Potentiometric Titration of Hydroperoxides and Peracids in Anhydrous

of Union Carbide and Carbon Corp., for the Atomic Energy Commission. Potentiometric Titration ofHydroperoxides and. Peracids in Anhydrous Ethylenediam...
1 downloads 0 Views 345KB Size
the chloride was calculated. The test results (Table I T ) are within 5% of those established by the Sational Bureau of Standards. B y the use of this procedure, 20 samples can be determined in 1 hour. Anions that precipitate silver in dilute nitric acid solutions interfere in the application of the method. Limit of Detection. T h e minimum concentration of halide in t h e solution used for luminosity measurements must approximate 0.5 y of chloride, 1 y of bromide, or 1.5 y of iodide per ml., in order for the results to be reasonably reproducible. Because the volume of the final solution may be restricted to 3 nil., the minimum amounts of chloride, bromide. and iodide on which a

uranium per ml.; no tests were made with more concentrated solutions.

reasonably accurate quantitative determination can be made by this method are, thus, approximately 1.5, 3, and 4.5 y, respectively. More dilute solutions can, of course, be concentrated by evaporation after being made alkaline, or an excess of silver nitrate may be added prior to evaporation to prevent loss of halide. The only limit to the amount of concentration that can he effected in this manner, is the viscosity of the concentrate. If the viscosity becomes such as to appreciably affect the rate of aspiration into the flame, the results mill be in error. I n the case of uranyl sulfate solutions, no reduction in luminosity due to yiscosity effects were observed for solutions containing up to 120 mg. of

LITERATURE CITED

(1) Dean, J. A,, ASAL. CHEM.27, 1224-0

119.55). (2) Dippel, W.A., Bricker, C. E., Furman, N. H., Ibid., 26, 553-9 (1954). (3) Gilbert, P. T., Jr., Ind. Labs. 3, No. 8, 41 (1952). (4) Honma, Rlinoru, ANAL. CHEJI. 27, 1656-9 (1955). . - - - - I

RECEIVED for review Kovember 21, 1955. Accepted September 5, 1956. Southeastern Regional Meeting, ACS, Columbia, S. C., November 1955. Work carried out under Contract KO. W-7405-eng-26 a t Oak Ridge National Laboratory, operated by Union Carbide Nuclear Co., a division of Union Carbide and Carbon Corp., for the Stomic Energy Commission.

Potentiometric Titration of Hydroperoxides and Peracids in Anhydrous Ethylenediamine A. J. MARTIN Polychemicals Deparfment, E. 1. du Pont de Nemours &

b Peracids, primary and secondary hydroperoxides, and hydrogen peroxide can b e titrated potentiometrically as weak acids with sodium aminoethoxide in anhydrous ethylenediamine. The titration curves can b e interpreted qualitatively and quantitatively. The potential a t half-neutralization can b e correlated with acid strength and structure of the peroxides. Mixtures of some peroxygen compounds give multibreak titration curves. Accuracy and precision are in the order of 270 relative.

Co., Inc., Wilmingfon, Del. hydroperoxide linkage and, hence, slightly acidic character. Hydrogen peroxide, while it does not strictly fall into any of these classifications, is important analytically because i t frequently appears as a n impurity in samples of organic peroxides. Its pK, value is 11.6 (2). I n 1948 MOSS,Elliott, and Hall (5) showed that very weak acids could be

aldehydes and ketones. Three of the six groups have acidic characteristics. The strongest are the peracids which exhibit ionization constants whose pKa values are approximately 3.5 units greater than those of the parent acid (2, 9). Hydroperoxides behave as very weak acids with pKa values of 11.6 to 12.8 (2, 3 ) . The peroxides of aldehydes and ketones commonly have the I

PERACETIC ACID

-

0.80

A

of organic peroxides by industry has increased very rapidly in t h e past few decades, there have been relatively few signifirant advances in the analysis of these materials. The most popular existing techniques of analysis are nonselective and are based on the reduction of the peroxide by agents such as iodide or ferrous ion (1, 8 ) . Discriminatory analyses of solutions containing one 01 more peroxides can sometimes be performed, but instruments such as the polarograph or the infrared spectrometer are generally required. Organic peroxides are usually classified into six groups according to the type of peroxide linkage (8). They are hydroperoxides, dialkyl peroxides, peroxy acids, peroxy esters, diacyl peroxides, and peroxy derivatives of LTHOUGH THE USE

I

2’

w

-I

9

I-

0.70-

W t

sa 0.60c

-\\\ ETHYL HYDROPEnOXlDE

V O L U M E TITRANT

Figure 1 . Potentiometric titration curves of weak acids in ethylenediamine 0 Potential a t half-neutralization

0 Inflection

point

VOL. 29, NO. 1 , JANUARY 1957

79

titrated potentiometrically in basic nonaqueous solution. The extension of this technique to peracids and hydroperoxides is logical and important in view of the lack of selective peroxide analyses. Several other solvents, titrants, and electrode s y s t e m have been reported for use in titrating weak acids (6, ?'), but for this work the original system was chosen-Le., antimony electrodes in ethylenediamine and sodium aminoethoxide as titrant.

E

0.551

2I

0.531

6-

I

4

I

!

I

6 e VOLUME T I T R A N T ( M L 1

I

IO

!

I

12

0

?

-a t-

Table 1. Effect of Structure and Acidity on Potential a t Half-Neutralization

Weak Acid Material Peracetic acid Phenol Hydrogen peroxide Ethyl hydroperoxide (primary) Cyclohexyl hydroperoxide (sec.) Isopropyl hydroperoxide (sec.) Cumene hydroperoxide (tert.) Butyl hydroperoxide (tert.)

W L K

4-

3 u

Potential pK, at Half- Value, Seutral- haueous ization, sdlution Volt

( 2 , 3, 4)

0.83 0.72 0.58

8.2 9.9 11.6

0.55

11.8

0.54

..

..

12.1

1-0 inflection

12,6

S O

inflection

12.8

EXPERIMENTAL

Materials. The peroxides used were obtained and purified if necessary a s follows. Hydrogen peroxide (90%) was used as received from D u Pont. Peracetic acid (Buffalo Electrocheniical Co.) was p u t into a n aqueous solution and the acetic acid impurity titrated with caustic. T h e peracetic acid was extracted into petroleum ether which was then washed and dried over calcium sulfate. Ethyl hydroperoxide was prepared by the reaction of hydrogen peroxide with diethyl sulfate. The reaction mixture n-as taken up in aqueous solution, and the excess hydrogen peroxide removed by tieatment with Dee0 enzyme system (Takamine Laboratory, Clifton, N. J.). This enzyme reacts specifically with hydrogen peroxide in a neutral buffered solution to form oxygen and water. The ethyl hydroperoxide was then extracted from the aqueous solution into petroleum ether. The ether solution, containing approximately 10% ethyl hydroperoxide by 15 eight, was washed and dried over calcium sulfate. Cyclohexyl hydroperoxide was prepared by air oxidation of cyclohexane and was purified by separation as the sodium salt, followed by regeneration and distillation. Cumene hydroperoxide was used as received from Hercules Ponder Co. tert-Butyl hydroperoxide (Lucidol Division, Sovadel-Agene Corp.) n-as 80

z

ANALYTICAL CHEMISTRY

A P P L I E D POTENTIAL (v.)

Figure 2. mixture A.

Hydrogen peroxide-ethyl hydroperoxide

Potentiometric titration curve in ethylenediamine

5. Polarographic analysis in aqueous 1 .OM sodium acetate-1 .OM acetic acid

carefully vacuum distilled to a purity greater than 99%. The ethanolamine and ethylenediamine were White Label, Eastman Kodak Co. Both were distilled twice before use, the ethylenediamine being distilled from sodium. The titrant was prepared by dissolving clean sodium in ethanolamine and diluting with ethylenediamine ( 5 ) . The solution, which was approximately 0.25N, was standardized against National Bureau of Standards' benzoic acid and also pure phenol. The titrant Kas stored in the automatic buret reservoir under nitrogen to prevent carbon dioxide contamination. Apparatus and Procedure. The titration cell consisted of a 200-ml., tall-form lipless beaker fitted with a three-hole rubber stopper. The buret

Table II.

tip, t h e indicator antimony electrode, and a pressure vent were inserted through t h e stopper. T h e reference antimony electrode was located inside the buret tip. Carbon dioxide contamination of the titrant was avoided by connecting a n Ascarite absorption tube to the top of the automatic buret. A Beckman Model H p H meter was used for potentialmeasurements. Many of the end points involved small potential changes, so the p H scale of the meter was used to advantage for making the readings rather than the millivolt scale. The apparatus was standardized every 4 hours by adjusting the potential of a half-neutralized solution of phenol in ethylenediamine to 0.72 volt. For the titration procedure 50 ml. of ethylenediamine was measured into the

Accuracy and Precision of Potentiometric Titrations of Peroxides in Ethylenediamine Av. llev.., % I -

Taken, Peroxide

mM

Recovery,

%

Hydrogen peroxide

7.12 3.56 1.78

97.0 100.9 97.1

Ethyl hydroperoxide

4.21 2.11

94.0 92.5

Cyclohesyl hydroperoxide

5.22 3.43

101.1 97.0

Peracetic acid

9.72 4.86 2.43

101.3 102.4 98 0

From

100%

From hfean

2.3

1.7

6.8

0.8

2.1

2.0

1.9

1.7

~~

Table 111.

lliytui e A B

~ ~ _ _ _ _ _ _ _ _ _ _ ~~

Accuracy of Potentiometric Titrations of Hydrogen Peroxide-Ethyl Hydroperoxide Mixtures

m V Found by Polarogi aphic Analysis H?Oz EtOOH 1 37 6.51 1 63 1 04

lipless beaker. il magnetic stirring bar was inserted and the sample was added. -4s rapidly as possible, the beaker was covered by the rubber stopper and the titration was then carried out in the usual may. Increments of 0.2 ml. were added near the inflection point. The data were plotted for the determination of the end point.

RESULTS AND CONCLUSIONS

Figuie 1 shows typical peroxide titration cuives. The phenol curve is included for comparison. Two characteristics of the curves are important-the potential a t half neutralization and the inflection point. I n aqueous potentiometric titrations the absolute strength of the acid being titrated can be measured by the p H of the solutions when half the acid has been neutralized. I n nonaqueous solutions there is no absolute scale of p H units; thus, it is difficult to calculate an absolute measure of acid strength from the titration curve. However, a n empirical relation exists for various acidic materials between the potentials a t half-neutralization and the acid strengths in the particular solvent used. Of course, experimental conditions must be reproduced precisely. Table I shows the inverse relation of the potential a t the half-neutralization point in ethylenediamine to the pKa value of the acid material measured in aqueous solution. Because the relationship is linear within experimental error, it is apparent that no unusual interactions are occurring between specific peroxides and solvent. The hydroperoxides also shorn a relationship between structure and potential a t half-neutralization. The acid strength of a peroxide when the peroxide linkage is located on a primary carbon atom is greater than when attached to a sec-

% ’ Recovery by Potentiometric Titration HzOz EtOOH 97 8 104 3 98 3 102 4

ondaiy carbon. The last two hydroperoxides listed in Table I, with the oxygen-to-oxygen linkage on a tertiary carbon atom, showed no inflections. The correlations between acid strength, structure, and potential a t half-neutralization are valuable in the interpretation of titration curves of unknon-n peroxides. The variation of acid strength with structure suggests the possibility of analysis of peroxide mixtures. If the acidities are different by a sufficient degree, such as analysis should be possible. I n the preparation of ethyl hydroperoxide, a solution containing both hydrogen peroxide and ethyl hydroperoxide was isolated. A portion of this material, when titrated potentiometrically, gave curve A of Figure 2. ilnother portion of the peroxide solution was taken up in aqueous acetic acid-sodium acetate buffer and analyzed polarographically. The polarogram is shown in curve B. The two analyses of the peroxide mixture agree within experimental error. A sample of crude 2,4-dimethylpentyl hydroperoxide also was analyzed. I n this case two inflection points were clearly visible on the titration curve, but only one wave was seen on the polarogram. The sample was probably a mixture of isomeric hydroperoxides whose acidities were different enough to allow discriminatory titration, but whose half-wave potentials were too similar to allow polarographic separation. Table I1 shows the precision and accuracy obtained in the titration of peroxides in ethylenediamine. The accuracy is based on an iodometric analysis (1) of the peroxides and is expressed as the average deviation from 1 0 0 ~ orecovery. The precision is indicated by the average deviation from the arithmetic mean. The large dis-

crepancy between these deviations for ethyl hydroperoxide can be explained by the presence of diethyl peroxide as an impurity in the hydroperoxide. Both peroxides will be detected by the iodine determination but only the hydroperoxide by the potentiometric titration. This points out that certain two-component peroxide mixtures can be analyzed for both peroxides by these two complementary analyses. Table I11 shows data from the analysis of mixtures of hydrogen peroxide and ethyl hydroperoxide. The potentiometric recovery compares well with the polarographic results. Interference in the potentiometric titration will arise from the presence of any material which obscures the peroxide inflection. Materials of acidity similar to the peroxide may interfere while those of sufficiently different acidity can be corrected for. Other studies on the titration of Teak acids have shown that mater in the solvents can be tolerated up to a 1% level (6). However, the work reported here bears out the findings of Aloss, Elliott, and Hall (6): For clear and reproducible titration curves, even the last traces of water should be removed. LITERATURE CITED

(1) Dickey, F. H., Raley, J. H., Rust, F. F . , Tresede, R. S., Vaughan, IT. E.. I n d . Eno. Chem. 41. 1673 (1949).’ Everett, A. J., Minkoff, G. J., Trans. Faraday SOC.49,410-14 (1953). Iiolthoff, I. M.,Medalia, 8 . I., J . Am. Chem. SOC.71, 3789 (1949). Lang, K. A , “Handbook of Chemistry,” p. 037, Handbook Publishers, Inc., Sandusky, Ohio, 1934. l l o s e , 31. L ,Elliott, J. H., Hall, R. T., ANAL.CHEJI.20, 784-8 (1948). Wollish, E. G., Schmall, Pifei, C. W,, 11..J . A m . Pharm. Assoc. (Sci. Ed.) 42, 509 (1953). ANAL. CHEX.2 8 , 679 (7) Riddick, J. -I., (1956). (8) Tobolsky, A. V., Mesrobian, R. B., “Organic Peroxides,” Interscience, n’ew York, 1954. (9) wolf, Robert, Bull. SOC. chim France 1954, 644-6. RECEIVED for review July 20, 1956. Accepted September 17, 1956. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., February 28, 19%.

VOL. 29, NO. 1, JANUARY 1957

81