Separation of Diethylbenzenes by Gas Chromatography Using Packed Columns SIR: Recent publications ( I , 2 ) provided data on successful gas liquid chromatographic separations of xylene isomers utilizing a cslumn of tetraalkylammonium bentonite (Bentone-34) modified with Silicone Oil or diisodecylphthalate and supported on Chromosorb, Celite, 01’ firebrick. A separation of diethylbenzene (DEB) isomers was also obtained on a similar column. Figure 1 shows a resolution of the constitients in a commercial DEB. Studies of the separation became desirable when it was found t h a t the stability of DEB-TBP (tri-nbutylphosphate) solutions against nitration by nitric acid varied widely with the source of the DEB. Nitration of the D E B was shown when unwanted fission products extracted along with uranium when the TE P-DEB solution was used to solvent extract a solution m-hich simulated a nitric acid dissolution of a spent reactor fuel element. Controlled nitration of solutions of T B P in pure ortho-, neta-, and paradiethylbenzenes proved the meta isomer to be most stable toward nitric acid, and the para isomer t o be least stable. It was, therefore, desired to correlate the compositions of several commercial DEB’S with their extraction behavior after nitration. EXPERIMENTAL
Apparatus. An Aerograph Autoprep Model A-700 (heated metalfilament detector) gar3 chromatograph equipped with a n L&N Speedomax H recorder was used in this work. T h e column consisted of a copper tube (20 feet long and 1/4-inch 0.d.) packed with 65- t o 80-merih Celite, onto which had been abso-bed Bentone-34 and Dom--Corning h-0. 550 Silicone Oil.
reta-DEB
~
Figure 1 .
Chromatogram of commercial diethylbenzene
Division, Houston, Texas). About 90 ml. of the benzene mas evaporated from the slurry, with stirring, a t room temperature. The Bentone-Silicone Oil mixture was then slurried in 200 ml. of benzene and then mixed with 40 grams of Celite. The solvent mas evaporated to dryness slowly, with constant stirring, under a stream of air a t room temperature. A straight length of copper tubing (20 feet long and 1’4-inch 0.d.) was filled with the dry material, plugged a t each end with glass wool, and then coiled t o fit the column housing. Finally, the column was conditioned at 150’ C. for about 30 minutes. DISCUSSION
The order of elution is: para-, meta-, and ortho-DEB. The proportions of
Bentone-34 and Dow-Corning 550 Silicone Oil were important to the gas liquid chromatographic identification of the isomers. If a greater proportion of Silicone Oil is used [for example, equal parts of Uentone-34 and Silicone Oil, the proportions recommended for optimal resolution of the xylene isomers (1)1, the metu-para resolution decreases. Also, if a higher proportion of Bentone34 is used, the ortho-mefu resolution decreases. No substrates other than Dow-Corning 550 were tested in combination with Bentone-34. Mixtures of the pure isomers of DEI3 were tested and proved the feasibility of using the chromatographs for quantitative analysis of each isomer to within 0.5%. The relative areas under the curves were determined by weighing. Subsequent results of analyses of several
Operating Conditions Column temperature, initial, C. fi3al. C. rise,’degrees per minute Injection-port temperature, O C. Detector temperature, ” C. Helium flow, cc. per minute Column pressure, p.s.1. Filament current, ma. Attenuation Sample size, pi. O
70
150 -3 225 225
Table 1.
Discussion of Isomers in Commercial Diethylbenzenes Composition, yo
35
50
175 16
10
Column Preparation. Two grams of Silicone Oil (Dow-Corning 550) were dissolved in 100 ml. of benzene and then slurried with 4.24 grams of Bentone-34 (National Lead Baroid
Source of DEB Koppers Plastic Division Dow Chemical Co.-Dowetch brand Dow Chemical Co., lot 05252 Eastman Organic Chemicals (technical, meta and para mixture) T-1931
ortho
10 9 5
17
meta 50 56 61 60
para 36 30 28 14
VOL. 35, NO. 1 1 , OCTOBER 1963
Unidentified 4
5 6 9
1759
commercial DEB’S are shown in Table I. No attempt was made to identify the materials that emerged just before and after the D E B peaks. CONCLUSIONS
Because this column gare good resolution of the D E B isomers it was
possible to correlate the stability of commercial DEB’S toward attack by nitric acid with their compositions. The stability of TBP-DEB mixtures, as measured by fission product extraction, increased with increased meta and decreased para concentrations and verified the results of tests with the pure isomers.
LITERATURE CITED
( I ‘ lfortimer, J. V., Gent, P. L., S a t w e 197, 789-90 (1963). ( z ~ $ ~ ~ $ ~ ; . S a m u e lF., *‘saL. 35,
CHARLES 4.B L ~ K EJR. ,
Oak Ridge xational LaboratorS-l
Oak Ridge, Tenn. I Operated by ~~i~~ Carbide carp. for the U.S. Atoniic Energy Commission.
Spectrophotometric Determination of Fatty Amines in Aqueous Solution SIR: Spectrometric analysis of aqueous solutions of fatty amines has recently been reported (9). This method was extended by this laboratory to field use for samples containing 0 to 1.0 p.p,m. amine. We found that by combining the buffer and dye reagents, a combination-single reagent could be used. The procedure requires less sample and shows increased sensitivity over field methods previously reported which use bromophenol blue as the dye and chloroform as the extraction solvent (1). EXPERIMENTAL
Combination Reagent. Dissolve 0.1 gram of methyl orange in deionized water. Dissolve 29.6 grams of sodium acetate trihydrate and 50 grams of potassium chloride in deionized water. Combine t h e t w o solutions, add 100 ml. of glacial acetic acid, and dilute to 500 ml. with deionized water.
Table 1.
Data for Standard Curve
ODA, p.p.m. Blank (0) 0.10 0.20 0.33
0.50
0.54 0.76 1.oo 1.09
Methyl orange procedure 96 86 r-
Previous procedure 1L 00
i l
TO
53 50 40 34 30
Transmittance at 420 n u , R.
1760
ANALYTICAL CHEMISTRY
94 Stj -( I
63 65
55 47 43
Procedure. T o a 250-ml. separatory funnel add 100 ml. of the sample t o be tested. Add 5 ml. of the cornbined reagent and set aside for 10 minutes. Add 20 ml. of ethylene dichloride and shake for 5 minuter. Allow t h e layers t o separate for approximately 5 minutes. Drain off t h e lower layer and measure transmittance immediately at 420 m,u, RESULTS AND DISCUSSION
Data for a typical standard curve are shown in Table I. The curve is linear over the range of 0 to 1 p.p.m. Table I also shows the sensitivity of previously used procedures and of this method. the latter being approximately 50% more sensitive. A single reagent system presents an advantage for field use. The buffer and dye reagents were combined in proportion so that 5 ml. of the single reagent was essentially equivalent to 4 ml. of buffer and 2 ml. of dye solution. This reagent is stable for a t least 1 month without refrigeration. Silverstein (2) reports the reaction of \-arious amines to this procedure. Since this modification is similar in its reaction with octadecylamine (ODA), interferences are the same. Sample handling and sampling techniques should be the same. Silverstein showed that morphoIine and cyclohexylamine gave slightly positive results. Table I1 shows this effect to be negligible except where a relative error of better than =klOyois required. There ib only a slight variation between choloroform and ethylene dichloride, the latter
Table (I. Effect of Neutralizing Amines on ODA Determination
Amount Cj-clo- Amount Morpho- hexylODh OD.4, line, amine, recovered, p.p.m. p.p.m. p.p.m. p.p.ni. 0.0 0.0 0.54 CI 54
0.0 0.02 0.54 0.55
10
500
10
50
0 54 0 54 Ethylene
500
dichloride extraction 0 ,54 Ethylene dichloride extraction 0 54 Chloroform extraction I 09 Chloroform evtraction
0.61 0 5S
0.57 0 50 0 95
being more sensitive. Ethylene dichloride is preferred also because of its higher boiling point, LITERATURE CITED
(1) Bednar, G., Dugo, R. J., Abstracts,
Division of Water, Sewage, and Sanitation Chemistry, 134th meeting .4CS, Chicago, Ill., September 1958. ( 2 ) Silverstein, It. M., ANAL.CHEX 3 5 , 154 (1963). R. A. LARRICK
Anal>-ticalLaboratory Dearhorn Chemical Co. Chicago 9, Ill.