Determination of Microgram Amounts of Isopropyl Alcohol in Aqueous Solutions SIR: A rapid and sensitive spectro photometric method for the determination of microgram amount.. of isopropyl alcohol in aqueous solutions. in the presence of large amounts of acetone, has been elaborated. Various methods for the determina tion of small quantities of isopropyl alcohol have been described. These include infrared absorption in nonaqueous solvents ( 2 ) , the oxidation t o acetone and the subsequent spectrophotometric determination of the latter ( 6 ) ,as well as spectrophotometric determinations in aqueous solutions using ceric ammonium nitrate as reagent (6, 7). None of these methods is applicable to the determinition of isopropyl alcohol in the microgram region. Moreover, the presence of acetone in overwhelming amounts interferes with the given methods. 4 spectrophctometric method for the determination of small amounts of methyl. ethyl, butyl, and amyl alcohols based on a spot test developed b y Feigl (S, 4 ) has been reported (8). This method applies the complex formed between vanadium(V) and 8-quinolinol (oxine) in acetic acid, which gives a red color in the presence of alcohols (1). This method ( 8 ) was found inapplicable t o the determination of microgram amounts of isopropyl alcohol. The vanadium-oxinate method could, however, be applied to the quantitative determination of microgram amounts of isopropyl alcohol after further modifications. The new procedure involves the preliminary preparation of a benzene solution of vanadium oxinate (SI 4 ) , the use of acetate buffer (8),and the salting out of the isopropyl alcohol-vanadiumoxinate complex by a saturated lithium sulfate solution into the benzene layer. This new technique provides a highly sensitive quantitative method with which as little as 1.5 pg. of isopropyl alcohol per ml. of original aqueous solution can be determined. Acetone and acetic acid do not interfere with this procedure, thus it may be applied as a standard quantitative test for isopropyl alcohol in acetone, isopropyl alcohol being a most common impurity in acetone.
grams of lithium sulfate and shake until in 6% acetic acid and 5 ml. of ammoa saturated solution is obtained. Adnium vanadate solution which contains 1 mg. of vanadium per ml. into a sepajust the p H to between 4 and 4.5. (If less than 10 ml. of solution are used, ratory funnel. Add 10 ml. of buffer make u p the volume t o 10 ml. with solution and 150 ml. of analytical grade benzene. Shake for 5 minutes. A4110~~ saturated lithium sulfate solution.) Add 10 ml. of the previously prepared vanato stand until the two layers separate dium oxinate reagent and shake for completely and discard the aqueous 20 minutes. After separation, discard layer. The solution should be prepared the aqueous layer. Shake the benzene fresh every day. solution with 10 ml. of 1N sodium hyLithium Sulfate Saturated Solution. droxide Rolution for 2 t o 3 minutes and Dissolve 34 grams of analytical grade discard the aqueous layer. Centrifuge lithium sulfate in 100 ml. of distilled the pink-colored benzene layer for 5 water. minutes to eliminate the turbidity. Apparatus. A Hilger spectrophoTransfer t o a 2-cm. light path cell and tometer was used. measure the absorbance a t X = 380 mp Procedure. Pipet 10 ml. of t h e in a qpectrophotometer. Calculate the aqueous solution containing from 1.5 isopropyl alcohol content of the sample t o 30 pg. of isopropyl alcohol per ml. from a standard calibration curve. into a separatory funnel. Add 3 t o 4 Calibration Curve. Weigh 0.3 gram of isopropyl alcohol, analytiral grade, transfer i t t o a 1-liter volumetric flask, and dilute with distilled water. Table I. Calibration Curve of StandPipet 1 ml. of this stock solution into a 100-ml. volumetric flask and make ard Isopropyl Alcohol Solutions up t o volume with distilled water. 10 ml. vanadium oxinate, 2-cm. cells This solution contains 3 pg. of isopropyl Isopropyl Absorbance alcohol per ml. Introduce 1 t o 10 ml. alcohol, With Without of this solution into separatory funnels pg. per ml. LitSOa Lips04 and continue as described under Procedure. 1.5 0. 025a ... A straight line calibration curve was 3.0 0.035 ... 6.0 0.060 obtained in the range 3 t o 30 fig. 12.0 0.125 0 025 If higher sensitivity is required, 15 18.0 0.17.5 0.045 ml. of vanadium oxinate may be used 24.0 0.240 0.055 for extraction and 4-cm. light path 30.0 0.300 0.070 cells for the absorbance readings. I n this way, 1.5 pg. of isopropyl alcohol a 15 ml. of vanadium oxinate, 4-cm. cells. per ml. of aqueous solution can be determined.
:
Table II. The Precision and Accuracy of the Method Isopropyl alcohol Taken, Found, Rel. std. pg./ml. pg./ml. Mean Std. dev. dev., % 6.4
10.0
936
f3.42
2.1
9.8
10.16
f0.332
f3.28
1.6
20.38
10.483
f2.37
1.86
9.6
20.6 20.8
Buffer Solution, PH 4.3.
ANALYTICAL CHEMISTRY
f0.208
10.6 10.2
19.8
To 100 ml. of glacial acetic acid add
53 grams of sodium acetate and 10 ml. of water. Heat gently t o complete dissolution. Vanadium Oxinate Solution. Transfer 5 ml. of a 2.5% 8-quinolinol solution
6.13 6.2 10.0 10.4 10.4 10.2
EXPERIMENTAL
Reagents.
Relative error, o/o
20.0
20.2 20.4 21.0 20.6 19.6
RESULTS A N D DISCUSSION
Absorbance Spectrum. T h e absorbance spectrum was examined in the range 300 t o 550 rnp. The spectrum obtained is similar t o t h a t obtained by Stiller (8) showing the same two inflections between 350 t o 400 mp and 450 to 485 mp. This indicates that the compound formed between isopropyl alcohol and vanadium oxinate is of the same nature as tha, of t h e other alcohols. When measured at 380 mp, the absorbances shown in Table I were obtained with the standard isopropyl alcohol solutions. Addition of Lithium Sulfate. With t h e addition of lithium sulfate t h e sensitivity of t h e reaction was increased 4 times (Table I ) . Lithium sulfate was found to be a superior salting-out agent compared with ammonium and sodium sulfate.
Effect of Time. T h e absorbance of two samples containing 30 and 60 p.13.m. of isopropyl alcohol, respectively, was measured every 5 minutes a n d after 1 hour, every 15 minutes. T h e results showed t h a t the color is stable for 12 hours, thus the exact time of reading after development of t h e color is not critical. Precision and Accuracy. T h e precision and accuracy of the method were found b y repeated analyses of these standard samples containing 6, 10, and 20 pg. of isopropyl alcohol per ml., respectively. The results are shown in Table 11. The relative standard deviation and relative error obtained decrease t o =tlOYo for quantities less than 3 pg. of isopropyl alcohol when 15 ml. of vanadium oxinate and 4-cm. light path cells are used.
LITERATURE CITED
(1) Buscarons, F., Marin, J., Claver, J , And Chim. Acta 3 , 310, 417 (1949). (2) Desnoyer, AI., Bull. Soc. Chim.France 1960, p. 1754. ( 3 ) Feigl, F., “Spot Tests in Organic Analysis,” 6th ed., p. 184. Elsevier, London, 1960. ( 4 ) Feigl, F., Stark, C., Microchzm. Acfa 1955, p. 996. (5) Ginther, C. R., Finch, R. C., ANAL. CHEM.32, 1894 (1960). (6) Reid, T’. W., Salmon, D. G., Analyst 80, 704 (1955). ( 7 ) Reid, V. W., Truelove. R. K., Ibid., 77, 325 (1952). ( 8 ) Stiller, M.,Anal. Chzm. Acta 25, 85 (1961).
MARIAXNA MANTEL Roreq Research Establishment Yavneh, Israel ~ ~ I I C H A E L AXBAR
Soreq Research Establishment, Yavneh, Israel and The Weizmann Institute of Science Rehovoth, Israel
Precise Microdetermination of Cobalt SIR: We wish to report a precise, microanalysis scherr e based on differential spectrophotometry for the determination of cobalt at the 2-pg. level in aluminum-cobalt wire monitors t h a t are used to measure the neutron flux in nuclear r e a c t m . This method provides for improved relative standard deviation of flux measurement data, previously limited to *6Yo by the nonuniform distribution of cobalt in the monitor wire and by the precision of the method of activation analysis. Application of isor opic dilution and solvent extraction based on the work of RGiiEka and Star? (6, 7 ) sho&-ed poor precision. Microcoulometric analysis failed because a satisfactory electrolyte could not be found. The only other alternative appeared to be a sensitive, highly precise spectrophotometric analysis. Nitroso-R salt (sodium-l-nitroso-2hydroxynaphthalene - 3, 6 - disulfonate) was selected as chelating agent because it is one of the mosl, sensitive (0.0019 pg. per sq. cm.) and selective spectrophotometric reagents For cobalt. Several other stable chelating; agents have been suggested recently for the determination of cobalt (1, 2, 6, 113). Only one (1) approximates the sensitivity and selectivity of nitroso-R salt. EXPERIMENTAL
Reagents. Deionized water was used for all reagent preparations. Buffer solution was 36 grams of disodium phosphate dihydrate, 6.2 grams of boric acid, a n d 500 ml. of 1.00N sodium hydrcxide per liter. Procedure. Each wire was weighed (-40 mg., 0.05 t o 0.1% Co, O . O l ~ o
impurities) a n d then dissolved in 3 ml. of 6 M hydrochloric acid. After adding phosphate and adjusting t h e p H to 3.5, 5 ml. of 10-gram-per-liter 1-nitroso-2-naphthol in 1: 1 acetic acid were added, and the solution was allowed to stand 1 hour as recommended by Saltzman (8). The complex was then extracted with three 8-ml. portions of chloroform. The organic matter was destroyed by fuming with 1 ml. of nitric acid and 0.75 ml. of perchloric acid. The mixture of nitric and perchloric acid contained 50 mg. of sodium nitrate to minimize sorption of cobalt on the glass vessel. After the l-nitroso2-naphthol was destroyed, the volumetric dilution of the residue to 10 ml. produced a clear solution. This result is in contrast to the work of Saltzman who reported the incomplete decomposition of 1-nitroso-2-naphthol by a nitric-perchloric acid mixture and the formation of a n oily yellow product. Aliquots were analyzed by a modified nitroso-R salt method. An aliquot containing approximately 2 pg. of cobalt was pipetted into a 10-ml. volumetric flask. One milliliter of 0.2M citric acid (to prevent metal hydroxide precipitation) and then 1.2 ml. of buffer were added. The p H was measured with test paper and adjusted to 7.6 & 0.4 with 1X sodium hydroxide, After addition of 200 pl, of 0.570 aqueous solution of nitroso-R salt, the solution was allowed to stand at room temperature 30 minutes. The p H value was chosen to ensure that the reaction with nitroso-R salt would be complete in the allotted time at ambient temperature. One milliliter of nitric acid was pipetted into the flask, the solution diluted, mixed, and the flask heated in boiling water for 10 h 0.1 minutes. Standards of 1 and 5
pg. of cobalt ( + l % re]. std. dev.) were prepared by the same method as the samples for references to adjust the Deckman Mode! D l i spectrophotometer to read 100 and 0% transmittance according to method IV of Reilley and Crawford ( 4 ) . Measurements were made a t 420 mp. RESULTS AND DISCUSSION
The average recovery for five synthetic samples containing 4 p g . of cobalt and 9,750 c.ps. of Co60 was 100.1%. The chloroform extractant was counted in a 3- X 3-inch N a I (Tl) crystal connected to a 256-channel analyzer. The relative standard deviation was h0.6% for a single extraction; the contribution of counting statistics to this error in recovery was =t0.4%. There is disagreement in the literature ( 2 , 3, 8, 9) about the optimum conditions for the nitroso-R salt method. These variables were examined to optimize the precision of the method. Stable Color Production. I n this modified method, maximum color development is attained in 15 t o 20 minutes without heating or exact timing. Heating in a water b a t h or boiling t h e solution t o develop t h e chelate complex as suggested in some methods ( 8 , 9) produced variations in color for a given standard that were easily detectable by method IV (4) in the 1 to 5 pg. range and could not be reproduced. Whether chloride ion can be tolerated (3) or not (9) in the determination is unimportant since chloride is removed in the fuming step after cobalt extraction. VOL. 36,
NO, 4, APRIL 1964
937
