passed through the ion exchange separation described above and analyzed for the heavy metals. Agreement between the results obtained with the original samples and those obtained after the

V O L U M E 2 4 , NO. 10, O C T O B E R 1 9 5 2

1645

chromate separation was quite satisfactory in the concentration ranges normally met in water treatment practice, although a loss of iron was apparent a t concentration ranges higher than those normally encountered. Erratic results were obtained when the iron content of the influent exceeded 12 p.p.m. This ion exchange separation has been in use in the laboratory for more than a year with satisfactory results. SEPR.ATIOii OF SILICA

In contrast t80the heavy metals separation, an anion exchange method was used for the elimination of the chromate interference wit,h the determination of silica. &inanion exchanger chloride has been used for the removal of chromate from water by a true exchange reaction ( 5 ) . By utilizing the chloride of a n-eakly basic anion exchange resin for this reaction, the chromate xas completely removed from the solution and vas replaced by an equivalent concentration of chloride, 17-hereas the silica present in the water passed through the exchanger unaffected. I t was t,hen determined by usual colorimetric procedures (1).

+ CrO4-- +RyC1.04 + 2C1R2CrOa + 2 0 1 1 - +21i O H + Cr04-R OH + H - + C1- --+ R C1 + H10 2R C1

lowed by about 500 ml. of distilled water a t the same rate. The regenerated exchanger Tvas then converted to the chloride salt by means of hydrochloric acid as described above. Experimental. This procedure was carried out with various chromate-free waters containing 3 to 100 p.p.m. of silicon dioxide. Yo difference was detectable between the analytical results for silica before and after passing the samples through the exchanger tube, provided that sufficient sample was passed to n-aste before taking a sample for analysis. Experiments indicated that the proper quantity to discard before sampling was 2.5 to 3 bed volumes. Table I1 shows results of one such test in which a water containing 10 p.p.m. of silicon dioxide was passed through an anion exchanger tube previoudy thoroughly rinsed with distilled water. Table 11. Displacement from .imberlite IK-4B (40-1111. resin bed

1 olunie)

Sample Before

Total Effluent Vol., hII. 0

Silica in Last 25 AI1 , P.P.M

(4)

3

50 75

(5)

6

5.5 9.5 10.0 10.0 10.0

(3)

This method is effective for dissolved silica present in the form of silicates. I n the concentration ranges normally met with in water, this probably covers materials variously referred to as crystalloidal and colloidal silica (3). In the presence of excess fluorides, dissolved silica would probably react as the fluosilicate anion and would be retained by the anion exchanger chloride, so t,hat the method would not be applicable in this case. The effluent obtained while passing the chromate-treated water through the anion exchanger cannot be used for heavy metals determinations because suspended iron will be filtered out by the resin, and copper will react with the anion exchanger to form a n insoluble cuprammine complex (4). As the anion exchange resin in the separation, Amberlite I R 4 B (Rohm & Haas Co., Philadelphia) and De-Acidite (Permutit Co., Xew York) have been used with equally satisfactory results. Other weakly basic anion exchange resins should also be satisfactory. Regenerative Procedure. IVhen this separation mas first adopted, it was carried out on a regenerative basis. The ion exchange tube was filled to a depth of a t least 15 cm. with anion exchange resin which had been thoroughly soaked in water. After backwashing, the exchanger was converted to the chloride with 570 hydrochloric acid. Twohundred milliliters of the acid were passed through the tube a t 10 ml. per minute, followed by distilled water a t the same rate until the effluent acidity to methyl orange dropped below 2 meq. per liter. The exchanger was now regenerated and rinsed. In carrying out the silica separation, a chromate-containing water sample, previously filtered t.hrough a Whatman No, 2 filter paper, was passed through the exchanger tube a t about 10 ml. per minute. In order to compensate for dilution of the silica in the sample by t.he distilled water occupying the voids in the exchanger column, the first 75 to 100 ml. of effluent were discarded. The next 10 to 25 ml. were collect’edin a clean container and used for the determination of silica by the usual procedures (1). The number of samples that can be run through an anion exchange tube before regeneration is necessary will, of course, depend upon the chromate content of the samples. -4s a precautionary measure, the authors have adopted the practice of regenerating after not more than 6 samples. In any event, if a yellow color appears in a sample effluent after the first 25 ml., the run should be stopped, and the anion exchanger should be regenerated. Regeneration was effected by treating the exchanger tube Tyith 1500 ml. of 1 to 1 ammonium hydroxide a t a rate of 10 ml. per minute. This large excess of reagent may be conveniently fed automatically by means of a flask containing a short glass tube at least 8 mm. in diameter inserted through a one-hole stopper. If this flask is inverted into the feed funnel or enlarged section of the ion exchanger tube, it will feed the reagent automatically without further attention. The ammonium hydroxide was fol-

1 2

?

25

100 I25 150

10.0

0.0

Single Use Procedure. The regenerative procedure becomes someiyhat burdensome when silica-chromate separations are required only infrequently. Under these circumstances, a modified procedure was adapted in which a smaller resin bed is used once and is then discarded. The anion exchanger was converted to the chloride in bulk. Generally about 250 grams were prepared by pouring the alkaliregenerated resin as received from the manufacturer into 3 or 4 inches of water in a 1-liter beaker and stirring until the resin was thoroughly wetted. About 500 ml. of 1to 1hydrochloricacidnere then added with continuous stirring. The acid should be added cautiou*ly as frequently there is a vigorous gas evolution a t this point. A few minutes after the acid is added, the supernatant liquid should be strongly acid to methyl orange. ilt least 10 minutes after the last acid addition, the beaker was decanted, and the resin was washed 8 to 10 times by decantation using distilled water and stirring vigorously after each addition of water. After the last rinse, the resin was transferred to a closed container for storage. Preparation of the anion exchanger chloride in bulk can also be carried out in a larger ion exchange tube so that a true countercurrent contact is obtained between the resin and the regenerant acid or rinse water. However, the batch method was satisfactory for this application. In the single use or “throw away” procedure, a small Buchnertype glass funnel was used (Corning No. 36060, capacity 15 ml., with fritted disk, 20 mm. in diameter, porosity C). The outlet of the funnel was placed in the mouth of a 250-ml. Erlenmeyer flask without a stopper. Enough anion exchange resin chloride was added to the funnel to make a layer about 1.5 cm. deep. The funnel was filled nith distilled water and allowed to drain. The distilled water rinse was repeated several times in order to remove any yellow color imparted to the water by the resin. The funnel was then filled with the chromate-containing sample and was permitted to drain into the same flask. This step was repeated and the filtrate was discarded. After transferring the funnel to a clean flask, two successive portions of chromate-containing water were poured through it. Because of the shallow resin bed, it is sometimes possible for some chromate to remain in the effluent a t this point. Therefore, the funnel was transferred to another clean flask, and the filtrate was passed through the resin a second time. Silica was determined colorimetrically on the effluent from this second filtration. The small portion of resin was discarded after each analysis. The fritted-glass disk had a tendency gradually to clog with resin fines. Therefore, after every 5 or 6 analyses, the Buchner funnel

ANALYTICAL CHEMISTRY

1646 was cleaned by passing a chromic acid cleaning solution through the disk and then rinsing until no trace of chromate color remained. The regenerative method has been in use in the laboratory for 3 years, and the single use method has been used for the past year with excellent results. The procedures are sufficiently simple and convenient t o be carried out by semiskilled technicians. LITER4TURE CITED

(1)

Am. Pub. Health Assoc., “Standaid Methods for the Exami-

(2)

(3) (4) (5)

nation of Water and Sewage,” 9th ed., pp. 47-9 (Cu), 53 (Fe) 46 (SiOn),New York, 1946. Kunin, Robert, ANAL.CHEY.,21, 87 (1949); 22, 64 (1950); 23, 45 (1951). ROY,C. J., A ~ LJ .. S C ~ 243, . , 393-403 (1945). Sussman, Sidney, “Ion Exchange-Theory and Application.” F. C. Nachod, ed., pp. 244-5, New York, Academic Press, 1949. Sussman, Sidney, Nachod, F. C., and Wood, William, Ind. Eng. Chem., 37,618-24 (1945).

RECEIVED for review February 27, 1952. Accepted July 14, 1952. Presented before the Meeting-in-Miniature of the Kew York Section, AMERICAN CHEMICAL SOCIETY, February 8, 1952.

Application of Absorption Spectrum of Ferric Acetate Complex to Determination of Iron WILHELJI REISS’, J. FRED HAZEL, AND WALLACE M. MCNABB Chemistry Department, University of Pennsylcania, Philadelphia, P a .

’HE purpose of the present work was t o study the absorption characteristics of the ferric acetate complex in acetic acid. Absorption spectra were determined in both the visible and the ultraviolet region. It was found that the properties of the complex in 50% acetic acid in the ultraviolet region could he applied t o the quantitative estimation of iron. S o reference was found in the literature to the determination of iron under thwe conditions. Riban (S) attempted to determine iron in neutral and weakly acid solutions by a ferric acetate color reaction, but wap unsuccessful because of the instability of the complex under thew conditions. Broda ( 1 )has determined acetic acid in an aqueous solution based upon a color reaction with ferric ions. rL

periods of months or which &-ereheated on a steam bath for several hours showed no significant change in the character or in the intensity of their absorption spectra. On the other hand, excess sulfuric acid, hydrochloric acid, and nitric acid were found to interfere with the formation of the ferric acetate romplw

m

-I

t/

.2

1

r

1.0

3.0

5.0

70

9.0

11.0

AcOH M/L Figure 2.

300

380

460

mu

540

620

Figure 1. Absorption S p e c t r u m of Ferric Acetate Complex

The absorption spectrum of the ferric acetate complex in 50.0% acetic acid (8.7 iM)has been determined and is shown in Figure 1. Maxima occur a t 463 t o 465 mp in the visible region and a t 337.5 mp in the ultraviolet region. It was found that the optical densities reached limiting values a t both these wave lengths when the acetic arid concentration corresponded to 50% (8.7 M). These facts are shown in Figures 2 and 3. The complex was found to be stable under these conditions. Solutions which were allowed to stand a t room temperature for 1

Pa.

Present address, Wyeth Institute of Applied Biochemistry, Philadelphia,

.4

Constant V o l u m e Titration w i t h Acetic Acid

2.0

6.0 10.0 MOLES AcOH/L

14.0

Figure 3. Constant V o l u m e Titration with Acetic Acid

Figure 4 shows that the optical density of the complex as determined a t 463 mp decreases linearly with the addition of increasing concentrations of nitric acid, This acid was used t o evaluate the effect of hydrogen ion concentration because the nitrate ion has little, if any, tendency toward complex formation with the ferric ion. The effect of sulfate and chloride anions on the color reaction was obtained by determining Beer’s law curves a t 465 mw with standard solutions of ferric nitrate, ferric chloride, and ferric sulfate in 50.0% acetic acid. The slopes of these curves can be