INDUSTRIAL A N D ENGINEERING CHEMISTRY
308
FromA, Figure 5 , it is apparent that the value 0.12/M lies in the curve section where ambient velocity pressure and M are unknowns. Selecting a trial value of 2cf = 1.60, 0.12/M = 0.075. For an ambient velocity pressure of 0.075 inch of water gage, A , Figure 5, indicates a value of 1.57 for M. For M = 1.57, 0.12/M = 0.077 inch of water gage. Further trials do not alter these figures significantly; hence, @
Ambient velocity pressure
=
0 12 = 0.08 inch of water gage 1.57
Ambient static pressure = 3.58 - 0.08 = 3.50 inches of water gage
ANGLEMEASUREMENT. Example Q. The correctly oriented and
leveled pitometer indicates a reading of 0 ’ on the scale when the airfoil chord is horizontal. In the flowing stream the angle scale indicates 7 when a sensitive differential inclined manometer connected to the two direction taps indicates no differential reading. The true ambient velocity pressure (Example 1) is 1.17 inches of water gage. From curve A , , Figure 8, the pitometer with wedge nose reads 2’high a t a velocity pressure of 1.17 inches of water gage. The correct angle of flow therefore is 7 - 2 = 5 O and the direction is from above to below horizontal. If the duct flow is reversed, the pitometer must be rotated 180” to face downstream. In this case the angle, as read, would be 2’ low (Figure 8, A*). Example 5. The correctly oriented and leveled pitometer indicates a reading of 356” when the pitometer chord is horizontal. Immersed in the flow stream, the angle scale reads 347” a t null reading. From front and rear readings, ambient velocity pressure is 0.20 inch of water gage. From AI, Figure 8, the pitometer reads 1 high; so the pointer should have read 346 ’. The angle of flow therefore is 356 - 346 = 10 O and the direction of flow is from below to above horizontal. O
Vol. 42, No. 2
LITERATURE CITED
(1) Aeronautical Research Committee, Reports and Memoranda. No. 71 (December 1912). (2) American Blower Co., Detroit, Mich., Bull. 35 (June 1917). (3) American Society of Mechanical Engineers, New York, “Fluid Meters, Their Theory and Application, Part 1,” 1937. (4) American Society of Mechanical Engineers, New York, “Power
Test Codes, Supplement on Instruments and Apparatus, Part 2, Pressure Measurements,” 1945. (5) Dalla Valle, J. M.,“Exhaust Hoods,” New York, Industrial Press, 1944. (6) Dalla Valle, J. M., “Studies in the Design of Local Exhaust Hoods,” Sc.D. thesis, Harvard University, 1930. (7) Dalla Valle, J. M., and Hatch, T., Trans. Am. SOC.iMech. Engi.
54, NO. B, 31-5 (1932). (8) Eckert, B., “Experiences with Flow Direction Instruments,” Jahrbuch der Deutschen Luftfahrtforschung, 1938; transla-
tion by Natl. Advisory Committee on Aeronautics, Tech (March 1941). (9) Ingram, F. R., Diez-Canseco, E., and Silverman, L., Heating, P i p i n g Air Conditioning (ASHVE Journal Section), 14, M e m o . 969
702 (1942). (10) Jacobs, E. K.,Ward, K. E., and Pinkerton, R. M., National Advisory Committee on Aeronautics, Rept. 460 (1946). (11) Martin, G., “Chemical Engineering,” London, Technical Press. 1928. (12) Ower, E., “Measurement of Air Flow,” London, Chapman C Hall, 1927. (13) Prandtl, L., and Tietjens, 0. G., “Applied Hydro- and beromechanics,” New York, McGraw-Hill Book Co., 1934. (14) Silverman, L., “Fundamental Factors in the Design of Exhaust Hoods,” Sc. D. thesis, Harvard University, 1943. (15) Van der Hegge Zijnen, B. G., Ingenieur (Utrecht), No. 38, Algemeen gedeelte (1929). (16) Whalen, F. G., University of Illinois, Engineering Expt. Station, Bull. 20 (March 1921). RECEIVED M a y 2,1949.
eneration of Chromic Acid Solutions by Cation Exchange RAYMOND L. COSTA Mutual Chemical Company of America, Baltimore, Md. Chromic acid anodic baths, copper stripping baths, and other chromic acid solutions that have been rendered useless by the accumulation of dissolved metals, such as aluminum, copper, and iron, are regenerated by passing through a cation exchange resin. The resin removes the metal ions by exchange with hydrogen ion, thus making the solution suitable for re-use and minimizing the waste disposal problem ordinarily involved in handling these solutions.
I
N RECEKT yeais, control of industrial wastes and abatement
of stream pollution have been subjects of increasing attention. Many laws have been enacted regulating the discharge of wastes into natural waters ( 1 ) . I n the metal-treating industries using chromic acid, such as aluminum anodizing, chrome plating, and copper stripping, satisfactory disposal of spent solutions may involve considerable expense. Dodge and Reams ( 3 ) have listed methods of treating wastes containing chromate ions. The most commonly used is the reduction of the chromate to trivalent chromium by sulfur dioxide, ferrous sulfate, sulfite, bisulfite, sulfide, or metal chips; the resulting solution is then treated with an alkali, lime, or limestone t o precipitate the metal content and neutralize the acidity of the solution before ultimate disposal. Another method, used to a lesser extent, is the precipitation of the chromate with lead or
barium salts; this is effective, but comparatively expensive, and. because of the poisonous nature of lead and barium, requires careful analytical control to avoid excess precipitant. Other investigators have attempted ion exchange methods for removal and recovery of chromate ions from wastes. A brief discussion of the theory of anion and cation exchange has been presented by Kunin ( 7 ) . Sussman, Nachod, and Wood (9) recovered chromate from dilute solutions by anionic resin exchange; they experienced a loss of chromate from reduction by the resin ranging from 10.5 to 24% of the total used in the tests; the solution recovered on regeneration of the resin contained 2.77% sodium chromate in a 13.4 N solution of ammonium hydroxide. Grindley (Q), using Zeo-Karb cation exchanger followed by Deminrolit B anion exchange resin, completely removed chromate from dilute solutions, but found it impossible t o restore the absorptive capacity of the resin by any simple treatment: his data indicate severe attack on the resins, as less than 7% of the total chromate was recovered, the rest having been reduced to trivalent chromium. Bliss (2)j citing examples from the copper and brass industrj , used cation exchange resins for recovery of copper, zinc, and chromium values from brass mill wastes containing bichromate and compared ion exchange with other types of processes for recovery of valuable materials from industrial waste solutions
February 1950
INDUSTRIAL AND ENGINEERING CHEMISTRY
During World War 11, in an attempt to relieve the acute shortage TABLEI. SUMMARY OF QUALITATIVE TESTS Capacity, of chromic acid, both anion and cation CrOa Fez08 CuO AlzOa, CrzOa Metal L./L. Reinarks exchange resins were evaluated in this G./L: G./L.’ G./L: G./L. G./L.’ pH Retention Resin laboratory as a means of recovering A. CHROME PLATINGBATHS valuable components of spent chromic 400 28.6 25.0 .. .. . . Poor . . Resin attacked . . Resin attacked acid solutions. However, no commer(lent .. .. cially available resin was found that Good .. satisfactorily resisted chemical attack Excellent , . by chromic acid. With the introducB. COPPERSTRIPPINGBATHS tion of oxidation-resistant sulfonic acidResin slightly attacked 200 .. 50 .. .. .. .. . Excellent Good Capacity to break-through 100 .. .. 25 .. i.4 type cation exchange resins (such as Capacity to break-through 100 2.5 . . . . . . Excellent 5.6 Amberlite IR-120, Permutit Q, and C. ANODIC BATHS Dowex 50) and the more stable modified 150 .. .. .. ., 15 Good Resin slightly attacked amine anion exchangers, the project was 106 17.3& 0:2 1:04 Excellent 2.6 Capacity approximately t o exhaustion reopened. An ion exchange method 84 .. .. 10.5a 0.2 0.60 Excellent 2.9 Capacity approximately was developed by means of which t o exhaustion 51 .. .. 8.4a 0 1.44 Excellent 3.2 Capacity approximately spent chromic acid solutions may be to exhaustion regenerated for re-use, thus minimizing D. MISCELLANEOUS the waste disposal problem in handling 100 .. .. .. 3.8 Poor . . Contained 42 grams NazO 109 .. .. ., 19.4 1:5 Poor in H,+ . , these solutions. and 31 grams ZnO per and Na In the early stages of the present excycles 1. a Estimated from p H (6). perimentation, it was found that chromate ions can be quantitatively removed from dilute solutions a t low pH by a modified amine anion exchange resin. ferric oxide per liter. Very little iron and copper were removed On regeneration of the resin with sodium hydroxide, a strongly from the solution, and a noticeable amount of chromic ion showed in alkaline solution of sodium chromate was recovered, presenting the effluent, indicating attack on the resin. By diluting the soluabout as much of a waste disposal problem as the original solution with an equal volume of water, it was found that, before the resin was saturated, about 50% of the dissolved metals were abtion. This phase of the work offered no promise with regard t o sorbed. By further dilution to 150 grams of chromium trioxide waste disposal or recovery of the valuable components of the soluper liter, the effective capacity of the resin was increased, and tion in a usable form, and was abandoned. about 75% of the dissolved metals were removed. At 100 grams of chromium trioxide per liter, almost quantitative removal of the QUALITATIVE TESTS metals was observed, and the ultimate capacity of the resin was about 12.5 grams cupric oxide and 14.3 grams ferric oxide per A qualitative study of the use of cation exchange resins indiliter of resin. COPPERSTRIPPING BATHS. A series of tests was run on a copper cated that the concentration of chromic acid must be limited. If stripping bath containing 200 grams of chromium trioxide and the pH of the solution is less than 0.1 (equivalent to 100 grams of 50 grams of cupric oxide per liter. Removal of copper from the chromium trioxide per liter if no other acid is present), the p H solution was good but the resin was attacked by the chromic equilibrium of the resin is shifted toward the regenerative conacid. The test soiution was diluted to 100 grams of chromium trioxide and 25 grams of cupric oxide per liter, a t which concentradition, in which case hydrogen ions tend to displace metal ions tion the resin retained almost all of the copper with negligible held by the resin, thus decreasing its effective metal capacity. I n attack on the resin; at the break-through point, indicated by a solutions containing more than 150 grams of chromium trioxide rather sharp increase in metal content of the effluent, the column per liter, there was chemical attack on the resin, as indicated by showed a capacity equivalent to 36 grams of cupric oxide per liter of resin. On testing a copper stripping bath containing 100 the presence of traces of trivalent chromium in treated solutions, grams of chromium trioxide and 2.5 grams of cupric oxide per the untreated material having contained no trivalent chromium. liter, it was found that, although removal of copper was good up At a chromic acid concentration of 100 grams per liter, it was to the break-through, the capacity was about 14 grams of cu ric found that cupric, aluminum, ferric, and chromic ions were oxide per liter of resin; the capacity of the resin was reducefby the lower pH of the influent solution, as decreasing p H tends to readily absorbed by the resin. approach the regenerative equilibrium of the resin, with hydrogen A t this point in the work, a column was set up for more comions replacing metal ions in the resin. plete evaluation of the resin. ANODIZING SOLUTIONS.Qualitative tests on anodizing solutions were begun with a solution containing 106 grams of chroThe resin bed was so arranged that either untreated solution or mium trioxide and approximately 17 grams of alumina per liter (pH regenerant acid could be admitted a t the to , and water run in a t 1.04). Although a small amount of alumina appeared in the the top or bottom, so that the column coufd be used, rinsed, reeffluent throughout the run, the amount removed was satisfactory; generated, rinsed, and backwashed in service. The column, about 2.6 liters of solution were treated per liter of resin before which was 1.375 inches in inside diameter and 48 inches high, had the column was exhausted (influent and effluent the same). An a glass wool plug a t the bottom, covered by a 2-inch layer of white anodic bath containing 51 grams of chromium trioxide and about sand, and a 29-inch bed of resin (700 ml.), as backwashed and 8 grams of alumina per liter (pH 1.44) showed complete redrained. A flow rate of 100 ml. per minute was used for service, moval of alumina during most of the run, with a rather sharp rinse, and regeneration. The solutions tested included chrome break-through of alumina in the effluent; the volume of solution plating baths, anodizing baths, and copper stripping baths, as these treated to exhaust the resin was 3.2 liters per liter of resin. A represent major sources of chromic acid waste solutions which bath containing 84 grams of chromium trioxide and about 10.5 present a disposal problem. After each test run, the column was grams of alumina per liter (pH 0.60) exhibited a volume capacity rinsed, regenerated (using 1 gallon of 7% sulfuric acid), rinsed, of about 2.9 liters per liter of resin. In the first and third runs backwashed, and rinsed, in accordance. with the resin manufacdescribed above, the influents contained 0.2 gram of chromic turer’s instructions. oxide per liter, and in the second, none; in all these tests, a trace The results of these qualitative tests are summarized in Table I of chromic ion was found in the effluent, the amount increasing as the run progressed, t o a maximum of about 0.5 gram of chromic and discussed below. oxide per liter. One additional run, using a solution of 150 grams CHROME PLATING BATHS. The first set of qualitative tests was of chromium trioxide and 15 grams of alumina per liter showed run on a simulated chrome plating bath, containing 400 grams of good alumina removal, and, as in the other tests, a trace of chromium trioxide, 25.0 grams of cupric oxide, and 28.6 grams of chromic ion in the effluent.
.
b
*
309
INDUSTRIAL AND ENGINEERING CHEMISTRY
310
Vol. 42, No. 2
QUAXTITATI\. E TESTS, USING ANODIC BATH
ge
In order to present more complete data as to the service life and operational characteristics of a cation exchange resin in chromic acid service, a series of 100 test runs was made, regenerating spent anodic solution. This was chosen as a typical servicing test, as the anodizing industry represents one of the largest single fields in which chromic acid waste presents a disposal problem. The spent solution used was obtained from a large anodizing plant in Maryland.
d
E6 9 i 2
L
4
,
8 2
3
2
0
4
12
16
20
24
28
32
56
A test column was set up for these runs, calculated to treat 1 gallon of anodic bath and approximately exhaust the resin in each Figure 1. Progress of Alumina Removal from Anodic Liquor, R u n 99 cycle. The column was 1.5 inches in 1130 ml. (0.04 cubic foot) of cation exchange resin inside diameter and 60 inches high, and had a glass wool and sand base as in the smaller column. The resin bed depth was 39 inches, or 0.04 cubic foot, as MISCELLANEOUS CHROMIC ACIDSOLTTTIONS. A single test run backwashed and drained. The flow rate for service, rinse, and rewas made on a chromic acid solution contaminated only with generation was 100 ml. per minute; rinsing after regeneration was continued until the effluent was free of sulfate, as tested with trivalent chromium (150 grams of chromium trioxide and 3.8 grams of chromic oxide per liter). Although some chromic ion barium chloride (about 3 gallons of rinse water). The regenerant was retained by the resin, the major portion came through in the used in each cycle was 2 gallons of 7% sulfuric acid. During the efiluent. It is apparent from the current work, and has also been test, samples were taken of the first, fifth, tenth, and every successive fifth cycle, by collecting the treated effluent from the observed by Bliss (Z), that trivalent chromium is not readily picked up by cation exchange resins at room temperature; this is time the water in the column was displaced by the influent soluprobably due to the fact that the trivalent chromium in chromic tion until the rinse water had displaced the effluent. After the acid solutions is not all present as simple metal ions, but partly as collected effluent was mixed, a portion was withdrawn for complex chromi-chromate anions (6,6). analysis. (Using specific gravity as a direct guide to the strength I n a series of tests with spent acidified sodium bichromate sohof the solutions, the initial and final effluents were arbitrarily set tions, such as result from the production of conversion coatings on as 1.007 specific gravity; the feed solution was 1.063 specific zinc, the resin was ineffective in removing metal contaminants. gravity.) The effluent of the 99th run was collected in 100-mi. portions, and analyzed to show the progress of a typical run. When operating in the hydrogen cycle, the concentration of sodium ions was so high that no metal ot)her than sodium was removed from the solution; when operating in the sodium cycle, the The over-all average of the 100 test runs n-as as follows: high concentration of sodium ion in the solution shifted the reInfluent Effluent generative equilibrium of the resin so that no other metal ion was retained by the resin in any appreciable quantity. 0.65 0.34 V O L U M E O F EFFLUENT, 100 ML.
E%a, g. per liter AlzOa, g. per liter crzo3, g. per liter CuO, g. per liter MnO, g. per liter
EXCHANGER CAPACITY
72.6 7.74 0.9 0.19 0.006
69.4 1.52 0.3 0.04
0,001
In Table I, exchanger capacities are given in volumes of liquid treated to exhaust one volume of resin. In more common ion exAnalyses of samples from selected runs are given in Table I1 change practice, capacity data are given as gram-milliequivalents The complete data of run 99 are given in Table 111, and the of ion per liter of resin, calculated to the initial break-through. progress of alumina removal is shown diagrammatically in Figure 1. The only data in the present work which can be satisfactorily expressed in these terms are from the 99th run described below in During the course of the 100 test runs, there was no noticeable “Quantitative Tests.” To the breakthrough point in this run, the resin exhibited a capacity of 1290 me. of aluminum ion per liter of resin; to TABLE 11. ANALYSISOF FEEDAND PRODUCT SOLUT~ONS TAKEN DURIXG ~OO-CYCLE TEST the end of the run (approximat’ely CrOs, AlzOa, CrzOa, complete exhaustion) the resin showed an additional capacity of 240 me., or a total capacity of 1530 me. of aluminum ion per liter of resin. The total aluminum i n t h e influent was 2110 me. These values are lower than normally expected of this type of resin. The low pH of the solutions tested is responsible for this decreased capacity. However, the value of the solutions being treated and recovered is so much higher than is experienced in water treatment, or other more common ion exchange processes, that the process as described is attractive from an economic viewpoint.
Run
KO. 1 5 10 15 20
25 30 35 40 45 50 55 60 G5 70 75 80 85 90 95
100
Grams per Liter Influent Effluent 72.4 69.6 70.6 70.1 70.4 70.0
72.6
72.3
73.0
69.4 69.0 68.2 67.3 68.0 69.3 69.4 69.5 69.6 69.9 70.0 69.4 69.8 69.1 70.1 69.4
Grams per Liter Influent Effluent 1.20 7.80 2.40 2.39 2.20 0.33
7.63
7.61
7.90
1.02 1.66 1.44 1.49 1.40 1.22 0.51 1.57’ 2.17 1.67 1.78 1.54 2.23 1.21 2.05 0.41
Grams per Liter Influent Effluent 0.3 0.9 0.2 0.2 0.2 0.3
0.9
0.9
0.8
0.3 0.2 0.3 0.2 0.3 0.2 0.2 0.3 0.3 0.2 0.2 0.5 0.2 0.3 0.3 0.3
PH
Influent 0.67
Effluent 0.36 0.42 0.33 0.32 0.38
0.88
0.35 0.33 0.32 0.32 0.32 0.32 0.32 0.37 0.34 0.33 0.33 0.32 0.37 0.37 0.37 0.83
0.65
0.62
February 1950
INDUSTRIAL AND ENGINEERING CHEMISTRY
TABLE111. CONCENTRATION O F CONTAMINANTS DURlNG RUN99, PROCESS~NG 1 GALLONOF SPENTANODICSOLUTION CrOa, G./L.
AlzOs,
G./L.
CrzOa, G./L.
73 0
Influent 7.90 Effluent
0.8
Effluent, MI.
. *
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 2800 2900 3000 3100 3200 3300 3400 3500 3600 3700 3800 3900
34.1 67.9 69.1 70.7 71.3 71.5 71.9 71.9 71.9 71.9 72.0 72.1 72.1 72.1 72.0 71.9 72.0 72.1 72.1 72.1 3 72.1 72.1 72.1 72.1 72.2 72.3 72.3 72.3 72.2 72.1 72.2 72.3 72.3 72.3 72.3 72.3 72.1 47.4 12.8
, .
0
..
.. 0.2 .. ..
0
.. , .
.. .. .. 0' 0 0 0
n
0 0 0 1.50
.. 0.2
..
d.i
..
6.i
..
0.2
.. ..
0 '5 0.5
..
4172
....
0: 25
..
7:2l 7:34
..
..
O.b ..
0'. 8
..
I'.i
..
1:1
.. ..
PH 0.62 0.71 0.42 0.36 0.32 0.3i 0.30 0.29 0.29 0.28 0.28 0,28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.28 0.29 0.30 0.33 0.35 0.38 0.41 0.45 0.48 0.50 0.52 0.55 0.57 0.59 0.61 0.62 0.73 1.07
deterioration of the resin, either in appearance or chemical behavior. DISCUSSION
As can be Seen by inspection Of the data from run 99, much higher degree Of purification can be Obtained by treating less solution per cycle. I n anodizing, where a moderate amount of alumina in solution is not harmful, it may be more economical to operate the resin to 6xhaustion in each cycle. A convenient check on the exhaustion of the column may be obtained by p H determinations on the influent and effluent, as p H is directly affected by alumina content in anodic liquors (6, IO); in other liquors, the replacement of metal ions by hydrogen ions will also reduce the pH. When the p H of the effluent and influent are the same, the resin capacity is exhausted. In operation of the resin as in the test runs, there is a loss of approximately 4% of the chromic acid to rinse water, when the influent and effiuent volumes are the same, corresponding to 4% dilution. This loss may be reduced where the process can stand further dilution by decreasing the concentration requirements in determining the beginning and end of effluent flow and swjtbhing to rinse water. Considering the operation of the resin column in treating anodic waste solution as typical, and considering the qualitative evaluation of the resin as described above, a similar approach may be used in the regeneration of copper stripping and chrome plating baths, as long as the concentration of chromic acid is in the range of 100 grams per liter or less; where it is higher, dilution is necessary to increase the effective capacity of the resin and minimize chemical attack. In many processes where water addition is necessary to make up for loss by evaporation, diluting water may be added to that portion of the solution withdrawn for treatment with the resin, thus accomplishing two purposes at once. As a guide to the approximate material cost for the operation
311
of an ion exchange column for the regeneration of a chromic acid solution, the following figures are given, based on the spent anodic solution used in the 100 test runs, and an installation to treat 250 gallons per cycle. Resin required 10 cubic feet at $26.80 per cu. ft. $268.00n Throughput per cycle 250 gal. of anodic solution a t 0.606 lb. of CrOs ger gal. 151 lb. CrOi 500 gal. of 7 % H?SO4 equivalent t o 330 lb. of 66 BB. HzSOdb Assume 4% dilution represents 4% loss of CrOa Material cost per cycle (excluding resin) $6.60C 330 lb. of 66" BB. HzSOd a t $0.02 350 lb. of limestone, to neutraliee acid for disposal, a t $3.50 per ton 0.61
-_
$7.21 Based on 100 cycles, the resin cost is $2.68 per cycle: the total material cost is therefore 59.89 per cycle. Value of recovered chromic acid figured at 8O%d of current price is (151 6)(0.80)($0.26) = $30.10 per cycle,
-
Based on April 1949 quotation on Amberlite IR-120 b Regenerant acid requirements given here are high. Using 50% less acid, about 90% of the capacity of the resin will be regenerated (from resin manufacturer's data). C Based on carboy quotation, 1.c.I. d Based on 80% removal of alumina. a
I n all probability, the useful service life of the resin would be far greater than 100 cycles, thus decreasing the cost of the resin proportionately. The use of limestone for the neutralization of waste acid has been discussed by Reidl (8). The regenerant acid efficiency may be increased by recycling the last portion of the acid from the preceding regeneration before using fresh acid. By recycling part of the acid, the concentration of dissolved metals in the acid may be increased to the extent that recovery of the metals may be feasible. I n the case of copper stripping baths, copper may be recovered electrolytically from the regenerant acid, thus making all the acid suitable for further use. Such subsequent processes would decrease the acid disposal problem and improve the over-all efficiency of the treating process. CONCLUSION
Chromic acid solutions that have been rendered useless by the accumulation of dissolved metals, such as alumina, copper, iron, etc., may be regenerated by treatment with a suitable oxidationresistant exchange resin, This treatment may be done periodically, regenerating the complete contentsof the tank needing treatment, or a small portion of the bath may be treated daily so as to maintain a nearly constant, predetermined concentration of all components, thus minimizing the size and cost of the necessary treating equipment, ACKNOWLEDGMENT
The author wishes to express his appreciation to W. H. Hartford, R. L. Copson, Marc Darrin, and George E. Best of the iMutual Chemical Company research staff for their assistance and suggestions. Thanks are due to Attilio Concordia, of Mutual's Analytical Laboratory, for his help in the analysis of all samples. LITERATURE CITED (1) Anable, A., and Kite, R. P., Chem. Eng. Progress, 44, 3 (1948).
(2) Bliss, H., Ibid., 44, 887 (1948). (3) Dodge, B. F., and Reams, D. C., Proc. Am. Electroplaters Soc., 1947,249-69. (4) Grindley, J., J . SOC.Chem. Ind., 64, 339 (1945). (5) Hartford, W. H., IND. ENG.CHEM.,34, 920 (1942). (6) Kasper, C., Bur. Standards J . Research, 9, 353 (1932); Research Paper 476. (7) Kunin, R., IND. ENG.CHEM.,41, 55 (1949). (8) Reidl, A. L., Chem. Eng., 54, No. 7, 100 (1947). (9) Sussman, S., Nachod, F. C., and Wood, W., IND.ENG.CHEM., 37,618 (1945). (10) Tarr, 0. F., Darrin, Marc, and Tubbs, L. G., Ibid., 33, 1575 (1941). RECFJVED August 20, 1949. Publication 65 of Research Laboratories, Mutual Chemical Company of America,
