CHROMIC ACID ANODIC BATHS Interpretation of Glass Electrode Measurements WIivSLOW H. HARTFORD Mutual Chemical Company of America, Baltimore, Md. asymmetric valence type of the solutions. Determination of the hydrolysis constant of the perchlorates has been employed as a check on the measurements. Of the compounds studied, chromic dichromate has been shown to possess exceptionally high hydrogen-ion activity in solution, which further increases with the passage of time. This serves to confirm the existence of the complex structure previously ascribed I O the compound. A similar complex structure exists to a lesser extent in the case of aluminum dichromate.
pH measurements of anodic baths and similar solutions indicate unusually high hydrogen-iod activity, which is further increased abnormally as the concentration of trivalent ion or of chromic acid is increased. This effect is of importance in the proper control of anodic baths. Measurements of the trivalent perchlorates show that the effect persists in dilutions as low as 0.001 molal and is probably due, not to rearrangement of complexes with liberation of acid, but rather to activity effects arising from the high ionic strength and
S
OLUTIOSS of chromic acid find use in the anodic treatment of aluminum. When first prepared, these solutions are substantially pure chromic acid, but on use aluminum and possibly trivalent chromium appear in increasing amounts, Since aluminum and chromic dichromates are usually assumed to be the compounds formed ( I , 8), solutions having that composition were used for study, chromic acid being added to approximate the compositions encountered in practice. The results observed when these solutions are tested with a glass-electrode p H meter present peculiarities; the purpose of this paper is to describe and, in so far as possible, to interpret these peculiarities. AS the complex nature of these solutions renders examination of their properties by physicochemical calculations difficult, solutions of chroinic and aluminum perchlorate were also prepared and studied.
Apparatus and Materials The glass electrode measurements were r a d e on a Beckman p H meter. Although the pH readings given by glass-electrode p H metere with solutions containing chromic acid a t lon p H may vary considerably with different electrode assemblies ( 7 ) ,measurements made with any single instrument are consistent. Since the readings obtained were used for comparison only, no corrections were made. Measurements in all cases were made a t temperatures within 2" of 25" C., the effect of this change in temperature has been shown to be less than the precision of the measurements, 0.02 pH. The chromic acid used was technical flake, testing 99.8 per cent Cr03, 0.04 per cent SO4, and 0.017 per cent Ci-203. The behavior of this material has been shown to be identical with that of c. P. chromic acid ( 7 ) . Chromic dichromate solutions were prepared by reducing chromic acid with 30 per cent hydrogen peroxide until more than 25 per cent of
the chromium was in the trivalent condition, boiling to destroy peroxide, analyzing for Cry' and Cr +-+, and adjusting to thecomposition of chromic dichroinatebyadding chromic acid. Aluminum dichromate solutions iyere prepared by treating chromic acid with a n excess of carefully washed hydrous alumina, filtering, oxidizing electrolytically any Cr present, analyzing, and adjusting to the composition A12(Cr,O?), by addition of the calculated quantity of G O a . Perchloric acid solutions were prepared by diluting c. P. perchloric acid to approximately the strength desired and then analyzing for acid content. Chromic perchlorate solution was prepared by completely reducing chromic acid with hydrogen peroxide in the presence of the calculated quantity of perchloric acid, the solution being kept dilute to prevent possible reduction of perchloric
FIGURE 1. pH
OF
CHROMICAND PERCHLORIC ACIDS AT VARIOUSCONCENTRATIOKS,
920
25" c.
INDUSTRIAL AND ENGINEERING CHEMISTRY
August, 1942
92 1
Table I1 give.- the iewltC oi >cries of euperiiiiriits in \\ hlch hnon n amounts of chromic acid \\ere added to chromic dichromate ~ i i t laluminum d~chroinate -olutiona. The apparent ratio oi c'i' to~r--'or=il''-\\asc~lculuted as i n Table I, and the expected pH I f a s also calculated on the basis of an inciease in apparent free chromic mid equal to that added. 1 1 1 additional decrease in pH \\ as notcd I\ hirh, . althorigh ncurly independent of the chroniic acid added, in the range studied is roughly proportional to the Conceiitrtitioii of trivalent ion; this indicatcs that these ions, nith their high ionic >trength, are the primary cause of the effects noted. It is theiefoie unlikely tliat the pH is a true iiicasure of the chemically uncombined chromic acid, although pH has been found of importance in predicting the electrochemicul behavior of these solutions ( 1 7 ) . Similarly the apparent ratio of Cry' to Cr-TT and -4l--', while mathematically a function of the composition of the solution, is determined within experimental error through measurement of pH. This relation, shown graphically in Figure 2, is apparent from examination of Tables I and 11. il
CHROMIC ACID ANODICBATHS acid. The solution was then boiled to destroy excess peroxide and analyzed for Cr+++ and perchloric acid. No chloride was present in the finished solution. Aluminum perchlorate solution was prepared by treating hydrous aluminum oxide with the calculated quantity of perchloric acid. The composition of the solution was checked by analysis before use.
Chromic Acid, Chromic Dichromate, and Aluminum Dichromate
TABLE I. pH
OF
CHROMICAND ALUMINUMDICHROMATE SOLUTIONS
Concn., Mole/Liter Crz(Crz0r)a Cr 0.357 0.714 0.1785 0.357 0.0892 0.1785 0.0446 0.0892 0.0223 0.0446 0.0223 0.0112 + + +
The following pH readings for the chromic acid used in this work are quoted (7) and are shown in Figure 1: 1.00 0.10
0.50 0.40
0.20 0.79
0.10 1.09
0.05 1.39
0.02 1.80
0.01 2.08
pH readings were obtained for solutions of chromic dichromate and aluminum dichromate, as indicated in Table I. The solutions were prepared by dilution of the most concentrated solution. Concentration of the dilute solutions gave check results and indicated complete reversibility of the dilution effects. Buzzard and Wilson (1) assumed that the free chromic acid content of such solutions is determined by the pH reading. Using this assumption, an apparent ratio of combined Cr"/Cr+++ or combined Cr"'/Al+++ was calculated according to the relationship: CrVi - free Crvi from pH apparent ratio = Cr+++ or Al+++ * Table I shows that at concentrations above 0.05 M , the apparent ratio of Cry' to Cr+++ or A1+++ decreases rapidly with increase in concentration. This is equivalent to stating that the pH readings are lower than were expected in these solutions, due either to ( a ) increased hydrolysis or complex anion formation, with liberation of chromic acid, or (b) greatly increased activity of the hydrogen ion in these solutions as a consequence of the high ionic strength and complex interionic forces.
Concn., Mole/Liter AIz(CrzO7)a Al+++ 1.364 0.682 0.341 0.682 0.1705 0.341 O.OS53 0.1705 0.0426 O.OS53 0.0213 0.0426 0.0107 0.0213 0
Cr,i 2.142 1.071 0.536 0.268 0.134 0.0668
Crvi
4.092 2.046 1.023 0.512 0.256
0.128 0.064
pH 0.23 0.68 1.04 1.39 1.67 1,95
Apparent Ratio Combded Crvi/Cr 1.9: 2.27 2.37 2.45 2.42 2.36
pH 0.18 0.92 1.44 1.82 2.14a 2.43Q 2.724
Apparent Ratio Combined Crvi/Al+ 2.40 2.78 2.87 2.90 2.90 2.90 2.90
+ + +
+ +
Values for free Crvi estimated by extrapolation.
pH measurements were also made on a large number of solutions of similar composition, prepared by various methods. Trivalent chromium was introduced by means of thermal decomposition, electrolytic reduction, and reduction with hydrogen peroxide, methyl alcohol, and oxalic acid; aluminum was introduced through dissolving pure aluminum powder in chromic acid, followed by electrolytic oxidation, and through the anodic oxidation of aluminum plates. I n all cases exceptionally high hydrogen-ion activity was noted, and the relation shown in Figure 2 was found to hold closely. Solutions were allowed to stand at room temperature until
INDUSTRIAL AND ENGINEERING CHEMISTRY
922
tions, n e v e r t h el e s s the complex nature of the d i c h r o m a t e s of t r i v a l e n t metals does not rule out the possibility that hydrogen ion is liberated through a rearrangement of complexes in the more concentrated solutions. Accordingly, measurements were made in an analogous manner with p e r c h l o r i c a c i d , chromic perchlorate, and a l u m i n u m perchlorate. Perchloric acid was chosen since it c a n n o t f o r m anionic complexes of itself, and since its aluminum and chromic salts yield aluminum ion and the simple violet chromic ion a t room t e m p e r a t u r e (15). Perchloric acid also lends itself better to study of dilute solutions, since insoluble basic chromates precipitate from the dichromate solutions a t pH values above 3.
EQUIPMENT FOR CHROMIC ACID ANODICTREATMENT IN A LARQE FACTORY
a constant pH value was obtained. This usually required from 6 t o 24 hours. A general tendency was noted for the pH to decrease on standing, indicating a gradual rearrangement of the structure of the solute. This effect was particularly pronounced with chromic dichromate solutions. Although the results obtained indicate that the effects described result from the high ionic strength of these soluACID TO DICHROMATE TABLE 11. EFFECTOF ADDINQCHROMIC SOLUTIONS Concn., Mole/Liter Added Total Cr2(CrzOi)a CrOa C r + + + Crvi
0.1785 0.1785 0.1785 0.1785 O.OS92 0,0892
0 0.100 0.200 0.400
0,0892 ‘0.0892
0.200 0.400
0.0446 0.0446 0.0446 0.0446
0
0.100 0 0,100 0.200 0.400
0.357 0.357 0.357 0.357 0.1785 0.1785 0.1785 0.1785 0.0892 0.0892 0.0892 0.0892
Concn., Mole/Liter Added AI+++ Alz(Crz0r)s CrOs
a
1.071 1.171 1.271 1.471 0.636 0.636 0.736 0.936 0.268 0.368. 0.468 0.668
Total Crvi
Interpolated from data in Table I.
Apparent Ratio Combinled DeCrvi/ Calcd. crease C r + + + 2.27 O:k4 0166 2.14 0.43 0.12 1.87 0.28 0.12 1.71 .0.06 2.37 2.19 Of76 0.60 0.08 2.06 0.39 0.08 1.85 2.45 0:92 0:64 2.32 0.69 0.04 2.16 0.44 0.04 1.98
P H Values Obsvd.
0.68 0.48 0.31 0.16 1.04 0.70 0.52 0.31 1.39 0.88 0.65 0.40
..
Apparent Ratio P H Values Combi\as observed with the peichlorate solutions, confirming the original assumption that they would be free from complex formation. Since anodic oxidation is a result of the potential of oxygen discharge a t the anode, which in turn is a function of hydrogen-ion activity, p H would be expected to be an important variable in the control of anodic baths. This has been shown to be the case ( I , 17). However, since baths high in chromic acid and aluminum content have been shown to possess proportionately greater hydrogen-ion activity, consideration must be given to this fact in determining the most economical operation of the anodic bath. The laboratory work of Tarr, Darrin, and Tubbs (17 ) has considered this point thoroughly, and the work has been confirmed by service records of anodic baths in commercial operations (18).
Acknowledgment The author wishes to express his appreciation to 0. F. Tarr and Marc Darrin, of the Mutual Chemical Company staff,
Vol. 34, No. 8
for their assistance and tuggestions. Many thanks are due also to Kalter C. Schumb of Massachusetts Institute of Technology for suggestions during the early part of the work, and especially to Charles Kasper of the National Bureau of Standards for his help with the manuscript and valuable suggestions.
Literature Cited Buzzard, R. W., and Wilson, J. H., J . Research Natl. Bur. Standaids, 18, 53 (1937); Research Paper 961. Chaudhury, P. C. R., J . Indian Chem. Soc., 16, 652 (1939). Ibid., 18, 102 (1941). Cupr, V., Collection Czrchoslou. Chem. Commun., 1, 467 (1929). Ewing, D. T., Hardesty, J. O., and Kao, T. H., Mich.Eng. Expt Sta., Bull. 18 (1928). Haring, H. E., Chem. 6c Met. E ~ Q .32, , 692 (1925). SSAL. ED.,14,174 (1942). Haitford, W. H., IND.ENC.CHE~LI., Kasper, C., Bur. Standards J . Research, 9,353 (1932) : Xeseai ch Paper 476. King, A., “Inorganic Preparations”, p. 104, New York, D. Van Nostrand Co., 1936. Larnh, A. B., and Fonda, G. R., J . Am. Chem. SOC.,4 3 , 1154 (1921). Lamb, A. B., and Jacques, A . G., I b X , 60, 1215 (1938). La Mer, V. K., and Mason, C. F., Ibid., 49,410 (1927) Mellor, J. W., “Comprehensive Treatise on Inorganic and Theoretical Chemistry”, Vol. 11, p. 385, New Yoik, Longmans. Green and Co., 1922. Ollard, E. A., J . Electroplaters’ Depositors’ Tech. Sac., 3 , 5 11929). Ro~th,R., “Zur Kenntnis der Perchlorate”, Mdnchen, 1910. Scatchard. G.. J . A m . Chem. Soc.. 47. 696 (1925). (17) Tarr, 0. F., Darrin, M., and Tubbs, L. G:, IN;),ENG.CHEM., 33, 1575 (1941). (18) Tubbs, L. G , and Hartford JV. H., private communications.
Liquid Sulfur Dioxide as Solvent Medium for Chemical Reactions IQUID sulfur dioxide JOHN ROSS, J. H. PERCY, R. L. BRAiXDT1, readily available a noncorhas long been used inrosive, nonflammable, readily A. I. GEBHART, J. E. MITCHELL, AND dustrially for solvent distilled, and virtually anhySEYMOUR YOLLES extraction of aromatics from drous liquid polar solvent Colgate-Palmolive-Peet Company, Jersey City, N. J. saturated hydrocarbons (4)in which can be handled in open the petroleum industry. In flasks in a good hood- or 1922 Grob and Adams (7) draught closet, or in closed showed that a solution of sulfur trioxide in liquid sulfur divessels under a pressure of a few pounds. oxide could be used for sulfonating aromatic hydrocarbons. We describe below a few types of simple chemical reactions Burkhardt and Lapmorth ( 1 ) employed this solvent in the in which we have used liquid sulfur dioside as a solvent resulfation of phenols with sulfur trioxide, and later Daimler action medium. Other chemical reactions and modifications and Platz (3) used sulfur trioxide in liquid sulfur dioxide of them will occur to chemists when the simplicity of handling for preparing carbyl sulfates from alcohols and olefins. Jander this solvent is more generally known. and his collaborators (8) examined the solubility, elecTypes of Reactions trolytic conductance, and dissociation of various materials in liquid sulfur dioxide. Although liquid sulfur dioxide has FRIEDEL-CRAFTS.Although there are several materials long been known as an excellent solvent for a wide variety of that can be used as diluents or as solvent media in which t o chemical materials, except for sulfonation very little has perform the Friedel-Crafts type of chemical reactions, there been reported as to its advantageous use as a solvent in which has been a need for a convenient, and readily accessible solvent to carry out other organic reactions. for use at low temperatures. Sparks and Field (16) describe Sulfur dioxide as supplied commercially contains 0.1 to the use of low-boiling alkyl halides such as ethyl chloride as 0.005 per cent water. Without any unusual or elaborate solvents for the aluminum chloride. However, the solubilit’y precautions, refrigeration-grade liquid sulfur dioxide can be of aluminum chloride in ethyl chloride is less than 4.4 per tapped from an inverted cylinder into a Dewar flask reaction cent below 0 C. vessel to give a liquid containing 0.01 per cent water as Liquid sulfur dioxide is an excellent solvent medium for the measured by the Fischer reagent (6, 14). We have thus Friedel-Crafts type of reaction, since it has the virtues of 1 Present address, Coffee Products Corporation, N e w York, N. Y. nonflammability, low reactivity, and high solvent properties
L
