A Study of Oxidations Using Copper(ll1) Reagents DONALD A. KEYWORTH' and K. G. STONE K e d z i e Chemical Laboratory, M i c h i g a n State College, f a s t Lansing, M i c h .
The determination of organic compounds on a micro scale by oxidation with potassium diperiodato cuprate (111) and potassium ditellurato cuprate(II1) in alkaline media was reported by Beck. Reagent solutions are easy to prepare and appear to be stable but no good method could be found for end-point detection. Data are presented to show that periodate and tellurate ions participate in the oxidation of many compounds, but reproducible results are difficult to obtain. The oxidation of cyanide ion and thiosulfate ion is almost quantitative. Sew methods employing copper(II1) reagents were studied to demonstrate the usefulness of the application.
potassium hydroxide dissolved in the smallest amount of water possible. At this point the mixture will be deep green (periodate) or will have a dark green precipitate (tellurate). Add a total of 60 grams of solid potassium persulfate in small portions at 1-minute intervals. Boil for 15 to 20 minutes to decompose the excess potassium persulfate, cool to room temperature, dilute to 1 liter, and store in a polyethylene bottle. T h e final solution is deep brown in color and deposits no precipitate on standing. The brown solution on evaporation to dryness yields a brown residue which may be powdered and stored. Re-solution of the residue in distilled water gives a solution which has all the characteristics of the unevaporated solution. END POINT DETECTION IN TITRATIONS
Because the copper(II1) solutions are so intensely colored, a self-indicating reagent appeared t o be available. Beck had reported this to be the case in microtitrations (3,4), although he did state that in some titrations colored precipitates appeared. I n titrations with 0.05M potassium diperiodato cuprate(III), blue or green solutions and greenish yellow precipitates were almost alxays found before the solution turned dirty brown, showing a large excess of reagent. .4 qualitative study was therefore made on the behavior of copper(I1) periodate solutions. When potassium periodate is added t o a solution of copper(I1) sulfate, a yellow-green solid is formed. As a solution of potassium hydroxide is added to the solid, the solid slowly dissolves yielding a deep green solution, and as the amount of potassium hydroxide increases, the solution finally becomes dark blue, Addition of water to the dark blue solution reverses the effects showing that the hydroxide ion concentration is the determining factor. This is not surprising in view of the ionization constants for periodic acid ( 5 , 6). I n titrations with 0.05M potassium ditellurato cuprate(II1) a greenish yellow precipitate is almost always present before an excess of reagent is present. Qualitatively, a greenish yellow precipitate is formed when a solution of telluric acid is added to a solution of copper(I1) sulfate, and this precipitate is soluble only when the concentration of potassium hydroxide is several times greater than that normally present in a titration. Because intensely colored solutions and precipitates are found, the end point is very difficult to detect visually a t ordinary concentrations in spite of the intensely colored reagent which otherwise should be self-indicating. -4series of titrations were then tried in an attempt to use a potentiometric end point. Since the solutions used contained 0.2 to 0.5.11 potassium hydroxide, the possible electrodes are limited, and measurements were finally made with platinumcalomel electrodes. I n titrations of sodium tartrate and potassium ferrocyanide, the observed potential graduallj increased up to the point \\here a visual excess of copper(II1) reagent was present. S o sharp increase was observed in the vicinity of the equivalence point. Therefore, attempts to use a potentiometric end point were abandoned. Just beyond the equivalence point in titrations with copper(II1) reagents both copper(I1) and copper(II1) would he present. Thus with both parts of an electrochemical couple present the dead-stop end point method should be applicable (21). A sharp end point x a s found v,-ith 100 to 200 mv. applied between platinum electrodes using both diperiodato cuprate(II1) and ditellurato cuprate(II1) for the oxidation of sodium thiosulfate in sodium bicarbonate solution and potassium cyanide, potassium ferrocyanide, and potassium arsenite in 0.5M potassium hydroxide. .4s a result, the dead-stop technique was used for all titrations of substances which had reasonable reaction rates in preliminary titrations.
T
HE use of potassium diperiodato cuprate(II1) and potassium
ditellurato cuprate(II1) as volumetric reagents has been described b y Beck on a micro scale (3,4)and especially for biochemical applications (1, 2). The results suggest that an evaluation of the reagents at macro concentration levels should be made. Compounds containing copper(II1) have been known since 1844, and the early work has been reviewed by Urtiss ( 1 2 ) . .\lore recently the presence of complex ions in solutions containing copper(II1j, periodate or tellurate ions, and hydroxide ions has been recognized b y the isolation of compounds (8, 9 ) , by the elimination of the possibility of a peroxidic species ( 7 ) ,and by the magnetic susceptibility characteristic of the odd-electron copper (111) state (9). The copper(II1) complex ions are stable in potasbium and sodium hydroxide solutions, but unstable in ammonium, lithium and tetramethylammonium hydroxide solutions (12). Lister ( 7 )reported that the use of stannate, stibnate, and selenate ions did not yield stable copper(II1) complex ions. Beck ( 3 , 4) found that titrations using copper(II1) are best carried out in solutions containing potassium hydroxide. The work which is described here includes the preparation of solutions, methods for detecting the end point in titrations, and studies of the oxidation of some selected compounds. Unless otherwise indicated, all calculationc are based on the assumption that the copper(II1)-copprr(I1) system is the only active redox couple PREPARATIO\ OF COPPER(II1) SOLUTIONS
Copper(I1) has been oxidized to copper(II1j in alkaline solution by electrolysis (12), nith hypochlorite ( 7 ) , and with potassium persulfate (3, 4, 8, 9, I d ) , the latter in the presence of periodate or tellurate ions. All solutions in this work were made by the persulfate method. Attempts to prepare stable copper(II1) complexes with phosphate, perchlorate, arsenate, chromate, plumbate, molybdate, and tungstate ions were unsuccessful. I n an attempt to reduce the cost of the periodate reagent, potassium iodate mas used as a starting material in place of potassium periodate. T h e increased amount of potassium persulfate required was found to give an unstable solution which deposited potassium sulfate slowly for over 2 weeks in both the storage bottle and the buret. When t h r alkaline copper(II1) solution is stored in glass bottles, the bottle is badly attacked, and, as the alkali is consumed, precipitates containing copper are formed. If the solution is stored in polyethjlene bottles, no apparent decomposition occurs. The fol1olT ing procedure was found to yield stable solutions containing about 0.053f diperiodato cuprate(II1) or ditellurato cuprate(II1) complexes. T o !IO0 ml. of boiling distilled water in a 2-liter beaker equipped with a mechanical stirrer, add 12.5 grams of copper( 11) sulfate pentahydrate. When dissolved, add 57.5 gmmq of potassium periodate or 44 grams of telluric acid (HZTeOd 2H20). Carefully add a solution of 67.5 grams of 1
Presrnt address, U. 8. Army
a33
834
ANALYTICAL CHEMISTRY OXIDATIONS OF INORGANIC SUBSTANCES
The general procedure used in the investigation with inorganic substances was to pipet volumes of solutions of known concentration into beakers and to add concentrated potassium hydroxide solution until the mixture was 0 . 5 X in potassium hydroxide. Preliminary titrations were made with copper(II1) reagent solutions using visual end points in order to observe the formation of precipitates and to check the stoichiometry. For cases which appeared to be favorable, further titrations were made using the dead-stop end point method. The oxidation of potassium iodide, potassium iodate, and sodium bisulfite was very slow and not suitable for volumetric use. The oxidation of potassium ferrocyanide by diperiodato suprate(II1) was reasonably rapid. On visual observation duplicate samples required 33 and 35 ml. of reagent solution. The dead-stop method gave a sharp break a t the end point, but the reproducibility of duplicate titrations was not improved. Calculation of the number of equivalents of oxidizing agent consumed per mole of ferrocyanide gave values in the range 4 to 4.5. These values do not correspond to any known stoichiometry. Because the precision was poor and the oxidation products were unknown, no further work was done with this system. Since Beck had reported ( 3 )the oxidation of arsenite to arsenate by diperiodato cuprate(III), this system was investigated further. Aliquots of arsenic( 111) solution in 0.5M potassium hydroxide solution which should have consumed 20 ml. of diperiodato cuprate(II1) solution were found to require only 3 to 8 ml. using the visual end point. I n most cases blue or green precipitates or colored solutions interfered badly. The reverse titration-Le., arsenic(II1) added to copper( 111)-did not improve the results. The use of the dead-stop method improved the results somewhat, 16.7 to 19 ml. being required compared to 25 ml. expected. Essentially the same results were found using ditellurato cuprate (111)solution with the exception that the precision was somewhat better. When some of the supernatant liquid from one of the diperiodato cuprate(II1) oxidations was tested with nitric acid and silver nitrate solution, an appreciable amount of iodate was found. It must be concluded that not only the copper(II1) but also the periodate is oxidizing the arsenic(II1). Therefore quantitative results would not be expected. Beck ( 3 ) reported that potassium cyanide was oxidized t o potassium carbonate and nitrate by copper(II1) reagents. With 0.05M diperiodato cuprate(II1) and ditellurato cuprate(II1) the oxidation of cyanide in 0.5M potassium hydroxide was moderately rapid, but the visual detection of the end point \vas difficult since precipitates and intense colors were present. The dead-stop end point gave a sharp break with either reagent. The ditellurato cuprate(II1) reagent gave better reproducibility and was therefore used for the rest of the work. It was found that 50-ml. aliquots of 0.01M cyanide solution required 21.8 & 0.2 ml. of approximately 0.05M copper(II1) solution. These results suggest that two equivalents of oxidizing agent are required per mole of cyanide ion and that the product might be cyanate ion:
KCN
+ 0 +KCNO
Cyanate ion is known t o hydrolyze in potassium hydroxide solution to yield ammonia and carbonate. The oxidation was therefore run in a Kirk microdiffusion cell with liessler's reagent in the diffusion cup. A strong test for ammonia n a s found and the test carried out in the same way without cyanide ion was negative. The evidence found is all in agreement with the consumption of two equivalents of oxidizing agent per mole of cyanide ion. Beck (3)also reported the oxidation of thiosulfate ion. I n this work the oxidation of thiosulfate ion by both diperiodato cuprate (111) and ditellurato cuprate(II1) was very slow in 0.561 potassium hydroxide. I n saturated sodium bicarbonate solution, thiosulfate ion is oxidized erratically by the periodate reagent because periodate also is an oxidizing agent. I n the same medium however, the tellurate reagent oxidizes thiosulfate ion with the
Table I. Oxidation of Potassium Cyanide and Sodium Thiosulfate by KKu(TeO& KCN, .M Trial 1
Trial 2
0 1221
0.01221
Na&Oa .bf 0 1377
0.01377
L'alcd 0 0489
O.04R5
Cu(II1). S L From Erom KCh Nad3~0~ 0 0626 0 0484 0 OF19 0 0496 0 0617
0 0482 0 0483
0.0408 0,0408 0.0382
0.0357 0,0339
consumption of approximatel). 1 equivalent per mole and the end point is best detected with the dead-stop method. The re0.1 ml. when 25 1111. were producibility of the titration was being used. Since the preliminary experiments with potaseiuni cyanide and sodium thiosulfate seemed promising, a cyclic experiment was designed to completely check the titrations of these two substances. A sample of copper(I1) dulfate pentahydrate was analyzed iodometrically and found to be 99.8% pure on the basis of the hydrate. A4solution of ditellurato cuprate(II1) was prepared from a known weight of the copper(I1) sulfate using an excess of oxidizing agent t,o be certain that all the copper was in the + 3 state. A solution of potassium cyanide ivas standardized with silver nitrate, and a solution of sodium thiosulfate was standardized iodometrically against potassium iodate. Aliquots of the thiosulfate solution were titrated in saturated sodium bicarbonate solution to a dead-stop end point. Aliquots of the cyanide solution were titrated in 0.5M potassium hydroxide solution to a dead-stop end point. In all cases the total volume of the solution titrated was 100 nil. and the volume of copper(II1) reagent varied from 15 to 40 nil. The normality (equal to the molarity) of the copper(II1) solution was calculated from these titrations assuming 2 equivalents per mole for the cyanide and 1 equivalent per mole for the thiosulfate. The results are given in Table I. These results indicate that the concentration plays an important role, particularly with the cyanide xhere, a t higher concentration, the oxididation is less than 2 equivalents per mole and a t lower concentrations more than 2 equivalents per mole are required. It must be concluded that these oxidations are not satisfactory for general use.
*
OXIDATIONS OF ORGANIC SUBSTANCES
I n an attempt to apply copper(II1) oxidations to organic analysis, experiments were tried with cinnamic acid, malonic acid, acetone, and ethyl alcohol. .In all cases the rate of oxidation by diperiodato cuprate(II1) was too slow a t room temperature to be of any value in volumetric analysis. At higher temperatures the rates were greater, but the volume of reagent solution consumed had no relation to the amount of compound added. The oxidation of glucose has been extensively used l)y Beck ( 2 , 4) for the estimation of blood sugar levels. Applying the method to the titration of 0.01.U glucose in 0.5.1' potassium hydroxide with 0.0511f diperiodato cuprate(II1) was found to be highly colored impractical. The rate of oxidation was very s10iv7., solutions and precipitates n ~ r eformed, and the process was not stoichiometric. If some of the supernatant solution which should contain only periodate was added to dilute nitric acid, a white precipitate was formed on the addition of silver nitrate showing the presence of iodate. Therefore, there are two competing oxidations, one by coppcr(II1) and the other by periodate. No combination of temperature, alkalinity, and glucose concentration could be found which gave reproducible, stoichiometric results. The oxidation of glucose by ditellurato cuprate(II1) gave the same limitations as found using the diperiodato cliprate (111). Tellurite could be detected in the presence of tellui,nte in 6 N hydrochloric acid hy rrciuction to tellurium metal with sulfurous acid formed on adding sodium bisulfite. Thc, pre-
V O L U M E 27, NO. 5, M A Y 1 9 5 5
835
cipitate formed i n oxidations of glucose with ditellurato cuprate (111) contained tellurite. Therefore, again competing oxidations were present, which indicated that further work with glucose was futile. The oxidation of tartrate ion starting with disodium tartrate dihydrate b y either diperiodnto cuprate(II1) or ditellurato cuprate(II1) was found t o be subject to the same limitations as the oxidation of glucose. The rate of oxidation was reasonable, hut highly colored solutions and precipitates prevented the use of a visual end point. No break was found in the plot for the potentiometric titration and titrations with the dead-stop end point were not reproducible. In addition iodate was found in the mixture from diperiodato cuprate(II1) oxidation. In connection with the oxidation by the complexing anion, the 2 to 1 ratio found in the copper(II1) complex is not necessarily found in the compound which can be isolated containing copper G ) ~ been isolated (11). For esample, Cug(I06)*and H ? C U ~ ( I Ohave (10). Hence periodate is avai1:tk)le for oxidation as soon as any c.oppcr(II1) is reduced if not hefore.
product composition, and the possible competition via tlie oxidizing action of the complexing anion. A t the milligram level these difficulties may well be minimized so that empirical methods yield satisfactory results ( S , 4 ) .
CONCLUSION
REcrrrrn for reriew September 9, 1954. Accepted Xoremher 1 6 , 1954. Presented before the Division of Analytical C h e n ~ i s t r ya t t h e 126th hfeeting of the AMERICASCIIEI\IICAL SOCIETY, N e i r T o r k , N 1. Abstracted in part from the X S . Thesis submitted b y Donald -1.Ke>-\vortli, J u n e 1!354.
1pplic‘:ttionc: of copper(II1) oxidation a t the millimole level are liiiiitcd 1)v the difficulty of enti-point detection, the uncertainty of
LITERATURE CITED
(1) Beck, G., Anal. Chim. Acta, 9 , 241-7 (1953). (2) Beck, G., Mikrochemie aer Mikrochem. Acta, 35, 169-73 (1950); 38, 1-10 (1951); 39, 22-9, 147-51 (1952); 40, 258-63 (1953). (3) Ibid., 36, 245-50 (1950). (4) Ibid., 3 8 , 152-9 (1951). (5) Crouthamel, C. E., Hayes, A. AI., and Martin, D. S.,J . Am. Chem. Soc., 73, 82-7 (1951). (6) Crouthamel. C. E., IIeek, H. V.,Martin, D. S.,and Banks, C. Si.. Ibid.,71, 3031-5 (1949). (7) Lister, 11. W., Can. J . Chem., 31, 638-52 (1953). (5) Nalaprade, L., Compt. rend., 204, 979-80 (193T). (9) Nalatesta, L., Gam. chim.ital., 71, 467-74, 550 (1941). (10) Smith, G. F., “Periodic Acid and Iodic Acid,” 5th ed.. p. 5, G. F. Smith Chemical Co., Columbus, Ohio, 1950. (11) Stone, K. G., and Scholten, H. G., Ax.4~.CHEM.,24, 671-4 (1952). (12) Urtiss, hl., Rec. tray. chim., 44, 425-34 (1925).
Determination of Zirconium in Magnesium Alloys Using p-Bromo- or p-Chloromandelic Acid ROLAND A. PAPUCCI,
F. C.
Broeman and Co., Cincinnati 70, O h i o , and
JOSEPH J. KLINGENBERG, X a v i e r
University, Cincinnati 7, O h i o
Successful application of p-bromo- and p-chloromandelic acids to the determination of zirconium in steel and aluminum alloys suggests a similar applicaLion to zirconium-containing magnesium alloys. Using these reagents, a rapid and reliable procedure was developed which can be applied to all types of magnesium alloys.
B R T I M E conditions and the constant search for high temperature alloys for use in the turbojet industry have led to a more complete study of magnesium base alloys. T h e addition of small amounts of zirconium to magnesium and magnesium alloys was found to improve the operating temperatures and the grain structure without affecting the creep resistance or machinability. T o meet this and other neiT metallurgical advances, accurate nnd more rapid methods for the determination of zirconium in sinal1 concentrations are needed. The phosphate method ( I ) is subject to error in t,he low concentration range and requires excessive time especially when the zirconium content is lower than 0.25Yo, The feasibility of using p-bromo- or p-chloromandelic acid for the determination of zirconium in steels and in aluminum alloys has been demonstrated ( 2 , 3 ) . The development of a method utilizing these reagents which would combine speed with accuracy and which could be applied to all types of magnesium alloys was undertaken. Such a method might also compare favorably with the alizarin red S colorimetric method (4)which is also available for the determination of zirconium in magnesium. Zirconium occurs in commercial magnesium alloys in acidsoluble and acid-insoluble forms. The strength characteristics of the specific alloys are related to the soluble zirconium content of the alloy. The acid-insoluble zirconium is usually small in
conip:trison to the acid-soluhle content. I n some cases the determination of the total zirconium conteqt is desired: in others both the amount of acid-solulile and acid-insoluble zirconium. Consequently two procedures n ere developed. One procedure describes the determination of acid-Foluble zirconium only. The other describes the determination of the soluble and insoluble forms. Both are applicable to all types of magnesiuni :illoys. PROCEDURE
Determination of Soluble Zirconium. A sample of 0.25 to 2.0 grams (amount depending on the zirconium content) is placed in a 250-ml. beaker. An amount of hydrochloric acid (1 t o 4) corresponding to 80 ml. per gram of alloy dissolved is added The beaker is covered with a watch glass and warmed. When the reaction is complete, the contents of the beaker are cooled to room temperature and the watch glass is rinsed with small amounts of n-ater. The solution should be clear. If a residue can
Table I.
Determination of Acid-Soluble Zirconium in Magnesium Alloys
Zirconium Present,
Zirconium Found, % p-Chloromandelic acid
Sarnple
%
Phosphate
Alizarin red S
1
0.44
0.42
0.44
2
0.46
0.45
0.45 0.44
3
0.48
0.46 0.47
0.47
4
0.95
0.9.5 0.95
0.97 0.97
0.44 0.44 0.47 0.47 0.47 0.48 0.48 0.47 0.48 0.98 0.97
0.98 0.98 0.98
p-Bromomnndeiic acid 0 4.5 0 45 0 46 0.49 0 48 0.98 0.98 0 98
