equilibrium shifts to halo complexes because of the low solvating power of acetone. Recently work similar t o our unpublished work regarding the equilibria of cobalt-halo complexes in acetone medium was reported ( 4 ) . Cobalt-8-quinolinol-halide complexes show maxima at 365 and 700 mp. The absorbance of the cobalt-8-quinolinol complex is a t a maximum when the acidity of the solvent is greater than 0.75N in HCl. Therefore all the solutions were prepared in 1N HClacetone. Beer’s law was studied by using solid 8-quinolinol complexes; i t holds below 20 pg. per ml. of metal radical. The molar absorptivities are given in Table I. Beer’s law also holds good for mixtures of 8-quinolinol complexes of cobalt and nickel in the same concentration range. The accuracy of the experiment was kept constant by using a 4-cm. path for absorbance measurements a t 700 mp. For the comparative analysis, samples of known composition (British Drug Houses, Analar) were used. The exact
Table 111.
Sample I I1 I11 IV V
Analysis of Samples Using 8-Quinolinol Co, Fg./lO ml. Ni, pg./lO ml.
Taken 17.6
Found Taken Found 16.4 10.4 10.2 44.0 44.5 10.4 10.2 88.0 89.1 10.4 10.5 105.6 106.5 10.4 10.2 132.0 133.7 10.4 10.3
concentrations of the stock solutions were determined with sodium diethyldithiocarbamate (3). From Table I1 it is obvious that Beer’s law does not hold good above 50 pg. of cobalt. The same stock solutions were used for the analysis using 8-quinolinol (Table 111). It will always be necessary to separate cobalt and nickel from the interfering radicals like iron and chromium, which form complexes in the same p H range. It is possible t o extend the present method to iron-cobalt and chromiumcobalt systems, as all these complexes
absorb at 365 mp while only cobalt also absorbs a t 700 mp. I n the present method, the solution of simultaneous equations are simplified, as only cobalt absorbs at 700 mp while the absorptivity of the nickel species is equal t o zero at 700 mp. ACKNOWLEDGMENT
Thanks are due to S. K. K. Jatkar for his interest in the work and for facilities given during the present work. LITERATURE CITED
(1) Am. Chem. Soc., “Reagent Chemicals,” p. 180, 1950.
(2) Berg, R., 2. anal. Chem. 76, 191 (1929). (3) Chilton, J. M., ANAL.CHEM.25, 1274 (1953). (4) Fine, D. A., J. Am. Chem. SOC.85, 1139 (1962). (5) Gustin, V. K., Sweet, T. R., ANAL. CHEW33,1942 (1961). (6) Lingane, J. J., Kerlinger, H., IND. ENG.CHEM.,ANAL.ED.13,77 (1941). (7) . , Yoe, J. H., Jacobs, W, D.. Anal. chim. Acta 20,‘332 (1959) RECEIVEDfor review March 29, 1962. Accepted September 21, 1962.
Determination of Trace Amounts of C& Acetylenes in Hydrocarbons. Hydration to Carbonyls and Determination by UItravio let Spectrophotometry of 2,4-Dinitrophenylhydrazones M. W. SCOGGINS and H. A. PRICE Phillips Petroleum Co., Bartlesville, Okla.
F A method is described for the deacetylenes with termination of C& both terminal and internal acetylenic bonds in concentrations as low as 5 parts per million. Total acetylenes are determined by catalytic hydration to carbonyls and conversion of the carbonyls to the corresponding 2,4-dinitrophenylhydrazones which are preferentially extracted and measured spectrophotometrically. All interferences tested except that of isoprene can b e eliminated. In the concentration range of 5 to 500 p.p.m., the 4 and 5 carbon atom acetylenes give an average recovery of 96%.
I
petrochemical processes, a knowledge of the amount of alkyne hydrocarbons in process streams is important. Alpha acetylenes are usually determined by adding a n excess N MANY
48
ANALYTICAL CHEMISTRY
of metallic ion and titrating either the acid formed in the reaction or the excess metallic ion. Dialkylacetylenes are rarely determined, especially if present in very low concentration, because suitable methods are not available. Wagner, Goldstein, and Peters ( 5 ) determined total Cd and CS acetylenes in liquid hydrocarbons by catalytic ketal formation followed by hydrolysis of the ketals in a hydrosylamine hydrochloride solution. The liberated hydrochloric acid was a measure of the acetylenes. This procedure is timeconsuming and not suitable for acetylene determinations in the parts per million range. Siggia (3) hydrated acetylenes t o carbonyls with a mercuric sulfatesulfuric acid catalyst and determined the carbonyl derivatives by the hydroxylamine hydrochloride procedure. The method is accurate and free of inter-
ferences but is not suitable for microgram quantities of acetylenes. Toren and Heinrich (4) suggested the extraction of 2,4-dinitrophenylhydrazones with iso-octane as a possible general procedure for the separation and determination of carbonyl compounds. Lohman ( 1 ) determined several aldehydes and ketones by condensation with 2,4-dinitrophenylhydrazine, selective estraction of the 2,4-dinitrophenylhydrazone from excess reagent with hesane, and measurement of the absorbance of the solution at 340 mp. I n this procedure, acetylenes in cyclohexane solution are hydrated to carbonyls with a mercuric sulfatesulfuric acid catalyst. Mercuric ion is removed by addition of sodium chloride and the carbonyls are extracted with 2,4-dinitrophenylhydrazine (DNPH) in sulfuric acid solution. The carbonyl
2,4-dinitrophenylhydrazones are selectively extracted into the cyclohexane phase and measured spectrophotometrically a t 340 mp. The reactions involved in the determination are:
Table 1.
Acetylene Butyne-2 Butyne-2 Butyne-2 Butyne-2 Butyne-2 Butyne-2 Pentyne- I}
Total Acetylenes by Hydration
Solvent Cyclohexane Cyclohexane Cyclohexane Cyclohexane Cyclohexane cyclohexane
Acetylene, p.p.m. Added Found 4.4 4.1, 3.9, 4.1 6.4 6.2, 6.2, 6 . 0 22.1 23.8,20.5,22.6 66.0 67.0,68.0,67.0 132 120, 114, 116 475, 500, 475 502
Recovery, % 91.0 98.4
100.8 101.8 88.8 96.2 Av.
06.1
+
R-cH,-C-R (H)
Table II. Effect of Aromatics and Olefins I
Acetylene Butvne-2 Pentyne-2 But%e-2 Butyne-2
H
)
a
The 2,4-dinitrophenylhydrazine acid salt is insoluble in cyclohexane. Acetylenes with internal and external acetylenic bonds react in this manner; therefore the procedure measures total acetylenes. Alpha-acetylenes can be determined by silver nitrate treatment of the original sample and titration of the liberated nitric acid. Betaacetylenes are obtained by difference. EXPERIMENTAL
Reagents and Apparatus. 2,4-Dinitrophenylhydrazine solution (DNP H ) is prepared by saturating 3 M &So4 with 2,4-dinitrophenylhydrazine. T h e solution should be kept cold and prepared fresh daily. Mercuric sulfate catalyst is prepared by dissolving 30 grams of HgSOl in 6.30 ml. of water containing 20 ml. of H2S04. Absorption measurements were made 011 a Beckman D U spectrophotometer in I-cm. cells. Procedure. Dilute a quantity of sample containing not more than 100 pg, of acetylenes to 10 ml. with spectrograde cyclohexane in a 40-ml. screwcap vial equipped with a polyethylene gasket in the cap. Make a blank determination simultaneously with 10 ml. of cyclohexane. Add 10 ml. of catalyst solution and allow the solutions to react for 1 hour a t room temperature with continuous mixing. Add approximately 0.5 gram of sodium chloride to remove the mercuric ion from solution as the slightly ionized mercuric chloride. A white precipitate appears a t this point if the sample contains an appreciable quantity of olefins. If the precipitate is not removed, results will be low. Remove the DreciDitate as follows: Centrifuge the vials until the solid mass settles to the bottom of the vial. Carefully withdraw 5 ml. of each phase and transfer it to another vial. Add 5 ml. of cyclohexane to bring the organic phase back to its original volume. Proceed now as if the solid did not form. L
A
Solvent Cvclohexane. hexene-1 Cyclohexane; hexene-1 Cyclohexane, benzene
Added 26.5 26.5 13.0
Acetylene, p .p,m, Found, av.a 25.8 27.0 12.3
Range 1.5 0.0 1.1
Average of at least three determinations.
Add 10 ml. of D N P H reagent solution and allow the mixture to react for 30 minutes a t room temperature with continuous mixing. Allow the two phases to separate and withdraw portions of the cyclohexane phase of the sample and blank. Measure the absorbance of the sample us. the blank in I-cm. cells a t 340 mp. Determine the acetylene concentration from a previously prepared calibration curve. Calibration Curve. Prepare a solution of methyl ethyl ketone in cyclohexane to contain 10 pg. of ketone per ml. of solution. Transfer aliquots of the solution to 40-ml. screw-cap vials and dilute to 10 ml. Add 10 ml. of D N P H reagent solution and continue as outlined in the above procedure. Plot the absorbance us. the corresponding carbonyl concentration. I n using the curve, the acetylene concentration is obtained from the carbonyl concentration by multiplying by the proper molecular weight ratios. DISCUSSION
The data in Table I indicate the extent to which the CrC6 acetylenes are quantitatively hydrated and recovered as carbonyls in the microgram concentration range. The values given in the last column indicate typical recoveries from prepared blends of acetylenes in cyclohexane. The data indicate an accuracy and precision of 1 0 . 8 and 0.5 p,p.m., respectively, for levels less than 100 p.p.m. For the entire range studied, the accuracy and precision are =t6.2 and 2.5 p.p.m., respectively. The hydration reaction employed was similar to that described by Siggia (3) except for temperature and catalyst concentration. The carbonyl determination described by Toren and
Heinrich (4) was modified by omitting the alcohol and substituting sulfuric acid for phosphoric acid. These changes result in more stable solutions and lower blank values. Suspected interference of mercuric ion with the carbonyl determination was confirmed by experiment. A finely divided solid that did not settle formed slowly in the cyclohexane phase when the catalyst solution and the D K P H reagent were added together in the presence of cyclohesane. Addition of chloride ion after the hydration step removed the mercuric ion from solution as the very slightly ionized mercuric chloride. Table I1 s h o w the effect of aromatics and olefins on the method. Olefins in concentrations up to 5% do not interfere if the solid olefin-HgC12 complex is removed from the solution by R-ithdrawing an aliquot of each of the two phases after the addition of sodium chloride and centrifugation. When this step was omitted, the olefin-acetylene blends (Table 11) analyzed approximately 13 p.p.m. acetylene. An aliquot of both phases is necessary, as carbonyls containing less than 5 carbons are almost completely miscible with aqueous solutions. Benzene in concentrations up to 10% does not interfere seriously. I n the presence of isoprene, the mercuric sulfate catalyst produced a bright yellow precipitate that was quite soluble in the cyclohexane phase and led to high results. All attempts to modify this condition failed. Other conjugated systems should behave similarly. Carbonyls and compounds which hydrolyze to carbonyls constitute an interference. This interference can be VOL. 35, NO. 1, JANUARY 1963
49
eliininated, however, by applying a correction after a spectrophotometric measurement of carbonyl without the hydration step. I n the present method, butanone was used to prepare the calibration curve. Other carbonyls be used, holyever, as the at 340 nlp of saturated-carbonyl 2.4-dinitro-
phenylhydrazones ( I , 2 ) are independent of the carbon chain length up to a t least 10 carbon atoms. LITERATURE CITED
(1) Lohman, F. H., ANAL. CHEM.30,
972 (1958). (2) Roberts, S. D., Green, Charlotte, J . A m . Chein. SOC. 68, 214 (1946).
(3) Siggia, Sidney, ANAL.CHEM.28, 1481 (1956). (4) Toren, P. E., Heinrich, B. J., Ibzd.,
27,1986 (1955).
( 5 ) Wagner, C. D., Goldstein, T., Peters, E. D., Ibid., 19, 103 (1947).
RECEIVEDfor review June 28, 1962. Accepted November 5, 1962. Division of Analytical Chemistry, 134th Meeting, ACS, Chicago, Ill., September 1958.
Chemical Microdetermination of Phenyl- and ToIyIsuIfonyIurea Derivatives in BIood WERNER KERN' Deparfment o f Chemical Research, Hoffmann-La Roche, Inc., Nutley, N. J .
b A novel quantitative spectrophotometric micromethod for the determination of phenyl- and tolylsulfonylurea derivatives in blood i s described. Following their selective extraction from the blood with organic solvents at pH 5, the compounds are subjected to nitration, followed by reduction to aromatic amines which are diazotized and coupled with N(1 -naphtyl) ethylenediamine to produce azo dyes which are measured spectrophotometrically. Depending on the specific analog the limits of sensitivity of the assay in blood were 5 to 26 pg. per ml. Blood levels of 5 representative analogs are reported, measured in rats as a function of time after oral doses of 100 mg. per kg. The general applicability of the principle of this procedure to the determination of aromatic compounds i s indicated.
T
HE A D J - C ~ Tof
phenyl- and tolylsulfonylurea compounds as hypoglycemic agents of clinical importance has called for analytical methods of sufficient sensitiT-ity to measure concentrations of these compounds in the blood of laboratory animals and human patients following administration of therapeutic doses. I n contrast t o their precursor carImtamide, a sulfonylurea derivative, the phenyl and tolyl analogs lack reactivc functional groups on 11-hich a sensitive spectrophotometric assay could be based readily. Several procedures have been reported for tolbutamide ( 2 , 5, II), a tolyl analog. and also one for chlorpropamid (IZ),a p-chlorophenyl analog. These methods are based on the strong absorption of arylsulfonyl derivatives in the ultraviolet rpgion near 220 to 230 1 Present address, Semiconductor and Materials Division, Radio Corp. of America, Sonierville, N. J.
50
ANALYTICAL CHEMISTRY
nip. Selective extraction and absorption procedures have been proposed by these authors to reduce the high level of contaminants in blood which absorb in this wavelength region. The sensitivity, specificity, and reproducibility of the spectrophotometric determination could thereby be improved. Since the completion of our own studies in 1957, several spectrophotometric procedures have been published (3, 8, IO) and have demonstrated their usefulness in clinical blood level studies. A search conducted in our laboratory for a suitable spectrophotometric method, during which several reactions were investigated, led to a different approach. The resulting method reported in this paper is based on the introduction of a reactive functional group into the phenyl or tolyl moiety of the sulfonylurea derivative. This is achieved by controlled nitration; the nitro product is then reduced to an aromatic amine and reacted to an azo dye by diazotization and coupling. The absorbance of the solution is measured spectrophotometrically to determine the quantity of the drug originally present. Using a solution of potassium nitrate in concentrated sulfuric acid as a nitrating agent, and
Table I.
Compounds Tested
"Roche" ReferKame of compound number ences l-Cyclohexyl-3phenylsulf onylurea 1 (71, (9) 1-Neobornvl-3-p-told. " sulf onylirea 11 (1) l-n-Butyl-3-p-tolylsulfonylurea I11 (9) l-Cyclohexyl-3-ptolylsulf onylurea IV (9) 1-Tetrahydrofurfuryl3-p-tolylsulf onylurea (6)
stannous chloride or metallic zinc in an acidic aqueous medium as reducing agents, it became possible to carry out this two-step reaction with microgram quantities of the compounds and to obtain quantitative and reproducible results. The optimal experimental conditions were established with the phenyl derivative 1-cyclohexyl-3-phenylsulfonyl urea. Subsequently, the same procedure gave equally good results with a series of other analogs, all tolyl derivatives. The compounds tested are listed in Table I. The analytical sensitivity varied with the compounds and ranged from 5 to 26 pg. per ml. in blood. Specificity was achieved by a selective twostep extraction based on the weakly acidic properties of the sulfonylurea moiety of these compounds. Applying this method, blood levels in rats of the five analogs listed in Table I were determined following oral administration of the drugs; the results of the study are included in this paper. Beyond the specific application and usefulness of the reported method in studies of sulfonylurea compounds, the elaboration of procedures permitting the reproducible and quantitative introduction of a reactive amine function into phenyl derivatives on a microgram scalp represents a contribution of potentially broad applicability in analytical chemistry. EXPERIMENTAL
All reagents used are of chemically pure grade. Citrate buffer: 1111, p H 4.65. Ethylene chloride (1,2-dichlorocthane) : distilled and washed with IN ammonia and distilled water. Kitration mixture: prepared by dissolving 10 grams of potassium nitrate in concentrated sulfuric acid to give 100 ml. of solution; the mixture is stored at room temperature and has good stability if kept dry. Reducing mixture: prepared daily by disSpecial
Reagents.
