homogenize the two layers; the area under its deflection curve does not enter into the calculations. The factors range from 0.76 for butanol t o 1.27 for n-butyl acetate, emphasizing the desirability of internal calibration for mixtures involving various functional groups. The resolution in these analyses, as shown in the figures, is excellent. I n one case only-the resolution of methyl acetate and acetone- care is required in construction of reproducible base lines for the determination of the appropriate areas. The average reproducibility in these analyses ranges from st0.5 to &4.4% of the contained constituent, This range is in agreement with general experience in this laboratory. Generally in the absence of interference among the deflection maxima, the reproducibility (average deviation, per cent of contained constituent) is in the range of less than h0.5 for 60% or greater composition, and about h 1 0 for the region of 1.0% content. The lower limit of detection for these constituents varies from 100 to 500 p.p.m., depending somewhat on the retention times. Usually, the sensitivity decreases with increasing retention time. CONCLUSIONS
Gas chromatographic techniques are applicable in the separation and analysis of mixtures obtained in the study of azeotropes. This feature is particularly desirable in the determination of the number of constituents present in fractions from distillations of mixtures of unidentified constituents. The technique is useful in conjunction with other instrumental analyses. Because azeotropes do not interfere, gas chromatography is readily adapted t o the analysis of mixtures involved in studies of the composition of azeotropes by analytical distillation.
Table II.
Gas Chromatographic Analysis of a Mixture of Methanol, Acetone, and Methyl Acetate
Av. Dev., Analysis, Area Constituent AV. Max. dev. Methanol 25.2 10.8 Acetone 7.9 -0.4 Methyl acetate 66.9 sto.9 Perkin-Elmer Vapor Fractometer, Model 154. 0 Four determinations. Table 111.
IV.
Av. dev. 0.5 0.2 0.6
Contained Constituent 2.0 2.5 0.9
Gas Chromatographic Analysis of a Mixture of 2-Propano1, Methyl Ethyl Ketone, and Chloroform
Analysis, Area Constituent Av. Max. dev. 2-Propanol 27.9 -0.5 Methyl ethyl ketone 64.1 $0.6 Chloroform 8.0 -0.2 Perkin-Elmer Vapor Fractometer, Model 154. 4 Five determinations. Table
70 ?f
%a
%a
Av. dev. 0.3 0.3 0.1
Av. Dev., 76 of Contained Constituent 1.1 0.5 1.2
Gas Chromatographic Analysis of a Mixture of Water, 1-Butanol, n-Butyl Acetate, and Di(n-butyl) Ether (Reproducibility data, four scans on sample of known composition)
Average Deviation Known Analyzed Calibration Constituent Weight, % Area, % Factor" Water 30.0 36.3 0.83 1-Butanol 15.0 19.7 0.76 n-Butyl acetate 45.0 35.4 1.27 Di(n-butyl) ether 10.0 8.6 1.16 Perkin-Elmer Vapor Fractometer, Model 154. a Known weight percentage divided by area percentage. LITERATURE CITED
(1) Browning, L. C., Watts, J. O., ANAL.
CHEW29, 24-7 (1957). (2) James, A. T , Martin, A. J. P., Analyst 77, 915-32 (1952). (3) James, A. T.9 A. J. p., Biochem. J. 50, 679-91 (1952). (4) James, D. E., Phillips, C. S. G., J. Chem. SOC.1953, 1600-11.
Absolute
% 1.4 0.4 1.2 0.3
% pf
contained constituent 3.9 2.0 3.4 3.5
(5) Phillips, C. S. G., Discussions Faradalp SO^. 1949,241-8. (6) Ray, N. H., J . A p p l . Chem. 4, 21-5 (January 1954). (7) TamPlin, s., Grimm, R. C.9 Crandall, J. W., unpublished data. RECEIVEDfor review April 29, 1957. Accepted October 24, 1957. Division of Analvtical Chemistry, 132nd Meeting, ACS; New York, N. Y., 1957.
w.
fer ;t- B uty I Hypochlorite for Detection of Nitrogenous Compounds on Chromatograms DANIEL P. SCHWARTZ and MICHAEL J. PALLANSCH Eastern Utilization Research and Development Division, Agricultural Research Service, o f Agriculture, Washington, D. C.
b A large number of nitrogen-containing compounds of biological interest may be readily detected on paper chromatograms by forming their N-chloro derivatives with tertbutyl hypochlorite and subsequently spraying with starch-potassium iodide.
U. S. Department
The procedure is less tedious and hazardous than other chlorinating techniques.
S
METHODS have been reported for forming and detecting on filter paper the N-chloro derivatives EVERAL
of organic nitrogen compounds. Chlorine gas and a starch-potassium iodide solution were used in the original procedure of Rydon and Smith (?'), whereas Reindel and Hoppe (5, 6) used chlorine dioxide and chlorine as N-chlorinating reagents and benzidine or o-toluidine as VOL. 30, NO. 2, FEBRUARY 1958
219
indicators. However, these methods are somewhat inconvenient and may leave a high background. Sodium hypochlorite has been used as a n A'chlorinating agent, but this procedure is admittedly almost entirely specific for mono-substituted amides (4). For these reasons the authors attempted to develop a rapid, simple, and convenient method for forming LV-chloroderivatives of a wide variety of nitrogen-containing compounds of biological interest. It has been found that tert-butyl hypochlorite is a n ideal reagent for this purpose. Of 98 nitrogen-containing conipounds tested, only five were unreactive. EXPERIMENTAL
Reagents. tert-Butyl hypochlorite was synthesized b y t h e method of Teeter and Bell ( 8 ) and purified by distillation in a n all-glass apparatus. The fraction distilling a t 78" at atmospheric pressure was collected and stored a t 5" in the dark in a glass-stoppered bottle. The spray reagents used t o detect the spots formed on the chromatograms b y nitrogen-containing compounds were prepared as follows: SOLUTION I. 1 part of tert-butyl hypochlorite t o 99 parts of 10 volume yo glacial acetic acid in 1,2-dichloroethane (Eastman Kodak K h i t e Label, or equal). This solution was stable for a t least 1 month a t room temperature in ordinary laboratory light. SOLUTION 11. 1% potassium iodidelYo soluble starch containing 5 mg. % mercuric iodide as a preservative. Paper Chromatography. All chromatograms were run on 8-inch-square sheets of W h a t m a n KO. 1 paper p r e m s h e d three times in water by ascending technique. Unwashed paper gave rise t o a blue-black band about inch wide and about 2 inches from t h e upper edge of t h e paper after chromatography. Solutions of the compounds studied were made in mater, 0 5 V hydrochloric acid, or LV ammonium hydroxide. Samples were spotted 2 em. apart a t the origin, using a microliter pipet. Hydrochlorides were neutralized on the sheet with 1 pl. of 2 5 ammonium hydroxide. Where dilution v a s necessary, microliter volumes were pipetted onto a polyethylene strip and diluted t o the desired concentration with water. The chromatograms were developed by the ascending technique in 9 X 10 X 3 inch museum jars, with the solvent system methanol-water-pyridine (8020-4). When the solvent reached the upper edge of the paper (about 2.5 hours), the papers were removed, dried overnight a t 40" t o 50", and sprayed with Solution I. The sheets were allowed to stand from 5 t o 10 minutes in a hood with a good draft, then sprayed lightly and evenly with Solution 11. Under these conditions, conipounds forming N-chloro derivatives, when in the proper concentration, were immediately evident as blue or purpleblue spots on a colorless background. The circumference of the spots was
220
ANALYTICAL CHEMISTRY
traced immediately and the enclosed areas were measured with a planimeter, It was undesirable t o aerate the papers for more than 10 minutes after spraying with Solution I, as many of the compounds tested (especially most of the common amino acids) apparently form unstable N-chloro derivatives. Under no conditions should the chromatograms be heated after chlorination, Paper Electrophoresis. T h e protein zones on ionograms produced b y conventional apparatus can be speedily located in a qualitative fashion b y the technique described. After electrophoresis the paper strips were heated in a circulating air oven a t 130" C. for 20 minutes. Chlorination and spraying mere carried out as for conventional paper chromatograms, The response t o tert-butyl hypochlorite of a number of naturally occurring protein mixtures and purified proteins was studied n i t h a commercial paper electrophoresis apparatus of the Durruni type ( 2 ) . Ten-microliter samples of the protein t o he studied were placed on the Whatman 3 NM paper strips in the apparatus and subjected to electrophoresis for 16 hours a t a current of approximately 0.625 ma. per strip. Human blood serum, concentrated milk serum, and solutions of crystalline plactoglobulin, a-lactalbumin, catalase, urease, and lipase were studied in a n acetate buffer of p H 5.4, a phosphate buffer of p H 7.5, and a ),orate huffer of pH 8.6. RESULTS A N D DISCUSSION
Results of the study are presented in Table I. Concentrations are given in micrograms for approximately the lower limits of detection per square centimeter under the conditions of this investigation. The extremely low specificity of the reagent renders it practically useless for the detection of any one cIass of nitrogen-containing compounds. However, the reagent can be applied successfully after ninhydrin, and, theoretically, can be used after any reagent which n-ill not react with it, oxidize potassium iodide to free iodine, or interfere with the formation of the starch-iodine complex. Despite its low specificity the reagent is, in most instances, relatively sensitive toward the majority of compounds within a given class, except the nitrogen-containing vitamins, When compared to ninhydrin as a reagent for amino acids on paper, the tert-butyl hypochlorite-starch-potassium iodide spray is noticeably superior for detecting the basic amino acids and tryptophan. According to Auclair and DubreuiI ( I ) , the lowest detectable quantities of arginine, lysine, histidine, and tryptophan with ninhydrin on twodimensional chromatograms are 2.0, 1.5, 7 . 5 , 2.0 y , respectively. Thus, as shoivn in Table I, the hypochlorite
reagent is approximately four times as sensitive. This increased sensitivity may be used advantageously on amino acid chromatograms by applying Solutions I and I1 prior to treatment with ninhydrin. The spots due to the starchiodine complex fade with time (depending on concentration of the amino acid) and the paper may then be sprayed with ninhydrin to obtain the purple spots from the regenerated amino acids. Duplicate chromatograms may be run and sprayed separately for comparison. Although the sulfur-containing amino acids cysteine, cystine, and methionine failed to give spots a t high concentrations, lanthionine, cystathione, homocystine, cysteic acid, methionine sulfone, and methionine sulfovide were readily detectable. For the monoamino, monocarboxylic amino acids, the chain length, branching, and position of the amino group apparently influence the reactivity of the amino group to .l--chlorination, in some instances. Thus, much higher concentrations of leucine and isoleucine were needed for detection than with valine. Sorvaline nnd norleucine were relatively less reactive than leucine and isoleucine, and y-amino-n-butyric acid was approuimately 15 times as reactive as the corresponding a amino acid. S o spot was given by a-aniinoisobutyric acid even in concentration as high as 5.5 y, nhereas the beta isomer reacted a t 1.1 y. The relative insensitivity of the alpha form of these pairs to Schlorination might he used as a means of differentiating the isomers in either pair, as they are sometimes difficult to separate on paper. Tyrosine has been reported to be unreactive when chlorinated with chlorine gas ( 7 ) or chlorine dioxide (6). I n the present procedure, however, it was easily detected. The present method failed to detect hippuric acid, even in very high concentrations, whereas Rydon and Smith ( 7 ) reported a positive response with their reagent. This may be due to the relatively large size of tert-butyl hypochlorite, and the close proximity of the benzene ring to the reactive group in hippuric acid. All protein mixtures analyzed produced patterns similar to those obtained with the conventional bromophenol blue dye (3) as a protein stain. Because of the transient nature of the color produced by tert-butyl hypochlorite-starch-potassium iodide, it is of little value for quantitative no&. As barbituric acid reacts with the reagent, veronal buffers cannot be used. The reagent is more sensitive than bromophenol blue. This extreme sensitivity was a serious handicap in analyzing mixtures in which the component with highest electrophoretic mobility 1% as strongly adsorbed on the
Table I.
Approximate Lower Limits of Detection of Some Nitrogenous Compounds on Filter Paper after Chromatography
Puriiiw, Pyrimidines, and Ikrivatives Adenine hdenosine Caffeine Cytosine Guanine C;uanosiiio Hypoxant hiiie
Amount Detected, -//Sq.Cm.
2-~Irth~l-5-ethos3.1iiietli~-l-~-
amiriopyrimidinc
blis cellaneo II s Creatinine Ethanolamine Glucosamine Hippuric acid
1 5 3 7 8 8 1 1 1 4 1 5 1 2
DJ
0 0 0 0 '30 1 0 0 8
Orotic arid Theobromine Thiot hj-mine Thymine Uracil I-ric acid S a n t hine
0 9 0 6
AiminoAcid and Peptide Esters C~lvcineethyl ester Gltcine methyl ester G1ycylglyc:inc methyl ester Serine ethyl ester Tyrosine methyl ester
Amino Acids a-.llaiiine @-Alanine a-.%mino-n-butyric acid r--4miiio-n-butyric acid a-bminoisohutyric acid
0.7 I .4 0.9 1.0 1.1
Proteins" nlasc a-I,actal\)uniiii 8-Lactoglohuliii Iipase (xiieat gci,m) ('at
Urease
11iscellaneous Ace t y1 tryptophan p.4minohenzoic acid Barbituric acid Creatine Creatine phosph;ttc
1
1 I
p-A4ininoisobutyricacid -4rginine iisparagine Aspartic acid Citrulline Cysteine
3.5-4.0 ' 7.1 15.0 0.6 0.8 1.7
0 4 0 7 1 0 T O -spot,
Indole llethylguariidine Phosphoethanolaminc Pyridoxine Sulfadiazine Sulfanilamide Sulfanilic acid Tyramine L-rea
5 4 4 9
6-l fet hJ.1t hiouracil
Amount Detected, r/Sq. Cm.
Cystine
y
7 11 0 0 13 7 2 1 0 3 1 6 9 0 0 2
-4riiount Detected, -//Sq.Cm. 1 0 3 3 35 2 r s o spot,
55 y 1.1 0 5 1 2 2 8 0 4
s o Spot, TO y s o spot,
Cysteic acid Cyst athione 3,4-Dihydroxypheii~l~ila11i1ic
Glutamic acid Glutamine
50 "/ 1 0 1 4 3 .7 3 5 1 1
Amino Acid Glvciiie Gljxocyaniine Histidine Homocystiiie Hydrox) lysine Hydroxyproline Isoleucine Lanthionine Leucine Lvsine llethionine
Anioun t Detected, yjSq. Cm.
IIethionine sulfone Methionine sulfoxide Methyl histidine Sorleucine Norvaline Ornithine Phenylalanine Phosphoserine Proline Sarcosine Serine Tauiine Thiolhistidine Threonine Tryptophan Tyrosine Valine
1.7 2.0 1.8 0.6 0.5 31.2 13.0 0.4 15.0 0.4 s o spot, 51 -/ 1.5 3.6 1. 0 57.5 51.7 0.6 2.2 7.7 19.2 2.0 0.9 1.2 1.2 2.0 0.3 1. o
0.9
Peptides nL-Alanyl-DL-asparagiiie DL-A~any~-DL-methionille
Carnosine Glutathione Glycylglycirie G1vcylglycylgl~-ciiie Gficyl-L-leucine Glycyl-L-tryptophan 1,eucylglycylglycine
(7) Rydon, H.
5 .0 10.0 0.6 0.8 1.2 1.0 1.5 1 6 1.1
k., dniith, P. \V. G.,
S a t u w 169, 922 (1952).
(8) Teeter, H. A I . , Bell, I:. Syntheses 32, 20 (1952).
\\-., Org.
I~ECEIVEI) for revicn- Julv 5, 1957. cepted Scptemher 28, 1957.
hc-
Gas-Liquid Partition Chromatography of Fluorocarbons T. M.
REED Ill
Department o f Chemical Engineering, University of Florida, Gainesville, Fla.
b Several stationary solvent liquids for the gas-liquid partition chromatography of fluorocarbon mixtures have been compared experimentally. The results are presented for di(2-ethylhexyl) sebacate, n-hexadecane, CI(CF2CFC1)3CF2COOC2H5 (the ethyl ester of Kel-F acid 8 1 14), Kel-F No. 90 grease, perfluorokerosine, and perfluorotributylamine as the stationary solvent media. Fluorocarbon and chlorofluorocarbon media give better resolution of fluorocarbon mixtures
than do hydrocarbon media. A solvent phase of good general application for resolving fluorocarbon mixtures boiling below 150" C. i s CI(CF2CFCI)3CFzCOOC2Hs. Products boiling above 29" C., obtained from the reaction of CFBSF~and C~FS, were subjected to partition chromatography using these stationary solvent media. Compounds obtained as products from this reaction probably include three isomers of CbF12, two isomers of CCF14, two isomers of C7F16, and one isomer of CgF18.
of liytliocarboris a d of hydrocarbon derivativcs have been successfully sryarated by gas-liquid partition chroiiiatography n ith hydrocarbon-typr eulistaiices as the stationary solvent phasc. In addition, Evans and Tatloiv ( 6 ) Iiavc reported t h a t dinonyl phthalate on Celite at 80" C. in a toluiiiii 16 ftct long serves as a satisfactory medium for resolving conipletely a mixture containing dihydroperfluorocycloheuane (boiling point, 78" C.), inonohydroperfluorocyclohexenrs (hoil1x'rcRi.s
VOL. 30, NO. 2, FEBRUARY 1958
221
