Determination of Carbonyl Compound by Extraction of Its 2, 4

Determination of Carbonyl Compound by Extraction of Its 2,4-Dinitrophenylhydrazone. P. E. Toren, and B. J. Heinrich. Anal. Chem. , 1955, 27 (12), pp 1...
0 downloads 0 Views 343KB Size
Determination of a Carbonyl Compound by Its 2,6Dinitrophenyl hydrazone PAUL E. TOREN and B. J. HEINRICH Okla.

Research Division, Phillips Petroleum Co., Bartlesviile,

A spectrophotometric determination of butadiene-

agent absorbs in the same region as the 2,4-dinitrophenylhydrazone product and, thus, interferes with the spectrophotometric measurement of 2,4-dinitrophenylhydrazone concentration. The addition of alkali shifts the absorbance of the 2,Pdinitrophenylhydrazone into the visible region and removes the spectral interference caused by the reagent. If a method for separating the product from the reagent could be found, the reagent would no longer interfere with the direct measurement of the 2,Pdinitrophenylhydrazone concentration, and the formation and measurement of the unstable red color would no longer be necessary. During the evaluation of the published methods described, it was noticed that the 2,4-dinitrophenylhydrazoneof the carbonyl compound being studied is much more soluble in iso-octane (2,2,4-trimethylpentane) than is the 2,4dinitrophenylhydrazine reagent. This difference in solubility was utilized to develop an extraction procedure which permits the direct spectrophotometric measurement of 2,4-dinitrophenylhydrazone concentration without the addition of alkali. This procedure was developed as a method for the determination of 2.3,4,5-bis(AZ-butenylene)tetrahydrofurfural,a butadiene-furfural condensation product whose structure is shown in Figure 1 ( 1 ) . This compound is an effective fly repellent known commercially as R-11. The procedure is discussed with reference to this particular compound; however, it is believed that the principle of the extraction procedure described is generally applicable to the determination of carbonyl compounds.

furfural condensation product, a carbonyl compound, can be made by reaction of the compound with 2,4dinitrophenylhydrazine in a two-phase system composed of iso-octane and an alcohol-water-phosphoric acid mixture. The 2,4-dinitrophenylhydrazoneis selectively extracted into the iso-octane phase and its concentration determined by absorption measurement at 340 mp. The excess reagent remains in the aqueous phase and does not interfere.

T

HE use of 2,4dinitrophenylhydrazine for the identification

and determination of carbonyl compounds is well known. I t was not until relatively recently, however, that this reagent w u applied to the determination of parts per million quantities of carbonyl compounds. Two methods were published in 1951 based on the formation of 2,4dinitrophenylhydrazones followed by the addition of alcoholic potassium hydroxide to produce a red color which could then be measured spectrophotometrically.

REAGENTS

Eastman 2,4dinitrophenylhydrazine and Phillips Reference Fuel or Spectro Grade iso-octane are used without purification. Carbonyl free ethyl alcohol is prepared by refluxing 600 to 700 ml. of absolute alcohol for 2 hours with 5 grams of 2,4-dinitrophenylhydrazine and a few milliliters of concentrated hydro-

Figure 1. Butadiene-furfural condensation product (R-11)

In the method of Lappin and Clark ( 2 ) the 2,4dinitrophenylhydrazone is formed by the direct reaction of the carbonyl compound with 2,4-dinitrophenylhydrazine in slightly acid methanol solution. The evaluation of this method by the authors of the work reported here indicated that the reaction conditions specified do not give reproducible results. After trying a number of modifications of the published procedure, including changes in reaction time, acid strength, reaction temperature, and solvent, they concluded that this method is not suitable for their purposes. In the procedure of Pool and Klose ( 3 ) the 2,4dinitrophenylhydrazine reagent in benzene solution is adsorbed on an alumina column. The sample, also in benzene solution, is added to the column and the 2,4-dinitrophenylhydrazones formed are eluted by a further addition of benzene. Although it was found that the activity of the alumina column is critical, this procedure appeared to be reliable and could probably have been adapted t o the use of the work reported. During the evaluation of these methods, however, another approach presented itself which was more readily applicable to the particular problem. The red color formed by the addition of alkali to the 2,4dinitrophenylhydrazone solutions in the methods described is not stable and fades appreciably within a few minutes after its formation. The formation of the red color is necessary in these methods, however, because the 2,4-dinitrophenylhydrazine re-

1986

1.5

I

1.0 W

x

U

m a 0

m

m

< 0.5

n l

I

250

300 WAVE LENGTH

Figure 2. A.

I

350

400

(Mp)

Spectra of typical R-11 solution and blank

Absorption spectrum of R-ll-2,4-dinitrophenylhydrazone (approximately 6 X 10-6,M)

B. Blank with same reagents

1987

V O L U M E 27, NO. 1 2 , D E C E M B E R 1 9 5 5

0 Figure 3. blank) of

I

2 3 4 MOLAR CONCENTRATION 11-11

5

(x

8

105)

Relationship of absorbance (minus R-ll-2,4-dinitrophenylhydrazone us. concentration of R-11

0 Absorbance of blank = 0.27 0 Unpurified alcohol used in reagent, absorbance of blank

= 1.03

chloric acid, and then distilling off the alcohol. The 2,Pdinitrophenylhydrazine reagent solution is prepared daily by mixing 1 volume of a fresh saturated solution of 2,Cdinitrophenylhydrazine in carbonyl-free alcohol with 3 volumes of 1 to 2 phosphoric acid (a solution of 1 volume 85% phosphoric acid in 2 volumes of water).

aqueous phase consists completely of iso-octane, all the alcohol being found in the aqueous phase. The diagram also shows t,hat mixing an aqueous phase containing less than 50% alcohol with an iso-octane phase causes only a small decrease in the volume of the iso-octane. The effect of the composition of the aqueous phase on the relative solubility of the 2,4dinitrophenylhydrazone in the iso-octane was measured by determining the fraction of a known amount of 2,4dinitrophenylhydraeone retained by the iso-octane in the presence of aqueous solutions containing different proportions of water and alcohol. The results of these measurements are shown in Figure 5. When the aqueous phase contains 70% alcohol, less than half of the 2,4-dinitrophenylhydrazoneis retained by the iso-octane, with 50'% alcohol the retention is about SO%, and when the alcohol content of the aqueous phase is 25% or less, essentially all of the 2,4-dinitrophenylhydrazoneis retained in the iso-octane phase. Consequently, the aqueous phase used in this determination contains 25% ethyl alcohol. With the Cary recording spectrophotometer used to obtain the

100% ETOH

A

PROCEDURE

Add 10 ml. of iso-octane containing 1 to 20 p.p.m. of the carbonyl compound to a 40-ml. screw-cap vial containing 10 ml. of the 2,Pdinitrophenylhydrazine reagent solution. Run a duplicate procedure as a blank with 10 ml. of pure iso-octane in place of the sample. Allow the solutions to react for 30 minutes with continuous mixing. In this laboratory a motor-driven rotator is used which turns t e vials end over end a t about 50 r.p.m.) Remove the reagent phase with a suction tube and wash the octane phase for 10 minutes with an equal volume of 1 to 2 phosphoric acid. Withdraw portions of the octane phases of the sample and the blank and measure the absorbance of the sample versus the blank a t the wave length of maximum absorbance (340 mfi for R-11). Determine the carbonyl concentration from a calibration curve prepared from samples containing known amounts of the substance being determined.

6

RESULTS

Figure 4.

Figure 2 shows the spectra of a typical R-11 solution and blank after treatment by the procedure described. The blank involved in this method is high, but it has been found to be constant in a series of determinations using the same reagent solutions. This is illustrated in Figure 3, which is a plot of the absorbance of the sample minus that of the blank versus carbonyl concentration. The linearity of the plot shows that the blank, although high, was constant over the entire series of measurements. The point on Figure 3 marked with a equare was obtained from a measurement in which the reagent solution had been made up with unpurified alcohol. The blank had an absorbance (versus iso-octane) of more than 1.0 as compared with a normal blank of about 0.3, but the difference between the absorbances of sample and blank was such that the point fell on the same calibration line.

Ethyl alcohol-iso-octane-water,

DISCUSSION

A number of solubility relationships are involved in this method, the first of which is that among the three solvents themselves. -4qualitative phase diagram of the system iso-octaneethyl alcohol-water was constructed (Figure 4) showing that in the presence of even a relatively small amount of water, the non-

Phase diagram temperature 30' C.

~~.

-

1988

ANALYTICAL CHEMISTRY

results above, duplicate blanks from a given reagent solution are reproducible to about 0.02 absorbance units. For this reason, the maximum sensitivity claimed for this determination is about 5 X 10-OM R-11, which gives an absorbance of about 0.1. The sensitivity of the method can be increased by concentrating the carbonyl compound in the iso-octane solvent before performing the analysis. This procedure is being used to determine concentrations of 1 X 10-eJ4 R-11 in milk by extracting 4 volumes of milk with one of iso-octane. The data plotted in Figure 3 show that the calibration line for R-11 is independent of the reagent solution as long as the simultaneous blank determination is also made. Consequently, a single calibration is all that is required for the analysis of any number of samples over any period of time. The work described is concerned with only one compound, and the general applicability of this procedure to all carbonyl compounds has not been experimentally verified. However, a number of semiquantitative experiments with other aldehydes and ketones indicate that extraction by iso-octane under the conditions of this method is probably a general property of 2,4-dinitrophenylhydrazones. In adapting this procedure t o the determination of a differeht carbonyl compound, a new calibration is necessary, because the absorption maxima and molar absorptivities of the 2,4-dinitrophenglhydrazones vary with the structure of the carbonyl compound ( 4 ) . The rate of

reaction may also vary and the reaction time should be determined when calibrating for a different carbonyl compound. ACKNOWLEDGMENT

The authors wish to express their appreciation to the Phillips Petroleum Co. for making possible the publication of this paper. NOTE:

Since the presentation of this paper, it has been found that in some instances-for example, the determination of traces of acetone in hydrocarbons-it is possible to use 1 t o 2 phosphoric acid alone in the aqueous phase, thus eliminating the need for repurified alcohol. The rate of the reaction may be decreased somewhat by this change, so the mixing time should be checked to be certain that the reaction is complete in the allotted time. LITERATURE CITED

(1) Hillyer, J. C., Swadesh, S., Leslie, M. L., and Dunlap, A . P., Ind. Eng. Chem., 40,2216-20 (1948). (2) Lappin. G. R., and Clark, L. C ,ANAL.CHEM.,23, 541 (1951) (3) Pool, 11.F..and Klose, A. A . , J . Am. Oil Chemists' Soc., 28, 215 (1951). (4) Roherts, J. D., and Green, C., J . A m . Chem. SOC., 68,214 (1946).

RECEIVED for review May 12, 1955. Accepted September 6. 1955. Sixth Annual Pittsburgh Conference on Analytical Chemistry a n d .Ipplied Spectroscopy, Pittsburgh, P a . , March 1953.

Sensitivity of Bromine-Bromide Potentiometric End Point WILLIAM C. PURDY, EUGENE A. BURNS, and L. 8. ROGERS Department of Chemistry and Laboratory for Nuclear Science, Massachusetts Institute of Technology, Cambridge 39, Mass.

The sensitivity of the bromine-bromide potentiometric end point has been investigated with respect to the factors contributing to the reagent blank in a coulometric titration. The size of the blank often decreased with an increase in the surface area of the indicator electrode, an increase in the rate of stirring the solution, and a decrease in the current passed by the potentiometer, thereby indicating that polarization was taking place. The blank also depended upon other experimental factors such as the concentrations of reagents, the electrode material, and the electrode pretreatment. The size of the potentiometric blank could be diminished by proper choice of the experimental conditions so that it closely approximated the corresponding amperometric blank.

A

(4)and amperometric (9) methods have been employed in coulometric titrations, the former have the inherent advantage of being more readily adaptable to successive determinations of a number of substances in a mixture by observing in a single titration separate end points for each, On the other hand, potentiometry is usually limited to systems containing a reversible half-cell reaction whereas amperometry is not. The relative sensitivities for the two methods have not been compared using a reversible half-cell reaction, although it is interesting to note that one group of investigators switched from classical potentiometry ( d ) , to a polarized potentiometric system when higher sensitivity was required ( 3 , 11, 1 2 ) . It has long been known that potentials for solutions more dilute than about l O - 7 J f frequently do not correspond to those calculated with the aid of the Kernst equation ( 5 ) . In fact, the LTHOUGH both potentiometric

observed potential has usually been found to be independent of the concentration of the substance in question. Though factors such as losses by adsorption onto walls of the container, hydrolysis, and colloid formation may contribute to the end result, control of the electrode potential may often pass over to competing reactions because polarization occurs during the measurement. The potential of a system, although presumably measured with no current flowing, is probably rarely ever measured under ideal conditions, because of the limited sensitivity of the galvanometer and the finite number of turns on the slide-wire used for balancing the bridge. In instruments of the vacuum-tube voltmeter type, a small grid current must be passed to obtain a signal. The purpose of the present study, therefore, was to examine the factors that might influence the sensitivity of a classical potentiometric end point-Le., the magnitude of the blank-on the assumption that such a potentiometric measurement would involve the passage of a small current. Then, under the same conditions as for the potentiometry, the amperometric sensitivity was determined for the same electrode, and the results were compared. EXPERIMENTAL DETAILS

Reagents and Solutions. Stock solutions of 1M sulfuric acid tvere prepared by dilution of the 96% reagent. A stock solution of 2M sodium bromide was made by dissolving the weighed amount of sodium bromide in 1 liter of distilled water. These stock solutions were used in preparing the 0.1M sulfuric acid and 0.2M sodium bromide ordinarily used as the electrolyte. A standard 0.05143M solution of trivalent arsenic was prepared by dissolving 1.2718 grams of arsenic trioxide in 1M sodium hydroxide, neutralizing with 1.V sulfuric acid, and diluting to 250 ml. Apparatus. A cell of the type reported by Myers and Swift